gnupic: Thread: Trouble getting cvs-version to fly

Hi all
I have downloaded the latest release (0.9.8) and have been able to get
it to compile.  I am having problems getting the cvs version that I am
using to compile properly.  The error message is:
cod.c: In function `init_DirBlock':
cod.c:158 parse error before "VERSION"
I am sure that it is a minor problem, but the question is really this.
Why can't the cvs version support the entire code branch (lib, doc,
examples,....)?  In order to get it to compile to this point I really
had to manipulate subdirectories from the latest code release and the
cvs versions.
Thanks
Mike

Mike Crowe wrote:
> 
> Hi all
> I have downloaded the latest release (0.9.8) and have been able to get
> it to compile.  

Good.  

> I am having problems getting the cvs version that I am
> using to compile properly.  The error message is:
> cod.c: In function `init_DirBlock':
> cod.c:158 parse error before "VERSION"
> I am sure that it is a minor problem, 

Strange.  

What you are refering to as a problem, has to be a cvs checkout
problem.  Maybe config.h is not found?

> but the question is really this.
> Why can't the cvs version support the entire code branch (lib, doc,
> examples,....)?  

Not sure what you mean.  

I haven't committed any changes to cvs since the 0.9.8 release.  I don't
think Scott has either.  

I use the cvs version to generate the distributions.  They are
identical.  Sounds like cvs checkout problems.

> In order to get it to compile to this point I really
> had to manipulate subdirectories from the latest code release and the
> cvs versions.

Thats the problem.

You don't have to manipulate anything in cvs to get it to compile. 

The tree is defined by configure.in and the root Makefile.am.  If you
move things around without altering their content, the project will not
compile.

The cvs structure could be improved to make it a little more developer
friendly, but it is not broken.

Maybe the instructions on the home page are flawed.  I will review them
this weekend and find/fix the problem.  

Until then...

Sounds like you want to mess with the sources.  I recommend the
following:

tar -zxf gpasm-0.9.8.tar.gz
cp -R gpasm-0.9.8 gpasm-0.9.8.orig
"make your changes to gpasm-0.9.8"
diff -ur gpasm-0.9.8.orig gpasm-0.9.8 > mypatch.diff

Please let me know if you have any questions.

> Thanks
> Mike
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: ####@####.####
> For additional commands, e-mail: ####@####.####

On Fri, 20 Jul 2001, Craig Franklin wrote:

> I haven't committed any changes to cvs since the 0.9.8 release.  I don't
> think Scott has either.

No I haven't. Also, I checked out 0.9.8 and compiled just fine for me.
(Although I did not do a fresh checkout, I just did an update).


Incidently, 0.9.8 is givening me a "mismatched ENDIF" warning on the
"END" directive. 0.9.5 (which is the last version I used to assemble
this particular file)  did not. I'm was going to take a look at this
weekend,  unless you (Craig) know of something that may've changed in
this area. This particular program has nested if/endif's too.

Scott

0.9.7 enforced a new rule.  If the main source file ended during a IF
statement, before ENDIF is encountered, a warning is generated.  

Send me a copy of the file that genenerates the warning and I will look
into it.

Scott Dattalo wrote:
> 
> On Fri, 20 Jul 2001, Craig Franklin wrote:
> 
> > I haven't committed any changes to cvs since the 0.9.8 release.  I don't
> > think Scott has either.
> 
> No I haven't. Also, I checked out 0.9.8 and compiled just fine for me.
> (Although I did not do a fresh checkout, I just did an update).
> 
> Incidently, 0.9.8 is givening me a "mismatched ENDIF" warning on the
> "END" directive. 0.9.5 (which is the last version I used to assemble
> this particular file)  did not. I'm was going to take a look at this
> weekend,  unless you (Craig) know of something that may've changed in
> this area. This particular program has nested if/endif's too.
> 
> Scott

To simplify the checkout of gpasm, an alias in cvs has been
established.  The alias is "project".  The instructions for using this
alias are on the gpasm homepage.

http://gpasm.sourceforge.net/#Download

Craig Franklin wrote:
> 
> Mike Crowe wrote:
> >
> > Hi all
> > I have downloaded the latest release (0.9.8) and have been able to get
> > it to compile.
> 
> Good.
> 
> > I am having problems getting the cvs version that I am
> > using to compile properly.  The error message is:
> > cod.c: In function `init_DirBlock':
> > cod.c:158 parse error before "VERSION"
> > I am sure that it is a minor problem,
> 
> Strange.
> 
> What you are refering to as a problem, has to be a cvs checkout
> problem.  Maybe config.h is not found?
> 
> > but the question is really this.
> > Why can't the cvs version support the entire code branch (lib, doc,
> > examples,....)?
> 
> Not sure what you mean.
> 
> I haven't committed any changes to cvs since the 0.9.8 release.  I don't
> think Scott has either.
> 
> I use the cvs version to generate the distributions.  They are
> identical.  Sounds like cvs checkout problems.
> 
> > In order to get it to compile to this point I really
> > had to manipulate subdirectories from the latest code release and the
> > cvs versions.
> 
> Thats the problem.
> 
> You don't have to manipulate anything in cvs to get it to compile.
> 
> The tree is defined by configure.in and the root Makefile.am.  If you
> move things around without altering their content, the project will not
> compile.
> 
> The cvs structure could be improved to make it a little more developer
> friendly, but it is not broken.
> 
> Maybe the instructions on the home page are flawed.  I will review them
> this weekend and find/fix the problem.
> 
> Until then...
> 
> Sounds like you want to mess with the sources.  I recommend the
> following:
> 
> tar -zxf gpasm-0.9.8.tar.gz
> cp -R gpasm-0.9.8 gpasm-0.9.8.orig
> "make your changes to gpasm-0.9.8"
> diff -ur gpasm-0.9.8.orig gpasm-0.9.8 > mypatch.diff
> 
> Please let me know if you have any questions.
> 
> > Thanks
> > Mike
> >
> > ---------------------------------------------------------------------
> > To unsubscribe, e-mail: ####@####.####
> > For additional commands, e-mail: ####@####.####
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: ####@####.####
> For additional commands, e-mail: ####@####.####