Watch, Follow, &
Connect with Us

Please visit our new home
community.embarcadero.com.


Welcome, Guest
Guest Settings
Help

Thread: JOEL ON SOFTWARE



Permlink Replies: 154 - Last Post: Oct 26, 2017 11:16 AM Last Post By: Tony Caduto Threads: [ Previous | Next ]
Roy Lambert

Posts: 1,056
Registered: 10/21/99
JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 11, 2017 8:06 AM
I finally forced myself to read to the end. Whilst there are some valid points if he writes code like he writes blogs I would hate to have to maintain it.

The summary for those who don't want to waste part of their life:

1. unicode is different to ansi
2. using ansi for different languages was a mess
3. unicode is a different type of mess

I think that's it.

Roy Lambert
Van Swofford

Posts: 389
Registered: 6/28/03
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 11, 2017 8:15 AM   in response to: Roy Lambert in response to: Roy Lambert
Roy Lambert wrote:

I finally forced myself to read to the end. Whilst there are some
valid points if he writes code like he writes blogs I would hate to
have to maintain it.

The summary for those who don't want to waste part of their life:

1. unicode is different to ansi
2. using ansi for different languages was a mess
3. unicode is a different type of mess

I think that's it.

Roy Lambert

Well said Roy.

--
Cheers,
Van

"Good judgment comes from experience, and a lot of that comes from bad
judgment." - Will Rogers
Rudy Velthuis (...


Posts: 7,591
Registered: 9/22/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 11, 2017 9:43 AM   in response to: Van Swofford in response to: Van Swofford
Van Swofford wrote:

Roy Lambert wrote:

I finally forced myself to read to the end. Whilst there are some
valid points if he writes code like he writes blogs I would hate to
have to maintain it.

The summary for those who don't want to waste part of their life:

1. unicode is different to ansi
2. using ansi for different languages was a mess
3. unicode is a different type of mess

I think that's it.

Bad summary. Unicode is far less of a mess than Ansi.
--
Rudy Velthuis http://www.rvelthuis.de

"Future historians will be able to study at the Jimmy Carter
Library, the Gerald Ford Library, the Ronald Reagan Library,
and the Bill Clinton Adult Bookstore." -- George Carlin
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 11, 2017 3:01 PM   in response to: Rudy Velthuis (... in response to: Rudy Velthuis (...
Rudy,

| Unicode is far less of a mess than Ansi.

I don't think so. I think Unicode is a pile-of-cr*p!

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-11 15:00:58
Rudy Velthuis (...


Posts: 7,591
Registered: 9/22/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 11, 2017 11:16 PM   in response to: Quentin Correll in response to: Quentin Correll
Quentin Correll wrote:

Rudy,

Unicode is far less of a mess than Ansi.

I don't think so. I think Unicode is a pile-of-cr*p!

You may think so, but it isn't.

--
Rudy Velthuis http://www.rvelthuis.de

"Comedy is nothing more than tragedy deferred."
-- Pico Iyer, Time
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 12, 2017 12:20 PM   in response to: Rudy Velthuis (... in response to: Rudy Velthuis (...
Rudy,

| | > Unicode is far less of a mess than Ansi.
| |
| | I don't think so. I think Unicode is a pile-of-cr*p!
|
| You may think so, but it isn't.

But it IS! We'll just have to disagree on this issue. <g>

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-12 12:18:57
Dalija Prasnikar

Posts: 2,281
Registered: 11/9/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 12, 2017 1:07 PM   in response to: Quentin Correll in response to: Quentin Correll
Quentin Correll wrote:
Rudy,

| | > Unicode is far less of a mess than Ansi.
| |
| | I don't think so. I think Unicode is a pile-of-cr*p!
|
| You may think so, but it isn't.

But it IS! We'll just have to disagree on this issue. <g>

I guess all your textual data can be represented with 7-bit ASCII ;-)

--
Dalija Prasnikar
https://twitter.com/dalijap
https://plus.google.com/+DalijaPrasnikar
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 13, 2017 11:09 AM   in response to: Dalija Prasnikar in response to: Dalija Prasnikar
Dalija,

| I guess all your textual data can be represented with 7-bit ASCII ;-)

MIne? Yep... I'm very simple-minded. ;-)

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-13 11:08:18
Luigi Sandon

Posts: 345
Registered: 10/15/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 14, 2017 5:54 AM   in response to: Quentin Correll in response to: Quentin Correll
| Unicode is far less of a mess than Ansi.

I don't think so. I think Unicode is a pile-of-cr*p!

Human languages and their written representations - including specific forms like math and music - are a mess. They were never designed to fit nicely a computer program, and make the developer life easier.

Unicode is the best attempt till now for being able to show and process them in a computer program, without using ad-hoc and proprietary encodings.
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 14, 2017 10:57 AM   in response to: Luigi Sandon in response to: Luigi Sandon
Luigi,

| > | Unicode is far less of a mess than Ansi.
| |
| | I don't think so. I think Unicode is a pile-of-cr*p!
|
| Human languages and their written representations - including
| specific forms like math and music - are a mess. They were never
| designed to fit nicely a computer program, and make the developer
| life easier.

| Unicode is the best attempt till now for being able to show and
| process them in a computer program, without using ad-hoc and

| proprietary encodings.

Yeah,... I know. ;-) But ANSI is simpler and more straight-forward
than Unicode.

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-14 10:56:00
Dalija Prasnikar

Posts: 2,281
Registered: 11/9/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 14, 2017 12:39 PM   in response to: Quentin Correll in response to: Quentin Correll
Quentin Correll wrote:
Luigi,

| > | Unicode is far less of a mess than Ansi.
| |
| | I don't think so. I think Unicode is a pile-of-cr*p!
|
| Human languages and their written representations - including
| specific forms like math and music - are a mess. They were never
| designed to fit nicely a computer program, and make the developer
| life easier.

| Unicode is the best attempt till now for being able to show and
| process them in a computer program, without using ad-hoc and

| proprietary encodings.

Yeah,... I know. ;-) But ANSI is simpler and more straight-forward
than Unicode.

Only if you don't have to convert from one ANSI encoding to other,
or you don't have to use characters that span over several different
encodings inside same text.

To make things worse, ANSI was not the only standard, so for instance
Croatian text had several different ones. Handling those was such
nightmare, that people used to skip writing Croatian characters in emails,
web pages and other places where they were not of utmost importance
because using them and converting from one to the other, or reading
text through wrong encoding was a nightmare.

Not to mention, that with each encoding you would also need particular
font files for that encoding. DTP was also huge mess because of that.

--
Dalija Prasnikar
https://twitter.com/dalijap
https://plus.google.com/+DalijaPrasnikar
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 14, 2017 11:24 PM   in response to: Dalija Prasnikar in response to: Dalija Prasnikar
Dalija


To make things worse, ANSI was not the only standard, so for instance
Croatian text had several different ones. Handling those was such
nightmare, that people used to skip writing Croatian characters in emails,
web pages and other places where they were not of utmost importance
because using them and converting from one to the other, or reading
text through wrong encoding was a nightmare.

I don't think that should have been a concern, after all the EU will soon decide on a common language and ban all others <G>

Not to mention, that with each encoding you would also need particular
font files for that encoding. DTP was also huge mess because of that.

That seems to imply I don't need font files for unicode or that all font files contain every unicode character (woops sorry code point)

Roy
Markus Humm

Posts: 4,978
Registered: 11/9/03
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 14, 2017 11:42 PM   in response to: Roy Lambert in response to: Roy Lambert
Am 15.07.2017 um 08:24 schrieb Roy Lambert:
Dalija


To make things worse, ANSI was not the only standard, so for instance
Croatian text had several different ones. Handling those was such
nightmare, that people used to skip writing Croatian characters in emails,
web pages and other places where they were not of utmost importance
because using them and converting from one to the other, or reading
text through wrong encoding was a nightmare.

I don't think that should have been a concern, after all the EU will soon decide on a common language and ban all others <G>

Maybe, but that won't be English now... ;-)

Greetings

Markus
Dalija Prasnikar

Posts: 2,281
Registered: 11/9/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 15, 2017 3:23 AM   in response to: Roy Lambert in response to: Roy Lambert
Roy Lambert wrote:

Not to mention, that with each encoding you would also need particular
font files for that encoding. DTP was also huge mess because of that.

That seems to imply I don't need font files for unicode or that all font files contain every unicode character (woops sorry code point)

You have Unicode font files. I don't know about every Unicode character,
because I never used them, but certainly contain all European - western
and eastern languages and their special character sets.

So you can have one font instead of many, and when someone adds
another language into translation, you don't have to change your design
because the font you had does not cover that language.

--
Dalija Prasnikar
https://twitter.com/dalijap
https://plus.google.com/+DalijaPrasnikar
Rudy Velthuis (...


Posts: 7,591
Registered: 9/22/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 15, 2017 5:58 AM   in response to: Dalija Prasnikar in response to: Dalija Prasnikar
Dalija Prasnikar wrote:

Quentin Correll wrote:
Luigi,

> Unicode is far less of a mess than Ansi.


| I don't think so. I think Unicode is a pile-of-cr*p!

Human languages and their written representations - including
specific forms like math and music - are a mess. They were never
designed to fit nicely a computer program, and make the developer
life easier.

Unicode is the best attempt till now for being able to show and
process them in a computer program, without using ad-hoc and
proprietary encodings.

Yeah,... I know. ;-) But ANSI is simpler and more straight-forward
than Unicode.

Only if you don't have to convert from one ANSI encoding to other,
or you don't have to use characters that span over several different
encodings inside same text.

To make things worse, ANSI was not the only standard, so for instance
Croatian text had several different ones. Handling those was such
nightmare, that people used to skip writing Croatian characters in
emails

Even in German, which is not very complicated (only a few Umlauts
etc.), people often handled ä as ae, ö as oe and ß as ss, etc.

--
Rudy Velthuis http://www.rvelthuis.de

"Put your hand on a hot stove for a minute, and it seems like an
hour. Sit with a pretty girl for an hour, and it seems like a
minute. THAT'S relativity." -- Albert Einstein?

Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 15, 2017 10:42 AM   in response to: Dalija Prasnikar in response to: Dalija Prasnikar
Dalija,

Yeah,... but as a USA-American English mono-linguistic person I'm sure
that you understand my limitations and therefore my preference(s). ;-)

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-15 10:38:34
Dalija Prasnikar

Posts: 2,281
Registered: 11/9/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 15, 2017 11:58 AM   in response to: Quentin Correll in response to: Quentin Correll
Quentin Correll wrote:
Dalija,

Yeah,... but as a USA-American English mono-linguistic person I'm sure
that you understand my limitations and therefore my preference(s). ;-)

I do understand them :)

We would have much simpler job if there would be only one language.

Unfortunately, it is not. And we have to handle that mess somehow.
Nothing is ever perfect, but Unicode sorted out a lot of mess.

--
Dalija Prasnikar
https://twitter.com/dalijap
https://plus.google.com/+DalijaPrasnikar
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 15, 2017 1:01 PM   in response to: Dalija Prasnikar in response to: Dalija Prasnikar
Dalija,

| but Unicode sorted out a lot of mess.

On a more serious note: I DO understand that. <g>

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-15 13:00:26
Dalija Prasnikar

Posts: 2,281
Registered: 11/9/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 15, 2017 1:09 PM   in response to: Quentin Correll in response to: Quentin Correll
Quentin Correll wrote:
Dalija,

| but Unicode sorted out a lot of mess.

On a more serious note: I DO understand that. <g>

:)

--
Dalija Prasnikar
https://twitter.com/dalijap
https://plus.google.com/+DalijaPrasnikar
Rudy Velthuis (...


Posts: 7,591
Registered: 9/22/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 16, 2017 12:59 PM   in response to: Dalija Prasnikar in response to: Dalija Prasnikar
Dalija Prasnikar wrote:

Quentin Correll wrote:
Dalija,

Yeah,... but as a USA-American English mono-linguistic person I'm
sure that you understand my limitations and therefore my
preference(s). ;-)

I do understand them :)

We would have much simpler job if there would be only one language.

Pfft! Then they would invent more and more emojis, expecting every text
to be able to contain those too. <g>

--
Rudy Velthuis http://www.rvelthuis.de

"The question of whether a computer can think is no more
interesting than the question of whether a submarine can swim."
-- Edsger Dijkstra
Brandon Staggs

Posts: 633
Registered: 3/3/01
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 14, 2017 4:24 PM   in response to: Quentin Correll in response to: Quentin Correll
"Quentin Correll" wrote on Fri, 14 Jul 2017 10:57:00 -0700:

But ANSI is simpler and more straight-forward
than Unicode.

