www.digitalmars.com         C & C++   DMDScript  

c++ - Linux remarks

reply Mark Evans <Mark_member pathlink.com> writes:
I know the Linux subject has been raised before, but since DMC prides itself on
speed and compact executables, I felt it appropriate to post some comments from
another developer regarding the GNU C++ tools on Linux.  He is porting code from
MSVC6 to GCC.

So it appears that DMC could really help a lot of Linux folks if it were ever
ported.

Still thanks for such a wonderful compiler for Windows.  I am really gratified
by the progress being made on namespaces and templates.

Mark

================================================

Another issue, is the complete and utter sluggishness of g++. Holy CRAP!! those
of you who know me have heard me get on my rant as to why is VC6 so friggin
slow. Damn, VC6 is a friggin sprinter compared to the turtle that is g++.
Compiling ... takes around 40 minutes. It takes 15 on VC6. And the  output for
g++ is also nothing short of a incredible! The .so with the -ggdb flag turned on
(this is supposed to  put in gdb compliant debug information), weighs in at a
whopping 36 MB!!!  Thats right, not 3.6 MB, but 36MB, I had to double check the
size myself  before I beleived it. And running "strip" on it reduces it to a
"mere" 7MB in size.  The comparable data for VC6 (the applicationKit_d.dll,
map, idb, and pdb) weighs in at 17MB, and the size of the image itself  is 1.9.
Sigh....
Jan 02 2003
parent reply "Walter" <walter digitalmars.com> writes:
If I can convince more of the djgpp and cygnus g++-on-windows people to
switch to DMC++, then a good case can be made for the viability of a linux
version.

DMC++ is small and fast because its heritage is as a DOS compiler where
speed & size were incredibly important.

"Mark Evans" <Mark_member pathlink.com> wrote in message
news:av27oh$vde$1 digitaldaemon.com...
 I know the Linux subject has been raised before, but since DMC prides
itself on
 speed and compact executables, I felt it appropriate to post some comments
from
 another developer regarding the GNU C++ tools on Linux.  He is porting
code from
 MSVC6 to GCC.

 So it appears that DMC could really help a lot of Linux folks if it were
ever
 ported.

 Still thanks for such a wonderful compiler for Windows.  I am really
gratified
 by the progress being made on namespaces and templates.

 Mark

 ================================================

 Another issue, is the complete and utter sluggishness of g++. Holy CRAP!!
those
 of you who know me have heard me get on my rant as to why is VC6 so
friggin
 slow. Damn, VC6 is a friggin sprinter compared to the turtle that is g++.
 Compiling ... takes around 40 minutes. It takes 15 on VC6. And the  output
for
 g++ is also nothing short of a incredible! The .so with the -ggdb flag
turned on
 (this is supposed to  put in gdb compliant debug information), weighs in
at a
 whopping 36 MB!!!  Thats right, not 3.6 MB, but 36MB, I had to double
check the
 size myself  before I beleived it. And running "strip" on it reduces it to
a
 "mere" 7MB in size.  The comparable data for VC6 (the
applicationKit_d.dll,
 map, idb, and pdb) weighs in at 17MB, and the size of the image itself  is
1.9.
 Sigh....
Jan 02 2003
parent reply Mark Evans <Mark_member pathlink.com> writes:
You would be asking those people to re-write all their make scripts and such.
Not likely.  They're using Cygwin for a reason:  it lessens code maintenance
tasks.  It provides a standard Unix environment on Windows.

So you would be asking them to drop their beloved Unix environment.  The only
way around that would be a version of DMC which uses Cygwin headers and the
POSIX-compliant Cygwin.DLL.  That's effectively a Linux port anyway; might as
well do the real thing!

What migration exists today is mostly from Windows to Linux, not the other way.
So targeting Linux/GNU folks on Windows seems a non-starter.

The people to target are Windows folks migrating into cross-platform
Windows/Linux development -- much as Borland did with Delphi-cum-Kylix.

Right now, with G++ supported on both platforms, it's a rather obvious choice
for cross-platform work (flaws notwithstanding).  All DMC need do is offer an
alternative for cross-platform work.

I venture that 80% or more of the cross-platform groups in existence are using
MSVC6 for Windows support and G++ for Linux support.  This includes OpenOffice,
complete with their horrific build scripts.  The chances of those scripts being
rewritten for another Windows compiler are nil.

On the other hand, if they had a true cross-platform free compiler, they might
consider it.  Why they didn't use G++ in the first place is a mystery to me.  It
may revolve around various SDK requirements on the Windows side (they use
certain Microsoft SDKs and even require the Microsoft assembler for certain
things).

Note too that Mac OS X is just a tweaked BSD Unix, so DMC for Linux would work
on Macs as well.  That's two platforms for the price of one port.

Mark


Walter says...
If I can convince more of the djgpp and cygnus g++-on-windows people to
switch to DMC++, then a good case can be made for the viability of a linux
version.
Jan 02 2003
next sibling parent reply "Walter" <walter digitalmars.com> writes:
"Mark Evans" <Mark_member pathlink.com> wrote in message
news:av2in5$15p3$1 digitaldaemon.com...
 You would be asking those people to re-write all their make scripts and
such.
 Not likely.  They're using Cygwin for a reason:  it lessens code
maintenance
 tasks.  It provides a standard Unix environment on Windows.
While I'm sure those are a significant chuck of Cygwin users, I don't see that issue come up as the reason - I've never seen them say they feel trapped into using Cygwin because of the scripts.
 So you would be asking them to drop their beloved Unix environment.  The
only
 way around that would be a version of DMC which uses Cygwin headers and
the
 POSIX-compliant Cygwin.DLL.  That's effectively a Linux port anyway; might
as
 well do the real thing!
To me, that's Cygwin's greatest drawback. If you're going to write a professional quality product and ship it on Windows, it ought to be a Windows app from the ground up. I've found I've been able to tell a Cygwin built app just by running it. It has a linuxee feel to the user interface (even the command line ones), the file handling isn't quite compatible with windows file systems, the command line string handling isn't right, it doesn't handle windowy things like Ctrl-Break right, etc.
 Right now, with G++ supported on both platforms, it's a rather obvious
choice
 for cross-platform work (flaws notwithstanding).  All DMC need do is offer
an
 alternative for cross-platform work.
I've written professional products that work on both linux and Windows. I've never felt a need to use Cygwin to make that happen on the Windows side. But I can see the appeal of using the same compiler on both platforms.
 I venture that 80% or more of the cross-platform groups in existence are
using
 MSVC6 for Windows support and G++ for Linux support.  This includes
OpenOffice,
 complete with their horrific build scripts.  The chances of those scripts
being
 rewritten for another Windows compiler are nil.
<g> Sometimes I think I'm the only one who shuns complex build scripts. I've seen some that used bizarre combinations of Perl, bash, C, awk, what-have-you. Just to build the program, you had to find/load tons of other software (and of course the scripts were sensitive to particular versions of all that software, arggh). When I finally figured out what all those things were doing, it was easy to replace them with a straightforward makefile. Now, granted, DMC++'s runtime library has some execrable build scripts for the RTL, but I've been rewriting them as I incidentally need to do other work on them. Don't get me wrong, I'd love to make a full-up linux version of DMC++. The trouble is I am completely maxed out supporting 4 professional level compilers. I do the tech support too (and I'm very grateful when others here step in and help out!) Ask other compiler vendors how many engineers they assign to just one product <g>. Doing a linux port is not easy. It would have to be binary compatible with gcc's output, and would have to have much of the extensions added that gcc has.
Jan 02 2003
next sibling parent reply Mark Evans <Mark_member pathlink.com> writes:
While I'm sure those are a significant chuck of Cygwin users, I don't see
that issue come up as the reason - I've never seen them say they feel
trapped into using Cygwin because of the scripts.
I don't see why else they would be building their code under Cygwin except the cross-platform aspect and the scripts. Otherwise they'd just support two platforms natively, like OpenOffice -- which also uses Cygwin anyway, for the scripts!
To me, that's Cygwin's greatest drawback. If you're going to write a
professional quality product and ship it on Windows, it ought to be a
Windows app from the ground up.
Right, and DMC already does that. The Linux is what's missing.
I've written professional products that work on both linux and Windows. I've
never felt a need to use Cygwin to make that happen on the Windows side. But
I can see the appeal of using the same compiler on both platforms.
Yes it's a no-brainer really.
Don't get me wrong, I'd love to make a full-up linux version of DMC++. The
trouble is I am completely maxed out supporting 4 professional level
compilers. I do the tech support too (and I'm very grateful when others here
step in and help out!) Ask other compiler vendors how many engineers they
assign to just one product <g>.
As has been mentioned before, there are dozens of people ready willing and able to help under nondisclosure terms. So far you've nixed that idea every time it has been floated. And there is never any talk of just open-sourcing all this stuff which would be ideal for everyone involved.
Doing a linux port is not easy. It would have to be binary compatible with
gcc's output, and would have to have much of the extensions added that gcc
has.
You'd be amazed at what would happen if DMC were open-sourced! I know, I know, licensing issues....well, I don't see what interest Symantec has in keeping all this under wraps at this late date. Mark
Jan 02 2003
next sibling parent reply Christof Meerwald <cmeerw web.de> writes:
On Fri, 3 Jan 2003 05:30:39 +0000 (UTC), Mark Evans wrote:
 You'd be amazed at what would happen if DMC were open-sourced!
Hmm, the source-code for OpenWatcom has been released a few months ago and AFAIK no one has even touched the C++ compiler source code. Why should it be different for DMC? bye, Christof -- http://cmeerw.org JID: cmeerw jabber.at mailto cmeerw at web.de ...and what have you contributed to the Net?
Jan 03 2003
parent reply Mark Evans <Mark_member pathlink.com> writes:
Christof Mee wrote:
Hmm, the source-code for OpenWatcom has been released a few months ago and
AFAIK no one has even touched the C++ compiler source code.
Two or more is always greater than one, which is what DMC has today. The jump from one person to two is a huge quantum leap. It won't take 100 developers on DMC to improve its prospects as implied by your remark. (No offense intended Walter, we all know you're the greatest, but even Superman can't be in two places at once.)
Why should it be different for DMC?
OpenWatcom has been on my radar screen for years now. One reason I use DMC is that OpenWatcom has never done a real open source release! So your premise is flawed in that sense. They're entangled in licensing and copyright issues. They have not released all the source code AFAIK. You still need the old Watcom commercial stuff --- which is no longer for sale! What they have released to date are "patches" over top of the legacy Watcom commercial distro. An equivalent situation expressed in our terms would be that "OpenDMC" would only work for owners of original Symantec C++ software. So the size of the OpenWatcom development team at this point is limited to - people who own original Watcom tools before Watcom stopped selling - and are interested in the open source release - and have time to devote to the project - and have skills to write a C/C++ compiler which limitations thin the herd rather substantially. Believe me, if Walter open-sourced DMC for a Linux port, the crowd would gather. If at all feasible, a good time to do that might be after he's finished with the current round of improvements on namespaces and templates. In the Mythical Man-Month there is the idea of a system architect serving as a central switchboard for development teams. He doesn't do everything himself, but supervises to ensure quality, consistency, and other Good Things. This arrangement works well. It has kept Python going for over a decade (Guido is the switchboard). I think a similar model could work for an open-source release of DMC. Besides that would give Walter more time to work on his D project. Mark
Jan 03 2003
next sibling parent reply "Walter" <walter digitalmars.com> writes:
"Mark Evans" <Mark_member pathlink.com> wrote in message
news:av4maj$2b6p$1 digitaldaemon.com...
 Besides that would give Walter more time to work on his D project.
Trying to manage an open source project would very possibly result in less time for me <g>. And besides, management is not one of my core competencies :-(
Jan 03 2003
parent reply "Robert M. Münch" <robert.muench robertmuench.de> writes:
"Walter" <walter digitalmars.com> schrieb im Newsbeitrag
news:av4oho$2c9n$1 digitaldaemon.com...

