PiTiVi in Ubuntu 11.10 and beyond

Jeff Fortin nekohayo at gmail.com
Wed May 25 18:26:52 UTC 2011


Hello there,

You may know me as one of the main contributors of the pitivi project. I
do bug triaging, testing, community management, documentation writing,
design, website maintenance, coding, etc.

It has come to my attention that the desktop team has decided at UDS to
remove Pitivi from the default selection of apps installed with Ubuntu.
I would like to have a better understanding of what led to this
decision, and hopefully revert it.

As far as I can tell, there is no recording of the session other than a
couple of notes on
http://summit.ubuntu.com/uds-o/meeting/desktop-o-default-apps/ , which I
shall quote entirely below for the record:

======= quote =======
Nominations off the CD:
 * PiTiVi (video editing)
  - Apple doesn't have an editor (this should be a motivation to keep
it, in fact)
  - We took off gimp because we don't need an editor for common cases
  - Still a bit crasfhy and not prime-time (2.5 stars in
software-center)
  - package itself is 265kb
  - poorly maintained (hasn't migrated to gtkbuilder)
  - Sounds like we want it off then
====== end quote =====

Since we were left in the dark and notified only after the fact (through
news sites and the session notes above), we've been wondering what has
been going on. Can I get a better explanation than that? These points
struck me as a bit unconvincing. But before I start, let me first say
that we have a new release coming up by the end of the week. More
details further below.


> "We took off gimp because we don't need an editor for common cases"

You took off GIMP because, as I remember it, you considered that apps
like Shotwell (F-Spot, back then) would be able to handle the basic
photo correction needs of most users. Unless you're telling me that
Ubuntu now aims to be a distro that only focuses on consuming culture
instead of producing it?
Video editing is becoming increasingly popular with the rise of video
sharing sites such as YouTube, Facebook, Vimeo, and many people now take
culture production into their own hands using dirt-cheap hardware (such
as point-and-shoot photo cameras or cellphones capable of
high-definition recording). Video is now part of the web, and video
production should be part of Ubuntu.


> "a bit crashy and not prime-time"

Could you be more precise? All applications have bugs. Bugs in video
editors are particularly difficult to troubleshoot, even moreso when 9
users out of 10 reporting bugs in LP don't provide the requested
information and the bug cannot be reproduced.
FWIW, while it certainly isn't perfect, it is certainly good-enough for
users to make videos such as http://www.pitivi.org/?go=showcase . 
Besides, the upcoming release is going to fix a ton of very annoying
bugs, but more details on that further below.


> "2.5 stars in the software center"

You base your claim of "poor user reception" upon *ten* user ratings?
When we have thousands of pitivi users? See also:
https://www.ohloh.net/p/pitivi
http://wiki.pitivi.org/wiki/Praise

Please consider:
- Ten ratings is a ridiculously low number for statistical significance.
- There are inherent sampling problems with a voluntary rating system
like yours, where a vocal minority of dissatisfied users will go rate
the app negatively while a silent majority of users are happy with the
app and don't spend time writing reviews all over the place.

Some more insight into how statistics can be misleading:
http://jeff.ecchi.ca/blog/2010/10/17/video-editors-usage-statistics-duty-calls/


> "poorly maintained (hasn't migrated to gtkbuilder)"

Is that the only thing you base yourself upon to say that an application
is "poorly maintained"? 
- First, there already is a development branch that tackles the
gtkbuilder migration in pitivi. It's not finished yet, as pitivi is not
a trivial application (and we had a big UI redesign in the past few
months), those things take time and it was not our absolute first
priority, but since there is an active contributor working on it, I can
guarantee that it will be done by Ubuntu 11.10.
- Secondly, there have been over 900 commits since last year. While we
are not the fastest moving project around in the FOSS world, I trust
that we are moving at a reasonable speed given our limited manpower and
the difficulty of the task at hand.

Whatsmore,
- There are *four* students working on pitivi for this Summer of Code,
hopefully tackling all the major projects/missing features that have
been the traditional subject of criticism towards pitivi. I haven't had
the time to blog about this yet because I was busy with studies, LGM,
work, and preparing the release.
- A new release maintainer has been appointed (which should help at
releasing more often). Over the past few weeks, we have been furiously
merging branches and patches, fixing bugs, and testing pitivi thoroughly
to push out a new release out of the door in the coming days. The
announcement to remove pitivi from Ubuntu's default selection came right
in the middle of that.


On that note, I am deeply disappointed by the complete lack of
contributions and support we had from Canonical and Ubuntu. Not only are
we told that we are "bad at maintaining our software", we are somehow
expected to churn out perfect releases at a rapid pace (we are being
considered "unmaintained" because we haven't managed to magically become
the perfect video editor within *one* year of being included in
Ubuntu?), and yet we have not seen any patches coming from @canonical or
@ubuntu addresses (except one 1-liner patch that "removes the version
number from the title bar").

Given that pitivi was in Main, I would have expected some sort of
support from Canonical. I'm not saying you folks should have been fixing
all our bugs and doing the development work for us, but my definition of
support is basically about being good open source citizens: "identify
the key problems that your users/customers are experiencing with the
software, and actually create fixes for them in cooperation with
upstream". It is not "just package whatever point-release that comes
from upstream, if any".

Instead, the burden on us was increased. After a while, I discovered
that I had to single-handedly triage all the bugs in Launchpad because
nobody else was doing it otherwise. Yet, you won't find a better-triaged
bug tracker than upstream pitivi (and even Launchpad, when I have the
time).

I'm getting tired of us being flamed for "not doing enough" when we're
doing our best with limited manpower and not getting any help.

We were not even contacted about this whole situation, the only
interaction we get from Ubuntu is "Have you got a release out? Feature
freeze is next week" once every six months, when we're lucky. Hell,
Edward Hervey himself was present at UDS this year and nobody even
bothered to ask him about this! Whatsmore, nobody ever told us that the
gtkbuilder migration was such a huge deal for you guys. Frankly, I
expect better communication than this from now on.


Sorry if the ending of this post sounds negative, but some things need
to be said and I want this to be a mutually beneficial relationship. I
want the situation to improve.

We have a new release coming with tons of bug fixes and features (see:
http://thiblahute.blogspot.com/2011/05/pitivi-pre-release.html ), great
projects for the summer, and we would really appreciate help from you
guys. Spend some time looking at the #pitivi channel... you won't find
many channels that are so lively and welcoming for newbies, users and
potential contributors. I spend insane amounts of time trying to attract
contributors by lowering their barriers to entry by having good
documentation, helping them get started, etc.

Jean-François, on behalf of the pitivi team





More information about the ubuntu-desktop mailing list