3.9 feature voting round

> stéphane ducasse wrote:
> ....
>> I do not understand why you want to oppose traits and tweak.
>> There is no such an opposition, after this is clear that there  
>> will  be some work (but the problem is not with traits)
>> to remove Morphic and use Tweak but we already said to andreas  
>> long  time ago that we are ready to help.
>> Apparently as usual you are talking but without knowing.
>
> Stephane,
>
> the last thing Andreas said regarding 3.9 with Traits in relation  
> to Tweak was, that because of a series of errors he ran into caused  
> by the class/metaclass changes by Traits and given the pain he  
> experienced while porting Tweak to 3.8 he didn't want to  
> investigate these errors further, and he didn't even mention making  
> another port of Tweak to 3.9.
>
> Moreover he gave an interesting and *skeptical or cautious*  
> overview about what Traits is and what it is not in his opinion.
>
> lists.squeakfoundation.org/pipermail/squeak-dev/2006-January/
> 099311.html
>
> You responded to this post but as usual you ignored its  
> implications. In fact you keep on ignoring Andreas' complaints  
> about what you stuff into the update stream since a long time.  
> Moreover, you try to turn down every skeptical opinion about  
> Traits. At least you ignore it.
>
> But, you could proof me wrong, here and now, and I will withdraw my  
> assertion that you behave ignorantly. Just agree to the idea that  
> Squeak makes from now on a minimalistic release Squeak-Core, with  
> Squeak-Traits and Squeak-Tweak (and perhaps others) on top of it.

> Cees De Groot wrote:
> ....
>
>> - Advise, I repeat: advise, the 3.9a team. They're entirely free to
>> completely ignore the outcome of this stuff, although if many many
>> many people vote for something and they don't do it, it'd be nice if
>> they'd give a reason;
>
>
> A few people play leaders and they feel free to completely ignore the
> complete community. You described exactly the sad state of Squeak today.

> And there have been skeptical opinions from
> important persons of this
> community which is on the brink to break into two
> parts right now. One
> idea to solve this problem was to have a *small*
> core Squeak and
> building on that special distributions or "flavors".
> This way a
> contradiction between Traits and Tweak, which
> apparently exists - be it
> temporarily, practically or whatsoever, could be
> circumvented. We could
> have Squeak-Core and on top of that Squeak-Traits
> and Squeak-Tweak,
> which of course does not exclude a merge or
> redistribution at a later time.

>
>
> Martin Wirblat a écrit :
>> Cees De Groot wrote:
>> ....
>>
>>> - Advise, I repeat: advise, the 3.9a team. They're entirely free to
>>> completely ignore the outcome of this stuff, although if many many
>>> many people vote for something and they don't do it, it'd be nice if
>>> they'd give a reason;
>>
>>
>> A few people play leaders and they feel free to completely ignore the
>> complete community. You described exactly the sad state of Squeak  
>> today.
>

> Marcus Denker wrote:
>> How would a better state look like?
>
> One of the things that I'd really like to see is a better (as in:  
> more transparent and better documented) decision making process. I  
> think that this is at the heart of what upsets many people  
> (including myself) - the fact that nobody really knows who has  
> decided to do X or Y and by whose authority.
>
> Take 3.9 for example - to me, it's not surprising that some actions  
> are perceived as somebody driving their personal agenda if even the  
> big ticket items on the 3.9 TODO list have no clear indication how  
> they ended up there and who decided by which means (other than  
> screaming the loudest) that they'd get into that release. (the same  
> is true for any past-SqC release btw - I am simply choosing 3.9  
> because it's the current one and -for me- one of the most confusing  
> in this regard)
>
> And that of course is part of the state of the community right now  
> - namely that the processes by which these decisions happen are  
> complete mysteries even to people who like to think of themselves  
> (perhaps wrongly) as part of the "initiated" (like myself). I can  
> honestly say that I have no idea whatsoever what the decision  
> processes look like at this point.
>
> Cheers,
>   - Andreas
>

