[Freedombox-discuss] Report third hackfest

Jonas Smedegaard dr at jones.dk
Tue Mar 8 18:09:10 UTC 2011


On Tue, Mar 08, 2011 at 11:36:11AM -0500, Les Orchard wrote:
>On 3/8/11 7:40 AM, Jonas Smedegaard wrote:
>>It is obvious to me, however, that there won't be one revision of 
>>FreedomBox - multiflavored or not. There will be corrections, security 
>>updates, and feature updates. And before all that there will be 
>>development drafts not at all usable for consumption by our target 
>>userbase.
>Maybe FreedomBox should have its own unstable / stable distros? Not 
>forks so much as overlays of packages on existing Debian distros. Just 
>thinking conceptually here. I'm not clued on how these things work, 
>though, so I don't really know what I'm talking about.

That may speed up some kinds of development, yes.

In my experience, esp. with Skolelinux and Sugarlabs, such creation of 
an own distribution framework - even if intended to sync tightly with 
another "real" distribution - ends up stealing too much time, focus, and 
is thus more damage than good IMO.

I maintain a pile of unofficial packages at debian.jones.dk (more than 
1000 packages last I counted them).  I intent to backport some(!) 
packages there from Debian unstable to be usable with Debian stable.

I make no promise to backport everything that everyone wants, and the 
things I do backport I do not promise to maintain for eternity.  I 
consider it a development aid, not a resource of production quality!

I backport only packages already (in the process of becoming) part of 
Debian officially.

I deliberately do *not* sign my archive, as I want it very clear that 
even though I have this cool label of being "official Debian Developer" 
that is *not* a promise that whatever crap I throw together magically is 
of the same wuality as Debian: Debian takes _years_ to mature!


>>If you want a FreedomBox production-ready in a month, then don't use 
>>Sid. And don't use Pagekite or NetSukuku or WebID or P2PSIP or 
>>P2P-DNS. Avoid _all_ the new cool stuff - use _only_ boring oldschool 
>>stuff!
>Some of the new stuff like Pagekite seems interesting to have in a 
>FreedomBox. An unstable FreedomBox distro could have PageKite—maybe 
>even before Debian unstable—but stable FreedomBox would only have it 
>once it's gone through the same release process any other Debian 
>package does

Good luck with that.

It is no doubt great fun and a good learning experience to try setup and 
maintain a distribution - just mind you that this is what it is!

I strongly, very very strongly, recommend to use Debian: get all code 
packaged rather than seek ways to circumvent that,

Feel free to play with upstream code e.g. to figure out it suitable at 
all to waste any more time on packaging and integrating with FreedomBox, 
but throw away such premature testing - don't try encapsulate it at 
stuff it into FreedomBox - that is A Bad Way!

When packaged and in Debian, it makes sense to try backport it also for 
Debian stable, e.g. for easier and more widespread testing by fellow 
developers (and crazy non-developers not caring about reliability).



>>Personally I believe that Semantic Web, and graphing math applied to 
>>it, was the key to the success of Google and Facebook, and can be the 
>>key to the success of decentralized tools as well. So I spend/waste 
>>time on what I believe to be "the next cool thing" - WebID. And I 
>>contribute the way I am good at: by packaging already invented and 
>>already coded pieces for Debian, and have it included into Debian 
>>officially.
>I need to spend some time wrapping my head around WebID, but that 
>contribution sounds great to me. Anything new that doesn't yet have a 
>Debian package, yet could contribute to a FreedomBox, should get 
>wrapped up in a package and go through the usual Debian process for 
>inclusion.

Congratulations!

I was wondering when someone used my own praching against me.  You did 
it! :-)


Yes, I totally agree: WebID needs to be packaged too before it is 
sensible to consider it concretely for inclusion into FreedomBox.

I poudly announced last sunday (a little too soon, as they are still 
waiting in the NEW queue so not truly there yet) that it was packaged, 
but tests same day demonstrated that my initial packaging if not crap 
then at some other place on the scale _below_ the "production ready" 
marker.  Those packages need some work yet - if nothing else then at 
least documentation on how it is used properly.

Then _another_ package needs to be created, containing a working WebID 
CGI script and surrounding things like beautification and (debconf 
preseedable) configuration.

So all in all this is very much a good demonstration of how things are 
_not_ ready when they themselves are ready.

Similarly I will expect Pagekite and all other wonderful yet unpackaged 
technologies to take _more_ time to be usable for FreedomBox even 
_after_ being production quality as upstream projects and _after_ being 
officially packaged into Debian.