Uh, no.

--
Brandon Staggs
StudyLamp Software LLC
http://www.studylamp.com

Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 15, 2017 10:44 AM   in response to: Brandon Staggs in response to: Brandon Staggs
Brandon,

| | But ANSI is simpler and more straight-forward | than Unicode.
|

| Uh, no.

I guess I should have added ...to me. <g>

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-15 10:43:26
Brandon Staggs

Posts: 633
Registered: 3/3/01
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 16, 2017 6:16 AM   in response to: Quentin Correll in response to: Quentin Correll
"Quentin Correll" wrote on Sat, 15 Jul 2017 10:44:01 -0700:

|| But ANSI is simpler and more straight-forward | than Unicode.

| Uh, no.

I guess I should have added ...to me. <g>

Well, sure, that is easily inferred. :-)

But really, ANSI is only easier if you only have need of one or two
scripts and you know exactly what you will always be working with.
Beyond that, Unicode is vastly easier, for the simple reason that you
(usually) do not have to care what language is in there in order to
properly work with it. Working with ANSI with multiple languages is
much more cumbersome once you get beyond a single language... In my
experience. :-)

--
Brandon Staggs
StudyLamp Software LLC
http://www.studylamp.com
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 16, 2017 7:42 AM   in response to: Brandon Staggs in response to: Brandon Staggs
Brandon

But really, ANSI is only easier if you only have need of one or two
scripts and you know exactly what you will always be working with.
Beyond that, Unicode is vastly easier, for the simple reason that you
(usually) do not have to care what language is in there in order to
properly work with it. Working with ANSI with multiple languages is
much more cumbersome once you get beyond a single language... In my
experience. :-)

If I wish to restrict myself to one language only why am I being forced into using something that is suitable for many?

Out of interest how many computer languages are unicode? Is it possible to develop software in unicode? I am curious about this.

Roy
Markus Humm

Posts: 4,978
Registered: 11/9/03
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 16, 2017 8:29 AM   in response to: Roy Lambert in response to: Roy Lambert
Am 16.07.2017 um 16:42 schrieb Roy Lambert:
Brandon

But really, ANSI is only easier if you only have need of one or two
scripts and you know exactly what you will always be working with.
Beyond that, Unicode is vastly easier, for the simple reason that you
(usually) do not have to care what language is in there in order to
properly work with it. Working with ANSI with multiple languages is
much more cumbersome once you get beyond a single language... In my
experience. :-)

If I wish to restrict myself to one language only why am I being forced into using something that is suitable for many?

Out of interest how many computer languages are unicode? Is it possible to develop software in unicode? I am curious about this.

Hm?
Most current development environments for Windows are Unicode enabled:
Java, C++ and most others as well.

In Delphi you can also use unicode in your variable and method names.
And if you're using English only it's no big issue, as the first 127
code points of Unicode are the 7 bit ASCII ones anyway. Sos where's your
real issue?

Greetings

Markus

Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 16, 2017 11:47 PM   in response to: Markus Humm in response to: Markus Humm
Markus

I. Sos where's your
real issue?

Its not an issue I was just curious. I obvoiosly did not make the question clear enough. Are there an computer languages out tehre where CODE (not valiable names) can be develped using unicode?

Roy
Dalija Prasnikar

Posts: 2,281
Registered: 11/9/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 17, 2017 1:42 AM   in response to: Roy Lambert in response to: Roy Lambert
Roy Lambert wrote:
Markus

I. Sos where's your
real issue?

Its not an issue I was just curious. I obvoiosly did not make the question clear enough. Are there an computer languages out tehre where CODE (not valiable names) can be develped using unicode?

Following is Swift code - you can use any Unicode characters you want in all identifiers
you declare. Of course, reserved words, and framework identifiers are in English, but
I doubt that you would find any common language out there where they are not in English.

Point is, they could also be Unicode - all infrastructure is there.

class 💩💩💩💩 {
func 💩💩💩(😎: Int, 🐯: Int) -> Int {
return 😎 + 🐯;
}
}

let 🐔 = 3;
let 😥 = 🐔 + 2;
let 💩 = 💩💩💩💩();
print(💩.💩💩💩(🐔, 🐯:😥));

--
Dalija Prasnikar
https://twitter.com/dalijap
https://plus.google.com/+DalijaPrasnikar
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 17, 2017 2:13 AM   in response to: Dalija Prasnikar in response to: Dalija Prasnikar
Dalija

Following is Swift code - you can use any Unicode characters you want in all identifiers
you declare. Of course, reserved words, and framework identifiers are in English, but
I doubt that you would find any common language out there where they are not in English.

That's what I was wondering - thinks for phrasing it better than I managed.

Point is, they could also be Unicode - all infrastructure is there.

I suppose I should be grateful that they're not. I have enough trouble reading others code without having it in a, to me, foreign language.

Roy
Luigi Sandon

Posts: 345
Registered: 10/15/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 19, 2017 11:51 AM   in response to: Dalija Prasnikar in response to: Dalija Prasnikar
Dalija Prasnikar wrote:
I doubt that you would find any common language out there where they are not in English.

In retrospect, it's a pity Wirth didn't use German in Pascal, and Hejlsberg didn't add a few Danish extensions... <G>
Rudy Velthuis (...


Posts: 7,591
Registered: 9/22/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 19, 2017 3:32 PM   in response to: Luigi Sandon in response to: Luigi Sandon
Luigi Sandon wrote:

Dalija Prasnikar wrote:
I doubt that you would find any common language out there where
they are not in English.

In retrospect, it's a pity Wirth didn't use German in Pascal, and
Hejlsberg didn't add a few Danish extensions... <G>

Problem would have been the Umlauts:

während x < 10 tue
anfang
{ etc... }

Although the Swiss might perhaps write:

waehrend x < 10 tue

--
Rudy Velthuis http://www.rvelthuis.de

"Never mistake motion for action."
-- Ernest Hemingway (1899-1961)
Dalija Prasnikar

Posts: 2,281
Registered: 11/9/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 19, 2017 4:07 PM   in response to: Rudy Velthuis (... in response to: Rudy Velthuis (...
Rudy Velthuis (TeamB, MVP) wrote:
Luigi Sandon wrote:

Dalija Prasnikar wrote:
I doubt that you would find any common language out there where
they are not in English.

In retrospect, it's a pity Wirth didn't use German in Pascal, and
Hejlsberg didn't add a few Danish extensions... <G>

Problem would have been the Umlauts:

während x < 10 tue
anfang
{ etc... }

Although the Swiss might perhaps write:

waehrend x < 10 tue

I believe that was the whole point of Luigi's comment :)

--
Dalija Prasnikar
https://twitter.com/dalijap
https://plus.google.com/+DalijaPrasnikar
Rudy Velthuis (...


Posts: 7,591
Registered: 9/22/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 17, 2017 2:13 AM   in response to: Roy Lambert in response to: Roy Lambert
Roy Lambert wrote:

Markus

I. Sos where's your
real issue?

Its not an issue I was just curious. I obvoiosly did not make the
question clear enough. Are there an computer languages out tehre
where CODE (not valiable names) can be develped using unicode?

If the language does not have non-ASCII operators and keywords,
probably not.

Except for APL, I don't know of such a language. And probably some
versions of Forth, Lisp and such languages. There code can havy any
name too.
--
Rudy Velthuis http://www.rvelthuis.de

"The company doesn't tell me what to say, and I don't tell them
where to stick it." -- Unknown
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 17, 2017 5:41 AM   in response to: Rudy Velthuis (... in response to: Rudy Velthuis (...
Rudy

Except for APL, I don't know of such a language. And probably some
versions of Forth, Lisp and such languages. There code can havy any
name too.

I loved APL, but back in the days I was using it it certainly wasn't unicode. I was the first in the group to discover how to print lowercase letters. Major breakthrough in user friendliness!

Roy
Rudy Velthuis (...


Posts: 7,591
Registered: 9/22/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 18, 2017 10:13 AM   in response to: Roy Lambert in response to: Roy Lambert
Roy Lambert wrote:

Rudy

Except for APL, I don't know of such a language. And probably some
versions of Forth, Lisp and such languages. There code can havy any
name too.

I loved APL, but back in the days I was using it it certainly wasn't
unicode.

But it uses weird syntactical constructs and many non-letter characters.

--
Rudy Velthuis http://www.rvelthuis.de

"Behind every great fortune there is a crime."
-- Honore de Balzac (1799-1850)
John Treder

Posts: 342
Registered: 8/2/02
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 18, 2017 5:15 PM   in response to: Rudy Velthuis (... in response to: Rudy Velthuis (...
Rudy Velthuis (TeamB, MVP) wrote:

But it uses weird syntactical constructs and many non-letter characters.

--
Rudy Velthuis http://www.rvelthuis.de

You can say that again! I took a course in APL in the late 60s when I was at IBM. Lots of math power, and a crib sheet was absolutely needed!
Then, in the 80s, I used (didn't write) an electromagnetic design program that was written in APL, for designing hard driver VCMs. It put some numbers around the rules of thumb, but didn't account for "end turns" at all, which made it quite conservative.

--
nhoJ
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 19, 2017 1:25 AM   in response to: Rudy Velthuis (... in response to: Rudy Velthuis (...
Rudy

But it uses weird syntactical constructs and many non-letter characters.

Weird - no. Many of them were very sensible, but different. One that I've always felt is far better than precedence rules is linear execution order (with things inside brackets done first).

Non-letter characters - yes if you restrict yourself to the letters of the English alphabet only.

Roy
Brandon Staggs

Posts: 633
Registered: 3/3/01
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 16, 2017 12:38 PM   in response to: Roy Lambert in response to: Roy Lambert
"Roy Lambert" wrote on Sun, 16 Jul 2017 07:42:28 -0700:

If I wish to restrict myself to one language only why am I being forced into using something that is suitable for many?

Huh? I don't understand. Which languages are you concerned aren't
served by Unicode?

--
Brandon Staggs
StudyLamp Software LLC
http://www.studylamp.com
Rudy Velthuis (...


Posts: 7,591
Registered: 9/22/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 16, 2017 1:04 PM   in response to: Roy Lambert in response to: Roy Lambert
Roy Lambert wrote:

Brandon

But really, ANSI is only easier if you only have need of one or two
scripts and you know exactly what you will always be working with.
Beyond that, Unicode is vastly easier, for the simple reason that
you (usually) do not have to care what language is in there in
order to properly work with it. Working with ANSI with multiple
languages is much more cumbersome once you get beyond a single
language... In my experience. :-)

If I wish to restrict myself to one language only why am I being
forced into using something that is suitable for many?

Because otherwise you are on your own. If everyone uses new technology,
you may have to do that too.

Just like it is hard to get support or parts for, say, tape decks,
video disks or Betamax recorders.

--
Rudy Velthuis http://www.rvelthuis.de

"We hang the petty thieves and appoint the great ones to public
office."
-- Aesop
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 16, 2017 11:47 PM   in response to: Rudy Velthuis (... in response to: Rudy Velthuis (...
Rudy

Just like it is hard to get support or parts for, say, tape decks,
video disks or Betamax recorders.

I thought your objected to using strawman arguments.

Roy
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 16, 2017 11:00 AM   in response to: Brandon Staggs in response to: Brandon Staggs
Brandon,

| Working with ANSI with multiple languages is
| much more cumbersome once you get beyond a single language...

I do understand that. However, fortunately for me, I'm
mono-linguistic. <g>

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-16 10:58:59
Arthur Hoornweg

Posts: 380
Registered: 6/2/98
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 20, 2017 4:16 AM   in response to: Brandon Staggs in response to: Brandon Staggs
Brandon Staggs wrote:

Beyond that, Unicode is vastly easier, for the simple reason that you
(usually) do not have to care what language is in there in order to
properly work with it.

Unfortunately Unicodestrings/Widestrings are UTF-16 and there are characters that don't fit in a widechar.

Brandon Staggs

Posts: 633
Registered: 3/3/01
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 20, 2017 5:55 AM   in response to: Arthur Hoornweg in response to: Arthur Hoornweg
"Arthur Hoornweg" wrote on Thu, 20 Jul 2017 04:16:49 -0700:

Unfortunately Unicodestrings/Widestrings are UTF-16 and there are characters that don't fit in a widechar.

