[Bf-committers] Wondering about the review of the command port patch...

[Dietrich Bollmann]

On Fri, 2008-09-19 at 09:56 +0200, Ton Roosendaal wrote:
> Hi,
> 
> You can reply in the tracker if you have questions.

My responses on the tracker have been pretty much ignored.
Also I didn't want to open the patch again.  If you don't want 
the command port you don't want it.  I just wanted to express 
my opinion about a bad review.

By the way, it would be nice to have another blender repository 
where people are allowed to open their own branches for experimentation.  
You could take what you want and people would be free to work on their 
stuff without worrying if the official blender developers are interested 
or not.  No need to spend another week for writing a patch after having 
been asked to do so (me) which is just ignored after thanks to a bad 
reviewer (me again).

For the moment it is quite annoying to play with the sources without 
having the possibility to submit incremental changes.  I finished by 
having my own svn repository in order to commit my changes and imported 
the blender sources as "vendor drop" - a huge and unnecessary waste of time.  
And a very old-fashioned way of organising an open source project in my opinion.

> The fact that patch submitters do so much work is unfortunately not a  
> guarantee we do the the same for them... that's just not practical with  
> 100+ open patches to look at.
> I'm also not impressed or intimidated by coding work in itself. A patch  
> should never be a blackmail version of a feature request. :)

...and you are creating a lot of frustration and loosing a lot of 
enthusiastic coders by doing it this way.

Actually I needed a command port for doing my own stuff (before I was 
using Maya which has a command port available) and later thought that it 
would be nice to give away my code for other people with the same needs.  
In my opinion a command port should be part of any serious program anyway - 
but of course, you are the "benevolent dictator" or how this job is called :)  
...not an easy thing to do, I know...

> I wanted to go over almost all open patches, and then come with a  
> proposal for more clear guidelines, so submitters know better what they  
> can expect, and to have active Blender developers involved with reviews  
> without them feeling obliged to spend hours on reviews.

It would be nice to have more payed developers for Blender who could work 
on the non-fun-only stuff.  A money problem of course.  For the moment 
everybody works for his own fun - me being a good example :)  ...and nobody 
is interested to work on other peoples patches - myself a good example again :)

> Further I'm also human and can make mistakes! Maybe your patch is  
> great, but then try to find people in the active developer/user  
> community to review and use it?

Anyway.  Never mind.  Blender is great even with you being human :)

...and one day, I am sure, you will have a command port anyway.  
Until than I use my own branch...

And now, after everybody hates me:

Thanks for your hard work - and good luck :)

Dietrich

> -Ton-
> 
> ------------------------------------------------------------------------
> Ton Roosendaal  Blender Foundation   ton at blender.org    www.blender.org
> Blender Institute BV  Entrepotdok 57A  1018AD Amsterdam The Netherlands
> 
> On 19 Sep, 2008, at 5:30, Dietrich Bollmann wrote:
> 
> > Hi Ton,
> >
> >> Comment By: Ton Roosendaal (ton)
> >> Date: 2008-09-18 19:49
> >>
> >> Message:
> >> A review gave following considerations;
> >>
> >> It's basically a good idea, but the patch is gigantic, changing a lot  
> >> of internals,
> >> which is unknown if that's going to be maintained. For an option that  
> >> not many people
> >> will use too problematic. My reason to close this one for now.
> >
> > No, the changes to the internals are minimal.
> >
> > It is the NEW functionality in dir source/blender/commandport which is
> > huge.
> >
> > I think the reviewer currently is too busy to look at the patch itself
> > and just wanted to get rid of an item in his todo list.  Probably he  
> > was
> > scared away by the number of lines and decided before looking into the
> > code itself.  What exactly did he write and who is he?  I would like to
> > have a chance to react to his comments.
> >
> > Even if somehow disappointed about the way my patch was treated from  
> > the
> > beginning, of course I accept if you don't want it.  But I would like  
> > to
> > have it rejected with arguments which are motivated by the code itself
> > and not by the fact of everybody being very busy - which I understand  
> > of
> > course very well!
> >
> > Please do not understand me wrong.  I know that everybody is very busy
> > and I know that everybody has to manage his time.  Sometimes it is good
> > to follow the first impression.  Still I would be happier with real
> > arguments - at least I could learn what exactly the blender coders
> > didn't like about my code.
> >
> > Here is a list of the c source files which have been manipulated:
> >
> > egrep '^Index: .*\.c'
> > blender-command-port-patch.2008-07-12.2.46.r15538.txt \
> >   | egrep -v 'source/blender/commandport'
> >
> > Index: source/creator/creator.c
> > Index: source/blender/python/BPY_interface.c
> > Index: source/blender/src/mainqueue.c
> > Index: source/blender/src/editscreen.c
> > Index: source/blender/src/usiblender.c
> > Index: source/blender/src/eventdebug.c
> >
> > I append the corresponding diffs.
> >
> > Thanks for your understanding,
> >
> > Sincerely, Dietrich
> >
> >> The good news is that a similar functionality (commandline python  
> >> editing) is on
> >> the wishlist in Blender itself. After that, an external console  
> >> should be easy.
> >
> > If it was the patch would be shorter :)
> [...snip...]