[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Anyone got a working copy of sp around?



"Adam P. Harris" <apharris@onshore.com> writes:

> BTW, I'm the Debian maintainer of jade.  I would much like to know why 
> jade + egcs do not get along; I've been in correspondance with James
> Clark (upstream maintainer) as well; I think he would be pleased to
> know what the situation is right now; I've notified him eg++ doesn't
> get along with jade.

Perhaps I'll bring this up on the egcs list.

> > It's specifically the build---apparently this is one slight bug that
> > egcs-1.0.3 gets us that 1.0.2 did not have.  It shows up on the i386
> > as well.
> 
> I can confirm that, although I cannot confirm that there was *ever* a
> version of eg++ that didn't raise this problem.
> 
> Nor am I sure it's a bug in jade or a bug in egcs.

I can't say for absolutely certain myself---I suspect egcs, though, as
I *think* the last working copy was compiled with egcs successfully. I
would guess using version 1.0.1.

> I assume you guys on Alpha have only eg++?  Is there any way you can
> build with g++ 272 vintage?

Well, the gcc maintainer just took the patches out of gcc that were
needed to get it to build on the alpha at all, because they were
conflicting with something for the intel.

So the answer would be something like, "no". :-)

> Yes, this probably holds for both jade and nsgmls (same source pkg).
> I don't recall seing too many situations where the eg++ bug tickles
> nsgmls though.

That's actually the problem we're having on the alpha.

> Assuming you all do not have eg++, but you do think that an older
> version of jade (i.e., pre 1.1) I have source snapshots available in
> CVS dating back to jade-1.0.1-1.

Could you give me the .diff.gz file for 1.1-5?  I've been trying to
build with a recent snapshot of egcs, and having no luck because I
don't remember what the changes were that made the link work with
egcs, and I've had *0* time to work on it.

Simply having a set of egcs-compatible diffs would work wonders.

> Possibly -fno-excpetions might help?  But that's pretty drastic and
> evil.

I'd rather not go there.

> I'm not on this mailing list so you'll have to let me know by email
> or what-have-you.  Please keep me up-to-date on how you all are able
> to work-around.

We'll let you know if we find anything.

Mike.


--
To UNSUBSCRIBE, email to debian-alpha-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org


Reply to: