www.digitalmars.com         C & C++   DMDScript  

D - Let's organize ourselves - a plan for D APIs.

reply Achilleas Margaritis <axilmar b-online.gr> writes:
It will be a real pity if D did not succeed as a programming language. In order 
for it to succeed, it needs a good collection of APIs which cover the most
usual 
needs for writing applications.

We are a lot of people in here that are interested about D. I suggest we 
organize ourselves and design the APIs and libraries which Digital Mars will 
approve as the official D APIs.

I am saying all this because many people have individually started to code 
libraries, especially GUIs. It would be a waste of resources for each one of us 
to try individually, since APIs are a lot of work. Therefore, I suggest we make 
some sort of 'committee', by whoever is willing to participate, and produce a 
set of specifications for the APIs that should exist.

After we finish specifications, we can divide the work in chunks and proceed to 
the implementation either individually or in groups.

What does everybody think on this ?
Apr 20 2004
next sibling parent reply "Matthew" <matthew.hat stlsoft.dot.org> writes:
I am still of the same opinion as when Lars suggested this several months
ago. (My post was "Re: D standard library group proposal", on the 4th Feb
2004. I don't know how to use the web interface to search, I'm afraid.)


"Achilleas Margaritis" <axilmar b-online.gr> wrote in message
news:c640sa$lsh$1 digitaldaemon.com...
 It will be a real pity if D did not succeed as a programming language. In

 for it to succeed, it needs a good collection of APIs which cover the most

 needs for writing applications.

 We are a lot of people in here that are interested about D. I suggest we
 organize ourselves and design the APIs and libraries which Digital Mars

 approve as the official D APIs.

 I am saying all this because many people have individually started to code
 libraries, especially GUIs. It would be a waste of resources for each one

 to try individually, since APIs are a lot of work. Therefore, I suggest we

 some sort of 'committee', by whoever is willing to participate, and

 set of specifications for the APIs that should exist.

 After we finish specifications, we can divide the work in chunks and

 the implementation either individually or in groups.

 What does everybody think on this ?

Apr 20 2004
next sibling parent reply J C Calvarese <jcc7 cox.net> writes:
Matthew wrote:
 I am still of the same opinion as when Lars suggested this several months
 ago. (My post was "Re: D standard library group proposal", on the 4th Feb
 2004. I don't know how to use the web interface to search, I'm afraid.)

I found your message: D/23166. I think your points are as valid now as when you wrote it. By the way, if you find the message in your newsreader, you can probably "View Source" or "View Headers" to find out the message number in 2 seconds. For example: Path: digitalmars.com!not-for-mail ... X-Complaints-To: usenet digitalmars.com NNTP-Posting-Date: Tue, 20 Apr 2004 20:31:42 +0000 (UTC) X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 6.00.2800.1409 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1409 Xref: digitalmars.com D:27944 The "Xref" in the last line of the header has the magic number. So the link to this example message is: D/27944 Just in case you wanted to know... -- Justin http://jcc_7.tripod.com/d/
Apr 20 2004
parent reply "Matthew" <matthew.hat stlsoft.dot.org> writes:
Cool. Thanks for the info.

Lars/Achilleas/JC/anyone, does someone want to see if they can prize an opinion
out of Walter on this one?

"J C Calvarese" <jcc7 cox.net> wrote in message
news:c64ilb$1jhh$1 digitaldaemon.com...
 Matthew wrote:
 I am still of the same opinion as when Lars suggested this several months
 ago. (My post was "Re: D standard library group proposal", on the 4th Feb
 2004. I don't know how to use the web interface to search, I'm afraid.)

I found your message: D/23166. I think your points are as valid now as when you wrote it. By the way, if you find the message in your newsreader, you can probably "View Source" or "View Headers" to find out the message number in 2 seconds. For example: Path: digitalmars.com!not-for-mail ... X-Complaints-To: usenet digitalmars.com NNTP-Posting-Date: Tue, 20 Apr 2004 20:31:42 +0000 (UTC) X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 6.00.2800.1409 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1409 Xref: digitalmars.com D:27944 The "Xref" in the last line of the header has the magic number. So the link to this example message is: D/27944 Just in case you wanted to know... -- Justin http://jcc_7.tripod.com/d/

Apr 21 2004
parent reply J C Calvarese <jcc7 cox.net> writes:
Matthew wrote:
 Cool. Thanks for the info.
 