But it's not UCS-2, where that is a real problem. It's UTF-16, which
is, like UTF-8, a variable-length encoding. There aren't any *code
points* that can't be encoded in UTF-16. That just means that some
code points (characters <> code points) take two widechars to store.

Anyway, as I said, usually you do not have to care what is in there.
If you're concerned about text that falls outside of the basic
multilingual plane and you want to parse the string yourself, then you
should be aware of surrogate pairs.

Unicode is not that hard, really, IME. But I wouldn't say something
obviously non-useful like "text is text." :-)

--
Brandon Staggs
StudyLamp Software LLC
http://www.studylamp.com
Rudy Velthuis (...


Posts: 7,591
Registered: 9/22/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 20, 2017 7:11 AM   in response to: Brandon Staggs in response to: Brandon Staggs
Brandon Staggs wrote:

But I wouldn't say something
obviously non-useful like "text is text." :-)

Are you saying text <> text? <g>

Actually, I really wonder why people (usually those that abuse strings
to store non-text data) argue about what is text and what is not. It is
bloody obvious, IMO.

I find debating that just as silly as debating about the meaning of
"is".
--
Rudy Velthuis http://www.rvelthuis.de

"Victory goes to the player who makes the next-to-last mistake."
-- Chessmaster Savielly Grigorievitch Tartakower (1887-1956)
Brandon Staggs

Posts: 633
Registered: 3/3/01
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 20, 2017 7:40 AM   in response to: Rudy Velthuis (... in response to: Rudy Velthuis (...
"Rudy Velthuis" wrote on Thu, 20 Jul 2017 07:11:58 -0700:

Brandon Staggs wrote:

But I wouldn't say something
obviously non-useful like "text is text." :-)

Are you saying text <> text? <g>

In a sense. Having dealt with what happens when you expect text to be
text and it turns out that text is no longer the same text even though
it looks the same, I know better. You may think that text should not
be regarded as binary data, but if you expect your filenames to be
found by Windows, you had better treat your text as binary data,
because otherwise that text you thought was just text can turn into
text that no longer matches the text Windows told you was there a
moment ago.

But text is usually text.

Actually, I really wonder why people (usually those that abuse strings
to store non-text data) argue about what is text and what is not. It is
bloody obvious, IMO.

I agree that one should not treat string variables as convenient blob
storage, and I suspect that the vast majority of "pain" people
experience with migrating to Unicode comes down to treating string
variables as blobs, or just not being interested in learning what
happens when different string types are cast to other string types.

I find debating that just as silly as debating about the meaning of
"is".

TFileName is a string. But a Windows filename should really be
regarded as binary data. So it isn't silly. When you let that string
be treated as nothing but text, its encoding can easily be changed
(depending on what you are doing with it, like putting it into an XML
tree that is saved to UTF-8) and the link to the actual file can be
lost because it is impossible to recover the original encoding in many
cases. But I have explained this before.

--
Brandon Staggs
StudyLamp Software LLC
http://www.studylamp.com

Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 20, 2017 8:38 AM   in response to: Brandon Staggs in response to: Brandon Staggs
Brandon

But text is usually text.

I like that one - much better than Rudy's "text is text" <G>

Roy
Rudy Velthuis (...


Posts: 7,591
Registered: 9/22/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 21, 2017 2:21 PM   in response to: Brandon Staggs in response to: Brandon Staggs
Brandon Staggs wrote:

"Rudy Velthuis" wrote on Thu, 20 Jul 2017 07:11:58 -0700:

Brandon Staggs wrote:

But I wouldn't say something
obviously non-useful like "text is text." :-)

Are you saying text <> text? <g>

In a sense. Having dealt with what happens when you expect text to be
text and it turns out that text is no longer the same text even though
it looks the same, I know better. You may think that text should not
be regarded as binary data, but if you expect your filenames to be
found by Windows, you had better treat your text as binary data

I vaguely remember that someone had problems finding filenames because
they had a different composition. Was that you?

Well, when searching, or collating, or something like it, you should be
sure that all the texts you have have the same composition mode, indeed.

That does not mean that one text is not the same as the other. It
merely means that they are encoded differently (different composition
mode), and that makes **comparing** them slightly more complicated than
a simple binary compare.

So yes, text is still text, not just "usually". But comparing texts
with possibly different composition modes (unnormalized texts) requires
a little more work than simply comparing bytes or words.
--
Rudy Velthuis http://www.rvelthuis.de

"The object of war is not to die for your country but to make
the other bastard die for his."
-- General George Patton (1885-1945)
Brandon Staggs

Posts: 633
Registered: 3/3/01
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 22, 2017 7:42 AM   in response to: Rudy Velthuis (... in response to: Rudy Velthuis (...
"Rudy Velthuis" wrote on Fri, 21 Jul 2017 14:21:43 -0700:

It
merely means that they are encoded differently (different composition
mode), and that makes **comparing** them slightly more complicated than
a simple binary compare.

Do you have any idea how long it would take to compare all of the file
names on a system to the one you are looking for, normalizing every
one and trying to figure out collisions? You aren't really thinking
this through.

If you ask for a file name you get a TFileName (a string-- and here's
the first problem -- that WideString is now cast into a string, where
data can now be lost). If you shove it into an XML tree as a string
you can't always control how it is streamed. When you get it back and
ask Windows for the file, you may or may not get the file, even if it
is there, and manually searching the entire system's file set is not
exactly feasible. It's a real problem and the only way to fix it is
to NOT treat the filenames as text -- which is very counterintuitive
since Delphi takes the file name and stores it in a string.

Anyway, text is usually just text. Not always. Anyone who says
otherwise has a limited scope of experience on the subject.

--
Brandon Staggs
StudyLamp Software LLC
http://www.studylamp.com
Rudy Velthuis (...


Posts: 7,591
Registered: 9/22/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 22, 2017 8:18 AM   in response to: Brandon Staggs in response to: Brandon Staggs
Brandon Staggs wrote:

"Rudy Velthuis" wrote on Fri, 21 Jul 2017 14:21:43 -0700:

It
merely means that they are encoded differently (different
composition mode), and that makes comparing them slightly more
complicated than a simple binary compare.

Do you have any idea how long it would take to compare all of the file
names on a system to the one you are looking for, normalizing every
one and trying to figure out collisions? You aren't really thinking
this through.

Yes, I am. One generally doesn't compare the full name, i.e. most
comparisons stop at the first or second code point anyway.

I don't know what you mean with "figuring out collisions"? It is not
uncommon that there are multiple files with the same name. Just list
all of them.

ISTM that iterating over a harddisk or subdirectory is much slower than
the file name comparison, unless you can use journaling structures like
Everything does.

--
Rudy Velthuis http://www.rvelthuis.de

"If you kill one person you are a murderer. If you kill ten
people you are a monster. If you kill ten thousand you are
a national hero." -- Vassilis Epaminondou
Arthur Hoornweg

Posts: 380
Registered: 6/2/98
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 25, 2017 3:43 AM   in response to: Brandon Staggs in response to: Brandon Staggs
Brandon Staggs wrote:

Anyway, as I said, usually you do not have to care what is in there.
If you're concerned about text that falls outside of the basic
multilingual plane and you want to parse the string yourself, then you
should be aware of surrogate pairs.

IMO UTF-8 is the best encoding for storage/transport (ascii backward compatible, compact storage except for Asian languages, easy debugging/viewing, no endianness issues). UTF-32 is the best for processing (directly indexable, character fits into 32-bit register) but very wasteful, hence less suitable for storage.

UTF16 isn't ideal for anything. First there's the big endian/little endian issue, secondly it's wasteful, thirdly it lacks direct codepoint indexability and most programmers aren't even aware of that.

Rudy Velthuis (...


Posts: 7,591
Registered: 9/22/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 25, 2017 5:40 AM   in response to: Arthur Hoornweg in response to: Arthur Hoornweg
Arthur Hoornweg wrote:

UTF16 isn't ideal for anything.

It is ideal for talking to OSes. Most platforms Delphi supports, except
Linux, have a UTF-16 interface: Windows (32 bit and 64 bit), OS X, iOS
and Android.

UCS4 (UTF-32) is simply too space hungry. UTF-8 is not the default on
most platforms.

--
Rudy Velthuis http://www.rvelthuis.de

"Always be wary of any helpful item that weighs less than its
operating manual."
-- Terry Pratchett (Jingo)
Rudy Velthuis (...


Posts: 7,591
Registered: 9/22/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 25, 2017 5:42 AM   in response to: Arthur Hoornweg in response to: Arthur Hoornweg
Arthur Hoornweg wrote:

thirdly it lacks direct
codepoint indexability and most programmers aren't even aware of that.

How often do you actually need direct code point indexability?

--
Rudy Velthuis http://www.rvelthuis.de

Muir's Law: When we try to pick out anything by itself, we find
it hitched to everything else in the universe.
Brandon Staggs

Posts: 633
Registered: 3/3/01
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 25, 2017 7:40 AM   in response to: Rudy Velthuis (... in response to: Rudy Velthuis (...
"Rudy Velthuis" wrote on Tue, 25 Jul 2017 05:42:05 -0700:

thirdly it lacks direct
codepoint indexability and most programmers aren't even aware of that.

How often do you actually need direct code point indexability?

Not only that, but how often do you need it before you have looked
at the text to determine what is in it? If you are parsing the text
to use it, you are already going through the text so there isn't much
value in being able to go right at a string index without
consideration for it being part of a surrogate pair. And even if you
don't care what came before or after that, you can easily test if the
widechar holds part of a surrogate pair. And in practice, you can
often ignore this entirely without any problems in algorithms that
directly access widechars.

I am sure there are scenarios where direct indexability actually does
present compelling value, but I think the perceived value is much
greater than the practical value. It's just hardly ever an issue.

--
Brandon Staggs
StudyLamp Software LLC
http://www.studylamp.com
Rudy Velthuis (...


Posts: 7,591
Registered: 9/22/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 25, 2017 8:50 AM   in response to: Brandon Staggs in response to: Brandon Staggs
Brandon Staggs wrote:

"Rudy Velthuis" wrote on Tue, 25 Jul 2017 05:42:05 -0700:

thirdly it lacks direct
codepoint indexability and most programmers aren't even aware of
that.

How often do you actually need direct code point indexability?

Not only that, but how often do you need it before you have looked
at the text to determine what is in it? If you are parsing the text
to use it, you are already going through the text so there isn't much
value in being able to go right at a string index without
consideration for it being part of a surrogate pair. And even if you
don't care what came before or after that, you can easily test if the
widechar holds part of a surrogate pair. And in practice, you can
often ignore this entirely without any problems in algorithms that
directly access widechars.

I am sure there are scenarios where direct indexability actually does
present compelling value, but I think the perceived value is much
greater than the practical value. It's just hardly ever an issue.

Exactly.

--
Rudy Velthuis http://www.rvelthuis.de

"The man who goes alone can start today; but he who travels with
another must wait till that other is ready."
-- Henry David Thoreau (1817-1862)
Rudy Velthuis (...


Posts: 7,591
Registered: 9/22/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 20, 2017 6:52 AM   in response to: Arthur Hoornweg in response to: Arthur Hoornweg
Arthur Hoornweg wrote:

Brandon Staggs wrote:

Beyond that, Unicode is vastly easier, for the simple reason that
you (usually) do not have to care what language is in there in
order to properly work with it.

Unfortunately Unicodestrings/Widestrings are UTF-16 and there are
characters that don't fit in a widechar.


That is why he said "(usually)". There are exceptions of course.

But you probably meant that there are codepoints that do not fit in one
WideChar. Such codepoints are encoded as surrogate pairs.

--
Rudy Velthuis http://www.rvelthuis.de

Meskimen's Law: There's never time to do it right, but always
time to do it over.
Rudy Velthuis (...


Posts: 7,591
Registered: 9/22/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 15, 2017 5:56 AM   in response to: Quentin Correll in response to: Quentin Correll
Quentin Correll wrote:

Luigi,

> Unicode is far less of a mess than Ansi.


| I don't think so. I think Unicode is a pile-of-cr*p!

Human languages and their written representations - including
specific forms like math and music - are a mess. They were never
designed to fit nicely a computer program, and make the developer
life easier.

Unicode is the best attempt till now for being able to show and
process them in a computer program, without using ad-hoc and
proprietary encodings.