>Maybe FreedomBox needs its own unstable sandbox distro where new 
>stuff gets packaged and played with until it makes it into Debian 
>proper? Again, not a fork, fully intending to flow into Debian proper 
>once the kinks are worked out.

He he, at least you have the curtesy of calling it a distro.  Sugarlabs 
kept saying for a looong time that their "Sugar on a Stick" was not a 
distro.

No, you cannot persuade me to call that a wise approach.  But I am not 
your daddy telling you what to do - good luck trying if you must.


>>You can try put a system together *today* containing WebID. That will 
>>be a system built from an _unstable_ distribution with _unofficial_ 
>>parts. Which means highly risky to release to others due to e.g. no 
>>guaranteed upgrade path or security bugfixes provided.
>That sounds like an unstable FreedomBox distro to me.

To me Debian is the distro, not FreedomBox.

To me FreedomBox is a distro-without-own-infrastructure-as-such - i.e. a 
Debian Pure Blend.  We have plenty of challenges already dealing with 
issues specific to FreedomBox - we should *not* also take upon us the 
issues generic to all distros.


>>You can try put a system together *tomorrow* containing WebID. That 
>>will be a system built from an _stable_ distribution with _unofficial_ 
>>parts. Which means somewhat risky to release to others, because key 
>>parts only potentially has upgrade path and security bugfixes 
>>provided.
>That sounds like a stable FreedomBox distro to me - eg. Debian stable, 
>just with preset package list.

This is the stage Skolelinux is at now.

In order to maintain that stage, they run an own infrastructure with 
translation tools, bugtracker and DVD build routines.

A german Skolelinux-based project set out to do some things different. 
Now, 3 years later, that german project stopped after municipal funding 
ran out, and their collected bugreports are - as far as I am aware - 
lost due to lack of man power and/or interest in passing bugs upstream 
to Skolelinux and/or Debian.

No, I don't find it sensible to use that approach.



>>You can try put a system together *today* containing Pagekite or 
>>NetSukuku or [your favorite tool here]. That will be a system built 
>>from (possibly a stable) distribution with _non-packaged_ parts. Again 
>>risky to release to others because all or some parts lac upgrade path 
>>and security bugfixes. Also more difficult for peer Freedom fighters 
>>(and possibly difficult for yourself too) to reliably replicate (i.e. 
>>not copy result but mimic the process) due to those non-packaged 
>>parts.
>Why non-packaged? If there's something that should be in the FreedomBox 
>(even if unstable), shouldn't packaging it up be one of the first 
>steps?

Sure, I wholeheartedly agree.  But that wouldn't be "today", then :-)

It just seems from conversations on this list that people insist on 
hacking without waiting for Debian packaging.  So I try describe how 
that, even if easier and faster at first, hurts long-term.

But heck - I am biased, being a Debian Developer.  So it might be that 
those wanting to bypass Debian with their pet projects do not concider 
my opinion relevant either :-P


>>Finally you can try put a system together *today* from only Debian 
>>stable (Squeeze). That will be a system built from an _stable_ 
>>distribution. Which means sensible to release to others, because all 
>>parts has upgrade path and a dedicated team provides security 
>>bugfixes.
>>
>>Obviously just making an image of Debian stable is too crude to 
>>senisbly call a FreedomBox 1.0. Lots of _other_ tasks you can do 
>>*today* other than imaging - this was just reflections on Sunday work 
>>on WebID.
>>
>>...or you can wait for Godot. Or Eben. Or a pile of money. Maybe 
>>the World is easier to fix tomorrow :-P
>Seems like getting *something* stable running with Debian on a plug 
>computer (or a VM pretending to be one) might be a good first step 
>toward bootstrap. It wouldn't be FreedomBox 1.0, but it would be a 
>stable base to get everyone on the same page and start throwing 
>unstable things at.

All of the above is doable in parallel.  What each of us find most 
relevant depend on whether we are ok working alone, and for how long.

I am ok working alone at first, then share my work with you guys, then 
work some more (with you or alone, depending on how well our different 
works fit together), and finally lean back as others (read: Debian 
infrastructure) keep my work alive for eternity.

Might be that my work isn't in the end called FreedomBox.  But it is 
Free, and it is part of Debian so others can use it too when next 
release of Debian comes out.


That final point above is also my point here: 
http://dr.jones.dk/fb/year/


  - Jonas

-- 
  * Jonas Smedegaard - idealist & Internet-arkitekt
  * Tlf.: +45 40843136  Website: http://dr.jones.dk/

  [x] quote me freely  [ ] ask before reusing  [ ] keep private
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.alioth.debian.org/pipermail/freedombox-discuss/attachments/20110308/1afc5d40/attachment.pgp>


More information about the Freedombox-discuss mailing list