 Lars/Achilleas/JC/anyone, does someone want to see if they can prize an opinion
 out of Walter on this one?

On the other hand, perhaps Walter will continue to neglect the library until such time that "child protective services" places Phobos in a foster home. There seems to be an accumulating collection of Phobos bugs in with simple fixes that aren't corrected in later releases. Not that they're major things, but it's discouraging. Of course, if the upcoming DMD 0.83 fixes these I won't have to re-post bug reports.
 "J C Calvarese" <jcc7 cox.net> wrote in message
 news:c64ilb$1jhh$1 digitaldaemon.com...
 
Matthew wrote:

I am still of the same opinion as when Lars suggested this several months
ago. (My post was "Re: D standard library group proposal", on the 4th Feb
2004. I don't know how to use the web interface to search, I'm afraid.)

I found your message: D/23166. I think your points are as valid now as when you wrote it. By the way, if you find the message in your newsreader, you can probably "View Source" or "View Headers" to find out the message number in 2 seconds. For example: Path: digitalmars.com!not-for-mail ... X-Complaints-To: usenet digitalmars.com NNTP-Posting-Date: Tue, 20 Apr 2004 20:31:42 +0000 (UTC) X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 6.00.2800.1409 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1409 Xref: digitalmars.com D:27944 The "Xref" in the last line of the header has the magic number. So the link to this example message is: D/27944 Just in case you wanted to know... -- Justin http://jcc_7.tripod.com/d/


-- Justin http://jcc_7.tripod.com/d/
Apr 21 2004
parent reply "Matthew" <matthew.hat stlsoft.dot.org> writes:
I think it's getting to the point where something needs to be done. I've four or
five Phobos things I'm bugging Walter about, but at the same time I'm causing
him
lots of work with significant language/compiler changes.

Maybe it's time for Lars' DSLG? I'm certainly still willing to volunteer in the
review capacity that was discussed in Feb.

"J C Calvarese" <jcc7 cox.net> wrote in message
news:c6759k$30n9$1 digitaldaemon.com...
 Matthew wrote:
 Cool. Thanks for the info.

 Lars/Achilleas/JC/anyone, does someone want to see if they can prize an


 out of Walter on this one?

On the other hand, perhaps Walter will continue to neglect the library until such time that "child protective services" places Phobos in a foster home. There seems to be an accumulating collection of Phobos bugs in with simple fixes that aren't corrected in later releases. Not that they're major things, but it's discouraging. Of course, if the upcoming DMD 0.83 fixes these I won't have to re-post bug reports.
 "J C Calvarese" <jcc7 cox.net> wrote in message
 news:c64ilb$1jhh$1 digitaldaemon.com...

Matthew wrote:

I am still of the same opinion as when Lars suggested this several months
ago. (My post was "Re: D standard library group proposal", on the 4th Feb
2004. I don't know how to use the web interface to search, I'm afraid.)

I found your message: D/23166. I think your points are as valid now as when you wrote it. By the way, if you find the message in your newsreader, you can probably "View Source" or "View Headers" to find out the message number in 2 seconds. For example: Path: digitalmars.com!not-for-mail ... X-Complaints-To: usenet digitalmars.com NNTP-Posting-Date: Tue, 20 Apr 2004 20:31:42 +0000 (UTC) X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 6.00.2800.1409 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1409 Xref: digitalmars.com D:27944 The "Xref" in the last line of the header has the magic number. So the link to this example message is: D/27944 Just in case you wanted to know... -- Justin http://jcc_7.tripod.com/d/


-- Justin http://jcc_7.tripod.com/d/

Apr 21 2004
next sibling parent reply "Kris" <someidiot earthlink.dot.dot.dot.net> writes:
I'm not a fan of the "committee" approach, but I wholeheartedly agree that
something needs to be done (just to offload Walter if nothing else); a small
committee might be the answer.

