Etienne Folio <etienne.folio@lrde.epita.fr> writes:

On Fri, Jun 5, 2009 at 17:28, Roland Levillain <roland@lrde.epita.fr> wrote:

| 2009-06-05  Etienne FOLIO  <folio@lrde.epita.fr> | |       Tests around projected histograms for the report. |       * folio/test/histo/classify_with_histo_rgb.cc: . |       * folio/test/histo/project_histo_add.cc: New. |       * folio/test/histo/project_screen.cc: New. |       * folio/test/histo/projected3d.cc: . |       * folio/test/value/comp.cc: New.

This is not cool.  You should refrain from committing dirty ChangeLog entries.  Please fix it.  (Re)read the LRDE Guidelines.

And why can't I see you patch in the olena-patches mailing list?

https://www.lrde.epita.fr/pipermail/olena-patches/2009-June/thread.html

Sorry, but Theo asked me to commit even dirty changes and garbage code and I really had nothing to say about them (even that it's unsorted testing stuff).

Sorry, but that's no excuse. You really didn't know how to describe what your patch does? Then what's the point in committing? How will someone else be able to utilize your work if you do not explain your changes?

About the patch in the ML, I have no clue... I use Tsuna's svn-wrapper and I thought it was automatic... Any ideas why ?

No, and that's not the problem. Patches must be sent to the list, whatever the method (even manually if needed). Tools are great, but their behavior is *not* the rule: if they have flaws, you should work around them. We are interested in the results, not the means you used. People used to send patches by hand before; tools admittedly simplified the process, but at the same time they have started to give a wrong feeling of safety. Hence the recurring unchecked and ugly ChangeLog entries, missing messages, etc. Committing is not just about saving your changes on a remote/shared machine. It's about sharing useful and reusable work. I know Théo sometimes urge students to commit, but please everybody, take five minutes to clean up your patches before sending them. Thank you in advance.