Re: Build process

On 2/24/07, Dave Raymer <[email protected]> wrote:
>
> On 2/23/07, Thiago Silva <[email protected]> wrote:
> >
> > On 2/23/07, [email protected] < [email protected]> wrote:
> > >
> > > I've been fighting this problem while working on getting the VM to
> > > build (and work) under linux -- I am going to get through the process
> > > by manually modifying the includeDB files; misery loves company.
> >
> > Why are you modifying the includeDB?
>
>
> Because its not correct.  GCC4.1.1 is complaining rabidly about incomplete
> definititions of structs that have had "pre-declared" via foward
> declarations, missing forward declarations in a number of cases, missing
> definitions of inlines, and "externs", the list goes on.  The most straight
> forward fix has been to simply add a few lines here and there to includeDB
> and includeDB2.

Ok. I didn't really came to experience what you did. Thanks for clarifying.

> > > I haven't been bitten by the primitive generator problem as of yet, but
> > > the code is not compiling yet.  GCC 4.1.1 is not happy with large
> > > amounts of the code and spending a good bit of time cleaning up things
> > > that the latest version of VS.NET lets slide right through.
> >
> > Have you completed the compilation?
>
>
> No, its mostly a sideline task at the moment.  I've got some research to
> complete before I have to have the VM, so I'm using it as a "get away"
> activity (when I need a break from reading papers or some other task);  I'm
> about half way through.

I'm happy to know about it.

> > I have a patch that makes gcc happy, but I can't be sure about its
> > correctness since my MingW/GCC build doesn't quite run the way I
> > expected.
>
>
> Your patch does not make GCC 4.1.1 happy, least wise not on Linux.
>

Oh, so you actually tried it :)

> > Also, imho, I think it might be specially difficult to get the VM
> > working on GNU/Linux with the current sources. Not only one has to
> > make a few non-trivial changes in the source in order to have it
> > compiling with GCC (such as some casts, and notably inline assembly
> > code, AFAIR) but also has to port the TASM assembly files and the
> > os.cpp so it uses the SO threads and so on. All this in one shot might
> > be too much to debug if things go wrong.
>
>
> > That's why I would suggest a compiler port (ie. make things work with
> > GCC first) before an OS port.
>
>
> Thiago, I appreciate your advice here, and I've been struggling to word this
> in some fashion that doesn't come off as arrogant and/or condescending.
> I've been doing this sort of development for 20 years now and my skills are
> up to the task, plus I've got you and the rest of the folks on the mailing
> list for questions if the need arises. Again, please don't take the above as
> "he thinks I'm an idiot" and throw your hands up declaring "Dave's a jerk" -
> The former couldn't be farther from the truth, and I do what I can to avoid
> generating unnecessary bad karma.
>

Don't worry, I don't think you sounded that bad :)
I'm just glad to see you are confident. On the other hand, I'm not
close to match your (and most of the subscribers on this list)
experience: 25 years old, here. 5 developing softwares. So, forgive me
if I say something naive. I'm trying my best to get my skills up to
the task.

--
Thiago Silva
Email: [email protected]
Jabber: [email protected]
Blog: http://sourcecraft.info/blog

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "Strongtalk-general" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to [email protected]
For more options, visit this group at http://groups.google.com/group/strongtalk-general?hl=en
-~----------~----~----~----~------~----~------~--~---