this needs wide distribution

[Xavi Drudis Ferran]

>  || On Fri, 22 Dec 2006 10:31:49 +0100 (CET)
>  || "Xavi Drudis Ferran" <xdrudis at tinet.cat> wrote:
>
>  xdf> To reject an extension to a free software program is trying to
>  xdf> make it non free.
>
> No. Freedom includes also the freedom to not want something.
>
> Many of the best Free Software projects are good because they have
> avoided feature bloat -- in other words have declined extensions.
>

Sure. That's why I added "You may simply not install it on your systems.".
you are free to choose the features of the free software you develop
and are even free to choose the features of non free software you develop
(no offense intended). That's obvious. Software patents prevent this
but this is another story.

Additionally you are free to remove features from free software (non-free
software does not enjoy this advantage), because
you are free to develop versions of it. But you can't choose for others
what features go in their versions of any free program. That's what I meant.

If you could reject what features other people put in their versions of a
program, then the program would not be free. Freedom menas you can take
your choices and others can too.

If I reject feature Z in free program A this is irrelevant to anyone else
(as it should be). Except when everyone gets their program from me to save
themselves the trouble of adding or removing features. But when I reject
feature Z in program A and people want it, then someone else may offer
a version of program A with feature Z included. If someone couldn't then
A wouldn't be free software (only recipients of the software get the
freedom, not the whole world, but let's not make it unnecessarily
complicated).

So I don't think calling that "I reject feature Z in program A" is very
meaningful in free software.

I don't think we should expect to change MS policy, so if they pay for
ODF->OXML conversion and make it available as free software, it will be
available (assuming it can be programmed which everyone seems to think it
can). Once it is available freedom allows you to choose whether to take
it in your version of OOo or not. Freedom allows you to distribute OOo
without ODF->OXML export, but it does not allow you to prevent MS, Novell
or the Pope to distribute OOo with OXML export.

If you convince the OOo team of not including it then people who want it
will simply get it from Novell, MS or the Pope. You can make it slightly
more expensive for them to mantain their version, that's all. And you risk
losing users which would defeat your purpose.

But you could also convince the OOo team of accepting the contribution
after changing it to try to educate the  user, highlight problems as they
appear, etc. Give warnings and title them "consumer information notice",
link them to the SELF project resources, highlight features in OOo not
present in OXML, highlkight bloat in OXML not present in ODF, give
additional warnings if the user tries to send it by email, even include a
text in the converted document warning of its inability to work and the
way to obtain an ISO 26300 or PDF/A version (as far as posible, you could
ask the user the necessary info about where it will be stored or
published), etc.

Then, if you manage to get the information accross, to convince users that
information is useful and not make it too much of a nuisance. Then maybe
users will prefer this version than the MS, Novell or catholic OOo. And
maybe we get something out of it.

It can also be decided than even thus this feature is bloat and OOo is
better without it. It's less code, less maintenance, likely less
requirements... You could even try to educate users without including
the feature. I don't know which is best. But I know that you can't decide
this feature won't be available to users. you can only decide it won't be
in the official OOo version (if you convince them).