Further, I don't see why this would be constrained to Phobos only ... for
example, I would imagine that some other independent projects might be
subject to the same scrutiny.

- Kris



"Matthew" <matthew.hat stlsoft.dot.org> wrote in message
news:c676j7$18n$1 digitaldaemon.com...
 I think it's getting to the point where something needs to be done. I've

 five Phobos things I'm bugging Walter about, but at the same time I'm

 lots of work with significant language/compiler changes.

 Maybe it's time for Lars' DSLG? I'm certainly still willing to volunteer

 review capacity that was discussed in Feb.

 "J C Calvarese" <jcc7 cox.net> wrote in message
 news:c6759k$30n9$1 digitaldaemon.com...
 Matthew wrote:
 Cool. Thanks for the info.

 Lars/Achilleas/JC/anyone, does someone want to see if they can prize



 opinion
 out of Walter on this one?

On the other hand, perhaps Walter will continue to neglect the library until such time that "child protective services" places Phobos in a foster home. There seems to be an accumulating collection of Phobos bugs in with simple fixes that aren't corrected in later releases. Not that they're major things, but it's discouraging. Of course, if the upcoming DMD 0.83 fixes these I won't have to re-post bug reports.
 "J C Calvarese" <jcc7 cox.net> wrote in message
 news:c64ilb$1jhh$1 digitaldaemon.com...

Matthew wrote:

I am still of the same opinion as when Lars suggested this several





ago. (My post was "Re: D standard library group proposal", on the 4th





2004. I don't know how to use the web interface to search, I'm





I found your message:




I think your points are as valid now as when you wrote it.


By the way, if you find the message in your newsreader, you can




"View Source" or "View Headers" to find out the message number in 2
seconds. For example:

Path: digitalmars.com!not-for-mail
...
X-Complaints-To: usenet digitalmars.com
NNTP-Posting-Date: Tue, 20 Apr 2004 20:31:42 +0000 (UTC)
X-Priority: 3
X-MSMail-Priority: Normal
X-Newsreader: Microsoft Outlook Express 6.00.2800.1409
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1409
Xref: digitalmars.com D:27944

The "Xref" in the last line of the header has the magic number. So the
link to this example message is:
D/27944

Just in case you wanted to know...

--
Justin
http://jcc_7.tripod.com/d/


-- Justin http://jcc_7.tripod.com/d/


Apr 21 2004
next sibling parent "Matthew" <matthew.hat stlsoft.dot.org> writes:
 I'm not a fan of the "committee" approach,

me either, but ...
 but I wholeheartedly agree that
 something needs to be done (just to offload Walter if nothing else); a small
 committee might be the answer.

... we're dying here at the moment. I think Lars group idea, with my modifications, will work well.
 Further, I don't see why this would be constrained to Phobos only ... for
 example, I would imagine that some other independent projects might be
 subject to the same scrutiny.

Aren't they independent?
 - Kris



 "Matthew" <matthew.hat stlsoft.dot.org> wrote in message
 news:c676j7$18n$1 digitaldaemon.com...
 I think it's getting to the point where something needs to be done. I've

 five Phobos things I'm bugging Walter about, but at the same time I'm

 lots of work with significant language/compiler changes.

 Maybe it's time for Lars' DSLG? I'm certainly still willing to volunteer

 review capacity that was discussed in Feb.

 "J C Calvarese" <jcc7 cox.net> wrote in message
 news:c6759k$30n9$1 digitaldaemon.com...
 Matthew wrote:
 Cool. Thanks for the info.

 Lars/Achilleas/JC/anyone, does someone want to see if they can prize



 opinion
 out of Walter on this one?

On the other hand, perhaps Walter will continue to neglect the library until such time that "child protective services" places Phobos in a foster home. There seems to be an accumulating collection of Phobos bugs in with simple fixes that aren't corrected in later releases. Not that they're major things, but it's discouraging. Of course, if the upcoming DMD 0.83 fixes these I won't have to re-post bug reports.
 "J C Calvarese" <jcc7 cox.net> wrote in message
 news:c64ilb$1jhh$1 digitaldaemon.com...

