Great compiler, but IDE is so outdated, I have read that someone is willing to create the GDB patch to work with DMC, any news on that? Automatically, as you have GDB support you have all modern IDE's working with DMC, that would rock. I know you just want to make more money and keep everything under control not going any way open with your compier, but honestly you can make more money if DMC get's compatible with other stuff.

"Ray Van Der Roon" <Ray_member@pathlink.com> wrote in message news:ddku3e$jm$1@digitaldaemon.com...
> Great compiler, but IDE is so outdated, I have read that someone is
willing to
> create the GDB patch to work with DMC, any news on that? Automatically, as
you
> have GDB support you have all modern IDE's working with DMC, that would
rock.
> I know you just want to make more money and keep everything under control
not
> going any way open with your compier, but honestly you can make more money
if
> DMC get's compatible with other stuff.

To do that, the linker would have to be rewritten. Unfortunately, that's not too practical, as it's entirely in assembler.

Ray Van Der Roon wrote:
> Great compiler, but IDE is so outdated, I have read that someone is willing to
> create the GDB patch to work with DMC, any news on that? Automatically, as you
> have GDB support you have all modern IDE's working with DMC, that would rock.
> I know you just want to make more money and keep everything under control not
> going any way open with your compier, but honestly you can make more money if
> DMC get's compatible with other stuff.

Personal swipes at Walter are counter-productive. Really. To be sure, there are parts of DMC that are outdated, but that doesn't mean there aren't ways to work around the problem. If you know the history surrounding DMC, you'll immediately understand that a lot of the code that Walter has to deal with is copywritten by other people and cannot be released into a GPL bazaar (bizzare?) or BSD core style of collaborative development. FWIW, I still have a Symantec C/C++ v.7 CD package at home -- I've considered selling it on e-bay as "NIB" (new in box).

If you truly want to help with getting DMC to play nicely with GDB, it's a matter of hacking the code in libbfd to recognize OMF format objects. It looks like there was code to do that at one time or another, but today, it's gone. I can't say there's a wealth of information out there on CodeView's debugging info, but OMF records are pretty nicely documented. Also, DMC has to conform to MS's name mangling scheme, otherwise, DMC wouldn't work with a large cross-section of software compiled on Win32.

The good news, if this effort can be made to happen, is that the GNU binary utilities could be used to upconvert OMF objects to MS ELF/DWARF objects.

Unfortunately, USAF program office travel tends to hamper my development and hacking hours currently, but I'm happy to offer advice and perhaps collaborate on making this happen.


-scooter

"Scott Michel" <scottm@aero.org> wrote in message news:dg4h8m$25iv$1@digitaldaemon.com...
> If you truly want to help with getting DMC to play nicely with GDB, it's a matter of hacking the code in libbfd to recognize OMF format objects. It looks like there was code to do that at one time or another, but today, it's gone. I can't say there's a wealth of information out there on CodeView's debugging info, but OMF records are pretty nicely documented. Also, DMC has to conform to MS's name mangling scheme, otherwise, DMC wouldn't work with a large cross-section of software compiled on Win32.

The CodeView debug format generated by DMC is as described in the Microsoft document:

"CV4 Symbolic Debug Information Specification"
Authors: Ross Barmoe, David Gray, Steve Smith, Dan Spalding, Jim Schaad, et
al.
Public Release
Revision 3.1 March 5, 1993
Languages Business Unit

Unfortunately, nobody seems to have ever put this information up on the internet.