Open projects list

Open projects list

[Stan Shebs]

I was poking around to see if named threads was written down anywhere
as a desired GDB feature, and couldn't find it on web pages or in
sources, which makes me wonder how many other longstanding wishlist
items aren't known about. I see that there are lots of feature
requests in bugzilla, but they are more about specific details,
rather than overall direction, and IMHO the bugzilla format is not
so helpful for larger or architectural changes that may need lengthy
background and explanation, like MI levels to support, reverse
execution infrastructure, native tracepoints, etc. GCC has a
whole collection of pages under http://gcc.gnu.org/projects/ for
this kind of thing.

Does anybody else think we need something like this?

Stan

Re: Open projects list

[Bob Rossi]

On Fri, Oct 07, 2005 at 12:25:00PM -0700, Stan Shebs wrote:
> I was poking around to see if named threads was written down anywhere
> as a desired GDB feature, and couldn't find it on web pages or in
> sources, which makes me wonder how many other longstanding wishlist
> items aren't known about. I see that there are lots of feature
> requests in bugzilla, but they are more about specific details,
> rather than overall direction, and IMHO the bugzilla format is not
> so helpful for larger or architectural changes that may need lengthy
> background and explanation, like MI levels to support, reverse
> execution infrastructure, native tracepoints, etc. GCC has a
> whole collection of pages under http://gcc.gnu.org/projects/ for
> this kind of thing.
> 
> Does anybody else think we need something like this?

Yeah, the more info the better. Seriously.

Bob Rossi

Re: Open projects list

[Daniel Jacobowitz]

On Fri, Oct 07, 2005 at 12:25:00PM -0700, Stan Shebs wrote:
> I was poking around to see if named threads was written down anywhere
> as a desired GDB feature, and couldn't find it on web pages or in
> sources, which makes me wonder how many other longstanding wishlist
> items aren't known about. I see that there are lots of feature
> requests in bugzilla, but they are more about specific details,
> rather than overall direction, and IMHO the bugzilla format is not
> so helpful for larger or architectural changes that may need lengthy
> background and explanation, like MI levels to support, reverse
> execution infrastructure, native tracepoints, etc. GCC has a
> whole collection of pages under http://gcc.gnu.org/projects/ for
> this kind of thing.
> 
> Does anybody else think we need something like this?

If we want something like this, a Wiki would be a better choice.  GCC
certainly found their wiki to be more effective than the projects
pages, which are mostly translations of old projects from the source
distribution.

Also, I find it somewhat ironic that you said "in bugzilla" since we're
still living in the dark ages of gnats.  I'm torn between setting up a
third independent bugzilla instance on sourceware, or going through the
pain of renumbering all the active references to PRs in the testsuite.
Open to suggestions.

-- 
Daniel Jacobowitz
CodeSourcery, LLC

Re: Open projects list

[Stan Shebs]

Daniel Jacobowitz wrote:

>On Fri, Oct 07, 2005 at 12:25:00PM -0700, Stan Shebs wrote:
>
>>I was poking around to see if named threads was written down anywhere
>>as a desired GDB feature, and couldn't find it on web pages or in
>>sources, which makes me wonder how many other longstanding wishlist
>>items aren't known about. I see that there are lots of feature
>>requests in bugzilla, but they are more about specific details,
>>rather than overall direction, and IMHO the bugzilla format is not
>>so helpful for larger or architectural changes that may need lengthy
>>background and explanation, like MI levels to support, reverse
>>execution infrastructure, native tracepoints, etc. GCC has a
>>whole collection of pages under http://gcc.gnu.org/projects/ for
>>this kind of thing.
>>
>>Does anybody else think we need something like this?
>>
>
>If we want something like this, a Wiki would be a better choice.  GCC
>certainly found their wiki to be more effective than the projects
>pages, which are mostly translations of old projects from the source
>distribution.
>
Heh, good idea, "Mr. 22,000-edits-on-Wikipedia" should have thought
of that...

Stan