Matthew wrote:

I am still of the same opinion as when Lars suggested this several





ago. (My post was "Re: D standard library group proposal", on the 4th





2004. I don't know how to use the web interface to search, I'm





I found your message:




I think your points are as valid now as when you wrote it.


By the way, if you find the message in your newsreader, you can




"View Source" or "View Headers" to find out the message number in 2
seconds. For example:

Path: digitalmars.com!not-for-mail
...
X-Complaints-To: usenet digitalmars.com
NNTP-Posting-Date: Tue, 20 Apr 2004 20:31:42 +0000 (UTC)
X-Priority: 3
X-MSMail-Priority: Normal
X-Newsreader: Microsoft Outlook Express 6.00.2800.1409
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1409
Xref: digitalmars.com D:27944

The "Xref" in the last line of the header has the magic number. So the
link to this example message is:
D/27944

Just in case you wanted to know...

--
Justin
http://jcc_7.tripod.com/d/


-- Justin http://jcc_7.tripod.com/d/



Apr 21 2004
prev sibling next sibling parent J Anderson <REMOVEanderson badmama.com.au> writes:
Kris wrote:

I'm not a fan of the "committee" approach, but I wholeheartedly agree that
something needs to be done (just to offload Walter if nothing else); a small
committee might be the answer.
  

from the wider community and hand a few fixes into Walter at once. Could take a bit of time of Walters shoulders.
Further, I don't see why this would be constrained to Phobos only ... for
example, I would imagine that some other independent projects might be
subject to the same scrutiny.

- Kris
  

-- -Anderson: http://badmama.com.au/~anderson/
Apr 21 2004
prev sibling parent resistor mac.com writes:
I think D is in fact in need of several things:

1) A template library.  Hopefully DTL will cover this.

2) A LIBC-style library.  This is more or less what Phobos is, but I think it
needs some restructuring.

3) A class library like JFC.

I think it's #3 that needs the most work at the moment.  Now, I know there's
some complications in 
this:  namely, since D is not a runtime'd language we won't have lots of runtime
libs hanging around, 
which is why I think it should work like this:

A D class library committee will define the interfaces, contracts, invariants,
and unit-tests for a full set 
of APIs like JFC or .NET's core classes.  Then individuals may write libraries
that conform to this 
specification, and perhaps extend it.

Example:  This hypothetical class library might specify a socket-wrapper API.  I
might write a 
networking library that implements the core API, but is optimized for game
writing.  It might have 
additional support for other features, but it is required to implement the core
API.  This might include 
things like requiring me to expose a Socket class in net.sockets.socket, and
specify a list of functions it 
must implement.

I guess it boils down to this:  we need to define a module layout and minimum
API requirements for 
libraries.  Then library authors are free to innovate and explore new features
to their hearts' content, 
but a programmer is guaranteed a minimum level of functionality no matter which
library he uses.

Admittedly, there is no way to force someone to implement the code API, but I
think that if 
implementing the core API spec were made an advertising point for a library,
people would be quick to 
do so.

*phew*  That was rather long winded, but I think I got the point across.

Owen

In article <c677g5$2mi$1 digitaldaemon.com>, Kris says...
I'm not a fan of the "committee" approach, but I wholeheartedly agree that
something needs to be done (just to offload Walter if nothing else); a small
committee might be the answer.

Further, I don't see why this would be constrained to Phobos only ... for
example, I would imagine that some other independent projects might be
subject to the same scrutiny.

- Kris



"Matthew" <matthew.hat stlsoft.dot.org> wrote in message
news:c676j7$18n$1 digitaldaemon.com...
 I think it's getting to the point where something needs to be done. I've

 five Phobos things I'm bugging Walter about, but at the same time I'm

 lots of work with significant language/compiler changes.

 Maybe it's time for Lars' DSLG? I'm certainly still willing to volunteer

 review capacity that was discussed in Feb.

 "J C Calvarese" <jcc7 cox.net> wrote in message
 news:c6759k$30n9$1 digitaldaemon.com...
 Matthew wrote:
 Cool. Thanks for the info.

 Lars/Achilleas/JC/anyone, does someone want to see if they can prize



 opinion
 out of Walter on this one?