> stéphane ducasse wrote:
>
>> The same that was in 3.6, 3.7, 3.8.
>
>
> Which, like I said, is equally bad in this regard.
>
>> We proposed a list of points that we would work on and waited for  
>> feedback.
>
>
> Sure. I didn't say there was no feedback, I said there was no
> transparent decision process. Or put differently, once you had that
> feedback, who made the decision to include those points and not others
> in the 3.9a list? To choose X over Y? Based on which authority did that
> person or group make the decision? And (arguably) most importantly (at
> least to avoid endless discussions) is there any kind of documentation
> (links to email threads for example) about the discussion and the decision?
>
> This whole discussion started by an accusation that soms people drive
> only their personal agenda - and the best to contradict this is to point
> to the decision process and say "look, here is where discussed this,
> here is vote/decision process, this is the result and it's explicity
> documented here for people just like you who don't read Squeak-dev daily
> and have not been following the discussion closely".
>
>> At least contrary to before, the list was explicit.
>
>
> The previous versions were actually no less explicit if you go over the
> mailing list archives but they were not as well documented so I see that
> as a definitive improvement.
>
> Cheers,
>   - Andreas
>

> Hi everyone. I personally think that Andreas Raab has a completely  
> legitimate complaint. I don't think the community has the tools to  
> make transparent decisions in a way that is participative, so the  
> best that can happen at the moment is a mailing list discussion  
> that sort of peters out and then someone decides what happened.
>
> These things take a lot of time, and in the end, the decision is  
> not really representative. I think this outcome is characteristic  
> of email - it is hard to do much better in this medium.
>
> In the context of the coming stepping down of the (self selected)  
> SqF Board, an election team has been formed, and a few of us think  
> that the first ingredient we need to be a community that makes  
> represented decisions, is a voting system. This consists of some  
> software (website that allows issues and alternative solutions to  
> be registered), some rules (who gets to vote?), and mostly the  
> participation of the community.
>
> Our current proposal for the voting software is at
> http://minnow.cc.gatech.edu/squeak/5835
>
> We are looking for comment on these requirements, and for someone  
> with web development skills to help implement.
>
> There are existing systems that provide part of the functionality,  
> for example CIVS at
> http://www.cs.cornell.edu/andru/civs.html
>
> The reason to roll our own solution here is that we want to  
> automate the whole process, in order to make it possible for anyone  
> to initiate a vote, with no hidden parts.
>
> About eligibility to vote, this is a more contentious issue I am  
> not addressing here, except to mention that we want to be able to  
> represent the wide Squeak community, and that we generally think  
> SqP is not a bad mechanism for registering voters.
>
> Daniel
> PS - the coming elections will probably be performed in CIVS, for  
> lack of a better system, but someone else will write more about  
> that, the next election is not my main concern.
>
>
> Andreas Raab wrote:
>> stéphane ducasse wrote:
>>> The same that was in 3.6, 3.7, 3.8.
>> Which, like I said, is equally bad in this regard.
>>> We proposed a list of points that we would work on and waited  
>>> for  feedback.
>> Sure. I didn't say there was no feedback, I said there was no  
>> transparent decision process. Or put differently, once you had  
>> that feedback, who made the decision to include those points and  
>> not others in the 3.9a list? To choose X over Y? Based on which  
>> authority did that person or group make the decision? And  
>> (arguably) most importantly (at least to avoid endless  
>> discussions) is there any kind of documentation (links to email  
>> threads for example) about the discussion and the decision?
>> This whole discussion started by an accusation that soms people  
>> drive only their personal agenda - and the best to contradict this  
>> is to point to the decision process and say "look, here is where  
>> discussed this, here is vote/decision process, this is the result  
>> and it's explicity documented here for people just like you who  
>> don't read Squeak-dev daily and have not been following the  
>> discussion closely".
>>> At least contrary to before, the list was explicit.
>> The previous versions were actually no less explicit if you go  
>> over the mailing list archives but they were not as well  
>> documented so I see that as a definitive improvement.
>> Cheers,
>>   - Andreas
>

>>> Anyway, I suggested last week on the SqF board list that we find  
>>> a 4.0
>>> team lead - 3.9a is approaching beta status, and I hope the beta
>>> period will be short so 4.0 development can start soon.
>>
>> '3.9' asVersion next ==> '3.10' asVersion
> Exactly. The 4.0 appellation needs to be reserved for a jump to a  
> much cleaner system that deserves a new release of the sources file.
>
> For a very long time we've been talking (waffling?) about making a  
> break and adopting a new compiled method format, cutting out a lot  
> of ancient backwards compatible dross, ... oh, lots. 4.0 really  
> ought to be that.

