[Freedombox-discuss] need for concensus or not

Sat Oct 9 16:43:25 UTC 2010

On Sat, Oct 09, 2010 at 04:41:06PM +0200, bertagaz at ptitcanardnoir.org 
wrote:
>This is my last mail in this subject, if I'm the only one to push some 
>kind of organization, I'll stop bother everyone with this. But...

I sure hope you do not keep that promise!

>On Sat, Oct 09, 2010 at 03:23:23PM +0200, Jonas Smedegaard wrote:
>> If we were hired in a company to reach a common goal, then yes, it 
>> really would make sense for us to agree ahead on what that common 
>> goal was.
>
>Well, compare with the debian project : it's not a company, but its dev 
>do have common goals.

Do we? What are those? Seriously.

My impression is we have common vision, but not common goals - and that 
this is a core principle for us and a reason our community is so alive.  

>So if we continue on comparing with the debian project, you would think 
>that like they have tons of documents describing their common goals, 
>and you have to read them before applying, hackers don't join the 
>project?

Ah, you mean the requirements of becoming a Debian Developer?

Well, Debian has many many more direct contributors than the Debian 
Developers - that is only the ones with direct access to the packaging 
infrastructure, and that needs consistency.

Yes, FreedomBox needs consistency too - but does it need additional 
consistency beyond the Debian framework which is its foundation?

Debian do not have consistency in choices like configfile format (like 
e.g. DD-WRT), or desktop (like e.g. Ubuntu).

If we were to code a system, like DD-WRT does, then we would be crazy 
not to coordinate our coding style and have a high level of consistency.  
I just don't see our project as coding, but as distro customization.

>> Specifically about hack vs. debconf (or sysadmin "dirty" hacks vs. 
>> distro style "clean" hacks, as I feel more accurately frame it), I 
>> believe I succeeded in raising my concerns (unlike some of my 
>> previous attempts at other teams). If some of us still favor 
>> sysadmin-style dirty hacks on top of Debian instead of distro-style 
>> clean hacks integrated with Debian, then let them do that - the rest 
>> of us can cherry-pick from their work in our slower pace. :-)
>
>But I'm not sure to see how we'll be able to merge and maintain a 
>project with different hack types. Sounds like a hell of a work for the 
>debian people here to try to maintain that.

Heh. I can't help you explain the sanity of sysadmin style hacking, as I 
do not fancy it myself. :-)

My point of the above is that there are room for both hacking styles.  
True, sysadmin style hacks cannot be consumed as-is in official 
packages, but are still valuable even if needing more polishing later.

>Actually I'm not telling we have to decide *everything*, nor make those 
>decisions unchangeable. But it seems to me that some of the direction 
>of this project have to be stated. Like what type of configuration 
>method it would adopt. Actually I'm not sure how to begin to dev. I see 
>a lot of people joining the list, wanting to contribute, but the only 
>thing they can do is to dig into the mailing list archive, which won't 
>bring them the beginning of an answer. So noone starts nothing.

I have noticed several remarks about lack of time to dig in.

I have also noticed remarks on wondering where to start. But also clear 
suggestions - e.g. by Lars - on concrete things to do which noone picked 
up yet.

What this project will "adopt" is what its members adopt!

Example: You sit down now and write a bunch of INI-style configuration 
files for bootstrapping ipv6 using quagga.  It works for you.

You then post your success on this list, but it then turns out that some 
don't like quagga, and some don't like INI files.  So you might end up 
having to maintain your work yourself, or try restructuring it to please 
more of your peers in this team.

You want the safety of that not happening.  Well, feel free to do it 
upside down, then: describe on this list that you want to work on ipv6 
and want input on which config style others consider sane for that, and 
which routing/tunneling/whatever mechanism to use, etc.

But why discuss how ipv6 should be configured before anyone have shown 
interest in working on that area?

When noone picked up some of the tasks mentioned on this list already, I 
dare say that it isn't because of fear their contribution won't be 
"adopted", but because of lack of hacking time, and because of personal 
interests in which areas to work in.

>I'm not sure the "let's do evything like everyone wants" will bring any 
>cohesion in this project, nor help people getting involved, finding a 
>place, etc...

What "people" you are referring to?

Do you need some concrete tasks to do yourself?

Or are you talking on behalf of others than yourself?

>But that's my opinion, and if I'm the only one to push that, I'll shut 
>up :)

Please do not shut up!  Your points are relevant, even if you might be 
the only one stating them.

  - 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/20101009/733c3614/attachment.pgp>