 Trying to manage an open source project would very possibly result in less
 time for me <g>. And besides, management is not one of my core
competencies
 :-(
First I can see why Walter doesn't want to release DMC as open-source. And it's his decision. Anyway, IMO there is an other way to "solve" this problem that doesn't need an open-source release. Walter, I'm sure you know some people around here for some time now. Further I think there are quite a lot of good developers hanging around here. Have you every thought to use their offer to help you by just giving the source-code to a small team of poeple? With this you extend on the eye-balls and don't loose anything. I would think about this step. Very simple to setup and not to much to manage... Robert
Jan 04 2003
parent "Matthew Wilson" <dmd synesis.com.au> writes:
Seems a reasonable start. However, it would require lots of management, and
I can certainly see why Walter is not keen.

"Robert M. Münch" <robert.muench robertmuench.de> wrote in message
news:av6cf9$bqt$1 digitaldaemon.com...
 "Walter" <walter digitalmars.com> schrieb im Newsbeitrag
 news:av4oho$2c9n$1 digitaldaemon.com...

 Trying to manage an open source project would very possibly result in
less
 time for me <g>. And besides, management is not one of my core
competencies
 :-(
First I can see why Walter doesn't want to release DMC as open-source. And it's his decision. Anyway, IMO there is an other way to "solve" this
problem
 that doesn't need an open-source release.

 Walter, I'm sure you know some people around here for some time now.
Further
 I think there are quite a lot of good developers hanging around here. Have
 you every thought to use their offer to help you by just giving the
 source-code to a small team of poeple? With this you extend on the
eye-balls
 and don't loose anything.

 I would think about this step. Very simple to setup and not to much to
 manage... Robert
Jan 04 2003
prev sibling parent reply Christof Meerwald <cmeerw web.de> writes:
On Fri, 3 Jan 2003 18:55:15 +0000 (UTC), Mark Evans wrote:
 Christof Mee wrote:
Why should it be different for DMC?
OpenWatcom has been on my radar screen for years now. One reason I use DMC is that OpenWatcom has never done a real open source release! So your premise is flawed in that sense. They're entangled in licensing and copyright issues. They have not released all the source code AFAIK. You still need the old Watcom commercial stuff --- which is no longer for sale! What they have released to date are "patches" over top of the legacy Watcom commercial distro.
That's not correct. They have released a "patch" for Watcom 11.0 which contains the complete Watcom 11.0c (except 3rd party files like Microsoft's Platform SDK which you can download from MS). But they have also released the complete source code to the compiler and all tools (see http://www.openwatcom.org/ftp/openwatcom/ or http://www.openwatcom.org/support/perforce_content.html)
 So the size of the OpenWatcom development team at this point is limited to
 - people who own original Watcom tools before Watcom stopped selling
AFAIK, the freely available Watcom 11.0c "patch" + Microsoft's Platform SDK is all you need (no previous version of Watcom is required)
 - and are interested in the open source release
 - and have time to devote to the project
 - and have skills to write a C/C++ compiler
and these aren't any different for an Open Source DMC. bye, Christof -- http://cmeerw.org JID: cmeerw jabber.at mailto cmeerw at web.de ...and what have you contributed to the Net?
Jan 03 2003
parent reply Mark Evans <Mark_member pathlink.com> writes:
Christof I do not accept at face value your vague assertion about the numbers
over at OpenWatcom.  Those are fairly recent (last half year?) developments.
And have you asked anyone how many people are involved?  So far it's just an
unsupported assertion on your part.

There are counterexamples anyway.  G++ comes to mind.  Why would DMC not attract
as many folks as that?  You've given no basis for the factual assertion or
hypothetical projections stemming from it.

You and Walter should argue with each other, not me.  One of you says he wants
to minimize participation (to the extreme of just himself <g>), while the other
claims that an open source launch is justified only by massive numbers.  I'm in
a position where I can't win.  If the numbers are too great, open source is bad
(Walter); if too small, it's equally bad (Christof).

Walter it's always possible to do a closed-source port with 1-2 other people you
consider "manageable."

Or you could simply designate a manager of your choice for an open-source
approach.  (Jan maybe.)  That's what most company owners do with their various
projects.

Or you could do a two-phased approach, first closed-, then open-source.  During
the closed-source phase the emphasis would be on code style cleanup and
consistency.

With D going full steam, I don't see a Linux port happening under other
auspices.  That to me is very sad.  The personal consequence for my ambitious
cross-platform work is that I'll have to standardize on Comeau, G++, or Borland
C++.  All of them support both platforms with excellent C++ standards compliance
(esp. Comeau).  None of them are as fast or cool as Digital Mars or have
anything like this level of personal support.

Thanks for listening. You're both terrific for chasing down the namespace /
template / C++ standards business.  Maybe after that end of things is squared
away you'll entertain different feelings about an open- or semi-open-source port
to Linux.  I appreciate all your hard work very much.

Mark
Jan 03 2003
parent reply "Walter" <walter digitalmars.com> writes:
"Mark Evans" <Mark_member pathlink.com> wrote in message
news:av59l0$2l38$1 digitaldaemon.com...
 There are counterexamples anyway.  G++ comes to mind.  Why would DMC not
attract
 as many folks as that?  You've given no basis for the factual assertion or
 hypothetical projections stemming from it.
For one reason, DMC will not be released under GPL, if for no other reason than the complex web of licenses on it now. It took me a year to get all the licenses in order to launch Digital Mars. D, however, is open source & GPL because I took pains to write it all from scratch with that in mind.
 Walter it's always possible to do a closed-source port with 1-2 other
people you
 consider "manageable."
It's been tried, twice now. They both failed likely because of my management skills.
 With D going full steam, I don't see a Linux port happening under other
 auspices.  That to me is very sad.  The personal consequence for my
ambitious
 cross-platform work is that I'll have to standardize on Comeau, G++, or
Borland
 C++.  All of them support both platforms with excellent C++ standards
compliance
 (esp. Comeau).  None of them are as fast or cool as Digital Mars or have
 anything like this level of personal support.
I'm not sure why you need to standardize on a compiler for cross-platform work. What issues come up that require it?
Jan 03 2003
next sibling parent reply "Nic Tiger" <nictiger progtech.ru> writes:
According to my experience you should teach your co-workers before you can
give them some work to do.
Since I think none of us has experience in writing optimizing C/C++
compilers, it would be rather difficult for Walter to share tasks with us.

But on the other side, I really can help in developing/maintaining such
things as utilities (sc, smake and so on) and run-time library.

I think we could help Walter in doing such helper things, while he will
develop compiler himself. I think he will do this best alone than in a group
of co-workers who still have to be taught at least the basic things about
writing compilers.

And I don't think that opening the sources will cure the problem. I think
only Walter can understand them in reasonable time.

Nic Tiger.

P.S. I really want Linux port of DMC, mainly because of gcc's terrible
inline assembler. I use inline assembler a lot and don't want to learn some
arbitrary "universal" assembler which gcc suggests instead of well known
Intel assembler style. I want to port part of my software to Linux and want
to use my favorite compiler - DMC.


"Walter" <walter digitalmars.com> ÓÏÏÂÝÉÌ/ÓÏÏÂÝÉÌÁ ×  ÎÏ×ÏÓÔÑÈ ÓÌÅÄÕÀÝÅÅ:
news:av5bqd$2m6t$1 digitaldaemon.com...
 "Mark Evans" <Mark_member pathlink.com> wrote in message
 news:av59l0$2l38$1 digitaldaemon.com...
 There are counterexamples anyway.  G++ comes to mind.  Why would DMC not
attract
 as many folks as that?  You've given no basis for the factual assertion
or
 hypothetical projections stemming from it.
For one reason, DMC will not be released under GPL, if for no other reason than the complex web of licenses on it now. It took me a year to get all
the
 licenses in order to launch Digital Mars. D, however, is open source & GPL
 because I took pains to write it all from scratch with that in mind.

 Walter it's always possible to do a closed-source port with 1-2 other
people you
 consider "manageable."
It's been tried, twice now. They both failed likely because of my
management
 skills.

 With D going full steam, I don't see a Linux port happening under other
 auspices.  That to me is very sad.  The personal consequence for my
ambitious
 cross-platform work is that I'll have to standardize on Comeau, G++, or
Borland
 C++.  All of them support both platforms with excellent C++ standards
compliance
 (esp. Comeau).  None of them are as fast or cool as Digital Mars or have
 anything like this level of personal support.
I'm not sure why you need to standardize on a compiler for cross-platform work. What issues come up that require it?
Jan 04 2003
next sibling parent reply "Robert M. Münch" <robert.muench robertmuench.de> writes:
"Nic Tiger" <nictiger progtech.ru> schrieb im Newsbeitrag
news:av7791$qa0$1 digitaldaemon.com...

 According to my experience you should teach your co-workers before you can
 give them some work to do.
 Since I think none of us has experience in writing optimizing C/C++
 compilers, it would be rather difficult for Walter to share tasks with us.

 But on the other side, I really can help in developing/maintaining such
 things as utilities (sc, smake and so on) and run-time library.
Hi, exactly. Who said that the most hard parts should be out-sourced? IMO Walter should look at all the little time killers that need to be done. I'm sure a lot can be out-sourced to some people here. Walter, you could start with easy things and see how it works out.
 I think we could help Walter in doing such helper things, while he will
 develop compiler himself. I think he will do this best alone than in a
group
 of co-workers who still have to be taught at least the basic things about
 writing compilers.
Well, managing tools, runtime library etc. could be done. Building up a validation suite etc. would help too. I'm not sure if it's necessary that others need to go to bare metal of compiler writing... And Walter, if you think that your management skills are not appropriate to handle this. Why not let someone do it? Keep a single point of contact to this person, which is the switchboard to the rest of the people. I don't expect a 10+ team here. So management would rather be easy. I think my management skills are much better than my pure compiler writing skills. Robert
Jan 05 2003
parent reply "Walter" <walter digitalmars.com> writes:
"Robert M. Münch" <robert.muench robertmuench.de> wrote in message
news:av94n3$1rsp$1 digitaldaemon.com...
 "Nic Tiger" <nictiger progtech.ru> schrieb im Newsbeitrag
 news:av7791$qa0$1 digitaldaemon.com...