Yeah,... I know. ;-) But ANSI is simpler and more straight-forward
than Unicode.

Actually, no, it isn't, unless you ONLY have to work with one single
codepage throughout your life. <g>

--
Rudy Velthuis http://www.rvelthuis.de

"Few things are harder to put up with than a good example."
- Mark Twain (1835-1910)
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 15, 2017 10:46 AM   in response to: Rudy Velthuis (... in response to: Rudy Velthuis (...
Rudy,

| ... unless you ONLY have to work with one single
| codepage throughout your life. <g>

Yeah! That's me and my preferential situation! <g>

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-15 10:44:26
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 12, 2017 12:40 AM   in response to: Rudy Velthuis (... in response to: Rudy Velthuis (...
Rudy

I finally forced myself to read to the end. Whilst there are some
valid points if he writes code like he writes blogs I would hate to
have to maintain it.

The summary for those who don't want to waste part of their life:

1. unicode is different to ansi
2. using ansi for different languages was a mess
3. unicode is a different type of mess

I think that's it.

Bad summary. Unicode is far less of a mess than Ansi.

Bad comment - I made no quantitative judgement.

Roy
Arthur Hoornweg

Posts: 380
Registered: 6/2/98
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 20, 2017 4:09 AM   in response to: Rudy Velthuis (... in response to: Rudy Velthuis (...
Rudy Velthuis (TeamB, MVP) wrote:
Bad summary. Unicode is far less of a mess than Ansi.

UTF16 is still a mess, because a character does not necessarily fit in a char. ("Character" in the sense of letter/glyph/codepoint).

Brandon Staggs

Posts: 633
Registered: 3/3/01
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 20, 2017 6:02 AM   in response to: Arthur Hoornweg in response to: Arthur Hoornweg
"Arthur Hoornweg" wrote on Thu, 20 Jul 2017 04:09:42 -0700:

UTF16 is still a mess, because a character does not necessarily fit in a char. ("Character" in the sense of letter/glyph/codepoint).

But it's not a "mess" and no encoding scheme can avoid the fact that
a character is not a code point. They are not synonymous and should
not be thought of as the same thing. A code point is not a character
is not a glyph.

You can't get around the fact that a character with a diacritical mark
can be represented either as a single code point or multiple code
points. This is not unique to UTF-16.

--
Brandon Staggs
StudyLamp Software LLC
http://www.studylamp.com
Rudy Velthuis (...


Posts: 7,591
Registered: 9/22/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 20, 2017 7:15 AM   in response to: Arthur Hoornweg in response to: Arthur Hoornweg
Arthur Hoornweg wrote:

Rudy Velthuis (TeamB, MVP) wrote:
Bad summary. Unicode is far less of a mess than Ansi.

UTF16 is still a mess, because a character does not necessarily fit
in a char. ("Character" in the sense of letter/glyph/codepoint).


It is less easy to use than UCS-4 (aka UTF-32), sure. But only
slightly. And wasting 3 bytes for most codepoints is a bit too much,
IMO.

UTF-16 is a nice compromise: no need for many multi-byte codepoints
like in UTF-8, and no need to waste 3 bytes for each code point either.

--
Rudy Velthuis http://www.rvelthuis.de

"We are condemned to kill time: thus we die bit by bit."
-- Octavio Paz
Brandon Staggs

Posts: 633
Registered: 3/3/01
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 20, 2017 7:45 AM   in response to: Rudy Velthuis (... in response to: Rudy Velthuis (...
"Rudy Velthuis" wrote on Thu, 20 Jul 2017 07:15:09 -0700:

UTF16 is still a mess, because a character does not necessarily fit
in a char. ("Character" in the sense of letter/glyph/codepoint).

It is less easy to use than UCS-4 (aka UTF-32), sure. But only
slightly. And wasting 3 bytes for most codepoints is a bit too much,
IMO.

Some people just assume that a fixed number of bytes per code point is
automatically easier to use. But the cases where that is actually
true are actually quite rare. One almost always needs to sequentially
scan text to make any use of an arbitrary position of it anyway.

https://en.wikipedia.org/wiki/UTF-32#Analysis

--
Brandon Staggs
StudyLamp Software LLC
http://www.studylamp.com
Remy Lebeau (Te...


Posts: 8,950
Registered: 12/23/01
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 24, 2017 11:47 AM   in response to: Arthur Hoornweg in response to: Arthur Hoornweg
Arthur Hoornweg wrote:

UTF16 is still a mess, because a character does not necessarily fit
in a char. ("Character" in the sense of letter/glyph/codepoint).

The same was true in MBCS ANSI as well. Plenty of languages that could
not fit a single "character"/glyph/whatever into a single AnsiChar.

--
Remy Lebeau (TeamB)
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 11, 2017 2:58 PM   in response to: Roy Lambert in response to: Roy Lambert
Roy,

| 1. unicode is different to ansi
| 2. using ansi for different languages was a mess
| 3. unicode is a different type of mess
|
| I think that's it.

I think you've nailed it! ;-) ROFL!

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-11 14:58:03
Arnaud Bouchez

Posts: 137
Registered: 8/2/15
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 11, 2017 11:56 PM   in response to: Roy Lambert in response to: Roy Lambert
Roy Lambert wrote:
I finally forced myself to read to the end. Whilst there are some valid points if he writes code like he writes blogs I would hate to have to maintain it.

3. unicode is a different type of mess

And we even didn't start talking with code points, graphs and encodings!
http://utf8everywhere.org/
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 12, 2017 12:40 AM   in response to: Arnaud Bouchez in response to: Arnaud Bouchez
Arnaud

Thanks for the link - I'll read it later. Just at first glance its at least readable without matchsticks to prop my eyes open.

Roy Lambert

Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 12, 2017 6:31 AM   in response to: Arnaud Bouchez in response to: Arnaud Bouchez
Arnaud

Thanks for the link. I wopuldn't say enjoyable reading but certainly informative. It certainly enlightens me about wenjie zhou's attitude.

Maybe Rudy should read it.

Roy Lambert

Dalija Prasnikar

Posts: 2,281
Registered: 11/9/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 12, 2017 9:44 AM   in response to: Roy Lambert in response to: Roy Lambert
Roy Lambert wrote:

It certainly enlightens me about wenjie zhou's attitude.

Actually it does not. AFAIK, he uses ANSI and it is more memory
efficient than and in any kind of Unicode encoding.

It is more in line with my attitude, though at time Delphi got Unicode
it made sense using UTF-16 as base string type.

But it also shows my point that processing data in UTF8 is reasonable
thing to do, converting only for display purposes. It is generally faster and
more memory efficient way.

Maybe Rudy should read it.

;-)

--
Dalija Prasnikar
https://twitter.com/dalijap
https://plus.google.com/+DalijaPrasnikar
Arnaud Bouchez

Posts: 137
Registered: 8/2/15
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 13, 2017 12:05 AM   in response to: Dalija Prasnikar in response to: Dalija Prasnikar
Dalija Prasnikar wrote:
But it also shows my point that processing data in UTF8 is reasonable
thing to do, converting only for display purposes. It is generally faster and
more memory efficient way.

This is exactly what we do for efficiency in our ORM/SOA/MVC framework: since we deal with JSON for data transmission, we use UTF-8 encoding everywhere, and only switch to string/UnicodeString when reaching the presentation layer.
It also has the benefit of forcing the business logic to be separated from the presentation layer... ;)
Rudy Velthuis (...


Posts: 7,591
Registered: 9/22/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 13, 2017 3:03 PM   in response to: Dalija Prasnikar in response to: Dalija Prasnikar
Dalija Prasnikar wrote:

Roy Lambert wrote:

It certainly enlightens me about wenjie zhou's attitude.

Actually it does not. AFAIK, he uses ANSI and it is more memory
efficient than and in any kind of Unicode encoding.

He uses something called GB2312 (or some such), which is actually a
two-byte-per-codepoint encoding. But he puts that in AnsiStrings.

That is not more memory efficient than any kind of Unicode encoding.
--
Rudy Velthuis http://www.rvelthuis.de

"If people are good only because they fear punishment, and hope for
reward, then we are a sorry lot indeed." -- Albert Einstein
Dalija Prasnikar

Posts: 2,281
Registered: 11/9/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 14, 2017 1:13 AM   in response to: Rudy Velthuis (... in response to: Rudy Velthuis (...
Rudy Velthuis (TeamB, MVP) wrote:
Dalija Prasnikar wrote:

Roy Lambert wrote:

It certainly enlightens me about wenjie zhou's attitude.

Actually it does not. AFAIK, he uses ANSI and it is more memory
efficient than and in any kind of Unicode encoding.

He uses something called GB2312 (or some such), which is actually a
two-byte-per-codepoint encoding. But he puts that in AnsiStrings.

That is not more memory efficient than any kind of Unicode encoding.

Yes, it is.

From Wikipedia:

"Compared to UTF-8, GB2312 (whether native or encoded in EUC-CN) is
more storage efficient: while UTF-8 uses three bytes[a] per CJK ideograph, GB2312
only uses two. "

https://en.wikipedia.org/wiki/GB_2312

So if he is fine with downsides, I can see why he prefers it over any kind
of Unicode.

--
Dalija Prasnikar
https://twitter.com/dalijap
https://plus.google.com/+DalijaPrasnikar
Arnaud Bouchez

Posts: 137
Registered: 8/2/15
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 11, 2017 11:57 PM   in response to: Roy Lambert in response to: Roy Lambert
Roy Lambert wrote:
1. unicode is different to ansi
2. using ansi for different languages was a mess
3. unicode is a different type of mess

4. Humanity is a mess
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 12, 2017 12:40 AM   in response to: Arnaud Bouchez in response to: Arnaud Bouchez
Arnaud

Roy Lambert wrote:
1. unicode is different to ansi
2. using ansi for different languages was a mess
3. unicode is a different type of mess

4. Humanity is a mess

That is my second close escape this morning. Just put my coffee cup down as I reached your post.

Roy
Van Swofford

Posts: 389
Registered: 6/28/03
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 12, 2017 5:39 AM   in response to: Arnaud Bouchez in response to: Arnaud Bouchez
Arnaud Bouchez wrote:

Roy Lambert wrote:
1. unicode is different to ansi
2. using ansi for different languages was a mess
3. unicode is a different type of mess

4. Humanity is a mess

Oh yeah! That's indisputable. :-)

--
Cheers,
Van

"Good judgment comes from experience, and a lot of that comes from bad
judgment." - Will Rogers
Alexandre Machado

Posts: 1,682
Registered: 8/10/13
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 12, 2017 2:51 PM   in response to: Roy Lambert in response to: Roy Lambert
I finally forced myself to read to the end. Whilst there are some valid points if he writes code like he writes blogs I would hate to have to maintain it.

Read what exactly? "Joel on software" is kind of vague....
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 12, 2017 10:50 PM   in response to: Alexandre Machado in response to: Alexandre Machado
Alexandre

Read what exactly? "Joel on software" is kind of vague....

https://www.joelonsoftware.com/2003/10/08/the-absolute-minimum-every-software-developer-absolutely-positively-must-know-about-unicode-and-character-sets-no-excuses/

I would say "enjoy" but I don't think you will

Roy
Alexandre Machado

Posts: 1,682
Registered: 8/10/13
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 13, 2017 2:28 PM   in response to: Roy Lambert in response to: Roy Lambert
Roy Lambert wrote:
Alexandre

Read what exactly? "Joel on software" is kind of vague....

https://www.joelonsoftware.com/2003/10/08/the-absolute-minimum-every-software-developer-absolutely-positively-must-know-about-unicode-and-character-sets-no-excuses/

I would say "enjoy" but I don't think you will

Roy

Ahhh... I think I will pass. I don't like spending much time on blogs where there is a picture of an ugly male staring at you all the time.
Maybe when Joel write something more interesting....
Rudy Velthuis (...


Posts: 7,591
Registered: 9/22/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 13, 2017 2:55 PM   in response to: Roy Lambert in response to: Roy Lambert
Roy Lambert wrote:

I finally forced myself to read to the end. Whilst there are some
valid points if he writes code like he writes blogs I would hate to
have to maintain it.

The summary for those who don't want to waste part of their life:

1. unicode is different to ansi
2. using ansi for different languages was a mess
3. unicode is a different type of mess

I think that's it.

If you think that's it, you didn't really understand it. The main
message is: it is stupid to try to avoid Unicode. Learn Unicode and
work with it.

