SeaMonkey vs Firefox&Thunderbird (localization perspective)
Huda Sarfraz
I am in the process of selecting open source applications for localization.
SeaMonkey 1.1 has been selected because it is a complete suite, but I need
some more information about it's relationship with Firefox&Thunderbird.
1. What is the relationship between the development process of
Firefox&Thunderbird and SeaMonkey? I know that they are both based on the
same backend code, but are their differences only limited to difference in
features? I read about the toolkit transition for SeaMonkey and SuiteRunner
here: http://wiki.mozilla.org/SeaMonkey:suiterunner, where one of the
reasons for the transition is stated as:
Using toolkit means we share more core files with Firefox, Sunbird,
Thunderbird etc, this means we benefit from their fixes to the core code,
and they benefit from ours.
Does this mean that until now fixes made to the core code were not shared
and so one could be more stable than the other (SM vs FF&TB)? And if the
development has branched out into two different directions, how will it be
merged again? Is all this documented in detail somewhere with maybe a
diagram? I found some information in this blog entry,
http://weblogs.mozillazine.org/seamonkey/ but if there is some more detail
documented somewhere it will be very helpful.
(I read in this forum post
http://forums.mozillazine.org/viewtopic.php?t=514029&sid=e66823f96f7919bfdea5ffd4d29df9eb
)
that some bugs were not fixed in SeaMonkey 1.1 because of the toolkit
transition phase, so they would not have to be re-fixed later, is this
true?)
2. I also read here:
http://www.mozilla.org/projects/seamonkey/news.htmlthat SeaMonkey
1.1 is "Powered by the same engine as Firefox 2", I'm assuming this means
that they both use Gecko 1.8.1 (http://developer.mozilla.org/en/docs/Gecko).
When the sharing of more core files is mentioned in the quote in my first
question, does it mean in addition to this?
3. I also noticed from http://developer.mozilla.org/en/docs/Gecko that TB
1.5 uses Gecko 1.8; some of our local langauge fonts are not rendered
correctly in Firfox 2.0 and SeaMonkey 1.1 (Gecko 1.8.1) and we might need to
do something about that in the future. Will the selection of SeaMonkey be
better in this regard because there we will have one version of Gecko to
deal with whereas with FF&TB both now and in the future they may be
different?
4. How can localization efforts for SM and FF&TB be co-ordinated or least
benefit from each other currently? This was mentioned briefly at #l10n some
days ago that they will be more co-ordinated in the future, but if we're
proceeding with SeaMonkey now will it help FF&TB?
5. Should the fact that SeaMonkey is not as well known as Firefox, and that
it has a smaller community be a concern and factor into the selection?
Thanks,
Huda.
(I posted similar questions in the mozillazine forums a few days ago and was
told that I should post to one of the newsgroups. I hope this is the right
place to ask.)
Robert Kaiser
newsgroup, see http://www.mozilla.org/projects/seamonkey/community.html
but as I'm reading here anyways, I'll reply to them here.
Huda Sarfraz schrieb:
> Does this mean that until now fixes made to the core code were not shared
> and so one could be more stable than the other (SM vs FF&TB)?
A big amount of core code is shared anyways (the critcial necko and
gecko components, for example) and therefore is not different in
stability or maintencance.
The parts that were forked by Firefox and developed into the "new
toolkit" SeaMonkey will be changing to, are some of the backend and
global definitions for UI elements, and some code affecting application
startup, profile management and such things.
> And if the
> development has branched out into two different directions, how will it be
> merged again? Is all this documented in detail somewhere with maybe a
> diagram? I found some information in this blog entry,
> http://weblogs.mozillazine.org/seamonkey/ but if there is some more detail
> documented somewhere it will be very helpful.
There are no detailed documentation bits other than the source code
itself. Shared and forked code is very intermingled right now, so it's
not even clearly describable in many cases. The best documentantion is
probably still http://wiki.mozilla.org/SeaMonkey:suiterunner - but you
know that one already.
And then, there are "suiterunner status report in the m.d.a.seamonkey
newsgroup - where talk about this actually really belongs.
Maybe the slides for my FOSDEM talk in less than two weeks will help you
a bit as well, once I have them ready to show to the public.
> (I read in this forum post
> http://forums.mozillazine.org/viewtopic.php?t=514029&sid=e66823f96f7919bfdea5ffd4d29df9eb
>
> )
> that some bugs were not fixed in SeaMonkey 1.1 because of the toolkit
> transition phase, so they would not have to be re-fixed later, is this
> true?)
This is all on the "I heard that..." basis, and mostly wrong actually.
Many things are not fixed just because noone did the work for it. We are
a small team of volunteers, all with their non-SeaMonkey day jobs, and
everybody can only do a certain amount of work.