On the other hand, perhaps Walter will continue to neglect the library until such time that "child protective services" places Phobos in a foster home. There seems to be an accumulating collection of Phobos bugs in with simple fixes that aren't corrected in later releases. Not that they're major things, but it's discouraging. Of course, if the upcoming DMD 0.83 fixes these I won't have to re-post bug reports.
 "J C Calvarese" <jcc7 cox.net> wrote in message
 news:c64ilb$1jhh$1 digitaldaemon.com...

Matthew wrote:

I am still of the same opinion as when Lars suggested this several





ago. (My post was "Re: D standard library group proposal", on the 4th





2004. I don't know how to use the web interface to search, I'm





I found your message:




I think your points are as valid now as when you wrote it.


By the way, if you find the message in your newsreader, you can




"View Source" or "View Headers" to find out the message number in 2
seconds. For example:

Path: digitalmars.com!not-for-mail
...
X-Complaints-To: usenet digitalmars.com
NNTP-Posting-Date: Tue, 20 Apr 2004 20:31:42 +0000 (UTC)
X-Priority: 3
X-MSMail-Priority: Normal
X-Newsreader: Microsoft Outlook Express 6.00.2800.1409
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1409
Xref: digitalmars.com D:27944

The "Xref" in the last line of the header has the magic number. So the
link to this example message is:
D/27944

Just in case you wanted to know...

--
Justin
http://jcc_7.tripod.com/d/


-- Justin http://jcc_7.tripod.com/d/



Apr 21 2004
prev sibling parent Lars Ivar Igesund <larsivar igesund.net> writes:
Matthew wrote:
 I think it's getting to the point where something needs to be done. I've four
or
 five Phobos things I'm bugging Walter about, but at the same time I'm causing
him
 lots of work with significant language/compiler changes.
 
 Maybe it's time for Lars' DSLG? 

It certainly is (IMHO, that is).
 I'm certainly still willing to volunteer in the
 review capacity that was discussed in Feb.

I would volunteer to sit in the group (for at least a few minutes a week), although I no longer believe that I'm able to head it. Maybe I could be a member of the group proper, then retire to be a reviewer after many years of hard work? ;) Seriously, dsource.org is the place for this. Not only does Phobos need a foster home, it need to be placed in a version control repository. Walter is still the boss of the language, but as we now have two compilers available, we more than ever need some central place to keep common code distributed with both (all) of them. But for anything of this to 'work', Walter needs to say that he accepts that a group is in 'charge'. Walter will be able to review and veto, of course. Well, you get my drift. We need a group, somewhere to keep phobos, fix phobos, and get it released. Lars Ivar Igesund
Apr 22 2004
prev sibling parent Lars Ivar Igesund <larsivar igesund.net> writes:
Actually, my and Achilleas suggestions are quite different. Where I
wanted to make a phobos group, Achilleas wants to standardize "all"
APIs, whether in phobos or not.

Lars Ivar Igesund

Matthew wrote:
 I am still of the same opinion as when Lars suggested this several months
 ago. (My post was "Re: D standard library group proposal", on the 4th Feb
 2004. I don't know how to use the web interface to search, I'm afraid.)
 
 
 "Achilleas Margaritis" <axilmar b-online.gr> wrote in message
 news:c640sa$lsh$1 digitaldaemon.com...
 
It will be a real pity if D did not succeed as a programming language. In

order
for it to succeed, it needs a good collection of APIs which cover the most

usual
needs for writing applications.

We are a lot of people in here that are interested about D. I suggest we
organize ourselves and design the APIs and libraries which Digital Mars

will
approve as the official D APIs.

I am saying all this because many people have individually started to code
libraries, especially GUIs. It would be a waste of resources for each one

of us
to try individually, since APIs are a lot of work. Therefore, I suggest we

make
some sort of 'committee', by whoever is willing to participate, and

produce a
set of specifications for the APIs that should exist.

After we finish specifications, we can divide the work in chunks and

proceed to
the implementation either individually or in groups.

What does everybody think on this ?


Apr 21 2004
prev sibling parent reply Ant <Ant_member pathlink.com> writes:
In article <c640sa$lsh$1 digitaldaemon.com>, Achilleas Margaritis says...