> Hi everyone. I personally think that Andreas Raab has a completely
> legitimate complaint. I don't think the community has the tools to make
> transparent decisions in a way that is participative, so the best that
> can happen at the moment is a mailing list discussion that sort of
> peters out and then someone decides what happened.
>
> These things take a lot of time, and in the end, the decision is not
> really representative. I think this outcome is characteristic of email -
> it is hard to do much better in this medium.
>
> In the context of the coming stepping down of the (self selected) SqF
> Board, an election team has been formed, and a few of us think that the
> first ingredient we need to be a community that makes represented
> decisions, is a voting system. This consists of some software (website
> that allows issues and alternative solutions to be registered), some
> rules (who gets to vote?), and mostly the participation of the community.

>
> Our current proposal for the voting software is at
> http://minnow.cc.gatech.edu/squeak/5835
>
> We are looking for comment on these requirements, and for someone with
> web development skills to help implement.
>
> There are existing systems that provide part of the functionality, for
> example CIVS at
> http://www.cs.cornell.edu/andru/civs.html
>
> The reason to roll our own solution here is that we want to automate the
> whole process, in order to make it possible for anyone to initiate a
> vote, with no hidden parts.
>
> About eligibility to vote, this is a more contentious issue I am not
> addressing here, except to mention that we want to be able to represent
> the wide Squeak community, and that we generally think SqP is not a bad
> mechanism for registering voters.
>
> Daniel
> PS - the coming elections will probably be performed in CIVS, for lack
> of a better system, but someone else will write more about that, the
> next election is not my main concern.
>
>
> Andreas Raab wrote:
>
>> stéphane ducasse wrote:
>>
>>> The same that was in 3.6, 3.7, 3.8.
>>
>>
>>
>> Which, like I said, is equally bad in this regard.
>>
>>> We proposed a list of points that we would work on and waited for  
>>> feedback.
>>
>>
>>
>> Sure. I didn't say there was no feedback, I said there was no
>> transparent decision process. Or put differently, once you had that
>> feedback, who made the decision to include those points and not others
>> in the 3.9a list? To choose X over Y? Based on which authority did
>> that person or group make the decision? And (arguably) most
>> importantly (at least to avoid endless discussions) is there any kind
>> of documentation (links to email threads for example) about the
>> discussion and the decision?
>>
>> This whole discussion started by an accusation that soms people drive
>> only their personal agenda - and the best to contradict this is to
>> point to the decision process and say "look, here is where discussed
>> this, here is vote/decision process, this is the result and it's
>> explicity documented here for people just like you who don't read
>> Squeak-dev daily and have not been following the discussion closely".
>>
>>> At least contrary to before, the list was explicit.
>>
>>
>>
>> The previous versions were actually no less explicit if you go over
>> the mailing list archives but they were not as well documented so I
>> see that as a definitive improvement.
>>
>> Cheers,
>>   - Andreas
>>
>

> I still have that old "voting" software I once wrote as a
> proof-of-concept of what we could do with SqP's web of trust. I
> thought it'd be neat to do a round of voting about what should land in
> v3.9, now that the 3.9a team is considering going to beta.
>
> The team also thought it'd be need to get some structured input, so I
> cleaned up the database and it's now available for voting.
>
> The idea is simple:
> - Go to http://de-1.tric.nl/seaside/sqp/list
> - Login with your Squeak People account
> - Create an issue or vote on issues.
> You will see the number of votes you have constantly updated. Note: I
> haven't yet added the possibility to delete issues, and if you add an
> issue, you have to allocate at least one vote to it. So think before
> creating issues! In fact, you can't even edit an issue - which is
> logical, otherwise you could rig the voting (editing an issue should
> logically imply losing all votes that were on it before the issue).
>
> The SqP angle is that the higher your ranking, the more votes you get
> - Observers get none, Apprentice gets 5, Journeyer 10, Master 15. This
> is an entirely unfounded distribution and I'm not going to debate it
> here ;-).
>
> Note: this voting round serves two purposes:
> - Advise, I repeat: advise, the 3.9a team. They're entirely free to
> completely ignore the outcome of this stuff, although if many many
> many people vote for something and they don't do it, it'd be nice if
> they'd give a reason;
> - Check whether this sort of utility is useful.
>
> The software can be found in the SqueakPeople repository on  
> SqueakSource.
>
> Have fun!
>
> Cees
>