--
Rudy Velthuis http://www.rvelthuis.de

"Faith is belief without evidence in what is told by one who
speaks without knowledge, of things without parallel."
-- Ambrose Bierce
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 14, 2017 12:00 AM   in response to: Rudy Velthuis (... in response to: Rudy Velthuis (...
Rudy

If you think that's it, you didn't really understand it. The main
message is: it is stupid to try to avoid Unicode. Learn Unicode and
work with it.

We've been through this before. As things stand I have no, zero, none at all, reason to use unicode. Any applications I write are restricted to English speakers and readers (and I'll be kind and allow Americans in there). Any conversion to unicode mandated by the OS is done for me. Why should I waste time and effort converting to unicode?

For some people there is a benefit, for others none. Why are you so determined to force everyone into the same straightjacket?

Roy
Brandon Staggs

Posts: 633
Registered: 3/3/01
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 14, 2017 5:57 AM   in response to: Roy Lambert in response to: Roy Lambert
"Roy Lambert" wrote on Fri, 14 Jul 2017 00:00:18 -0700:

Why should I waste time and effort converting to unicode?

I am sure this was addressed before, but I wasn't paying attention.
Are you sure there is that much effort involved? If you indeed are
only using plain English text, migrating to Unicode should be
virtually no work. The only difficulties come in when you are using
abandoned libraries or using String variables for binary buffers.

I had a large project that needed to be hand-tuned when the switch to
Unicode came all those years ago. I was using different ANSI
encodings and the like, but even then, that wasn't so painful and I am
glad I don't have to look back.

But if you are sure there are no benefits to upgrading your tools,
then sure, why bother putting the time into it?

--
Brandon Staggs
StudyLamp Software LLC
http://www.studylamp.com
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 14, 2017 11:29 PM   in response to: Brandon Staggs in response to: Brandon Staggs
Brandon


But if you are sure there are no benefits to upgrading your tools,
then sure, why bother putting the time into it?

Thank you for understanding.

Roy
Dalija Prasnikar

Posts: 2,281
Registered: 11/9/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 14, 2017 6:16 AM   in response to: Roy Lambert in response to: Roy Lambert
Roy Lambert wrote:

For some people there is a benefit, for others none. Why are you so determined to force everyone into the same straightjacket?

Because making two straightjackets is not viable option.

The whole world is Unicode, whether you like it or not, and
there is far more people needing it than ones that don't.

--
Dalija Prasnikar
https://twitter.com/dalijap
https://plus.google.com/+DalijaPrasnikar
Van Swofford

Posts: 389
Registered: 6/28/03
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 14, 2017 8:27 AM   in response to: Dalija Prasnikar in response to: Dalija Prasnikar
Dalija Prasnikar wrote:

Roy Lambert wrote:

For some people there is a benefit, for others none. Why are you so
determined to force everyone into the same straightjacket?

Because making two straightjackets is not viable option.

The whole world is Unicode, whether you like it or not, and
there is far more people needing it than ones that don't.

I believe what Roy is saying is that the whole world is not Unicode, at
least the part that he lives in, and whether you like it or not, he
intends to stay in the non-Unicode world. :-)

I'm pretty much in the same boat, since all the text in my software and
databases is pure ASCII. However, as I am rewriting the app for
multi-platform, I'm converting all strings to the current string
definition, with the exception that I go to/from UTF-8 for storage in
the database in order to keep the size down for quicker internet
transport. So, the new version will be Unicode, but only because it is
pretty trivial to make the transition as I rewrite it, and thankfully I
never fell into the trap of using strings to hold binary data.

--
Cheers,
Van

"Good judgment comes from experience, and a lot of that comes from bad
judgment." - Will Rogers
Dalija Prasnikar

Posts: 2,281
Registered: 11/9/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 14, 2017 10:37 AM   in response to: Van Swofford in response to: Van Swofford
Van Swofford wrote:
Dalija Prasnikar wrote:

Roy Lambert wrote:

For some people there is a benefit, for others none. Why are you so
determined to force everyone into the same straightjacket?

Because making two straightjackets is not viable option.

The whole world is Unicode, whether you like it or not, and
there is far more people needing it than ones that don't.

I believe what Roy is saying is that the whole world is not Unicode, at
least the part that he lives in, and whether you like it or not, he
intends to stay in the non-Unicode world. :-)

I get that ;-)

What I mean is that whole world with exception of English speaking
languages needs Unicode. There are always few exceptions even there
but for the rest Unicode is essential. And then there is English speaking
side, where Unicode is not needed on the first sight, but it depends on
software. If they handle data where it is possible to encounter even so much
as a single letter that falls out of ASCII range they would also need Unicode.

I'm pretty much in the same boat, since all the text in my software and
databases is pure ASCII. However, as I am rewriting the app for
multi-platform, I'm converting all strings to the current string
definition, with the exception that I go to/from UTF-8 for storage in
the database in order to keep the size down for quicker internet
transport. So, the new version will be Unicode, but only because it is
pretty trivial to make the transition as I rewrite it, and thankfully I
never fell into the trap of using strings to hold binary data.

UTF8 is pretty good choice. For first 7 bits, it completely overlaps with
ASCII. Of course, that requires some source code changes, but not
extreme ones.

--
Dalija Prasnikar
https://twitter.com/dalijap
https://plus.google.com/+DalijaPrasnikar
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 14, 2017 11:03 AM   in response to: Dalija Prasnikar in response to: Dalija Prasnikar
Dalija,

Good analysis!

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-14 11:02:59
Van Swofford

Posts: 389
Registered: 6/28/03
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 14, 2017 12:54 PM   in response to: Dalija Prasnikar in response to: Dalija Prasnikar
Dalija Prasnikar wrote:

What I mean is that whole world with exception of English speaking
languages needs Unicode. There are always few exceptions even there
but for the rest Unicode is essential. And then there is English
speaking side, where Unicode is not needed on the first sight, but it
depends on software. If they handle data where it is possible to
encounter even so much as a single letter that falls out of ASCII
range they would also need Unicode.

I've been fortunate up to now, because my problem domain is in the
aviation world, pilots specifically, and they have a mandate to use
English everywhere, regardless of their native language, because of air
traffic control communications. So I've always had an audience for my
software that speaks English, even if it isn't their language of
choice. However, I have a lot of Chinese customers, and I can't help
but think I might make some additional sales if I pander to the
customer a bit and offer several translations, including Chinese. So,
I'm building my new version with the capability of doing that, even if
I leave it in English for the initial release and come back later to
add translations.

UTF8 is pretty good choice. For first 7 bits, it completely overlaps
with ASCII. Of course, that requires some source code changes, but not
extreme ones.

Yes! I'm lucky in that all of the airport data that I use is in ASCII
text files, so when I process it and build the database for
distribution to customers, I can handle it any way I like. The changes
from my old Delphi 7 implementation have been minimal for string
handling, thankfully.

--
Cheers,
Van

"Good judgment comes from experience, and a lot of that comes from bad
judgment." - Will Rogers
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 14, 2017 11:34 PM   in response to: Van Swofford in response to: Van Swofford
Van

If the customer requires it, or you can use it to increase sales that's a good reason to go that route.

Roy Lambert

Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 14, 2017 11:29 PM   in response to: Dalija Prasnikar in response to: Dalija Prasnikar
Dalija


What I mean is that whole world with exception of English speaking
languages needs Unicode. There are always few exceptions even there
but for the rest Unicode is essential.

How did they manage before unicode if it was essential?

And then there is English speaking
side, where Unicode is not needed on the first sight, but it depends on
software. If they handle data where it is possible to encounter even so much
as a single letter that falls out of ASCII range they would also need Unicode.

Or the second sight, or the third sight.If its English it doesn't have those characters. If it encounters them its an error.

Roy
Markus Humm

Posts: 4,978
Registered: 11/9/03
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 14, 2017 11:45 PM   in response to: Roy Lambert in response to: Roy Lambert
Am 15.07.2017 um 08:29 schrieb Roy Lambert:
Dalija


What I mean is that whole world with exception of English speaking
languages needs Unicode. There are always few exceptions even there
but for the rest Unicode is essential.

How did they manage before unicode if it was essential?

Have you ever used printers back in the DOS days?
If not set to the right codepage they wopuld print "strange" chars at
all places where you had umlauts in your text or applications in text
mode using the extended line characters which then looked odd on all
code pages except 437 (which is the US one which in turn lacks qquite a
few umlauts etc.).

=> to answer your question: only with additional work

Greetings

Markus
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 16, 2017 12:32 AM   in response to: Markus Humm in response to: Markus Humm
Markus

Have you ever used printers back in the DOS days?
If not set to the right codepage they wopuld print "strange" chars at
all places where you had umlauts in your text or applications in text
mode using the extended line characters which then looked odd on all
code pages except 437 (which is the US one which in turn lacks qquite a
few umlauts etc.).

Weirdly enough I've used printers before the DOS days - some of us on these forums are really old farts!

Roy
Markus Humm

Posts: 4,978
Registered: 11/9/03
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 16, 2017 8:30 AM   in response to: Roy Lambert in response to: Roy Lambert
Am 16.07.2017 um 09:32 schrieb Roy Lambert:
Markus

Have you ever used printers back in the DOS days?
If not set to the right codepage they wopuld print "strange" chars at
all places where you had umlauts in your text or applications in text
mode using the extended line characters which then looked odd on all
code pages except 437 (which is the US one which in turn lacks qquite a
few umlauts etc.).

Weirdly enough I've used printers before the DOS days - some of us on these forums are really old farts!

Roy

I guess char set 437 was good enough for you and most printers were
shipping with that one active I guess.

But the issues come up when printers are set to different codepages than
the PCs using them and when the PC tries to print umlauts.

In English only texts these issues might not appear.

Greetings

Markus
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 16, 2017 11:52 PM   in response to: Markus Humm in response to: Markus Humm
Markus

I guess char set 437 was good enough for you and most printers were
shipping with that one active I guess.

But the issues come up when printers are set to different codepages than
the PCs using them and when the PC tries to print umlauts.

Ah - a youngster. I come from the days when the "active" character set was the band fitted to the printer. Of course that was before you could change the daisywheel or golfball <g>

Roy
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 17, 2017 11:04 AM   in response to: Roy Lambert in response to: Roy Lambert
Roy,

| Ah - a youngster. I come from the days when the "active" character
| set was the band fitted to the printer

I come from the era prior to printers having "bands"; they had
print-wheels. ;-)

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-17 11:03:06
Van Swofford

Posts: 389
Registered: 6/28/03
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 17, 2017 11:28 AM   in response to: Quentin Correll in response to: Quentin Correll
Quentin Correll wrote:

Roy,

Ah - a youngster. I come from the days when the "active" character
set was the band fitted to the printer

I come from the era prior to printers having "bands"; they had
print-wheels. ;-)

And I come from the era when printers had monks. :-)

--
Cheers,
Van

"Good judgment comes from experience, and a lot of that comes from bad
judgment." - Will Rogers
David Erbas-White

Posts: 200
Registered: 10/11/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 17, 2017 12:29 PM   in response to: Van Swofford in response to: Van Swofford
On 7/17/2017 11:28 AM, Van Swofford wrote:
Quentin Correll wrote:

Roy,

Ah - a youngster. I come from the days when the "active" character
set was the band fitted to the printer

I come from the era prior to printers having "bands"; they had
print-wheels. ;-)

And I come from the era when printers had monks. :-)

...and I used to work for that company... ("It's a miracle!")

(I wonder how many will get that reference <G>

David Erbas-White
Achim Kalwa

Posts: 35
Registered: 9/28/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Aug 8, 2017 1:10 AM   in response to: David Erbas-White in response to: David Erbas-White
David Erbas-White wrote:

Quentin Correll wrote:
And I come from the era when printers had monks. :-)

...and I used to work for that company... ("It's a miracle!")

(I wonder how many will get that reference <G>

Xerox.
https://www.youtube.com/watch?v=faH1FXPqymU
David Erbas-White

Posts: 200
Registered: 10/11/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Aug 8, 2017 8:51 AM   in response to: Achim Kalwa in response to: Achim Kalwa
On 8/8/2017 1:10 AM, Achim Kalwa wrote:
David Erbas-White wrote:

Quentin Correll wrote:
And I come from the era when printers had monks. :-)

...and I used to work for that company... ("It's a miracle!")

