Personal remarks on Epiphany/Galeon [Was: Release Team's Almost-Final Modules List]

[Jeff Waugh <jdub perkypants org>]

<quote who="Jeff Waugh">

>     galeon
> 
>       There was a lot of vocal support for Epiphany on the mailing list, but
>       little for Galeon. So, the consensus points to Epiphany. That was due
>       to a number of factors: The Epiphany project goals seem to be better
>       aligned with GNOME's goals, the Galeon developers do not seem to be
>       100% behind GNOME's goals; Epiphany has had regular releases for GNOME
>       2.3.x; the Epiphany hackers are working within the project to define
>       standards and code for toolbar editing and other functionality, etc.
>       While there are a lot of reservations about offending the Galeon
>       hackers, and great disappointment that the two projects have not been
>       able to cooperate, Epiphany does seem to have the consensus, and make
>       the most technical sense. That is not to say that Epiphany is without
>       faults, or that Galeon is not excellent software -> we do need to make
>       a choice at some stage: Right now there is strong support to add a
>       browser to the Desktop release, and strong support for that browser to
>       be Epiphany.

Unfortunately, I don't think we're making the right decision here, despite
it being the obvious consensus of input to date (and the fact that I agree
with all of the comments above re: technical merit). The pressure we're
under to ship a browser with the Desktop release has blinded us to important
issues we should consider, especially in the case of competing modules. So,
here's what I think:

  - Epiphany and Galeon are both high quality projects, based on the same
    code base, competing at a very even level. They both have strengths and
    weaknesses (on a purely technical level).

  - Galeon has a much larger user base, to whom its features are understood
    and familiar.

  - Neither project has had a stable point-oh release of its current series
    (I'm assuming here that the Galeon dudes still use the kernel versioning
    scheme, and we are waiting for their 1.4 or 2.0 release).

  - Epiphany has modified major features of the browser interface, such as
    the bookmarks system, and has not yet seen wide user testing of these
    changes. Personally, I think this particular feature needs some serious
    testing and attention before users will accept it. I don't believe that
    it's as good as a polished, traditional hierarchical system, nor is it
    as good as Safari's design for a flat system. I'm not just questioning
    the user interface here, btw. I'm questioning the design from the data
    model all the way up to the user model. :-)

  - It seems, at least to me, that Epiphany has not fully achieved Marco's
    vision for the 'perfect GNOME browser', though it is obvious that he is
    working very hard to do so. I mention this because with the modules
    freeze comes the feature freeze, so what you see in Epiphany today is
    what you'll see in Epiphany when 2.4 is released. I don't believe that
    we can say "yes, Epiphany is feature complete and ready to roll" right
    now.

  - If GNOME does not include a browser with 2.4, users and distributions
    will choose. It's not like people will suddenly not have a browser. :-)
    I think this is a crucial point because it lets us make a "Linus-style"
    lazy decision -> let the community decide. By the time we get to module
    selection for 2.6, I'm sure we'll see a clear winner.

  - If we make this choice right now, what happens to the other project? For
    instance, if we choose Epiphany, will Galeon simply disappear, or will
    the developers rally themselves to outclass Epiphany? If we choose
    Galeon, will Epiphany disappear, or will it prove to be more integrated
    and GNOME-like than Galeon? Because the browsers are on similar footing
    competitively, if one does not disappear, that means that GNOME is
    essentially competing with itself. What a horrific waste of developer
    talent and good will! What a difficult proposition for users who entrust
    us to 'bless' or choose the best software - what should they think?

My summarised suggestion:

  Let's defer our choice, to allow the two projects to "compete naturally"
  in the time leading up to 2.6. I believe that we will have an obvious
  winner by then, with stable releases of the browsers, and clear buy-in
  from users and vendors. They will decide anyway - let's let them decide on
  the merit of stable releases, not on the "decree" of the GNOME project. I
  would like to see this "Linus-style" lazy evaluation applied to competing
  modules in the future, too, if we are unfortunate to see it happen again.

Thanks very much to both the Epiphany and Galeon hackers, thanks to you
guys, we have two very excellent browsers that we have to choose from! :-)

- Jeff

-- 
linux.conf.au 2004: Adelaide, Australia         http://lca2004.linux.org.au/
 
   "Consensus is whatever the developers remember or agree with." - Paul
                            Vixie, Open Sources