Re: A patch for toplevel Makefile.in [and a gnatsweb problem]

Re: A patch for toplevel Makefile.in [and a gnatsweb problem]

[Phil Edwards]

Most of you don't know me; I'm one of the mumble[0] libstdc++ maintainers.
Benjamin Kosnik sent me a chunk of this

   http://sourceware.cygnus.com/ml/overseers/2000/msg00225.html

and said, "your turn in the barrel, monkey boy."[1]  So as crash pointed
out here,

   http://sourceware.cygnus.com/ml/overseers/2000/msg00163.html

I signed myself up and read through the archives for 2000.  I think I'm
somewhat up to speed now...



On a different subject, Jim Kingdon suggested that I send this gnatsweb
problem to overseers.  If gnatsweb is Not Your Problem[tm], the rest of
my babbling can be skipped.  If it is, then I would really appreciate any
assistance with this problem.  If overseers is not the appropriate place
for this kind of thing, then LART me and I will cheerfully accept it.

I've never used gnatsweb before.  What happens is that I can login okay,
and subsequent pages have

    User: pme
    Database: libstdc++
    Access: edit

on them, so that looks okay.  I can do an advanced query, leaving all
the fields blank to get a listing of all the PRs still open.  I can even
store queries.  But when I select any of the "number" links that will view
a PR, like

    http://sourceware.cygnus.com/cgi-bin/gnatsweb.pl?cmd=view&pr=33&database=libstdc++

or any of the "edit" links to try and change something, it sets a
"gnatsweb-test-cookie" and then I get kicked back to the generic login
screen.  And everything I try after that keeps kicking me back there.

FWIW, I'm using Netscrape 4.7 under NT[2].  Everything is enabled.


Phil
pme@sourceware.cygnus.com
devphil@users.sourceforge.net (don't shoot me)

[0]  At least five people, last I looked.
[1]  I'm paraphrasing; his actual words were along the lines of "would
     you please consider doing this?"
[2]  NT with speakers and a nice monitor makes a really sweet X terminal.

Re: A patch for toplevel Makefile.in [and a gnatsweb problem]

[Tom Tromey]

Phil> On a different subject, Jim Kingdon suggested that I send this
Phil> gnatsweb problem to overseers.  If gnatsweb is Not Your
Phil> Problem[tm], the rest of my babbling can be skipped.

I haven't seen or heard of your gnatsweb problem before.
I recommend contacting bug-gnats@gnu.org.
That is where the Gnats hackers live.

If they say that you should upgrade gnatsweb, report back to overseers.

Tom

Re: A patch for toplevel Makefile.in [and a gnatsweb problem]

[Phil Edwards]

Most of you don't know me; I'm one of the mumble[0] libstdc++ maintainers.
Benjamin Kosnik sent me a chunk of this

   http://sourceware.cygnus.com/ml/overseers/2000/msg00225.html

and said, "your turn in the barrel, monkey boy."[1]  So as crash pointed
out here,

   http://sourceware.cygnus.com/ml/overseers/2000/msg00163.html

I signed myself up and read through the archives for 2000.  I think I'm
somewhat up to speed now...



On a different subject, Jim Kingdon suggested that I send this gnatsweb
problem to overseers.  If gnatsweb is Not Your Problem[tm], the rest of
my babbling can be skipped.  If it is, then I would really appreciate any
assistance with this problem.  If overseers is not the appropriate place
for this kind of thing, then LART me and I will cheerfully accept it.

I've never used gnatsweb before.  What happens is that I can login okay,
and subsequent pages have

    User: pme
    Database: libstdc++
    Access: edit

on them, so that looks okay.  I can do an advanced query, leaving all
the fields blank to get a listing of all the PRs still open.  I can even
store queries.  But when I select any of the "number" links that will view
a PR, like

    http://sourceware.cygnus.com/cgi-bin/gnatsweb.pl?cmd=view&pr=33&database=libstdc++

or any of the "edit" links to try and change something, it sets a
"gnatsweb-test-cookie" and then I get kicked back to the generic login
screen.  And everything I try after that keeps kicking me back there.

FWIW, I'm using Netscrape 4.7 under NT[2].  Everything is enabled.


Phil
pme@sourceware.cygnus.com
devphil@users.sourceforge.net (don't shoot me)

[0]  At least five people, last I looked.
[1]  I'm paraphrasing; his actual words were along the lines of "would
     you please consider doing this?"
[2]  NT with speakers and a nice monitor makes a really sweet X terminal.

Re: A patch for toplevel Makefile.in [and a gnatsweb problem]

[Tom Tromey]

Phil> On a different subject, Jim Kingdon suggested that I send this
Phil> gnatsweb problem to overseers.  If gnatsweb is Not Your
Phil> Problem[tm], the rest of my babbling can be skipped.

I haven't seen or heard of your gnatsweb problem before.
I recommend contacting bug-gnats@gnu.org.
That is where the Gnats hackers live.

If they say that you should upgrade gnatsweb, report back to overseers.

Tom

Re: A patch for toplevel Makefile.in [and a gnatsweb problem]

[Tom Tromey]

>> If they say that you should upgrade gnatsweb, report back to overseers.

Phil> Whether this is enough recommendation to warrant an upgrade or
Phil> not is up to y'all; I'd say yes, but then I have something to
Phil> gain from it.  :-)

I looked at this.
The new gnatsweb relies on things our current Gnats can't do.
I don't have the time to upgrade our entire Gnats installation right
now.  Bummer.

