The Forgotten "F": A Tale of Fedora's Foundations

[Miloslav Trmač]

2014-04-22 21:46 GMT+02:00 Josh Boyer <jwboyer at fedoraproject.org>:

> On Tue, Apr 22, 2014 at 2:54 PM, Miloslav Trmač <mitr at volny.cz> wrote:
> > 2014-04-21 22:35 GMT+02:00 Josh Boyer <jwboyer at fedoraproject.org>:
> >
> >> I think the problem I have with this well-intentioned thread is that
> >> it's a broad reaction to a specific issue we're trying to sort out
> >> right now.  Webapps aren't new, the fact that a large portion of them
> >> aren't FOSS isn't new, and their usage in and interoperability with
> >> Fedora is not new.
> >
> >
> > Oh no, all of this is actually new.
>
> Your definition of "new" is... strange.
>
> > The change was so smooth and gradual that now, looking back, we see it as
> > inevitable and natural; but compared to the 1995/2000-time era, it has
> been
> > a drastic change.
>
> Yes.  Also, 1995 compared to 2000 had drastic changes even within that
> timeframe.  Even with your statements below, I have no idea how you
> consider this new (nor why 1995/2000 are relevant).
>

OK, let me rephrase "my definition of 'new'" :)  If there were a specific
break between the "old" and "new", there would have been an appropriate
time to discuss the changes, our goals, assumptions and methods.  But there
never was a specific break, so there never was an appropriate time, and
that discussion has, mostly, not happened.

You're right, this is not new *now*, "now" is just a random time and random
circumstances.  But looking back, new things *have clearly happened* and we
need to take the time to discuss the changes, our goals, assumptions and
methods, and this random time is just as good as any.

> If we take "FOSS" as a means to achieve some benefits (freedom from
> lock-in,
> > privacy, control) and not a goal in itself, the situation has changed to
> > such an extent that FOSS is not even close to giving the average desktop
> > user the expected benefits.
>
> Well, according to what you said above you can't even expect FOSS to
> give advantages where it isn't even being used.
>

Yes; but the historical assumption was that building a *local* full-stack
FOSS OS is sufficient / the right goal, and that assumption is false.

> Over the past 10 years, even those of us only installing FOSS have ended
> up
> > running an enormous amount of proprietary software.  That's, in
> retrospect,
>
> Perhaps s/running/using indirectly.


Running as well.  All the JavaScript, Java (with x86 binaries inside .jars,
even), Flash.  And also using indirectly, yes; Google Search is more
valuable to many people than quite a few locally-running programs.


> > a completely unintuitive, unexpected and undesired result[3], and keeping
> > exactly the same means to achieve the desired benefits (again, freedom
> from
> > lock-in, privacy, control) seems like sheer folly to me.
>
> I disagree with your conclusion here.  Or maybe I misunderstand what
> you're trying to say.  How is it unexpected or unintuitive that using
> a web service like gmail or twitter or facebook means you are now
> beholden to that service for that data?


Sure, these are expected results *when you ask this question*.

But from the wider FOSS community, or the Fedora community, is *not* asking
this question.  We take it for granted that attracting more contributors to
make more or better FOSS software to be run locally is what we should be
doing.

And so we have painful discussions about how important it is to have, only
have, allow, encourage, etc., FOSS software run locally, but *that doesn't
make as much difference to the benefits we are promising*.

If you critically examine most of the rhetoric and supposed benefits of
Open Source or Free Software, they are significantly limited in the world
of network effects and web services.

   - Better quality, higher reliability?  Both are increasingly up to the
   cloud host, or the connection, not up to software running locally.
   - More flexibility?  No, you can use exactly the same websites as
   everyone else.
   - Lower cost?  Well, yes, if you don't count being subject to
   advertising.
   - Full control over your data?  No, much of the data is on the internet
   somewhere.
   - Freedom from lock-in?  No; the ability to change a web browser or a
   .doc editor is trivial compared to the inability to escape Facebook.
   - Freedom to redistribute?  Only the invisible parts up to the web
   browser; not the services one is actually interacting with.
   - Freedom to modify? Same.

Is any of this "unexpected or unintuitive"?  I suppose not, but we arguing
about a FOSS desktop as if these limitations didn't exist or matter.

Claiming "because I use a FOSS desktop I should have all the freedoms
> and benefits of FOSS EVERYWHERE" is entirely disingenuous.


Claiming that pure FOSS is important because it will give users these
benefits is equally disingenuous.  For end-user desktops, AFAICS the
binding constraints for most of the claimed benefits of FOSS are no longer
significantly affected by locally running FOSS software exclusively

The Foundations aren't the problem.
>

Well, we could resolve the disconnect between expected benefits and the
methods / project identity in several ways.

   - We could focus on the expected benefits, and start looking for new
   ways to make a difference for these benefits for everyone.  In that case
   the Foundations would have to be modified or at least significantly
   extended.
   - We can somehow scale back the expected benefits, e.g. say that they
   only apply to servers, client-server pairs, or ~programmers.
   - We can just say that the Foundations are fine and we are doing this
   because we enjoy it (or because there are *other* benefits that we
   actually value, perhaps the ability to poke into things, learn from them,
   and hack them for custom uses), and only stop advertising some of the
   benefits.

I think what you're driving at is somewhat of a content issue, and
> we've never had a great story around content or services.  It's a very
> difficult nut to crack and we as a community are very behind.
>

It's not only "content", it's equally "software available over the
network".  Compare with
http://arstechnica.com/information-technology/2014/02/neither-microsoft-nokia-nor-anyone-else-should-fork-android-its-unforkable/?comments=1&post=26199423.
     Mirek
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.fedoraproject.org/pipermail/advisory-board/attachments/20140422/89430a6e/attachment-0001.html>