[...]

 I suggest we 
organize ourselves and design the APIs and libraries which Digital Mars will 
approve as the official D APIs.

[...]
After we finish specifications, we can divide the work in chunks and proceed to 
the implementation either individually or in groups.

What does everybody think on this ?

It's the right way to do it. It's not Walter's way of doing it. Seems that Walter's and Matthew are now in (I instist) *secret* conversations to come up with some thing for the DTL. Can you change that? Ant
Apr 20 2004
next sibling parent "Matthew" <matthew.hat stlsoft.dot.org> writes:
"Ant" <Ant_member pathlink.com> wrote in message
news:c64279$o2e$1 digitaldaemon.com...
 In article <c640sa$lsh$1 digitaldaemon.com>, Achilleas Margaritis says...

 [...]

 I suggest we
organize ourselves and design the APIs and libraries which Digital Mars


approve as the official D APIs.

[...]
After we finish specifications, we can divide the work in chunks and


the implementation either individually or in groups.

What does everybody think on this ?

It's the right way to do it. It's not Walter's way of doing it. Seems that Walter's and Matthew are now in (I instist) *secret* conversations to come up with some thing for the DTL.

It's not "secret", just "work in progress". Would you have us post sometimes several compiler updates a day on the newsgroup, of no benefit to anyone else? AFAIK, Walter collaborates with others as well, as the need arises. As for DTL, I am hoping for heaps of feedback on all kinds of aspects, once I've got the compiler support for the basic mechanisms I need.
Apr 20 2004
prev sibling parent reply Achilleas Margaritis <axilmar b-online.gr> writes:
Well, if anybody cares about APIs, let this thread be the start of it.

I am a newbie in this forum (although I read the newsgroup for a long time
now), 
and I don't have the slightest idea on what DTL is.

Even if I did, who am I to change anything ? :-)

I am just fed up with programming languages that get in the way. D does not,
but it does not have the APIs. It's a shame to see Java win the programming 
language battle, since it is clearly inferior to D.

Of course, Sun has gazillions of programmers behind Java...

Ant wrote:
 In article <c640sa$lsh$1 digitaldaemon.com>, Achilleas Margaritis says...
 
 [...]
 
 
I suggest we 
organize ourselves and design the APIs and libraries which Digital Mars will 
approve as the official D APIs.

[...]
After we finish specifications, we can divide the work in chunks and proceed to 
the implementation either individually or in groups.

What does everybody think on this ?

It's the right way to do it. It's not Walter's way of doing it. Seems that Walter's and Matthew are now in (I instist) *secret* conversations to come up with some thing for the DTL. Can you change that? Ant

Apr 20 2004
next sibling parent reply Ant <duitoolkit yahoo.ca> writes:
On Wed, 21 Apr 2004 00:27:08 +0300, Achilleas Margaritis wrote:

 Well, if anybody cares about APIs, let this thread be the start of it.
 
 I am a newbie in this forum (although I read the newsgroup for a long time
now), 
 and I don't have the slightest idea on what DTL is.

I guess it stands for D template Library. I don't know if you are old enough to know about a language called C++. The idea comes from the C++ Standard Template Library. Ant PS I seem to remember your name. however say you're new to the group. Did you post before here?
Apr 20 2004
parent reply Achilleas Margaritis <Achilleas_member pathlink.com> writes:
Oh, it did not cross my mind that DTL stands for the D Template Library.

Yes, I am old enough to remember C++!!! Hey, I work with C++ the last 7 years!!!
:-)

I posted a few times before, yes.

It takes much more than a DTL, I am afraid. 

Yesterday, I did a presentation of D to my colleagues. They were impressed by
the language, but they all asked the same question:

"What can I do with it ?"

Come on guys, it's the APIs that count. We all know this...without APIs, D is
doomed to failure.

Somebody will say that C++ did not fail, even if they were no APIs available.
Well, that is half the truth. There are lots of C++ apps around, each one based
on a specific toolkit provided either by the vendor of the O/S (WIN32/MFC,
Motif) or by a third party (Qt, WxWindows, etc). 