 According to my experience you should teach your co-workers before you
can
 give them some work to do.
 Since I think none of us has experience in writing optimizing C/C++
 compilers, it would be rather difficult for Walter to share tasks with
us.
 But on the other side, I really can help in developing/maintaining such
 things as utilities (sc, smake and so on) and run-time library.
Hi, exactly. Who said that the most hard parts should be out-sourced? IMO Walter should look at all the little time killers that need to be done.
I'm
 sure a lot can be out-sourced to some people here. Walter, you could start
 with easy things and see how it works out.

 I think we could help Walter in doing such helper things, while he will
 develop compiler himself. I think he will do this best alone than in a
group
 of co-workers who still have to be taught at least the basic things
about
 writing compilers.
Well, managing tools, runtime library etc. could be done. Building up a validation suite etc. would help too. I'm not sure if it's necessary that others need to go to bare metal of compiler writing... And Walter, if you think that your management skills are not appropriate
to
 handle this. Why not let someone do it? Keep a single point of contact to
 this person, which is the switchboard to the rest of the people. I don't
 expect a 10+ team here. So management would rather be easy. I think my
 management skills are much better than my pure compiler writing skills.
 Robert
I'd really like someone who was willing to manage a project to see DMD through to a linux version. It's a project worthy of anyone's management skills <g>. A good friend & colleague is working on the IDDE, but since he's unpaid, priority has to go to projects that pay the bills. I'm open to suggestion for anyone who wants to work on improving any of the rtl or ancilliary tools. A worthwhile project would be to convert all the library asm to using the compiler inline assembler as much as possible (most of it was written long before the compiler had a decent inline asm). Other things that will help a lot is just taking opportunities as they present themselves to spread the word about DMC/DMD, by posting about it, linking to it from web pages, asking people who release source libraries to support DMC, etc.
Jan 05 2003
next sibling parent reply "Nic Tiger" <nictiger progtech.ru> writes:
I can perform your task: convert ASM modules into inline-ASM in run-time
library.

But I'm curious what is the purpose?
To eliminate need of masm? Or what?
I doubt about converting c0 modules, it seems to me not possible.

Nic Tiger.

"Walter" <walter digitalmars.com> ÓÏÏÂÝÉÌ/ÓÏÏÂÝÉÌÁ × ÎÏ×ÏÓÔÑÈ ÓÌÅÄÕÀÝÅÅ:
news:avb524$2sku$1 digitaldaemon.com...
 "Robert M. Münch" <robert.muench robertmuench.de> wrote in message
 news:av94n3$1rsp$1 digitaldaemon.com...
 "Nic Tiger" <nictiger progtech.ru> schrieb im Newsbeitrag
 news:av7791$qa0$1 digitaldaemon.com...

 According to my experience you should teach your co-workers before you
can
 give them some work to do.
 Since I think none of us has experience in writing optimizing C/C++
 compilers, it would be rather difficult for Walter to share tasks with
us.
 But on the other side, I really can help in developing/maintaining
such
 things as utilities (sc, smake and so on) and run-time library.
Hi, exactly. Who said that the most hard parts should be out-sourced?
IMO
 Walter should look at all the little time killers that need to be done.
I'm
 sure a lot can be out-sourced to some people here. Walter, you could
start
 with easy things and see how it works out.

 I think we could help Walter in doing such helper things, while he
will
 develop compiler himself. I think he will do this best alone than in a
group
 of co-workers who still have to be taught at least the basic things
about
 writing compilers.
Well, managing tools, runtime library etc. could be done. Building up a validation suite etc. would help too. I'm not sure if it's necessary
that
 others need to go to bare metal of compiler writing...

 And Walter, if you think that your management skills are not appropriate
to
 handle this. Why not let someone do it? Keep a single point of contact
to
 this person, which is the switchboard to the rest of the people. I don't
 expect a 10+ team here. So management would rather be easy. I think my
 management skills are much better than my pure compiler writing skills.
 Robert
I'd really like someone who was willing to manage a project to see DMD through to a linux version. It's a project worthy of anyone's management skills <g>. A good friend & colleague is working on the IDDE, but since he's unpaid, priority has to go to projects that pay the bills. I'm open to suggestion for anyone who wants to work on improving any of the rtl or ancilliary tools. A worthwhile project would be to convert all the library asm to
using
 the compiler inline assembler as much as possible (most of it was written
 long before the compiler had a decent inline asm).

 Other things that will help a lot is just taking opportunities as they
 present themselves to spread the word about DMC/DMD, by posting about it,
 linking to it from web pages, asking people who release source libraries
to
 support DMC, etc.
Jan 06 2003
parent reply "Walter" <walter digitalmars.com> writes:
"Nic Tiger" <nictiger progtech.ru> wrote in message
news:avbtbb$9mt$1 digitaldaemon.com...
 I can perform your task: convert ASM modules into inline-ASM in run-time
 library.

 But I'm curious what is the purpose?
 To eliminate need of masm? Or what?
 I doubt about converting c0 modules, it seems to me not possible.
To eliminate dependency on masm. The trouble with masm is microsoft put out too many versions that have incompatible syntax. The end result is any random masm cannot be relied upon to produce a correct obj file. I have long since abandoned trying to be compatible with yet another iteration of masm. I use a particular masm that's over 10 years old, and never change it, so the results (bugs and all) are repeatable. That makes it hard for anyone else to rebuild the library, though. Of course, not all of the files can be converted, and it's debatable whether it is worth bothering with the 16 bit code.
Jan 07 2003
parent reply Christof Meerwald <cmeerw web.de> writes:
On Tue, 7 Jan 2003 18:45:57 -0800, Walter wrote:
 I use a particular masm that's over 10 years old, and never change it, so
 the results (bugs and all) are repeatable. That makes it hard for anyone
 else to rebuild the library, though.
I have written some documentation on how the runtime library can be built with the version of MASM that is included in the freely available Windows 98 DDK. It's available on http://cmeerw.org/prog/dm/rtl.html (including a quick&amp;dirty wrapper that converts between the masm386.exe and ml.exe command line syntax). bye, Christof -- http://cmeerw.org JID: cmeerw jabber.at mailto cmeerw at web.de ...and what have you contributed to the Net?
Jan 09 2003
parent "Walter" <walter digitalmars.com> writes:
"Christof Meerwald" <cmeerw web.de> wrote in message
news:avl4lf$2cfg$1 digitaldaemon.com...
 On Tue, 7 Jan 2003 18:45:57 -0800, Walter wrote:
 I use a particular masm that's over 10 years old, and never change it,
so
 the results (bugs and all) are repeatable. That makes it hard for anyone
 else to rebuild the library, though.
I have written some documentation on how the runtime library can be built with the version of MASM that is included in the freely available Windows
98
 DDK. It's available on http://cmeerw.org/prog/dm/rtl.html (including a
 quick&amp;dirty wrapper that converts between the masm386.exe and ml.exe
 command line syntax).
That's great! But I assume it is for only the 32 bit asm files?
Jan 10 2003
prev sibling next sibling parent reply Jan Knepper <jan smartsoft.us> writes:
 I'd really like someone who was willing to manage a project to see DMD
 through to a linux version. It's a project worthy of anyone's management
 skills <g>.
Well, I have put up the website for that at http://www.opend.org/ It there a link from http://www.digitalmars.com/ to it??? I have not checked in a long time to be honest as I have been burried with financial burden and really have to focus on making money and keep the server and newsgroups going.
 A good friend & colleague is working on the IDDE, but since he's unpaid,
 priority has to go to projects that pay the bills.
I recognize that... I have done quite some work on the IDDE code as well and do not think I am far away from having it decently fire up and actually getting it back in shape, but hey... The person working on in now has it name written all over the thing if I remember correctly.
 I'm open to suggestion for anyone who wants to work on improving any of the
 rtl or ancilliary
 tools. A worthwhile project would be to convert all the library asm to using
 the compiler inline assembler as much as possible (most of it was written long
 before the compiler had a decent inline asm).
That should be relatively easy to do... I think a good way to start with that would be to add all the RTL code to CVS as that enabled more than one person to work on the stuff at the same time. I have put the D front into CVS on opend.org, but have not seen many actually access the CVS. Of course I do not give unsupervised write permission... Same should be done for RTL and other ancilliary tools source. Also... I do have a version of IMPLIB that makes creating the system .LIB files from the system .DLL a little easier. It uses the compiler front end to parse the headerfiles and extract the function declarations. It uses an .INI file to define the stack sizes of the parameters and generates a .LIB with the proper n decoration where possible.
 Other things that will help a lot is just taking opportunities as they
 present themselves to spread the word about DMC/DMD, by posting about it,
 linking to it from web pages, asking people who release source libraries to
 support DMC, etc.
Have done that where ever I could I think... Support for it on other sites would be great though. For one I have patched Codejock's Xtreme Toolkit so that it compiles with DMC++. Jan
Jan 06 2003
next sibling parent reply Richard <fractal clark.net> writes:
In article <3E19A553.526FBC80 smartsoft.us>, Jan Knepper says...

Concerning open source - free project, I found the following /. an interesting
read:

http://interviews.slashdot.org/interviews/03/01/09/1216259.shtml?tid=156

Richard
Jan 10 2003
parent reply Jan Knepper <jan smartsoft.us> writes:
Richard,

There is all kinds of cute, interesting and also misleading reads about open
source and free projects.

I think it would be good to just go to http://www.sourceforce.org/ just see how
many project are out there and how many are actually financially balanced.
With financially balanced I mean that the people that work on the project get
more
than the satisfaction of working on the project, but actually are able to pay
their bills BECAUSE they work on the project.

I have spend and still spend a LOT of hours on open source and free project and
I
can be very honest in saying that so far it only has taken my time and cost me
money.

Currently the www.digitalmars.com is hosted on the internet server of my
company.
(One of my contributions to this project). Walter as well as I have seen a LOT
of
traffic, but I think Walter would agree that the financial balance has been
missing. I do not know how many CD sales Digital Mars does a month right now,
but
I know it is far from covering anything.

So tell me... When I look at the following financial picture:
    Mortage $1,200 / month.
    Phone $100 / month.
    Health Insurrance $442 / month
    Food $400 / month.
    Electric $ 200 / month

    TOTAL $2,342 / month

How open source - free project(s) that do not pay me a dime, better they costs
me
about an addition $950 a month! pay for my bills??? Answer is... They do not!

I always wonder why so many people seem to think that open source - free
project(s) are so great... I have not heard the pharmaceutical industry giving
their products away for free or even the formula's to what they developed.
Neither
have I heard that for the automotive industry or airplane industry, or
electronics
for that matter... So what am I missing here? Can somebody explain???

Thanks
Jan



Richard wrote:

 In article <3E19A553.526FBC80 smartsoft.us>, Jan Knepper says...

 Concerning open source - free project, I found the following /. an interesting
 read:

 http://interviews.slashdot.org/interviews/03/01/09/1216259.shtml?tid=156

 Richard
Jan 10 2003
parent reply "Matthew Wilson" <dmd synesis.com.au> writes:
I don't think you've missed anything, Jan. Those of us who have to pay our
own bills - through consultancy, contracting, journalism, whatever - do it
for intellectual stimulation and a kind of technical egalitarian spirit.
There's certainly no money in it.