(I wonder how many will get that reference <G>

Xerox.
https://www.youtube.com/watch?v=faH1FXPqymU

Wow - that was a blast from the past, I hadn't watched it since the
original presentations...

I'm not talking about the commercial per se, but the copier shown in the
commercial. It was the predecessor to the one I worked on for several
years, but looked almost identical (it was the Xerox 9700 laser
copier/printer -- the laser was bigger than today's entire printer
assembly).

David Erbas-White
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Aug 8, 2017 11:45 AM   in response to: David Erbas-White in response to: David Erbas-White
David,

| I'm not talking about the commercial per se, but the copier shown in
| the commercial. It was the predecessor to the one I worked on for
| several years, but looked almost identical (it was the Xerox 9700
| laser copier/printer -- the laser was bigger than today's entire
| printer assembly).

Me too, again. <g>

--

Q -- XanaNews 1.20-0cfde51 - 2017-08-08 11:45:02
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Aug 8, 2017 11:39 AM   in response to: Achim Kalwa in response to: Achim Kalwa
Achim,

| | (I wonder how many will get that reference <G>
|
| Xerox.
| https://www.youtube.com/watch?v=faH1FXPqymU

Thank you!

--

Q -- XanaNews 1.20-0cfde51 - 2017-08-08 11:39:05
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 18, 2017 10:45 AM   in response to: Van Swofford in response to: Van Swofford
Van,

| And I come from the era when printers had monks. :-)

<g>

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-18 10:45:09
Dan Barclay

Posts: 855
Registered: 11/9/03
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 18, 2017 7:03 AM   in response to: Quentin Correll in response to: Quentin Correll
Quentin Correll wrote:
Roy,

| Ah - a youngster. I come from the days when the "active" character
| set was the band fitted to the printer

I come from the era prior to printers having "bands"; they had
print-wheels. ;-)

But the bands were more fun, you can turn them into a chainsaw.

You could snatch a couple of job cards from someone else's account, make a short program that printed dashes or underlines across the page a bunch of times, then follow with a dozen or so form feeds. Put someone else's job card with the deck and drop it in the queue (a stacker then).

You could train operators to dive across the room to try and stop the printer before it balled up in paper. Do that a few times then every time they hear the sound of lines across the page they perk up.

Or so I'm told.

Dan
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 18, 2017 7:15 AM   in response to: Dan Barclay in response to: Dan Barclay
I bet we all remember 30cps dial up and modems as well <G>

Roy Lambert
Robert Evans

Posts: 112
Registered: 10/16/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 18, 2017 7:45 AM   in response to: Roy Lambert in response to: Roy Lambert
In article <926622 at forums dot embarcadero dot com>, Roy Lambert
<roy at lybster dot me dot uk> wrote
I bet we all remember 30cps dial up and modems as well <G>

Some of us learnt to program on teleprinters on 110 baud (10cps) dial-up
:)

--
Bob Evans
Dan Barclay

Posts: 855
Registered: 11/9/03
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 18, 2017 8:59 AM   in response to: Robert Evans in response to: Robert Evans
Robert Evans wrote:
In article <926622 at forums dot embarcadero dot com>, Roy Lambert
<roy at lybster dot me dot uk> wrote
I bet we all remember 30cps dial up and modems as well <G>

Some of us learnt to program on teleprinters on 110 baud (10cps) dial-up
:)

I thought I'd gone to heaven when I took a graduate course in network analysis and got an account with tty instead of being limited to punch cards.

Then I learned the absolute necessity of backups when the data center lost some tapes with a semester's work. When they finally recovered files nearly a week later, the first thing I did was have them dumped to punch cards. Those went in a box under my bed.

Dan
Robert Evans

Posts: 112
Registered: 10/16/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 19, 2017 12:22 PM   in response to: Dan Barclay in response to: Dan Barclay
In article <926639 at forums dot embarcadero dot com>, Dan Barclay <?@?.?.invalid>
wrote
Robert Evans wrote:
Some of us learnt to program on teleprinters on 110 baud (10cps) dial-up
:)

I thought I'd gone to heaven when I took a graduate course in network
analysis and got an account with tty instead of being limited to punch
cards.

Yes but paper tape was an absolute pain when you needed to edit a typo
in the middle of your source code! (Perhaps marginally less so on the
Olivetti TE300s than on Teletype ASR33s, although the tape punches of
the former tended to be unreliable.) To re-punch the whole tape, or cut
and splice, that was the question...

After a few months of timesharing (Fortran IV on a GE-635), I got a
vacation job at a large automotive component maker, programming a market
research suite in Fortran on a Univac 1100.

There was normally a significant wait before the ladies of the key-punch
department would finish with the weekly and monthly production jobs and
then get around to looking at my coding forms, so I acquired my own card
punch machine - the kind with 12 rubber-topped buttons, one for each
knife - and a good memory for all the code combinations :)

Then I learned the absolute necessity of backups when the data center
lost some tapes with a semester's work. When they finally recovered
files nearly a week later, the first thing I did was have them dumped
to punch cards. Those went in a box under my bed.

The trick with punched cards was to line them up in their box and then
draw a diagonal line across the top edge of the stack, thereby making it
obvious if some kind person had dropped the lot.

--
Bob Evans
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 21, 2017 10:48 AM   in response to: Robert Evans in response to: Robert Evans
Robert,

| Yes but paper tape was an absolute pain when you needed to edit a
| typo in the middle of your source code!

Ah,... another ol'timer! <g> I too remember those days.

| There was normally a significant wait before the ladies of the
| key-punch department would finish with the weekly and monthly
| production jobs and then get around to looking at my coding forms, so
| I acquired my own card punch machine - the kind with 12 rubber-topped
| buttons, one for each knife - and a good memory for all the code
| combinations :)

Ah, yes,... the IBM-001. <g>

https://www.google.com/search?q=ibm+001&rlz=1T4GUEA_enUS633US633&tbm=isch&imgil=EfegnFTeTAWosM%253A%253B2d7qz8Wzx34w3M%253Bhttp%25253A%25252F%25252Fvirgilmachine.blogspot.com%25252F2014%25252F12%25252Fretrotech-ibm-001-manual-keypunch.html&source=iu&pf=m&fir=EfegnFTeTAWosM%253A%252C2d7qz8Wzx34w3M%252C_&usg=__G8PFFhFzlb00FEFMn_ANSpTQHr0%3D&biw=1068&bih=737&dpr=1.25&ved=0ahUKEwi2k8vH9JrVAhUJh1QKHWPFCVAQyjcIXA&ei=RjtyWbbYB4mO0gLjiqeABQ#imgrc=EfegnFTeTAWosM:&spf=1500658518996

I too did a LOT of work "in the ol'days" on one of those. <g> And
then I finally got my own IBM-026. <g>

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-21 10:33:23

Robert Evans

Posts: 112
Registered: 10/16/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 24, 2017 3:18 AM   in response to: Quentin Correll in response to: Quentin Correll
In article <927108 at forums dot embarcadero dot com>, Quentin Correll
<qcorrell at pacNObell dot net> wrote
| I acquired my own card punch machine

Ah, yes,... the IBM-001. <g>

I think mine was the British ICT (later ICL) version:
http://www.computinghistory.org.uk/det/3415/ICL-Hand-Key-Punch-Card-Machine/

I too did a LOT of work "in the ol'days" on one of those. <g> And
then I finally got my own IBM-026. <g>

I loved the IBM 29 when a couple of years later I managed to get my
hands on one. Maybe not quite as solidly built as the 26 but very fast
and easy to use. <g>

--
Bob Evans
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 24, 2017 11:30 AM   in response to: Robert Evans in response to: Robert Evans
Robert,

| Bob Evans

Your name brings back some old memories of mine. Back in ancient-times
(Mid-to-late '50's.), when I worked for IBM, I worked with a Robert
(Bo) Evans. Now deceased. :-(

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-24 11:24:55
Robert Evans

Posts: 112
Registered: 10/16/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 24, 2017 1:27 PM   in response to: Quentin Correll in response to: Quentin Correll
In article <927431 at forums dot embarcadero dot com>, Quentin Correll
<qcorrell at pacNObell dot net> wrote
Your name brings back some old memories of mine. Back in ancient-times
(Mid-to-late '50's.), when I worked for IBM, I worked with a Robert
(Bo) Evans. Now deceased. :-(

I was but a kid in the 50's. My Dad was also a Bob but, as far as I
know, the nearest he ever got to the USA was chatting over ham radio to
his pals with 'W' and 'K' callsigns :-)

--
Bob Evans
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 18, 2017 10:49 AM   in response to: Robert Evans in response to: Robert Evans
Robert,

| | I bet we all remember 30cps dial up and modems as well <G>
|
| Some of us learnt to program on teleprinters on 110 baud (10cps)
| dial-up :)

Some of us learned to program before there were 110 baud teleprinters.
<g>

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-18 10:47:31
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 18, 2017 11:02 AM   in response to: Roy Lambert in response to: Roy Lambert
Roy,

| I bet we all remember 30cps dial up and modems as well <G>

Yep.

Anyone here besides me remember ABC and RUSH? <g>

[Jim Babcock shared his RAND Corporation office with me (I worked for
IBM.) many years ago. We still exchange Xmas cards.]

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-18 10:56:31
David Erbas-White

Posts: 200
Registered: 10/11/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 18, 2017 12:09 PM   in response to: Quentin Correll in response to: Quentin Correll
On 7/18/2017 11:02 AM, Quentin Correll wrote:
We still exchange Xmas cards.

Do they have holes punched in them?!? <G>

David Erbas-White

Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 18, 2017 2:11 PM   in response to: David Erbas-White in response to: David Erbas-White
David,

| | We still exchange Xmas cards.
|
| Do they have holes punched in them?!? <G>

<giggle>

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-18 14:10:13
Van Swofford

Posts: 389
Registered: 6/28/03
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 18, 2017 7:38 AM   in response to: Dan Barclay in response to: Dan Barclay
Dan Barclay wrote:

Or so I'm told.

LOL!

--
Cheers,
Van

"Good judgment comes from experience, and a lot of that comes from bad
judgment." - Will Rogers
Dan Barclay

Posts: 855
Registered: 11/9/03
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 18, 2017 9:08 AM   in response to: Van Swofford in response to: Van Swofford
Van Swofford wrote:
Dan Barclay wrote:

Or so I'm told.

LOL!

Decent use of Seymour Cray's CDC 6600 supercomputer. Or, that's what it seemed like at the time.

Dan

http://www.computerhistory.org/revolution/supercomputers/10/33

Edited by: Dan Barclay on Jul 18, 2017 11:30 AM
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 18, 2017 10:56 AM   in response to: Dan Barclay in response to: Dan Barclay
Dan,

| You could snatch a couple of job cards from someone else's account,
| make a short program that printed dashes or underlines across the
| page a bunch of times, then follow with a dozen or so form feeds.
| Put someone else's job card with the deck and drop it in the queue (a
| stacker then).

Naughty, naughty! ;-)

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-18 10:56:03
Markus Humm

Posts: 4,978
Registered: 11/9/03
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 17, 2017 12:20 PM   in response to: Roy Lambert in response to: Roy Lambert
Am 17.07.2017 um 08:52 schrieb Roy Lambert:
Markus

I guess char set 437 was good enough for you and most printers were
shipping with that one active I guess.

But the issues come up when printers are set to different codepages than
the PCs using them and when the PC tries to print umlauts.

Ah - a youngster. I come from the days when the "active" character set was the band fitted to the printer. Of course that was before you could change the daisywheel or golfball <g>

Roy

Ok, I started when needle printers were current.
But I'm not a youngster!
I can write batch files and use command line!

Greetings

Markus
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 16, 2017 11:02 AM   in response to: Roy Lambert in response to: Roy Lambert
Roy,

| - some of us on these forums are really old farts!

<raising hand>

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-16 11:01:53
Dalija Prasnikar

Posts: 2,281
Registered: 11/9/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 15, 2017 3:27 AM   in response to: Roy Lambert in response to: Roy Lambert
Roy Lambert wrote:
Dalija


What I mean is that whole world with exception of English speaking
languages needs Unicode. There are always few exceptions even there
but for the rest Unicode is essential.

How did they manage before unicode if it was essential?

Like said, it was a nightmare, sometimes you even had to manually go through
converted data and fix wrong conversions.

And then there is English speaking
side, where Unicode is not needed on the first sight, but it depends on
software. If they handle data where it is possible to encounter even so much
as a single letter that falls out of ASCII range they would also need Unicode.