When these companies needed to use C++ in order to write big applications, there
was nothing around so they invested and produced the toolkits mentioned above.
But now the situation is different. D comes at a time that Java/.NET reigns
supreme, and the C++ world is covered by the toolkits mentioned above. 

That is why, in my opinion, D needs a uniform set of cross-platform high-quality
APIs to begin with.

In article <pan.2004.04.20.22.57.02.785748 yahoo.ca>, Ant says...
On Wed, 21 Apr 2004 00:27:08 +0300, Achilleas Margaritis wrote:

 Well, if anybody cares about APIs, let this thread be the start of it.
 
 I am a newbie in this forum (although I read the newsgroup for a long time
now), 
 and I don't have the slightest idea on what DTL is.

I guess it stands for D template Library. I don't know if you are old enough to know about a language called C++. The idea comes from the C++ Standard Template Library. Ant PS I seem to remember your name. however say you're new to the group. Did you post before here?

Apr 21 2004
parent reply "Matthew" <matthew.hat stlsoft.dot.org> writes:
"Achilleas Margaritis" <Achilleas_member pathlink.com> wrote in message
news:c65fsp$2sm$1 digitaldaemon.com...
 Oh, it did not cross my mind that DTL stands for the D Template Library.

 Yes, I am old enough to remember C++!!! Hey, I work with C++ the last 7

 :-)

 I posted a few times before, yes.

 It takes much more than a DTL, I am afraid.

 Yesterday, I did a presentation of D to my colleagues. They were impressed by
 the language, but they all asked the same question:

 "What can I do with it ?"

 Come on guys, it's the APIs that count. We all know this...without APIs, D is
 doomed to failure.

 Somebody will say that C++ did not fail, even if they were no APIs available.
 Well, that is half the truth. There are lots of C++ apps around, each one based
 on a specific toolkit provided either by the vendor of the O/S (WIN32/MFC,
 Motif) or by a third party (Qt, WxWindows, etc).

 When these companies needed to use C++ in order to write big applications,

 was nothing around so they invested and produced the toolkits mentioned above.
 But now the situation is different. D comes at a time that Java/.NET reigns
 supreme, and the C++ world is covered by the toolkits mentioned above.

 That is why, in my opinion, D needs a uniform set of cross-platform

 APIs to begin with.

I don't imagine anyone would disagree with you. What libraries are you planning to write?
 In article <pan.2004.04.20.22.57.02.785748 yahoo.ca>, Ant says...
On Wed, 21 Apr 2004 00:27:08 +0300, Achilleas Margaritis wrote:

 Well, if anybody cares about APIs, let this thread be the start of it.

 I am a newbie in this forum (although I read the newsgroup for a long time



 and I don't have the slightest idea on what DTL is.

I guess it stands for D template Library. I don't know if you are old enough to know about a language called C++. The idea comes from the C++ Standard Template Library. Ant PS I seem to remember your name. however say you're new to the group. Did you post before here?


Apr 21 2004
parent reply Ant <Ant_member pathlink.com> writes:
In article <c65gso$4lv$1 digitaldaemon.com>, Matthew says...
"Achilleas Margaritis" <Achilleas_member pathlink.com> wrote in message
news:c65fsp$2sm$1 digitaldaemon.com...
 Oh, it did not cross my mind that DTL stands for the D Template Library.

 Yes, I am old enough to remember C++!!! Hey, I work with C++ the last 7

 :-)

 I posted a few times before, yes.

 It takes much more than a DTL, I am afraid.

 Yesterday, I did a presentation of D to my colleagues. They were impressed by
 the language, but they all asked the same question:

 "What can I do with it ?"

 Come on guys, it's the APIs that count. We all know this...without APIs, D is
 doomed to failure.

 Somebody will say that C++ did not fail, even if they were no APIs available.
 Well, that is half the truth. There are lots of C++ apps around, each one based
 on a specific toolkit provided either by the vendor of the O/S (WIN32/MFC,
 Motif) or by a third party (Qt, WxWindows, etc).

 When these companies needed to use C++ in order to write big applications,

 was nothing around so they invested and produced the toolkits mentioned above.
 But now the situation is different. D comes at a time that Java/.NET reigns
 supreme, and the C++ world is covered by the toolkits mentioned above.