My own motivations for doing STLSoft are
 - it's a lot easier to persuade clients to use STLSoft code (partly since
it's strictly header-only) than extracts from my company's public-domain
libraries, with all it's dependencies on .cpp and libs, etc.
 - it's a great source of journalistic information, witness my ongoing
blatherings in WDM and CUJ
 - it's a great way to refactor code that has real quality, but that is
obscured inside non-portable, ugly, proprietary libs/headers/cpp
 - it's a great learning experience. My understanding of the principles of
STL, the rules of C++, and the behavioural quirks of various compilers has
tripled at least

As for cash? Hmmm. No cash.

I used to very much enjoy Ron Burk's (creator and former editor of WDJ)
comments on the subject. He thought that people who gave away their IP were
nuts!

So we (some of us at least) are nuts, but at least we know we're nuts. :)

Matthew

BTW, you guys in the states pay an enormous amount in health insurance,
don't you?


"Jan Knepper" <jan smartsoft.us> wrote in message
news:3E1EF7D6.6B9A67E9 smartsoft.us...
 Richard,

 There is all kinds of cute, interesting and also misleading reads about
open
 source and free projects.

 I think it would be good to just go to http://www.sourceforce.org/ just
see how
 many project are out there and how many are actually financially balanced.
 With financially balanced I mean that the people that work on the project
get more
 than the satisfaction of working on the project, but actually are able to
pay
 their bills BECAUSE they work on the project.

 I have spend and still spend a LOT of hours on open source and free
project and I
 can be very honest in saying that so far it only has taken my time and
cost me
 money.

 Currently the www.digitalmars.com is hosted on the internet server of my
company.
 (One of my contributions to this project). Walter as well as I have seen a
LOT of
 traffic, but I think Walter would agree that the financial balance has
been
 missing. I do not know how many CD sales Digital Mars does a month right
now, but
 I know it is far from covering anything.

 So tell me... When I look at the following financial picture:
     Mortage $1,200 / month.
     Phone $100 / month.
     Health Insurrance $442 / month
     Food $400 / month.
     Electric $ 200 / month

     TOTAL $2,342 / month

 How open source - free project(s) that do not pay me a dime, better they
costs me
 about an addition $950 a month! pay for my bills??? Answer is... They do
not!
 I always wonder why so many people seem to think that open source - free
 project(s) are so great... I have not heard the pharmaceutical industry
giving
 their products away for free or even the formula's to what they developed.
Neither
 have I heard that for the automotive industry or airplane industry, or
electronics
 for that matter... So what am I missing here? Can somebody explain???

 Thanks
 Jan



 Richard wrote:

 In article <3E19A553.526FBC80 smartsoft.us>, Jan Knepper says...

 Concerning open source - free project, I found the following /. an
interesting
 read:

 http://interviews.slashdot.org/interviews/03/01/09/1216259.shtml?tid=156

 Richard
Jan 10 2003
next sibling parent reply Jan Knepper <jan smartsoft.us> writes:
Matthew Wilson wrote:

 I don't think you've missed anything, Jan. Those of us who have to pay our
 own bills - through consultancy, contracting, journalism, whatever - do it
 for intellectual stimulation and a kind of technical egalitarian spirit.
 There's certainly no money in it.
That's what I was afraid of...
 My own motivations for doing STLSoft are
  - it's a lot easier to persuade clients to use STLSoft code (partly since
 it's strictly header-only) than extracts from my company's public-domain
 libraries, with all it's dependencies on .cpp and libs, etc.
  - it's a great source of journalistic information, witness my ongoing
 blatherings in WDM and CUJ
  - it's a great way to refactor code that has real quality, but that is
 obscured inside non-portable, ugly, proprietary libs/headers/cpp
  - it's a great learning experience. My understanding of the principles of
 STL, the rules of C++, and the behavioural quirks of various compilers has
 tripled at least
Well, I have supported MFC for DMC++ for years I also used to have SGI STL available, but have no time for it anymore... At least not at this moment.
 As for cash? Hmmm. No cash.
Nop! Nothing EVER!!!
 I used to very much enjoy Ron Burk's (creator and former editor of WDJ)
 comments on the subject. He thought that people who gave away their IP were
 nuts!
They are... I only give 'part' of it away... <g>
 So we (some of us at least) are nuts, but at least we know we're nuts. :)
Yup...
 BTW, you guys in the states pay an enormous amount in health insurance,
 don't you?
Well, that actually for two people and about the best plan we can get... Jan
Jan 10 2003
parent "Matthew Wilson" <dmd synesis.com.au> writes:
Yes, Jan. I think you've hit on the key point: only give part of it away. In
a sense that's the strategy I take: I'm never going to put the technologies
that clients pay me serious money for out into the public domain.

It's kind of like the classic loss-leader marketing strategy: one's
open-source project gives people (hopefully) a confidence in the quality of
one's work, and then they may pursue that for specific paying tasks.

Hopefully, it's not as starkly cynical than that, in that the
free/open-source stuff is worthwhile on its own. If it's not, then you have
to wonder which software mega-corporation is lurking quietly behind the
venture (a la ActiveState).


"Jan Knepper" <jan smartsoft.us> wrote in message
news:3E1F2EE7.1F933ED4 smartsoft.us...
 Matthew Wilson wrote:

 I don't think you've missed anything, Jan. Those of us who have to pay
our
 own bills - through consultancy, contracting, journalism, whatever - do
it
 for intellectual stimulation and a kind of technical egalitarian spirit.
 There's certainly no money in it.
That's what I was afraid of...
 My own motivations for doing STLSoft are
  - it's a lot easier to persuade clients to use STLSoft code (partly
since
 it's strictly header-only) than extracts from my company's public-domain
 libraries, with all it's dependencies on .cpp and libs, etc.
  - it's a great source of journalistic information, witness my ongoing
 blatherings in WDM and CUJ
  - it's a great way to refactor code that has real quality, but that is
 obscured inside non-portable, ugly, proprietary libs/headers/cpp
  - it's a great learning experience. My understanding of the principles
of
 STL, the rules of C++, and the behavioural quirks of various compilers
has
 tripled at least
Well, I have supported MFC for DMC++ for years I also used to have SGI STL available, but have no time for it anymore... At least not at this moment.
 As for cash? Hmmm. No cash.
Nop! Nothing EVER!!!
 I used to very much enjoy Ron Burk's (creator and former editor of WDJ)
 comments on the subject. He thought that people who gave away their IP
were
 nuts!
They are... I only give 'part' of it away... <g>
 So we (some of us at least) are nuts, but at least we know we're nuts.
:)
 Yup...

 BTW, you guys in the states pay an enormous amount in health insurance,
 don't you?
Well, that actually for two people and about the best plan we can get... Jan
Jan 10 2003
prev sibling parent reply Richard <fractal clark.net> writes:
 I always wonder why so many people seem to think that open source - free
 project(s) are so great... I have not heard the pharmaceutical industry
giving
 their products away for free or even the formula's to what they developed.
Neither
 have I heard that for the automotive industry or airplane industry, or
electronics
 for that matter... So what am I missing here? Can somebody explain???
Sorry about the /. post. When I read the article, I focused in on a comparison with Walter and the DM project. I was also interested in the working methods, and wisdom from an obviously competent person - I glossed over the previous discussions in this thread about the philosophical question of why one should devote time to something that did not provide a cost-of-living level financial return. I’m sure many who read this group feel sympathy for your current difficulties. I know that I do. But as for offering a compelling explanation for why one should offer one’s IP for free public consumption, its really a question that one either already has an answer for, or will have to develop internally. Last night, I was having a late night beer at a bar with some friends, and I happened to have met the bartender a short time in the past on vacation. She was a Scottish girl, very pretty really, and I noticed that she was reading Zen in the Art of Motorcycle Maintenance by Pirsig. I thought it very odd because 2 weeks prior I had suggested a friend of mine borrow one of my copies – I happen to have two – for his trip abroad. He really is one of my best friends, and he seemed a little lost recently. After I gave a copy to him, I thought it might be a good idea to read it again. Coincidence and loose ramblings aside, what the answer to your question boils down to is that at some fundamental level, people are compelled to do things that in the absence of similar understanding by the observer are inexplicable. Most of the theory and early implementation of all the things we are using to have this conversation were evolved by people who did not, and do not benefit from them financially. I don’t think that means a person should avoid creating new things because precedent suggests that they don’t help with life transactions. Well, I’m hopeful that you will solve your current life equation. Best wishes for the rest of the year. Richard
Jan 10 2003
parent reply Mark Evans <Mark_member pathlink.com> writes:
Please start another thread if you want to debate open vs. closed source.

Let's all thank Linus Torvalds for giving away Linux to the masses and making
this discussion possible at all.

Mark
Jan 10 2003
next sibling parent Mark Evans <Mark_member pathlink.com> writes:
Getting back to the original theme of this discussion, a hard aspect of any
Linux port would be the debugger issue.  On Windows we use the Microsoft
CodeView format.  Debuggers on Linux expect something else.

The standard GNU debugger, GDB, expects the 'stabs' format, although Dwarf-II
has displaced stabs as the standard in GCC 3.1 and up.  The alternative WineDbg
can handle CodeView format on Linux, but whether that means only CodeView in a
pure Windows app running under WINE, or also native Linux apps, is unclear. I
suspect only WINE support.  Yet WineDBG is open source and could be 'tweaked.'
http://www.winehq.com/Docs/wine-devel/dbg-others.shtml
(see chart at bottom)

Which then would be the path of least resistance:

1. Leave DMC symbol generation alone, use non-standard WineDbg, possibly with
custom mods to handle native Linux apps.
2. Leave DMC symbols alone, write a converter from CodeView format to Dwarf-II,
and use standard GDB.
3. Change DMC itself to emit Dwarf-II on Linux, and use standard GDB.

Mark
Jan 11 2003
prev sibling parent reply Jan Knepper <jan smartsoft.us> writes:
Mark Evans wrote:

 Please start another thread if you want to debate open vs. closed source.

 Let's all thank Linus Torvalds for giving away Linux to the masses and making
 this discussion possible at all.
Well, thanks to him for sure, but also to BSD (Berkeley Software Distributions) which have been FREELY available for a very long time and probably a few others... http://www.freebsd.org/ http://www.openbsd.org/ http://www.netbsd.org/ Jan
Jan 11 2003
parent just passing thru <just_member pathlink.com> writes:
:o)



In article <3E203A35.88379ADE smartsoft.us>, Jan Knepper says...
Mark Evans wrote:

 Please start another thread if you want to debate open vs. closed source.

 Let's all thank Linus Torvalds for giving away Linux to the masses and making
 this discussion possible at all.
Well, thanks to him for sure, but also to BSD (Berkeley Software Distributions) which have been FREELY available for a very long time and probably a few others... http://www.freebsd.org/ http://www.openbsd.org/ http://www.netbsd.org/ Jan
Jan 17 2003
prev sibling parent "David Thomas" <dthomas cogitoinc.com> writes:
A first step to open source is to put it on SourceForge.net.  It's a lot
easier than trying to maintain similar capabilities on your own.