Or the second sight, or the third sight.If its English it doesn't have those characters. If it encounters them its an error.

What about people's names? If someone with foreign name comes along
and you have to store the name accurately.

--
Dalija Prasnikar
https://twitter.com/dalijap
https://plus.google.com/+DalijaPrasnikar
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 16, 2017 12:32 AM   in response to: Dalija Prasnikar in response to: Dalija Prasnikar
Dalija

Like said, it was a nightmare, sometimes you even had to manually go through
converted data and fix wrong conversions.

They managed therefore it wasn't essential - made live easier when it came along that I do accept.

What about people's names? If someone with foreign name comes along
and you have to store the name accurately.

No I wouldn't. I'd want to store it in a form I could recognise and pronounce. Its a silly argument - should a system written for Indian users store a Chinese name in Chinese characters?

Roy
Dalija Prasnikar

Posts: 2,281
Registered: 11/9/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 16, 2017 2:44 AM   in response to: Roy Lambert in response to: Roy Lambert
Roy Lambert wrote:
Dalija

Like said, it was a nightmare, sometimes you even had to manually go through
converted data and fix wrong conversions.

They managed therefore it wasn't essential - made live easier when it came along that I do accept.

We are talking about programming, don't we? If you have to fix stuff manually,
instead automatically through code, then this by no means can be called "managing".
Unicode IS essential,

What about people's names? If someone with foreign name comes along
and you have to store the name accurately.

No I wouldn't. I'd want to store it in a form I could recognise and pronounce. Its a silly argument - should a system written for Indian users store a Chinese name in Chinese characters?

My name is actually Prašnikar. So you don't have to be able
to pronounce it, you would probably fail to pronounce it anyway.
But for some use cases it is essential that it is correctly written.

Purchasing with credit card on Internet was also a nightmare.
Sometimes if I would write Prašnikar, credit card validation would fail,
sometimes it would fail if I would use Prasnikar.

If you never had to deal with special characters and their encodings,
you are probably not the right person to say how easy or hard was that
without the Unicode.

If you ever had to deal with them, we would not have this kind of
conversation right now.

--
Dalija Prasnikar
https://twitter.com/dalijap
https://plus.google.com/+DalijaPrasnikar
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 16, 2017 6:49 AM   in response to: Dalija Prasnikar in response to: Dalija Prasnikar
Dalija


But for some use cases it is essential that it is correctly written.

No. It needs to be wriiten/stored consistently - not correctly. That's nice but not essential.

Roy
Dalija Prasnikar

Posts: 2,281
Registered: 11/9/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 16, 2017 7:11 AM   in response to: Roy Lambert in response to: Roy Lambert
Roy Lambert wrote:
Dalija


But for some use cases it is essential that it is correctly written.

No. It needs to be wriiten/stored consistently - not correctly. That's nice but not essential.

It is essential if you have to exchange that data with others. The way
they might store it may differ and then you have inconsistencies.

While it is not essential that my name is correctly written here on forums,
it is essential that it is correctly written in my bank account and other
related data.

--
Dalija Prasnikar
https://twitter.com/dalijap
https://plus.google.com/+DalijaPrasnikar
Dalija Prasnikar

Posts: 2,281
Registered: 11/9/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 16, 2017 2:51 AM   in response to: Roy Lambert in response to: Roy Lambert
Roy Lambert wrote:

What about people's names? If someone with foreign name comes along
and you have to store the name accurately.

No I wouldn't. I'd want to store it in a form I could recognise and pronounce. Its a silly argument - should a system written for Indian users store a Chinese name in Chinese characters?

BTW, Croatian alphabet does not have letters Q W X Y.
Now imagine our alphabet ruled the computer world,
and you had to jump through the hoops to use those
characters.

Also, we quite manage to handle situations when someone's
name contains those characters. I am sure you would not
want me to write your name as Roj (that would be Croatian
equivalent) Or even better, for John, our version would be Džon.

--
Dalija Prasnikar
https://twitter.com/dalijap
https://plus.google.com/+DalijaPrasnikar
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 16, 2017 6:54 AM   in response to: Dalija Prasnikar in response to: Dalija Prasnikar
Dalija


BTW, Croatian alphabet does not have letters Q W X Y.
Now imagine our alphabet ruled the computer world,
and you had to jump through the hoops to use those
characters.

It doesn't and I don't - shrug

Also, we quite manage to handle situations when someone's
name contains those characters.

That's good - but you don't need unicode to do it <g>

I am sure you would not
want me to write your name as Roj (that would be Croatian
equivalent)

Are you sure - have you just substituted letters? There are two possible sources for Roy. One is Rory the other is roi I prefer the latter so just call me King Lambert

Roy
Dalija Prasnikar

Posts: 2,281
Registered: 11/9/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 16, 2017 7:08 AM   in response to: Roy Lambert in response to: Roy Lambert
Roy Lambert wrote:
Dalija


BTW, Croatian alphabet does not have letters Q W X Y.
Now imagine our alphabet ruled the computer world,
and you had to jump through the hoops to use those
characters.

It doesn't and I don't - shrug

I am just trying to explain how actually complicated things are
and fact that you don't care about Unicode is just pure luck.

Also, we quite manage to handle situations when someone's
name contains those characters.

That's good - but you don't need unicode to do it <g>

I am talking about difference between writing it correctly and
pronouncing it correctly. And again the fact we don't need Unicode
to handle Q W X Y letters is merely based in fact that we are all
using ASCII standard that has those letters covered with first 7-bits
and that this part is not subject of interpretation. And again, that is
just mere luck.

I am sure you would not
want me to write your name as Roj (that would be Croatian
equivalent)

Are you sure - have you just substituted letters? There are two possible sources for Roy. One is Rory the other is roi I prefer the latter so just call me King Lambert

I am sure, in Croatian Y is replaced with J.

--
Dalija Prasnikar
https://twitter.com/dalijap
https://plus.google.com/+DalijaPrasnikar
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 16, 2017 11:42 PM   in response to: Dalija Prasnikar in response to: Dalija Prasnikar
Dalija

I understand and accept that I'm in a different position to others. Yes its a matter of luck that computers were developed by English speakers and that influenced the choice of character set. Just think if the inventors had used one of the pictographic languages we'd be able to write half a program with one "character".

What annoys me is the way the zelots keep trying to ram unicode down my throat when I (me personally) don't need it. I can understand that for many it is a great boon but not for me. At best, for me, it would be neutral, more likely, an annoyance and possibly a problem.

Roy Lambert

Dalija Prasnikar

Posts: 2,281
Registered: 11/9/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 17, 2017 1:53 AM   in response to: Roy Lambert in response to: Roy Lambert
Roy Lambert wrote:
Dalija

I understand and accept that I'm in a different position to others. Yes its a matter of luck that computers were developed by English speakers and that influenced the choice of character set. Just think if the inventors had used one of the pictographic languages we'd be able to write half a program with one "character".

What annoys me is the way the zelots keep trying to ram unicode down my throat when I (me personally) don't need it. I can understand that for many it is a great boon but not for me. At best, for me, it would be neutral, more likely, an annoyance and possibly a problem.


You keep thinking there is a choice. Point is there is no choice, not
because someone didn't want to give it to you, but because giving
you the choice would be way, way too complicated, expensive and
time consuming for tool vendor.

I wish, even more than you do, that there could be Unicode switch,
so we don't have to waste time on such futile discussions. But there
is not, and it was not possible to do it.

And the zealots you are talking about, are not zealots, they are people
that desperately need Unicode support. All other commonly used tools
out there are Unicode, because that is what developers need.

And if you don't need Unicode, you can use those tools and you don't
even have to think about it. Just like you can use Unicode Delphi.
The only problem there is is converting old code to Unicode Delphi,
and that is it. And believe me, folks at Embarcadero have done
everything in their power to minimize changes in your code transition.

--
Dalija Prasnikar
https://twitter.com/dalijap
https://plus.google.com/+DalijaPrasnikar
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 17, 2017 2:18 AM   in response to: Dalija Prasnikar in response to: Dalija Prasnikar
Dalija

You keep thinking there is a choice. Point is there is no choice, not
because someone didn't want to give it to you, but because giving
you the choice would be way, way too complicated, expensive and
time consuming for tool vendor.

I wish, even more than you do, that there could be Unicode switch,
so we don't have to waste time on such futile discussions. But there
is not, and it was not possible to do it.

I do have to disagree with this. It was possible, it would not have been to complicated, but was obviously seen as to expensive.

And the zealots you are talking about, are not zealots, they are people
that desperately need Unicode support. All other commonly used tools
out there are Unicode, because that is what developers need.

They may do - I accept that. I don't and few of them seem able to even understand that let alone accept it.

And believe me, folks at Embarcadero have done
everything in their power to minimize changes in your code transition.

Wrong.

Roy
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 17, 2017 11:05 AM   in response to: Roy Lambert in response to: Roy Lambert
Roy,

| What annoys me is the way the zelots keep trying to ram unicode down
| my throat when I (me personally) don't need it. I can understand that
| for many it is a great boon but not for me. At best, for me, it would
| be neutral, more likely, an annoyance and possibly a problem.

Ditto!!!

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-17 11:05:29
Brandon Staggs

Posts: 633
Registered: 3/3/01
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 17, 2017 1:36 PM   in response to: Quentin Correll in response to: Quentin Correll
"Quentin Correll" wrote on Mon, 17 Jul 2017 11:05:40 -0700:

Roy,

| What annoys me is the way the zelots keep trying to ram unicode down
| my throat when I (me personally) don't need it. I can understand that
| for many it is a great boon but not for me. At best, for me, it would
| be neutral, more likely, an annoyance and possibly a problem.

Ditto!!!

Really, what are the rest of us missing? Is there some kind of army
out there making people use Unicode? Unicode "just is." I mean, it's
not like some gang of ASCII users went around bashing the heads in of
anyone still using EBCDIC and that's how it became a standard, and
neither is anyone forcing people to use Unicode as far as I can tell.
There is no viable alternative to Unicode -- UTF-8 in particular
drives the internet. You can't escape it and I haven't seen any clear
explanation of "the problem." UTF-8 in particular is perfect for
people who insist all they need is ASCII. Because the first 7 bit of
any character sequence in UTF-8 is... ASCII!

Shrug.

--
Brandon Staggs
StudyLamp Software LLC
http://www.studylamp.com
Brandon Staggs

Posts: 633
Registered: 3/3/01
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 17, 2017 1:41 PM   in response to: Brandon Staggs in response to: Brandon Staggs
"Brandon Staggs" wrote on Mon, 17 Jul 2017 13:36:08 -0700:

Because the first 7 bit of
any character sequence in UTF-8 is... ASCII!

Worded poorly. UTF-8 is 100% backward compatible with ASCII, since a
7-bit code point in UTF-8 happens to correspond exactly with all 7-bit
ASCII values. And even if someone actually means they use extended
ASCII, UTF-8 is still almost perfect for them. I just don't get the
resistance.

Maybe some people are confusing changes in the Delphi string types
(which happened forever ago in computer time) with the "issue" of
Unicode?

--
Brandon Staggs
StudyLamp Software LLC
http://www.studylamp.com
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 18, 2017 12:27 AM   in response to: Brandon Staggs in response to: Brandon Staggs
Brandon

We stand on either side of a great divide and communicate only when the wind blows a few random words across the divide.

Roy Lambert

Brandon Staggs

Posts: 633
Registered: 3/3/01
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 18, 2017 7:24 AM   in response to: Roy Lambert in response to: Roy Lambert
"Roy Lambert" wrote on Tue, 18 Jul 2017 00:27:37 -0700:

We stand on either side of a great divide and communicate only when the wind blows a few random words across the divide.

Heh.

Sometimes I don't want to put the time into something I am not sure
will help me directly, but often I hear enough from other people to
get past that and find I end up adding knowledge or tools that I wish
I had not gone so long without. One great example is Regex. Regex is
cryptic and can take a little while to really take it in fluently, but
once done, you have an extremely powerful tool at your disposal. My
only regret is not sitting down and spending time to learn it earlier.

Unicode is one of those things where not only is a small amount of
knowledge about it extremely helpful, resisting it would take more
effort than just learning it-- for me, of course. I'm glad it was
added to Delphi as the default string type, something that is frankly
a no-brainer in retrospect. There was no better way for them to
properly support Windows moving forward.

Anyway.