> 2. I also read here:
> http://www.mozilla.org/projects/seamonkey/news.htmlthat SeaMonkey
> 1.1 is "Powered by the same engine as Firefox 2", I'm assuming this means
> that they both use Gecko 1.8.1
> (http://developer.mozilla.org/en/docs/Gecko).
Right. The identical Gecko code (as versioned by the rv: and Gecko/xxx
parts of the UA string) is shared by SeaMonkey, Firefox and other
Mozilla products. This is one of the parts that thankfully never were
forked by the Firefox team.
> 3. I also noticed from http://developer.mozilla.org/en/docs/Gecko that TB
> 1.5 uses Gecko 1.8; some of our local langauge fonts are not rendered
> correctly in Firfox 2.0 and SeaMonkey 1.1 (Gecko 1.8.1) and we might
> need to
> do something about that in the future. Will the selection of SeaMonkey be
> better in this regard because there we will have one version of Gecko to
> deal with whereas with FF&TB both now and in the future they may be
> different?
FF 1.5.x and TB 1.5.x share the same Gecko with SeaMonkey 1.0.x, same is
true for FF 2.x, TB 2.x and SeaMonkey 1.1.x - only that Thunderbird has
not released a final 2.0 version yet.
FF3, TB3 and SeaMonkey 1.5 (or 2.0, that is still open for internal
discussion) will also share the same Gecko 1.9, which will have many
rendering improvements, being based on the cairo libraries (and pango
text rendering on Linux).
> 4. How can localization efforts for SM and FF&TB be co-ordinated or least
> benefit from each other currently? This was mentioned briefly at #l10n some
> days ago that they will be more co-ordinated in the future, but if we're
> proceeding with SeaMonkey now will it help FF&TB?
Currently, FF/TB L10n is maintained in cvs, and completely different
from SeaMonkey, which is maintained privately by localizers, not (yet)
backed by shared cvs repositories or such.
In the SeaMonkey 1.5/2.0 timeframe (Gecko 1.9), we're working to get the
cvs-based process of FF/TB also in place for SeaMonkey, which should
make us share the exactly same files for "toolkit" L10n.
The files of those "toolkit" parts are identicaly already in many cases
right now in current release, so with the help of some tools, you
probably can share them already (teams that localize all those apps can
probably tell you how to do that best).
In the future, they'll reside in a common place in the cvs, which will
make that even easier.
> 5. Should the fact that SeaMonkey is not as well known as Firefox, and that
> it has a smaller community be a concern and factor into the selection?
It probably depends on what you want to achieve and what your target
audience is.
SeaMonkey is more targeted at experts, advanced users and maybe even
business users who want to "install only one product to achieve
everything", while Firefox is strongly targeted at consumers who change
over from IE.
Therefore, Firefox will always have a bigger audience and will be known
to more people than SeaMonkey.
The smaller community has its merits and it pains, but, to make it
short, don't fear us to die just because of a smaller team - we already
have shown that we are a very dedicated bunch of people just by saving
the suite from dying almost 2 years ago...
> (I posted similar questions in the mozillazine forums a few days ago and
> was
> told that I should post to one of the newsgroups. I hope this is the right
> place to ask.)
This is the right place for L10n questions, SeaMonkey development
questions should better go to the SeaMonkey newsgroups, listed in the
document I referred to at the beginning of this post.
Robert Kaiser
SeaMonkey Council member
Gervase Markham
> 1. What is the relationship between the development process of
> Firefox&Thunderbird and SeaMonkey? I know that they are both based on the
> same backend code, but are their differences only limited to difference in
> features?
For the user, yes. Although that's not very helpful as, from a user
point of view, any two bits of software are only different in their
features :-)
> I read about the toolkit transition for SeaMonkey and SuiteRunner
> here: http://wiki.mozilla.org/SeaMonkey:suiterunner, where one of the
> reasons for the transition is stated as:
>
> Using toolkit means we share more core files with Firefox, Sunbird,
> Thunderbird etc, this means we benefit from their fixes to the core code,
> and they benefit from ours.
>
> Does this mean that until now fixes made to the core code were not shared
> and so one could be more stable than the other (SM vs FF&TB)?
Switching Seamonkey over to using toolkit will increase the amount of
shared code. Talking about relative stability is complicated because the
latest version of SM and the latest version of FF and TB might be from
different code branches.
> And if the
> development has branched out into two different directions, how will it be
> merged again?
There are no plans to merge SM and FF/TB; they address different
markets. The primary products of the Mozilla project are FF and TB;
however, some contributors wished to keep maintaining a suite, and so
they do.
Gerv