"Jan Knepper" <jan smartsoft.us> wrote in message
news:3E19A553.526FBC80 smartsoft.us...
 I'd really like someone who was willing to manage a project to see DMD
 through to a linux version. It's a project worthy of anyone's management
 skills <g>.
Well, I have put up the website for that at http://www.opend.org/ It there a link from http://www.digitalmars.com/ to it??? I have not
checked in
 a long time to be honest as I have been burried with financial burden and
really
 have to focus on making money and keep the server and newsgroups going.

 A good friend & colleague is working on the IDDE, but since he's unpaid,
 priority has to go to projects that pay the bills.
I recognize that... I have done quite some work on the IDDE code as well
and do
 not think I am far away from having it decently fire up and actually
getting it
 back in shape, but hey... The person working on in now has it name written
all
 over the thing if I remember correctly.

 I'm open to suggestion for anyone who wants to work on improving any of
the
 rtl or ancilliary
 tools. A worthwhile project would be to convert all the library asm to
using
 the compiler inline assembler as much as possible (most of it was
written long
 before the compiler had a decent inline asm).
That should be relatively easy to do... I think a good way to start with that would be to add all the RTL code to
CVS as
 that enabled more than one person to work on the stuff at the same time. I
have
 put the D front into CVS on opend.org, but have not seen many actually
access
 the CVS. Of course I do not give unsupervised write permission... Same
should be
 done for RTL and other ancilliary tools source.

 Also... I do have a version of IMPLIB that makes creating the system .LIB
files
 from the system .DLL a little easier. It uses the compiler front end to
parse
 the headerfiles and extract the function declarations. It uses an .INI
file to
 define the stack sizes of the parameters and generates a .LIB with the
proper n
 decoration where possible.

 Other things that will help a lot is just taking opportunities as they
 present themselves to spread the word about DMC/DMD, by posting about
it,
 linking to it from web pages, asking people who release source libraries
to
 support DMC, etc.
Have done that where ever I could I think... Support for it on other sites would be great though. For one I have patched Codejock's Xtreme Toolkit so that it compiles with
DMC++.
 Jan
Jan 29 2003
prev sibling next sibling parent reply Mark Evans <Mark_member pathlink.com> writes:
A good friend & colleague is working on the IDDE
Developing an IDE for Linux is wasted energy. There are good Linux IDEs already. It would be best to focus on the compiler itself, and let people integrate DMC with Eclipse, XEmacs, whatever. Besides, if you are going to port the IDE, you should port it to a cross-platform toolkit like wxWindows so you don't have to maintain two bodies of source code. There are plenty of Intel assemblers for Linux. NASM comes to mind. http://sourceforge.net/projects/nasm http://home.attbi.com/~fbkotler/ There are also cross-assemblers for PowerPC. A little digging should turn them up. http://www.penguinppc.org/ Mark
Jan 06 2003
next sibling parent reply Jan Knepper <jan smartsoft.us> writes:
Mark Evans wrote:

A good friend & colleague is working on the IDDE
Developing an IDE for Linux is wasted energy. There are good Linux IDEs already. It would be best to focus on the compiler itself, and let people integrate DMC with Eclipse, XEmacs, whatever.
I do not think the IDDE is going to be ported to Linux.. This is the original Win32 IDDE that needs some smacking around before it recompiles and actually loads again.
 Besides, if you are going to port the IDE, you should port it to a
 cross-platform toolkit like wxWindows so you don't have to maintain two bodies
 of source code.
wxWindows or Qt or... So far parts are written with just and only flat windows API. Other parts are written on MFC 3.20 (I think). It basically is a LOT of sources in different styles that are hooked together someway. Without too much documentation.... <g> I have spend quite some time on it, but have not gotten to the point where I could actually recompile it and use it. However, the person that currently works on it I think was on the original development team and should make better progress in less time than I did in basically foreign code as I call it. Jan
Jan 06 2003
parent Mark Evans <Mark_member pathlink.com> writes:
wxWindows or Qt or...
...Eclipse. The days of custom IDEs are numbered. The effort going into the Symantec IDE -- even on Windows -- would probably be better spent integrating with Eclipse. An Eclipse plugin suite would be portable to all platforms, Win/Linux/Mac. So you would kill three birds with one stone. Eclipse already has the C/C++ editing and project management stuff, necessary hooks for ISV debuggers, breakpoints in the C/C++ source, compilers, etc. Essentially it takes care of everything so that you can concentrate on core functionality. http://www.eclipse.org/articles/Article-API%20use/eclipse-api-usage-rules.html Based on your description of the Symantec IDE code I would say it's time for clean start anyway, even on Windows. Best regards, Mark
Jan 06 2003
prev sibling next sibling parent reply "Matthew Wilson" <dmd synesis.com.au> writes:
I double that: maybe the effort current being spent on the Win32 IDDE could
be better spent on integrating into existing commercial IDEs, or the growing
population of open-source IDEs?

"Mark Evans" <Mark_member pathlink.com> wrote in message
news:avcf5q$k9k$1 digitaldaemon.com...
A good friend & colleague is working on the IDDE
Developing an IDE for Linux is wasted energy. There are good Linux IDEs already. It would be best to focus on the compiler itself, and let people integrate DMC with Eclipse, XEmacs, whatever. Besides, if you are going to port the IDE, you should port it to a cross-platform toolkit like wxWindows so you don't have to maintain two
bodies
 of source code.

 There are plenty of Intel assemblers for Linux.  NASM comes to mind.
 http://sourceforge.net/projects/nasm
 http://home.attbi.com/~fbkotler/

 There are also cross-assemblers for PowerPC.  A little digging should turn
them
 up.
 http://www.penguinppc.org/

 Mark
Jan 06 2003
parent reply Jan Knepper <jan smartsoft.us> writes:
Matthew Wilson wrote:

 I double that: maybe the effort current being spent on the Win32 IDDE could
 be better spent on integrating into existing commercial IDEs, or the growing
 population of open-source IDEs?
Might be also a good thing to do... I know the Intel C++ compiler plugs right into Visual Studio's IDE. Obviously there must be some docco out there on how to do that. It might be a good thing for DMC++ and DMD for that matter to offer that somewhere in the near future. Jan
Jan 07 2003
next sibling parent reply Mark Evans <Mark_member pathlink.com> writes:
I know the Intel C++ compiler plugs right into Visual Studio's IDE.
Oh no, not that! Visual Studio is expensive and closed-source. Something open source and free would be the ticket. The best and most widely supported of these IDEs is Eclipse, which already handles the GNU C/C++ tools. If I'm buying MS tools anyway, I'll probably use MS compilers. Let alone the third- and second-world folks who can't afford MS tools. Mark
Jan 07 2003
parent reply "Walter" <walter digitalmars.com> writes:
"Mark Evans" <Mark_member pathlink.com> wrote in message
news:avf5oi$2375$1 digitaldaemon.com...

 Oh no, not that!  Visual Studio is expensive and closed-source.  Something
open
 source and free would be the ticket.  The best and most widely supported
of
 these IDEs is Eclipse, which already handles the GNU C/C++ tools.
 If I'm buying MS tools anyway, I'll probably use MS compilers.
 Let alone the third- and second-world folks who can't afford MS tools.
You're right. It's pointless to integrate with VS.
Jan 09 2003
parent Jan Knepper <jan smartsoft.us> writes:
 "Mark Evans" <Mark_member pathlink.com> wrote in message
 news:avf5oi$2375$1 digitaldaemon.com...

 Oh no, not that!  Visual Studio is expensive and closed-source.  Something
open
 source and free would be the ticket.  The best and most widely supported
of
 these IDEs is Eclipse, which already handles the GNU C/C++ tools.
 If I'm buying MS tools anyway, I'll probably use MS compilers.
 Let alone the third- and second-world folks who can't afford MS tools.
You're right. It's pointless to integrate with VS.
Pointless??? Jan
Jan 09 2003
prev sibling parent reply "Matthew Wilson" <dmd synesis.com.au> writes:
Intel & VS: It does. I use it. It's very pleasing to have a superb compiler
in a superb IDE. Visual C++'s compiler is so bad, it is such a juxtaposition
that it goes in the best IDE.
Docco: There isn't, but I know how to do it.
DMC++ & DMD (&other stuff as well): I am actually working on a project -
which I will be going back to as soon as I've finished the current manic
stint of writing - that will do the things you've mentioned, and a whole lot
more, and will not just work with VS's IDE (though that'll be the first
one). I'd rather not talk about it now (i) _really_ don't have time to get
into it, (ii) also not quite finished with the idea, but am planning to have
it done within the next couple of months (for one thing I have planned an
article on the subject which will have to be drafted before end of Feb), and
I have to go out and earn some money in March! So, if you guys can all be
patient, I might have a nice surprise in a few weeks.


"Jan Knepper" <jan smartsoft.us> wrote in message
news:3E1AF31F.C78847FE smartsoft.us...
 Matthew Wilson wrote:

 I double that: maybe the effort current being spent on the Win32 IDDE
could
 be better spent on integrating into existing commercial IDEs, or the
growing
 population of open-source IDEs?
Might be also a good thing to do... I know the Intel C++ compiler plugs right into Visual Studio's IDE. Obviously there must be some docco out there on how to do that. It might
be a
 good thing for DMC++ and DMD for that matter to offer that somewhere in
the near
 future.

 Jan
Jan 07 2003
next sibling parent "Matthew Wilson" <dmd synesis.com.au> writes:
btw, am also intending it to be cross-platform, but obviously starting with
Win32

"Matthew Wilson" <dmd synesis.com.au> wrote in message
news:avfaea$25k7$1 digitaldaemon.com...
 Intel & VS: It does. I use it. It's very pleasing to have a superb
compiler
 in a superb IDE. Visual C++'s compiler is so bad, it is such a
juxtaposition
 that it goes in the best IDE.
 Docco: There isn't, but I know how to do it.
 DMC++ & DMD (&other stuff as well): I am actually working on a project -
 which I will be going back to as soon as I've finished the current manic
 stint of writing - that will do the things you've mentioned, and a whole
lot
 more, and will not just work with VS's IDE (though that'll be the first
 one). I'd rather not talk about it now (i) _really_ don't have time to get
 into it, (ii) also not quite finished with the idea, but am planning to
have
 it done within the next couple of months (for one thing I have planned an
 article on the subject which will have to be drafted before end of Feb),
and
 I have to go out and earn some money in March! So, if you guys can all be
 patient, I might have a nice surprise in a few weeks.


 "Jan Knepper" <jan smartsoft.us> wrote in message
 news:3E1AF31F.C78847FE smartsoft.us...
 Matthew Wilson wrote:

 I double that: maybe the effort current being spent on the Win32 IDDE
could
 be better spent on integrating into existing commercial IDEs, or the
growing
 population of open-source IDEs?
Might be also a good thing to do... I know the Intel C++ compiler plugs right into Visual Studio's IDE. Obviously there must be some docco out there on how to do that. It might
be a
 good thing for DMC++ and DMD for that matter to offer that somewhere in
the near
 future.