Tom

Re: A patch for toplevel Makefile.in [and a gnatsweb problem]

[Jim Kingdon]

> Note to other sourceware admins: we run a locally-patched gnatsweb.

I've added a note to
http://sourceware.cygnus.com/sourceware/gnats.html - I can be kind of
a packrat with respect to keeping around records like that (if people
find these things helpful, do try to keep them up to date or otherwise
help, it gets discouraging if such pages are suspected of being
write-only).

> I sent this patch to the gnatsweb maintainer

Excellent!  Thanks for doing this.

> but I have no idea if it was incorporated.

Well, if you can succeed in working with the maintainers, you might
have the potential to save yourself merging work in the future :-)
(yeah, I know, this message is shameless evangelising, but I'm not
above that at times).

Re: A patch for toplevel Makefile.in [and a gnatsweb problem]

[Tom Tromey]

Phil> confirms the same problem, and says that the 2.7 beta of
Phil> gnatsweb doesn't suffer from this problem.

That's good enough for me.
I'll try to upgrade soon.

Note to other sourceware admins: we run a locally-patched gnatsweb.
Our version will turn URLs in text in the body of the PR into links
in the generated HTML.  I sent this patch to the gnatsweb maintainer
but I have no idea if it was incorporated.

Tom

Re: A patch for toplevel Makefile.in [and a gnatsweb problem]

[Phil Edwards]

Tom Tromey <tromey@cygnus.com>:
> Phil> On a different subject, Jim Kingdon suggested that I send this
> Phil> gnatsweb problem to overseers.  If gnatsweb is Not Your
> Phil> Problem[tm], the rest of my babbling can be skipped.
>
> I haven't seen or heard of your gnatsweb problem before.
> I recommend contacting bug-gnats@gnu.org.
> That is where the Gnats hackers live.
>
> If they say that you should upgrade gnatsweb, report back to overseers.

Here's my report back.

A couple other people have complained about this same problem on the
libstdc++ lists since I mentioned it here.  This message

   http://sourceware.cygnus.com/ml/bug-gnats/2000-03/msg00043.html

follows up my bug report to bug-gnats, confirms the same problem, and
says that the 2.7 beta of gnatsweb doesn't suffer from this problem.
Whether this is enough recommendation to warrant an upgrade or not is up
to y'all; I'd say yes, but then I have something to gain from it.  :-)

According to another recent message on bug-gnats (msg00045), the gnatsweb
maintainer has apparently moved with no forwarding address.  (Or could be
sick, etc.)


Phil

Re: A patch for toplevel Makefile.in [and a gnatsweb problem]

[Tom Tromey]

>> If they say that you should upgrade gnatsweb, report back to overseers.

Phil> Whether this is enough recommendation to warrant an upgrade or
Phil> not is up to y'all; I'd say yes, but then I have something to
Phil> gain from it.  :-)

I looked at this.
The new gnatsweb relies on things our current Gnats can't do.
I don't have the time to upgrade our entire Gnats installation right
now.  Bummer.

Tom

Re: A patch for toplevel Makefile.in [and a gnatsweb problem]

[Jim Kingdon]

> Note to other sourceware admins: we run a locally-patched gnatsweb.

I've added a note to
http://sourceware.cygnus.com/sourceware/gnats.html - I can be kind of
a packrat with respect to keeping around records like that (if people
find these things helpful, do try to keep them up to date or otherwise
help, it gets discouraging if such pages are suspected of being
write-only).

> I sent this patch to the gnatsweb maintainer

Excellent!  Thanks for doing this.

> but I have no idea if it was incorporated.

Well, if you can succeed in working with the maintainers, you might
have the potential to save yourself merging work in the future :-)
(yeah, I know, this message is shameless evangelising, but I'm not
above that at times).

Re: A patch for toplevel Makefile.in [and a gnatsweb problem]

[Tom Tromey]

Phil> confirms the same problem, and says that the 2.7 beta of
Phil> gnatsweb doesn't suffer from this problem.

That's good enough for me.
I'll try to upgrade soon.

Note to other sourceware admins: we run a locally-patched gnatsweb.
Our version will turn URLs in text in the body of the PR into links
in the generated HTML.  I sent this patch to the gnatsweb maintainer
but I have no idea if it was incorporated.

Tom

Re: A patch for toplevel Makefile.in [and a gnatsweb problem]

[Phil Edwards]

Tom Tromey <tromey@cygnus.com>:
> Phil> On a different subject, Jim Kingdon suggested that I send this
> Phil> gnatsweb problem to overseers.  If gnatsweb is Not Your
> Phil> Problem[tm], the rest of my babbling can be skipped.
>
> I haven't seen or heard of your gnatsweb problem before.
> I recommend contacting bug-gnats@gnu.org.
> That is where the Gnats hackers live.
>
> If they say that you should upgrade gnatsweb, report back to overseers.

Here's my report back.

A couple other people have complained about this same problem on the
libstdc++ lists since I mentioned it here.  This message

   http://sourceware.cygnus.com/ml/bug-gnats/2000-03/msg00043.html

follows up my bug report to bug-gnats, confirms the same problem, and
says that the 2.7 beta of gnatsweb doesn't suffer from this problem.
Whether this is enough recommendation to warrant an upgrade or not is up
to y'all; I'd say yes, but then I have something to gain from it.  :-)

According to another recent message on bug-gnats (msg00045), the gnatsweb
maintainer has apparently moved with no forwarding address.  (Or could be
sick, etc.)


Phil