 That is why, in my opinion, D needs a uniform set of cross-platform

 APIs to begin with.

I don't imagine anyone would disagree with you. What libraries are you planning to write?

Isn't that exactly what he is trying to avoid? First let's see what's needed and how it's should be implemented. Then the first that needs it would write it (for the platform s/he needs it). Ant
Apr 21 2004
parent reply Achilleas Margaritis <Achilleas_member pathlink.com> writes:
In article <c65nil$fei$1 digitaldaemon.com>, Ant says...
In article <c65gso$4lv$1 digitaldaemon.com>, Matthew says...
"Achilleas Margaritis" <Achilleas_member pathlink.com> wrote in message
news:c65fsp$2sm$1 digitaldaemon.com...
 Oh, it did not cross my mind that DTL stands for the D Template Library.

 Yes, I am old enough to remember C++!!! Hey, I work with C++ the last 7

 :-)

 I posted a few times before, yes.

 It takes much more than a DTL, I am afraid.

 Yesterday, I did a presentation of D to my colleagues. They were impressed by
 the language, but they all asked the same question:

 "What can I do with it ?"

 Come on guys, it's the APIs that count. We all know this...without APIs, D is
 doomed to failure.

 Somebody will say that C++ did not fail, even if they were no APIs available.
 Well, that is half the truth. There are lots of C++ apps around, each one based
 on a specific toolkit provided either by the vendor of the O/S (WIN32/MFC,
 Motif) or by a third party (Qt, WxWindows, etc).

 When these companies needed to use C++ in order to write big applications,

 was nothing around so they invested and produced the toolkits mentioned above.
 But now the situation is different. D comes at a time that Java/.NET reigns
 supreme, and the C++ world is covered by the toolkits mentioned above.

 That is why, in my opinion, D needs a uniform set of cross-platform

 APIs to begin with.

I don't imagine anyone would disagree with you. What libraries are you planning to write?

Isn't that exactly what he is trying to avoid? First let's see what's needed and how it's should be implemented. Then the first that needs it would write it (for the platform s/he needs it). Ant

Exactly. It's not what I plan to write, it is what it is needed to get D accepted by software houses. I think the following libs are needed: 1) collections 2) callbacks (signals and slots/events) 3) sockets 4) gui 5) database 6) filesystem 7) compression 8) xml The Java's APIs is a good start...Qt is also another good place to look for what core APIs need. Feel free to add anything you need.
Apr 22 2004
parent reply Ant <Ant_member pathlink.com> writes:
In article <c68570$1lht$1 digitaldaemon.com>, Achilleas Margaritis says...
Exactly. It's not what I plan to write, it is what it is needed to get D
accepted by software houses.

I think the following libs are needed:

1) collections
2) callbacks (signals and slots/events)
3) sockets
4) gui
5) database
6) filesystem
7) compression
8) xml

The Java's APIs is a good start...

oops... Walter doesn't like you anymore. He prefers the python libs.
Qt is also another good place to look for what
core APIs need.

There use to be a much more complete list on wiki, can anybody find it? Who was it from? Ant
Apr 22 2004
parent Achilleas Margaritis <axilmar b-online.gr> writes:
 oops... Walter doesn't like you  anymore.
 He prefers the python libs.

Hey, I don't like Java either. Does Walter like me now ? :-)
Apr 22 2004
prev sibling parent reply "Scott Egan" <scotte tpg.com.aux> writes:
How's this look for a start.  If we're going to start library development
then we should have an architecture.

Is there any way we can host some sort of model on wiki or dsource or dm?

We need to work out the way fwd first.
Apr 22 2004
parent Achilleas Margaritis <axilmar b-online.gr> writes:
Scott Egan wrote:

 How's this look for a start.  If we're going to start library development
 then we should have an architecture.
 
 Is there any way we can host some sort of model on wiki or dsource or dm?
 
 We need to work out the way fwd first.
 
 
 

I like it. At bottom layer, import libs of native apis. Above that, simple layer of objects. And above that, DFC (D Foundation Classes).
Apr 22 2004