 Jan
Jan 07 2003
prev sibling parent reply Mark Evans <Mark_member pathlink.com> writes:
Intel & VS: It does. I use it. It's very pleasing to have a superb compiler
in a superb IDE.
Oh yes, except for that $500-$1,000 price tag to get the MSVC IDE, plus $400 for the Intel compiler. I own MSVC6 and its IDE. Eclipse impresses me as being a lot more capable. Compared to MSVC6 the plugin support is awesome. MSVC6 just lets you swap out console tools. Eclipse lets you develop entire user interfaces for your plugins with the cross-platform SWT toolkit and Java. If you're supporting the MSVC IDE, by all means do so, that will leave everyone else free to work on better IDEs. Mark
Jan 07 2003
next sibling parent reply "Matthew Wilson" <dmd synesis.com.au> writes:
Whoa, Mark. Tone it down a bit.

I'm not suggesting that Visual Studio + Intel is the answer for everyone. I
am not even suggesting that it is the answer for me. I make use of seven
compilers, and treat them all (but one - guess which) as equally useful. Of
the six high-quality ones, Intel, Digital Mars and Metrowerks are head and
shoulders above the rest, as far as I am concerned.

In the development I was alluding to - of which I will be more forthcoming
when I have more time - a variety of environments will eventually be
targeted, including the leading free/open-source ones. However, I will start
with Visual Studio because it has a huge developer base, is a good IDE, and
is the IDE (note that I'm not saying compiler here) that I use and am most
familiar with.

I am neither proponent of, nor apologist for, Microsoft in general or Visual
Studio/Visual C++ in particular, and I kind of resent your reactionary
stance against something which is not representative of my opinions. Having
said that, however, I am not going to knock the product of a company -
however dubious/unpopular the commercial practises of that company may or
may not be - just because that is the vogue. Visual Studio's IDE is a good
one, certainly not the best, but of the widely used one's it surely stacks
up (Have you tried to use Borland's!?). Intel's compiler is a good one, not
the best, but one of them. It is churlish to suggest otherwise. They're not
for everyone, and no-one is suggesting that they are. (In fact I look
forward to checking out Eclipse in a couple of months.)

Let's try and keep this newsgroup in its current position: one of the few on
the web where aggressiveness, arrogance, pomposity and attitude are all left
at the door.

Matthew

"Mark Evans" <Mark_member pathlink.com> wrote in message
news:avfceh$26ib$1 digitaldaemon.com...
Intel & VS: It does. I use it. It's very pleasing to have a superb
compiler
in a superb IDE.
Oh yes, except for that $500-$1,000 price tag to get the MSVC IDE, plus
$400 for
 the Intel compiler.

 I own MSVC6 and its IDE.  Eclipse impresses me as being a lot more
capable.
 Compared to MSVC6 the plugin support is awesome.  MSVC6 just lets you swap
out
 console tools.  Eclipse lets you develop entire user interfaces for your
plugins
 with the cross-platform SWT toolkit and Java.

 If you're supporting the MSVC IDE, by all means do so, that will leave
everyone
 else free to work on better IDEs.

 Mark
Jan 07 2003
next sibling parent reply Jan Knepper <jan smartsoft.us> writes:
 Let's try and keep this newsgroup in its current position: one of the few on
 the web where aggressiveness, arrogance, pomposity and attitude are all left
 at the door.
Amen to that! Jan
Jan 07 2003
parent reply roland <--rv ronetech.com> writes:
Jan Knepper wrote:

Let's try and keep this newsgroup in its current position: one of the few on
the web where aggressiveness, arrogance, pomposity and attitude are all left
at the door.
Amen to that! Jan
yes. but just i must say i didn't feel _any_ aggressiveness, arrogance, pomposity and attitude on Mark's posts. anyway i second him: DM already has a correct IDE, there is no hurry for a new one, i agree, if the new one is open source, it's good, easing DMC's integration in M$'s IDE is not a priority as far as DM is concerned IMHO. roland
Jan 08 2003
parent reply Mark Evans <Mark_member pathlink.com> writes:
but just i must say i didn't feel _any_ aggressiveness, arrogance, 
pomposity and attitude on Mark's posts
Then you are more perceptive because there was none ;-). Everyone, please avoid reading strong emotions between the lines. I mean "reactionary" and all that -- come on guys, where do you get that stuff. And where did I accuse Matthew of being a Microsoft pointy-head. Sheesh. The things I put up with. What I tried to say, and reiterate here, is that MSVC IDE support is a good thing! I'm glad you're working on it because that work frees others to support superior IDEs. The more IDE support for DMC, the better. And you were given a specific technical reason for considering Eclipse superior to MSVC. MSVC lets you invoke different console tools in different sequences. That's all. Eclipse does that plus a whole lot more. It offers a vastly more elaborate plugin API complete with user interface development based on SWT and progammability based on Java. There are more reasons. An entire industry consortium supports Eclipse, including biggies like IBM, QNX, and high-profile names from the UML and XML worlds. Eclipse runs on half a dozen platforms. It offers syntax-highlighted editors with debug breakpoints that you can access from the plugin API. Heck I've used MSVC for years at various companies. My impression (confirmed by several conversations at these firms) is that people use it because it's all they know, not because they like it. The same is true of the MSVC6 compiler itself. It's just a default choice. Someone asked about Eclipse performance. All I can say is: quite reasonable and improving. The whole Eclipse project is focused on performance improvements. On my AMD 1.33 GHz system I have not noticed much difference from MSVC. Whatever price Eclipse pays in performance is gained in cross-platform support that MSVC never had, and never will. So by all means, do an MSVC IDE port. More power to you. For DMC and DMD however, I think Eclipse is a better alternative. Isn't life wonderful when we have choices. Regards, Mark Eclipse supported platforms to date --------------------------------------------------------- Stable Build: M4 (2.1 Stream) http://download.eclipse.org/downloads/drops/S-M4-200212181304/eclipse-news-M4.html http://download.eclipse.org/downloads/drops/S-M4-200212181304/index.php Windows 98/ME/2000/XP Mac OSX (Mac/Carbon) Linux RH 7.1/SuSE 7.1 (x86/Motif) Linux RH 7.1/SuSE 7.1 (x86/GTK 2) Solaris 8 (SPARC/Motif) QNX (x86/Photon) AIX (PPC/Motif) HP-UX (HP9000/Motif) (IRIX port is also started, check Eclipse newsgroups) Eclipse links --------------------------------------------------------- Plugin directory (there are others) http://eclipse-plugins.2y.net/eclipse/plugins.jsp CDT project (C/C++ support via GNU compiler and debugger) http://www.eclipse.org/cdt/ API usage http://www.eclipse.org/articles/Article-API%20use/eclipse-api-usage-rules.html
Jan 08 2003
parent reply "Matthew Wilson" <dmd synesis.com.au> writes:
Mark

I wasn't accusing you of being all of "aggressiveness, arrogance, pomposity
and attitude", just a subset thereof. ;)

Despite your protestations to the contrary, I can't get how your saying "I'm
glad you're working on it because that work frees others to support superior
IDEs" is anything other than rude. But I'm a big boy, and have certainly had
worse on newsgroups, so am happy to drop it; it's getting silly, and we're
inadvertantly conspiring to create the atmosphere I was wanting to avoid,
and which the Digital Mars newsgroup is renowned for not containing. I'm
sure your manner is not reflective of your attitude.

Not withstanding the unfortunate phrasing, you've ignored the fact that I
said Visual Studio was but _one_ of a set of tools I was targeting, so
you've not really listened. Nor have you addressed Jan's salient comments
about established user bases. Why deliberately ignore a very large (is it
the largest - I recall reading that somewhere. Of course if could be M$
propaganda ...) community who are likely more frustrated by the low quality
of the Visual C++ compiler than the somewhat less-than-optimal IDE? I cannot
believe that these people are going to respond better to being told to adopt
a new environment, where they have to learn a new UI with all that that
entails, than to being given a simple install and thereafter being able to
select one or more quality compilers instead of their lacklustre one at the
check of a box? I never have to touch the mouse in Visual Studio. How long
will it take before I can say the same in another IDE? (As far as I've been
able to try, it's nigh on impossible in C++ Builder, CodeWarrior and DM
IDDE. Conversely, switching between Intel's compiler and Visual C++ takes
less than a second, and it will be the same for DMC++.

You yourself said in another thread that

 "DMC++ has to be "ten times better" to dislodge people from using MSVC++"

Having made a tentative step, users may be encouraged to make further steps,
and eventually end up with a superior IDE to go with the superior
compiler(s) they now use. Seems to me a pragmatic position, reflecting human
(and corporate) nature and the reality of now.

Matthew

"Mark Evans" <Mark_member pathlink.com> wrote in message
news:avhv0b$isu$1 digitaldaemon.com...
but just i must say i didn't feel _any_ aggressiveness, arrogance,
pomposity and attitude on Mark's posts
Then you are more perceptive because there was none ;-). Everyone, please
avoid
 reading strong emotions between the lines.  I mean "reactionary" and all
that --
 come on guys, where do you get that stuff.  And where did I accuse Matthew
of
 being a Microsoft pointy-head.  Sheesh.  The things I put up with.

 What I tried to say, and reiterate here, is that MSVC IDE support is a
good
 thing!  I'm glad you're working on it because that work frees others to
support
 superior IDEs.  The more IDE support for DMC, the better.

 And you were given a specific technical reason for considering Eclipse
superior
 to MSVC.  MSVC lets you invoke different console tools in different
sequences.
 That's all.  Eclipse does that plus a whole lot more.  It offers a vastly
more
 elaborate plugin API complete with user interface development based on SWT
and
 progammability based on Java.

 There are more reasons. An entire industry consortium supports Eclipse,
 including biggies like IBM, QNX, and high-profile names from the UML and
XML
 worlds.  Eclipse runs on half a dozen platforms.  It offers
syntax-highlighted
 editors with debug breakpoints that you can access from the plugin API.
Heck


 I've used MSVC for years at various companies.  My impression (confirmed
by
 several conversations at these firms) is that people use it because it's
all
 they know, not because they like it.  The same is true of the MSVC6
compiler
 itself.  It's just a default choice.

 Someone asked about Eclipse performance.  All I can say is: quite
reasonable and
 improving.  The whole Eclipse project is focused on performance
improvements.
 On my AMD 1.33 GHz system I have not noticed much difference from MSVC.
 Whatever price Eclipse pays in performance is gained in cross-platform
support
 that MSVC never had, and never will.

 So by all means, do an MSVC IDE port.  More power to you.  For DMC and DMD
 however, I think Eclipse is a better alternative.  Isn't life wonderful
when we
 have choices.

