Hi!
intrigeri:
>> The question boils down to, when creating new features that require user
>> interaction or have user output, does it have have to provide
>> internationalization support to support as many languages as possible?
>
> Yes, we do insist on all Tails -specific software being i18n'd: see
> "2.6.4.1 Internationalization" in the PELD spec. I believe we are
> generally doing pretty well on this side, despite some serious
> glitches (e.g. suboptimal tails-greeter behavior for RTL languages).
>
> To further clarify, we're insisting *on i18n only* during initial
> development: actual l10n work (translation etc.) can be done later, so
> that's not seriously delaying things (one would not seriously think of
> writing production-ready end-user interfaces that are not i18n'd
> anyway; this is 2013).
>
> Therefore, "Whonix does not (yet) have the requirement to support as
> many languages as possible before a feature can be added", suggesting
> it may be different in Tails, sounds quite misleading to me:
> it sounds
> like lack of i18n might be a feature Whonix has, and Tails might not
> have...
Having multi language support is actually a feature, an advantage of
Tails. To reflect this better, I added it to a table:
https://sourceforge.net/p/whonix/wiki/Comparison with Others2/#features
> I find these expectations scary for the Whonix project (and for Whonix
> users), as it seems to imply that you'll have to forever remain the
> final decision maker in there (benevolent dictator for life, as they
> say). But well, that's obviously your call.
I see that quite differently. In a symbolic comparison... There may be a
political democratic party A. A person could either join party A and
contribute and try to influence their decision making process of party A
or if the person feels that their basic convictions are too different,
start a new party B.
Another comparison more on topic. Let's look at Debian and Ubuntu.
Debian people (I assume) strongly believe, that no proprietary software
should get installed by default installation. Ubuntu is values pragmatic
solutions more than freedom and decided to include proprietary software
by default installation. They also made other decisions, which I think
won't happen to Debian, such as Unity as default desktop, big UI
changes, big website redesign, dropping support for many platforms,
promoting commercial software in software center and more. Without
judging their course of action, it wouldn't have made sense to attempt
to talk Debian people into doing these things, even if they are the ones
who actually do the coding work.
No developer ever complaint about any decision making process in Whonix.
(A few users wanted to dictate what priorities should be without
offering help.) Past cooperations with anonymous and smarm (as Whonix
was still called TorBOX) worked out well. Current cooperations with new
contributors also working well. I'd speculate that might be because
people who get interested in contributing to a project are accepting its
implicit philosophy.
>>> c. By affirming that strongly that one has to "obey" else
>>> "contributions do not get merged", it suggests that there are
>>> precedents of this process. Reference needed.
>
>> The word "obey" has already been corrected.
>
>> The reason for pointing the quoted sentence above is to answer "Why
>> Whonix, why not contribute to Tails instead?" like questions.
>
>> Example, I like to see XChat in Tails:
>> https://mailman.boum.org/pipermail/tails-dev/2012-August/001442.html
>
>> Decision making process says no.
>
> Just to clarify, my understanding of what happened is quite different.
>
> First, the only reason you provided in this thread, in support for
> shipping XChat, was sharing anonymity set with Whonix. It's obviously
> not a sufficient one.
This was one reason. The other reason was, because people in Tails forum
asked how to do it and I thought why not share my findings.
> Second, decision making process was stalled at "Unless we have a very
> strong argument that proves us that XChat does things that Pidgin do
> not and that most of Tails users would benefit of". As far as I can
> tell, I've seen no such argument being made,
> so the decision making
> process has been stalled since then (and status quo remains until some
> progress is made).
This is correct.
Some statements without references:
- Maybe I could have made up my point why I believe that Pidgin is bad
choice as IRC client and why XChat is better.
- I felt many people in the forum requested it.
- It's obvious to me, that Pidgin isn't a comfortable IRC client and
that XChat is much better.
- I think the majority of people believes the same.
- There are more XChat than Pidgin users.
I could have invest time to provide all the references and reasons and
make a giant write up to prove something which is obvious to me and
risk, that there is still no interest. Rather, I safe the time to argue
with people who's convictions (without judgment) are so different, and
do something, which I feel to be more promising and productive.
>> I can do nothing but accept it.
>
> I beg to disagree: you're welcome to participate further in the
> decision making process, e.g. by explaining why, according to you,
> this change should be made.
Thanks. :) I'll do. When there are new topics, where I believe I can
contribute something useful, I'll do.
>> Now, when I dislike the decision and I strongly care about it, I can do
>> nothing other than either fork Tails or start a new project.
> ... or realize that, with persistence + remembered additional
> packages, XChat lovers can get their preferred stuff easily every time
> they boot Tails, so perhaps it's no big deal eventually?
Not sure about that. XChat requires some files in /home/ (to stop CTCP
an so on). Other things would require a persistent /etc/...
Remembered additional packages is/was one of many items, I was/am unable
to implement (+ unable to implement in state where it gets merged) and
unsure when you would do it.
> I very probably missed why you care that much about XChat, so I'm
> sorry I call this a "minor reason", but from my PoV, it seems uncommon
> for free software project to fork (or for rock bands to split) for
> this kind of relatively minor reasons.
XChat was just an example.
There are other examples, where I disagree with the decision making
process/consensus, points where I wouldn't even start a discussion, such
as disagreeing having Two-layered virtualized system being a wishlist
item rather than a most important item. My reasons for maintaining a
separate project instead of contributing to an existing one are really
not minor, as the updated FAQ entry on that topic hopefully reflects.
> I mean, if people routinely
> behaved this way, we would have a few dozens diverging forks of the
> Linux kernel.
Don't they routinely behave this way? There are more Linux distributions
out than distrowatch can review and add.
>> Problem is, I don't know any wiki host interested in this kind of topic,
>> where (registered) Tor users are free to edit, and where the involved
>> projects are willing to participate. If there was, I am sure the
>> involved projects all are mature enough to make it objectively correct
>> in no time.
>
> Personally, I'm not interested in working on this comparison (yet),
> so I don't care where it is hosted and who can edit it.
This surprises me. I thought reviewing that page and going into lengthy
discussion about that implicates that.
>> The faq entry is supposed to be a subjective answer, why I maintain a
>> separate project, which differences I personally care about.
>
> I suggest making this clear on that page, then.
Done. I also rewrote it and moved the comparison to its own page:
https://sourceforge.net/p/whonix/wiki/Comparison with Others2/#features
That page and the FAQ entry has been rewritten in most parts:
https://sourceforge.net/p/whonix/wiki/FAQ/#why-dont-you-merge-with-tails-and-join-efforts
It is now hopefully no longer offending/biased/adverting/etc.
>> The comparison page is supposed to be as objective/factual
>> as possible.
> Then perhaps it should not use the "Why don't you merge with Tails and
> join efforts?" FAQ as a reference (in "Many other differences"
> section), 'cause doing this makes at least me read that FAQ with the
> same expectations.
Agreed. And therefore removed that link.
Cheers,
adrelanos
