Re: [hwloc-devel] hwloc web site

2009-09-15 Thread Samuel Thibault
Jeff Squyres, le Tue 15 Sep 2009 20:07:22 -0400, a écrit :
> On Sep 15, 2009, at 7:14 PM, Samuel Thibault wrote:
> 
> >> I did a bunch of integration work for the hwloc web site -- could  
> >you
> >> guys review this before it goes live?  This is my development copy:
> >>
> >> http://www.open-mpi.org/~jsquyres/www.open-mpi.org/projects/hwloc/
> >
> >Could you perhaps already update the doc (html and pdf) with what we
> >have now, at least to fix the name and milestone?
> 
> I'm not sure what you mean -- can you be specific?

The PDF link brings a PDF with libtopology as title, and the
Documentation link too. In principle that should just be a matter of
updating both with what's generated from the current svn.

Samuel


Re: [hwloc-devel] hwloc web site

2009-09-15 Thread Jeff Squyres

On Sep 15, 2009, at 7:14 PM, Samuel Thibault wrote:

> I did a bunch of integration work for the hwloc web site -- could  
you

> guys review this before it goes live?  This is my development copy:
>
> http://www.open-mpi.org/~jsquyres/www.open-mpi.org/projects/hwloc/

Could you perhaps already update the doc (html and pdf) with what we
have now, at least to fix the name and milestone?




I'm not sure what you mean -- can you be specific?

FWIW, the hwloc site went live this weekend -- the site is now:

http://www.open-mpi.org/projects/hwloc/

--
Jeff Squyres
jsquy...@cisco.com



Re: [hwloc-devel] hwloc web site

2009-09-15 Thread Samuel Thibault
Hello,

BTW, thanks for your work on the website/trac/etc. integration, I hardly
have any time atm to do things like this :)

Jeff Squyres, le Fri 11 Sep 2009 09:41:03 -0400, a écrit :
> I did a bunch of integration work for the hwloc web site -- could you  
> guys review this before it goes live?  This is my development copy:
> 
> http://www.open-mpi.org/~jsquyres/www.open-mpi.org/projects/hwloc/

Could you perhaps already update the doc (html and pdf) with what we
have now, at least to fix the name and milestone?

Samuel


Re: [hwloc-devel] API, and make distcheck failure

2009-09-15 Thread Jeff Squyres

On Sep 15, 2009, at 2:07 AM, Brice Goglin wrote:


I just gave a deep look and fixed several minor things in the API. The
biggest thing I am not sure about is:
We currently have several functions that may find something by type or
by depth (for instance the number of objects, the n-th one, the next
one, ...). These functions are currently named
"hwloc_get_foo_by_depth(..., depth)" if looking by depth and
"hwloc_get_foo(..., type)" if by type. The latter may fail if there  
are

multiple levels with same types (e.g. Caches), the former always work
(unless depth is invalid). Should we keep these function name patterns
or change them to something else? like _by_depth/_by_type or
/_by_type?



I prefer being explicit -- so _by_type / _by_depth.  My $0.02.


Also, make distcheck currently fails here. It looks for index.html
inside the build dir while it should look inside the source dir.
However, it seems that it correctly looks inside the source dir when
building outside of make distcheck. Don't know what's going on...



I'm unable to recreate this.  Can you give me a recipe to recreate  
this problem?


--
Jeff Squyres
jsquy...@cisco.com



[hwloc-devel] API, and make distcheck failure

2009-09-15 Thread Brice Goglin
Hello,

I just gave a deep look and fixed several minor things in the API. The
biggest thing I am not sure about is:
We currently have several functions that may find something by type or
by depth (for instance the number of objects, the n-th one, the next
one, ...). These functions are currently named
"hwloc_get_foo_by_depth(..., depth)" if looking by depth and
"hwloc_get_foo(..., type)" if by type. The latter may fail if there are
multiple levels with same types (e.g. Caches), the former always work
(unless depth is invalid). Should we keep these function name patterns
or change them to something else? like _by_depth/_by_type or
/_by_type?

Also, make distcheck currently fails here. It looks for index.html
inside the build dir while it should look inside the source dir.
However, it seems that it correctly looks inside the source dir when
building outside of make distcheck. Don't know what's going on...

Brice


LC_ALL=C /usr/bin/w3m ../doc/doxygen-doc/html/index.html | sed -n -e 
'/^Introduction$/,$p' > ../../README
/bin/bash: ../../README: Permission denied
w3m: Can't load ../doc/doxygen-doc/html/index.html.
make[3]: *** [../../README] Error 1
make[3]: Leaving directory 
`/home/bgoglin/libtopology/build/hwloc-0.9.1a1-1/_build/doc'
ERROR: generating new README failed
ERROR: cannot continue
make[2]: *** [dist-hook] Error 1
make[2]: Leaving directory 
`/home/bgoglin/libtopology/build/hwloc-0.9.1a1-1/_build'
make[1]: *** [distdir] Error 2
make[1]: Leaving directory 
`/home/bgoglin/libtopology/build/hwloc-0.9.1a1-1/_build'
make: *** [distcheck] Error 1