 Regards,
 Mark


 Eclipse supported platforms to date
 ---------------------------------------------------------
 Stable Build: M4 (2.1 Stream)
http://download.eclipse.org/downloads/drops/S-M4-200212181304/eclipse-news-M 4.html
 http://download.eclipse.org/downloads/drops/S-M4-200212181304/index.php

 Windows 98/ME/2000/XP
 Mac OSX (Mac/Carbon)
 Linux RH 7.1/SuSE 7.1 (x86/Motif)
 Linux RH 7.1/SuSE 7.1 (x86/GTK 2)
 Solaris 8 (SPARC/Motif)
 QNX (x86/Photon)
 AIX (PPC/Motif)
 HP-UX (HP9000/Motif)
 (IRIX port is also started, check Eclipse newsgroups)


 Eclipse links
 ---------------------------------------------------------
 Plugin directory (there are others)
 http://eclipse-plugins.2y.net/eclipse/plugins.jsp
 CDT project (C/C++ support via GNU compiler and debugger)
 http://www.eclipse.org/cdt/
 API usage
http://www.eclipse.org/articles/Article-API%20use/eclipse-api-usage-rules.ht ml

Jan 08 2003
parent reply Mark Evans <Mark_member pathlink.com> writes:
I know how popular MSVC is, and I was not being rude or even trying.  You
misinterpreted and are going over the top with accusations.

As DMC and Eclipse improve, they are approaching that 10x mark.  Thank you
Walter Bright! :-)

Mark
Jan 08 2003
parent "Walter" <walter digitalmars.com> writes:
"Mark Evans" <Mark_member pathlink.com> wrote in message
news:avikeu$umv$1 digitaldaemon.com...
 As DMC and Eclipse improve, they are approaching that 10x mark.  Thank you
 Walter Bright! :-)
You're welcome! P.S. I agree that something has to be quite a bit better than established competition in order to convince people to switch, not just incrementally better. Right now I'm focussed on eliminating conformance problems in the compiler, this will go a long way towards removing barriers to converting to DMC++.
Jan 11 2003
prev sibling parent "Walter" <walter digitalmars.com> writes:
"Matthew Wilson" <dmd synesis.com.au> wrote in message
news:avfhlh$29un$1 digitaldaemon.com...
 Let's try and keep this newsgroup in its current position: one of the few
on
 the web where aggressiveness, arrogance, pomposity and attitude are all
left
 at the door.
It's true, the Digital Mars groups & posters have been great. In its 2 years of operation, I think only one message was yanked (because the poster had mistakenly posted copyrighted info). It's always been the case with DM and its previous incarnations (I ran the Zortech BBS for years). I like to think it is because DMC simply appeals to programmers with class.
Jan 11 2003
prev sibling next sibling parent reply Jan Knepper <jan smartsoft.us> writes:
Mark Evans wrote:

 Oh yes, except for that $500-$1,000 price tag to get the MSVC IDE, plus $400
for
 the Intel compiler.
Well, those are market prices and you would be surprized how many copies are being sold for those prices. Afterall, people that work on these products need to eat (and more than that). Whether the products are good or not... I happen to think that part of my current financial suffering is because people still COPY software that should have been bought. Some software has price tags... Is that abnormal?
 I own MSVC6 and its IDE.  Eclipse impresses me as being a lot more capable.
 Compared to MSVC6 the plugin support is awesome.  MSVC6 just lets you swap out
 console tools.  Eclipse lets you develop entire user interfaces for your
plugins
 with the cross-platform SWT toolkit and Java.

 If you're supporting the MSVC IDE, by all means do so, that will leave everyone
 else free to work on better IDEs.
I think it would be good to support the MSVC IDE and more... I think the great thing with supporting the MSVC IDE is that current MSVC user can use the DMC++ and DMD compilers without having to think too much... <g> I personally am not too fond of the MSVC IDE... Jan
Jan 07 2003
parent "Matthew Wilson" <dmd synesis.com.au> writes:
Amen to that! :)

"Jan Knepper" <jan smartsoft.us> wrote in message
news:3E1B4EB9.26167BD smartsoft.us...
 Mark Evans wrote:

 Oh yes, except for that $500-$1,000 price tag to get the MSVC IDE, plus
$400 for
 the Intel compiler.
Well, those are market prices and you would be surprized how many copies
are being
 sold for those prices. Afterall, people that work on these products need
to eat (and
 more than that). Whether the products are good or not...

 I happen to think that part of my current financial suffering is because
people
 still COPY software that should have been bought. Some software has price
tags... Is
 that abnormal?

 I own MSVC6 and its IDE.  Eclipse impresses me as being a lot more
capable.
 Compared to MSVC6 the plugin support is awesome.  MSVC6 just lets you
swap out
 console tools.  Eclipse lets you develop entire user interfaces for your
plugins
 with the cross-platform SWT toolkit and Java.

 If you're supporting the MSVC IDE, by all means do so, that will leave
everyone
 else free to work on better IDEs.
I think it would be good to support the MSVC IDE and more... I think the great thing with supporting the MSVC IDE is that current MSVC
user can
 use the DMC++ and DMD compilers without having to think too much... <g>
 I personally am not too fond of the MSVC IDE...

 Jan
Jan 07 2003
prev sibling parent "Robert M. Münch" <robert.muench robertmuench.de> writes:
"Mark Evans" <Mark_member pathlink.com> schrieb im Newsbeitrag
news:avfceh$26ib$1 digitaldaemon.com...

 Eclipse lets you develop entire user interfaces for your plugins
 with the cross-platform SWT toolkit and Java.
Just one question, as I don't like Java Apps at all. Isn't this thing horrible slow and fat? Robert
Jan 07 2003
prev sibling parent "Walter" <walter digitalmars.com> writes:
"Mark Evans" <Mark_member pathlink.com> wrote in message
news:avcf5q$k9k$1 digitaldaemon.com...
 Developing an IDE for Linux is wasted energy.  There are good Linux IDEs
 already.  It would be best to focus on the compiler itself, and let people
 integrate DMC with Eclipse, XEmacs, whatever.
You're right. Porting the IDE to Linux was never the plan, it is much too Windows-centric.
Jan 09 2003
prev sibling parent reply "Matthew Wilson" <dmd synesis.com.au> writes:
You're getting mentions in most of the articles I'm writing (and to be
published this year in CUJ & WDM), as well as some in a loftier venture that
is bubbling away at the moment. :)

You've also got a menu link on STLSoft's websites, which none of the other
compilers do yet. The only ones that will be getting the same are those that
take a similarly good-natured and mutually beneficial as do yourself, and I
can count them all on one hand. On three fingers in fact. Ironically,
they're the three best compilers. Go figure!



"Walter" <walter digitalmars.com> wrote in message
news:avb524$2sku$1 digitaldaemon.com...
 "Robert M. Münch" <robert.muench robertmuench.de> wrote in message
 news:av94n3$1rsp$1 digitaldaemon.com...
 "Nic Tiger" <nictiger progtech.ru> schrieb im Newsbeitrag
 news:av7791$qa0$1 digitaldaemon.com...

 According to my experience you should teach your co-workers before you
can
 give them some work to do.
 Since I think none of us has experience in writing optimizing C/C++
 compilers, it would be rather difficult for Walter to share tasks with
us.
 But on the other side, I really can help in developing/maintaining
such
 things as utilities (sc, smake and so on) and run-time library.
Hi, exactly. Who said that the most hard parts should be out-sourced?
IMO
 Walter should look at all the little time killers that need to be done.
I'm
 sure a lot can be out-sourced to some people here. Walter, you could
start
 with easy things and see how it works out.

 I think we could help Walter in doing such helper things, while he
will
 develop compiler himself. I think he will do this best alone than in a
group
 of co-workers who still have to be taught at least the basic things
about
 writing compilers.
Well, managing tools, runtime library etc. could be done. Building up a validation suite etc. would help too. I'm not sure if it's necessary
that
 others need to go to bare metal of compiler writing...

 And Walter, if you think that your management skills are not appropriate
to
 handle this. Why not let someone do it? Keep a single point of contact
to
 this person, which is the switchboard to the rest of the people. I don't
 expect a 10+ team here. So management would rather be easy. I think my
 management skills are much better than my pure compiler writing skills.
 Robert
I'd really like someone who was willing to manage a project to see DMD through to a linux version. It's a project worthy of anyone's management skills <g>. A good friend & colleague is working on the IDDE, but since he's unpaid, priority has to go to projects that pay the bills. I'm open to suggestion for anyone who wants to work on improving any of the rtl or ancilliary tools. A worthwhile project would be to convert all the library asm to
using
 the compiler inline assembler as much as possible (most of it was written
 long before the compiler had a decent inline asm).

 Other things that will help a lot is just taking opportunities as they
 present themselves to spread the word about DMC/DMD, by posting about it,
 linking to it from web pages, asking people who release source libraries
to
 support DMC, etc.
Jan 06 2003
parent "Walter" <walter digitalmars.com> writes:
"Matthew Wilson" <dmd synesis.com.au> wrote in message
news:avdt2a$1d40$1 digitaldaemon.com...
 You're getting mentions in most of the articles I'm writing (and to be
 published this year in CUJ & WDM), as well as some in a loftier venture
that
 is bubbling away at the moment. :)

 You've also got a menu link on STLSoft's websites, which none of the other
 compilers do yet. The only ones that will be getting the same are those
that
 take a similarly good-natured and mutually beneficial as do yourself, and
I
 can count them all on one hand. On three fingers in fact. Ironically,
 they're the three best compilers. Go figure!
And I thank you for that! And I'm not at all surprised that the best compilers are the ones that value their customers the most.
Jan 11 2003
prev sibling parent reply "Walter" <walter digitalmars.com> writes:
"Nic Tiger" <nictiger progtech.ru> wrote in message
news:av7791$qa0$1 digitaldaemon.com...
 According to my experience you should teach your co-workers before you can
 give them some work to do.
 Since I think none of us has experience in writing optimizing C/C++
 compilers, it would be rather difficult for Walter to share tasks with us.

 But on the other side, I really can help in developing/maintaining such
 things as utilities (sc, smake and so on) and run-time library.

 I think we could help Walter in doing such helper things, while he will
 develop compiler himself. I think he will do this best alone than in a
group
 of co-workers who still have to be taught at least the basic things about
 writing compilers.

 And I don't think that opening the sources will cure the problem. I think
 only Walter can understand them in reasonable time.
What has been an enormous help to me is Christof's (and others') help in getting STLport to work. It enables me to focus on the compiler rather than the morass that is STL. Boiling complex programs down to simple test cases is a bit of an art, and Christof is a master at it.
 P.S. I really want Linux port of DMC, mainly because of gcc's terrible
 inline assembler. I use inline assembler a lot and don't want to learn
some
 arbitrary "universal" assembler which gcc suggests instead of well known
 Intel assembler style. I want to port part of my software to Linux and
want
 to use my favorite compiler - DMC.
My brain hurts every time I try to use gcc's inline asm because the operands are backwards. Remember that old documentary about the guy who wore special goggles that turn the world upside down? After two weeks of that, his brain rewired itself so it was right side up again. Then, he took the goggles off. The documentary ended with a warning not to try that. That's what I feel like using gcc's asm <g>.
Jan 05 2003
parent reply "Nic Tiger" <nictiger progtech.ru> writes:
"Walter" <walter digitalmars.com> wrote in message
news:avb4f4$2s2r$1 digitaldaemon.com...
 "Nic Tiger" <nictiger progtech.ru> wrote in message
 news:av7791$qa0$1 digitaldaemon.com...
 According to my experience you should teach your co-workers before you
can
 give them some work to do.
 Since I think none of us has experience in writing optimizing C/C++
 compilers, it would be rather difficult for Walter to share tasks with
us.
 But on the other side, I really can help in developing/maintaining such
 things as utilities (sc, smake and so on) and run-time library.