--
Brandon Staggs
StudyLamp Software LLC
http://www.studylamp.com
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 19, 2017 1:20 AM   in response to: Brandon Staggs in response to: Brandon Staggs
Brandon

Sometimes I don't want to put the time into something I am not sure
will help me directly, but often I hear enough from other people to
get past that and find I end up adding knowledge or tools that I wish
I had not gone so long without. One great example is Regex. Regex is
cryptic and can take a little while to really take it in fluently, but
once done, you have an extremely powerful tool at your disposal. My
only regret is not sitting down and spending time to learn it earlier.

I keep thinking I really must learn regex. Two things always cross my mind - 1) I used APL, loved it, but think that's enough gobbledegook for one lifetime 2) for the use I'd make of it regex would (if I know the syntax) be faster to write but a lot slower to run, and it generally doesn't take to long to write & test purpose built code.

Unicode is one of those things where not only is a small amount of
knowledge about it extremely helpful, resisting it would take more
effort than just learning it-- for me, of course. I'm glad it was
added to Delphi as the default string type, something that is frankly
a no-brainer in retrospect. There was no better way for them to
properly support Windows moving forward.

I do have to agree with the "no-brainer", but probably not in the way you were thinking <g>

Roy
Mike Margerum

Posts: 573
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 19, 2017 8:46 AM   in response to: Roy Lambert in response to: Roy Lambert
I keep thinking I really must learn regex. Two things always cross my mind - 1) I used APL, loved it, but think that's enough gobbledegook for one lifetime 2) for the use I'd make of it regex would (if I know the syntax) be faster to write but a lot slower to run, and it generally doesn't take to long to write & test purpose built code.

The problem I have with regex is every time i go to use it, I
effectively have to relearn it. The syntax just doesn't stick. I guess
I just don't use it enough.
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 19, 2017 10:35 AM   in response to: Mike Margerum in response to: Mike Margerum
Mike

The problem I have with regex is every time i go to use it, I
effectively have to relearn it. The syntax just doesn't stick. I guess
I just don't use it enough.

That's probably why its faster for me to write custom code each time <g>

Roy
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 18, 2017 10:52 AM   in response to: Brandon Staggs in response to: Brandon Staggs
Brandon,

| Really, what are the rest of us missing? Is there some kind of army
| out there making people use Unicode? Unicode "just is."

It wasn't always. <g>

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-18 10:51:57
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 16, 2017 11:07 AM   in response to: Dalija Prasnikar in response to: Dalija Prasnikar
Dalija,

| BTW, Croatian alphabet does not have letters Q W X Y.

Hmmm,... Good thing I don't have to do business in Croatia. ;-)

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-16 11:06:38
Dalija Prasnikar

Posts: 2,281
Registered: 11/9/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 16, 2017 11:36 AM   in response to: Quentin Correll in response to: Quentin Correll
Quentin Correll wrote:
Dalija,

| BTW, Croatian alphabet does not have letters Q W X Y.

Hmmm,... Good thing I don't have to do business in Croatia. ;-)

No need to worry, we can handle those ;-)

My point was that it is common misconception that English
alphabet is some kind of baseline, and other languages
just have some characters extra (I am not covering non-latin ones)

Point is that Croatian comparing to English does have few characters
extra, but it is not just mere extension, because some English
characters are not part of our original alphabet nor our language
uses them.

Just adding more mess to the subject ;-)

--
Dalija Prasnikar
https://twitter.com/dalijap
https://plus.google.com/+DalijaPrasnikar
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 16, 2017 4:14 PM   in response to: Dalija Prasnikar in response to: Dalija Prasnikar
Dalija,

| Just adding more mess to the subject ;-)

<g>

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-16 16:14:19
Igor Savkic

Posts: 27
Registered: 8/26/01
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 17, 2017 1:45 AM   in response to: Dalija Prasnikar in response to: Dalija Prasnikar
Point is that Croatian comparing to English does have few characters
extra, but it is not just mere extension, because some English
characters are not part of our original alphabet nor our language
uses them.

Just adding more mess to the subject ;-)

Hey, you're the lucky one, on top of latin alphabet we have one more, cyrillic,
now that's a mess ;)
Joke aside, thanks to Unicode and UTF8, everything is much easier than 20 years
ago. Altough there are still of old printers in use and if you want to print in
high speed you still need to handle code pages and initialization sequences. And
of course there's still YUSCII, it's fascinating how resilient it is ;)
Dalija Prasnikar

Posts: 2,281
Registered: 11/9/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 17, 2017 2:04 AM   in response to: Igor Savkic in response to: Igor Savkic
Igor Savkic wrote:
Point is that Croatian comparing to English does have few characters
extra, but it is not just mere extension, because some English
characters are not part of our original alphabet nor our language
uses them.

Just adding more mess to the subject ;-)

Hey, you're the lucky one, on top of latin alphabet we have one more, cyrillic,
now that's a mess ;)
Joke aside, thanks to Unicode and UTF8, everything is much easier than 20 years
ago. Altough there are still of old printers in use and if you want to print in
high speed you still need to handle code pages and initialization sequences. And
of course there's still YUSCII, it's fascinating how resilient it is ;)

Yeah, having cyrillic is additional fun...

I am glad I don't have to handle YUSCII any more...

--
Dalija Prasnikar
https://twitter.com/dalijap
https://plus.google.com/+DalijaPrasnikar
Luigi Sandon

Posts: 345
Registered: 10/15/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 19, 2017 11:35 AM   in response to: Dalija Prasnikar in response to: Dalija Prasnikar
BTW, Croatian alphabet does not have letters Q W X Y.

And the Latin alphabet is not the only alphabet around... and today you may need to use more than one alphabet in the same application or document. Sure, if you live in a little local world you don't have such issues, if you live in a broader one they are quite common.
Luigi Sandon

Posts: 345
Registered: 10/15/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 19, 2017 11:42 AM   in response to: Roy Lambert in response to: Roy Lambert
No I wouldn't. I'd want to store it in a form I could recognise and pronounce. Its a silly argument - should a system written for Indian users store a Chinese name in Chinese characters?

Yes.
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 19, 2017 11:46 PM   in response to: Luigi Sandon in response to: Luigi Sandon
Luigi

No I wouldn't. I'd want to store it in a form I could recognise and pronounce. Its a silly argument - should a system written for Indian users store a Chinese name in Chinese characters?

Yes.

I have great sympathy for anyone using a system written to pander to that.

Do you seriously expect all users of all computer systems to be able to read/write/speak all languages? Just in case you hadn't realised that is what you're asking for.

Roy
Luigi Sandon

Posts: 345
Registered: 10/15/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 20, 2017 1:02 AM   in response to: Roy Lambert in response to: Roy Lambert
Do you seriously expect all users of all computer systems to be able to read/write/speak all languages? Just in case you hadn't realised that is what you're asking for.

Do you seriously expect all users of all computer systems to be able to read/write/speak *a single language only*? Globalization means there is a lot of people working everywhere in the world while needing to communicate and store data from a lot of other people in different countries using different languages. Any software that can't cope is of very limited use, and development tools can't ignore it. Creating separate ANSI and Unicode versions is a too big effort.
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 20, 2017 1:45 AM   in response to: Luigi Sandon in response to: Luigi Sandon
Luigi

Do you seriously expect all users of all computer systems to be able to read/write/speak all languages? Just in case you hadn't realised that is what you're asking for.

Do you seriously expect all users of all computer systems to be able to read/write/speak *a single language only*? Globalization means there is a lot of people working everywhere in the world while needing to communicate and store data from a lot of other people in different countries using different languages. Any software that can't cope is of very limited use, and development tools can't ignore it. Creating separate ANSI and Unicode versions is a too big effort.

Strangely enough I do not expect all users of all computer systems to read/write/speak only one language. I expect a significant proportion to do so. I also expect that a significant proportion of the polyglots to have English as a second language. I also expect that there will be regional variations more in dialect than written form. Whilst I'm certain there will be some out there who are equally fluent in Cantonese and Punjabi (and I mean unlike my equal fluency - zero) I doubt there are many.

Lets you myself and yourself as test cases. I read/write/speak English, a bit of schoolboy French - that's it. How about you?

Roy
Dan Barclay

Posts: 855
Registered: 11/9/03
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 24, 2017 9:50 AM   in response to: Roy Lambert in response to: Roy Lambert
Roy Lambert wrote:
Luigi

Do you seriously expect all users of all computer systems to be able to read/write/speak all languages? Just in case you hadn't realised that is what you're asking for.

Do you seriously expect all users of all computer systems to be able to read/write/speak *a single language only*? Globalization means there is a lot of people working everywhere in the world while needing to communicate and store data from a lot of other people in different countries using different languages. Any software that can't cope is of very limited use, and development tools can't ignore it. Creating separate ANSI and Unicode versions is a too big effort.

Strangely enough I do not expect all users of all computer systems to read/write/speak only one language. I expect a significant proportion to do so. I also expect that a significant proportion of the polyglots to have English as a second language. I also expect that there will be regional variations more in dialect than written form. Whilst I'm certain there will be some out there who are equally fluent in Cantonese and Punjabi (and I mean unlike my equal fluency - zero) I doubt there are many.

Lets you myself and yourself as test cases. I read/write/speak English, a bit of schoolboy French - that's it. How about you?

Aviation has a common language, because there simply must be one. That apparently hasn't been much of a problem even in very busy commercial airspace all over the world and even in military operations.

There apparently is more confusion between Britts and Americans on joint operations, though each communicates with everybody else just fine. Two peoples separated by a common language <g>.

Dan
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 24, 2017 11:31 AM   in response to: Dan Barclay in response to: Dan Barclay
Dan,

| Two peoples separated by a common language <g>.

<chuckle>

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-24 11:31:40
Roy Lambert

Posts: 1,056
Registered: 10/21/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 24, 2017 11:39 PM   in response to: Dan Barclay in response to: Dan Barclay
Dan

There apparently is more confusion between Britts and Americans on joint operations, though each communicates with everybody else just fine. Two peoples separated by a common language <g>.

+quite a lot <G>

Roy
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 14, 2017 11:02 AM   in response to: Van Swofford in response to: Van Swofford
Van,

| I believe what Roy is saying is that the whole world is not Unicode,
| at least the part that he lives in, and whether you like it or not, he
| intends to stay in the non-Unicode world. :-)
|
| I'm pretty much in the same boat, since all the text in my software
| and databases is pure ASCII.

That's also my personal situation. I don't use, need or want Unicode.
For ME it's a major PITA!

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-14 10:59:55
Brandon Staggs

Posts: 633
Registered: 3/3/01
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 14, 2017 4:27 PM   in response to: Quentin Correll in response to: Quentin Correll
"Quentin Correll" wrote on Fri, 14 Jul 2017 11:02:04 -0700:

I don't use, need or want Unicode.
For ME it's a major PITA!

I don't understand what could be the pain point for you. You have
said before that you don't think using strings as binary storage
buffers is good, so if you're not doing that, why is it such a pain to
have your strings stored in the native system's encoding?

--
Brandon Staggs
StudyLamp Software LLC
http://www.studylamp.com
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 15, 2017 10:57 AM   in response to: Brandon Staggs in response to: Brandon Staggs
Brandon,

| why is it such a pain to
| have your strings stored in the native system's encoding?

In thinking about trying to answer your question I realized that I
don't remember how or why. It's a decades old [DOS environment]
"position" I've lived with for a very long time. <g>

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-15 10:49:57
Quentin Correll


Posts: 2,390
Registered: 12/1/99
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Jul 14, 2017 11:02 AM   in response to: Roy Lambert in response to: Roy Lambert
Roy,

| As things stand I have no, zero, none at all, reason to use unicode.

Ditto!!!

--

Q -- XanaNews 1.20-0cfde51 - 2017-07-14 10:58:09
Tony Caduto


Posts: 2
Registered: 4/22/00
Re: JOEL ON SOFTWARE
Click to report abuse...   Click to reply to this thread Reply
  Posted: Oct 26, 2017 11:16 AM   in response to: Roy Lambert in response to: Roy Lambert
LOL, just because "JOEL" says it's so does not mean it is.
That guy is kind of a Xick.....replace the X with a D LOL

I was once interviewed by some dipshit manager that read his "guerrilla interview" tactics and believed every word that jerk said, worst interview of my entire life. I ended up just walking out.

--
Tony Caduto

Legend
Helpful Answer (5 pts)
Correct Answer (10 pts)

Server Response from: ETNAJIVE02