 I think we could help Walter in doing such helper things, while he will
 develop compiler himself. I think he will do this best alone than in a
group
 of co-workers who still have to be taught at least the basic things
about
 writing compilers.

 And I don't think that opening the sources will cure the problem. I
think
 only Walter can understand them in reasonable time.
What has been an enormous help to me is Christof's (and others') help in getting STLport to work. It enables me to focus on the compiler rather
than
 the morass that is STL. Boiling complex programs down to simple test cases
 is a bit of an art, and Christof is a master at it.


 P.S. I really want Linux port of DMC, mainly because of gcc's terrible
 inline assembler. I use inline assembler a lot and don't want to learn
some
 arbitrary "universal" assembler which gcc suggests instead of well known
 Intel assembler style. I want to port part of my software to Linux and
want
 to use my favorite compiler - DMC.
My brain hurts every time I try to use gcc's inline asm because the
operands
 are backwards. Remember that old documentary about the guy who wore
special
 goggles that turn the world upside down? After two weeks of that, his
brain
 rewired itself so it was right side up again. Then, he took the goggles
off.
 The documentary ended with a warning not to try that.

 That's what I feel like using gcc's asm <g>.
My opinion is: when I program PDP-11 family processors, I certainly will use 'MOV src, dest', but when I program Intel family processors I want use 'MOV dest, src' and not opposite. Since assembler is not portable by definition, I don't want any generalized form of it, I rather make different files and write different asm for different machines (processors) than make myself think in terms of 'generalized assembler'. So, I'm waiting for normal Intel assembler under Linux, like I have in DMC for DOSX and Win32. Nic Tiger.
Jan 06 2003
parent reply "Walter" <walter digitalmars.com> writes:
"Nic Tiger" <nictiger progtech.ru> wrote in message
news:avbu0d$9v1$1 digitaldaemon.com...
 My opinion is: when I program PDP-11 family processors, I certainly will
use
 'MOV src, dest', but when I program Intel family processors I want use
'MOV
 dest, src' and not opposite.

 Since assembler is not portable by definition, I don't want any
generalized
 form of it, I rather make different files and write different asm for
 different machines (processors) than make myself think in terms of
 'generalized assembler'.

 So, I'm waiting for normal Intel assembler under Linux, like I have in DMC
 for DOSX and Win32.
I'm an old '11 asm programmer too. But the assembler should look like the cpu manufacturer's documentation. And it does with DMC. I think you're quite right in seeing no point to a generalized assembler form.
Jan 06 2003
parent Jan Knepper <jan smartsoft.us> writes:
 Since assembler is not portable by definition, I don't want any
generalized
 form of it, I rather make different files and write different asm for
 different machines (processors) than make myself think in terms of
 'generalized assembler'.

 So, I'm waiting for normal Intel assembler under Linux, like I have in DMC
 for DOSX and Win32.
I'm an old '11 asm programmer too. But the assembler should look like the cpu manufacturer's documentation. And it does with DMC. I think you're quite right in seeing no point to a generalized assembler form.
I could not agree more... Just for this reason C was invented if I remember correctly... Jan
Jan 07 2003
prev sibling parent Mark Evans <Mark_member pathlink.com> writes:
For one reason, DMC will not be released under GPL
How that buttresses Christof's assertions (about the OpenWatcom labor pool) escapes me. Besides, if Christof is right, then you are happy, because the numbers will be small.
It's been tried, twice now. They both failed likely because of my management
skills.
Then take the full-open route. Open-sourcing DMC requires nothing of you, not even CVS. You just tarball the source, put a license on it, and announce "here it is, no support, sorry." If someone wants to port to Linux, they will. If no one does, we are no worse off. Presumably the source code is commented. As for compiler expertise there is more than enough of that in the Linux world.
It took me a year to get all the
licenses in order to launch Digital Mars.
Including the IDE, debugger, Windows header files, and Microsoft issues. The Linux port needs none of that. Just a console C++ compiler. Much of that code you have written since the acquisition anyway. Oh well I give up now. The strange thing about this conversation is the 100% level of user interest in Linux coupled with a 100% bizarre opposition to all practical suggestions towards achieving it. It's a crying shame, because the opportunity is so ripe. Mark
Jan 04 2003
prev sibling parent reply "Walter" <walter digitalmars.com> writes:
"Mark Evans" <Mark_member pathlink.com> wrote in message
news:av375v$1gek$1 digitaldaemon.com...
 You'd be amazed at what would happen if DMC were open-sourced!
The D compiler is open source. But it's a lot less embarassing than the DMC source, which shows the effects of successive waves of rewrites in different styles by myself :-(
Jan 03 2003
parent Mark Evans <Mark_member pathlink.com> writes:
The D compiler is open source. But it's a lot less embarassing than the DMC
source, which shows the effects of successive waves of rewrites in different
styles by myself :-(
Let the open source world help you whip it into shape then. This type of issue is the big problem with the OpenWatcom effort. They are forever fiddling with the unreleased portions hoping to sanitize them of legal, stylistic, or technical problems. I wish they would just do the legal stuff and then launch the dang thing. Mark
Jan 03 2003
prev sibling parent reply "Rajiv Bhagwat" <dataflow vsnl.com> writes:
1. Have you taken a look at CMake www.cmake.org for a portable approach
towards simplifying complex build scripts?

This is slightly off-topic request, but has anyone thought about creating
DMC prj/opn files for CMake?
Walter, how about modifying the IDE to use 'text' version of .opn file
instead of the current binary one?

I have requested this one earlier too, so that we can put the opn files in
cvs. This way, any changes in the settings could also be tracked. Without
this, the use of IDE is severely limited and only makefiles have to be used.
(Also, the opn files were 'touched' when IDE was closed even if the settings
were not changed....I don't know if that has been corrected.)


2. If the 'binary compatibility to gcc' is left aside, will the porting be
quicker? Why is the binary compatibilty a limiting factor?

3. If we can live without gcc 'features' under Windows, surely we can under
Linux! DMC features are good enough.


"Walter" <walter digitalmars.com> wrote in message
news:av304o$1cn1$1 digitaldaemon.com...
 "Mark Evans" <Mark_member pathlink.com> wrote in message
 news:av2in5$15p3$1 digitaldaemon.com...
 You would be asking those people to re-write all their make scripts and
such.
 Not likely.  They're using Cygwin for a reason:  it lessens code
maintenance
 tasks.  It provides a standard Unix environment on Windows.
While I'm sure those are a significant chuck of Cygwin users, I don't see that issue come up as the reason - I've never seen them say they feel trapped into using Cygwin because of the scripts.
 So you would be asking them to drop their beloved Unix environment.  The
only
 way around that would be a version of DMC which uses Cygwin headers and
the
 POSIX-compliant Cygwin.DLL.  That's effectively a Linux port anyway;
might
 as
 well do the real thing!
To me, that's Cygwin's greatest drawback. If you're going to write a professional quality product and ship it on Windows, it ought to be a Windows app from the ground up. I've found I've been able to tell a Cygwin built app just by running it. It has a linuxee feel to the user interface (even the command line ones), the file handling isn't quite compatible
with
 windows file systems, the command line string handling isn't right, it
 doesn't handle windowy things like Ctrl-Break right, etc.

 Right now, with G++ supported on both platforms, it's a rather obvious
choice
 for cross-platform work (flaws notwithstanding).  All DMC need do is
offer
 an
 alternative for cross-platform work.
I've written professional products that work on both linux and Windows.
I've
 never felt a need to use Cygwin to make that happen on the Windows side.
But
 I can see the appeal of using the same compiler on both platforms.


 I venture that 80% or more of the cross-platform groups in existence are
using
 MSVC6 for Windows support and G++ for Linux support.  This includes
OpenOffice,
 complete with their horrific build scripts.  The chances of those
scripts
 being
 rewritten for another Windows compiler are nil.
<g> Sometimes I think I'm the only one who shuns complex build scripts.
I've
 seen some that used bizarre combinations of Perl, bash, C, awk,
 what-have-you. Just to build the program, you had to find/load tons of
other
 software (and of course the scripts were sensitive to particular versions
of
 all that software, arggh). When I finally figured out what all those
things
 were doing, it was easy to replace them with a straightforward makefile.
 Now, granted, DMC++'s runtime library has some execrable build scripts for
 the RTL, but I've been rewriting them as I incidentally need to do other
 work on them.

 Don't get me wrong, I'd love to make a full-up linux version of DMC++. The
 trouble is I am completely maxed out supporting 4 professional level
 compilers. I do the tech support too (and I'm very grateful when others
here
 step in and help out!) Ask other compiler vendors how many engineers they
 assign to just one product <g>.

 Doing a linux port is not easy. It would have to be binary compatible with
 gcc's output, and would have to have much of the extensions added that gcc
 has.
Jan 03 2003
parent "Walter" <walter digitalmars.com> writes:
"Rajiv Bhagwat" <dataflow vsnl.com> wrote in message
news:av49jk$24n6$1 digitaldaemon.com...
 Walter, how about modifying the IDE to use 'text' version of .opn file
 instead of the current binary one?
It's a good idea, but there's a long list of higher priority things.
 2. If the 'binary compatibility to gcc' is left aside, will the porting be
 quicker? Why is the binary compatibilty a limiting factor?
To take advantage of all the tons of existing libraries out there. Without binary compatibility, DMC for linux will be its own island.
 3. If we can live without gcc 'features' under Windows, surely we can
under
 Linux! DMC features are good enough.
Unfortunately, plenty of source code under linux uses those features.
Jan 03 2003
prev sibling parent reply Christof Meerwald <cmeerw web.de> writes:
On Thu, 2 Jan 2003 23:41:25 +0000 (UTC), Mark Evans wrote:
 So you would be asking them to drop their beloved Unix environment.  The only
 way around that would be a version of DMC which uses Cygwin headers and the
 POSIX-compliant Cygwin.DLL.
There is one problem with Cygwin - it's covered by the GPL. See http://cygwin.com/licensing.html: "By default, all executables link against this library (and in the process include GPL'd Cygwin glue code). This means that unless you modify the tools so that compiled executables do not make use of the Cygwin library, your compiled programs will also have to be free software distributed under the GPL with source code available to all." bye, Christof -- http://cmeerw.org JID: cmeerw jabber.at mailto cmeerw at web.de ...and what have you contributed to the Net?
Jan 03 2003
parent Mark Evans <Mark_member pathlink.com> writes:
There is one problem with Cygwin - it's covered by the GPL.
Christof you're right. I do not advocate that DMC pay much attention to Cygwin for the simple reason that DMC already supports Windows. What I advocate is a straightforward native Linux port (and Mac, to the extent that it can be easily covered by a Linux port, though it is a different chip). If something like Cygwin were desired, the route to take would be Mingw. It calls native Windows APIs and doesn't use the Cygwin DLL. Mark
Jan 03 2003