[dojo-contributors] Google Groups management

Dustin Machi dmachi at dojotoolkit.org
Mon May 11 18:36:10 EDT 2009


On May 11, 2009, at 5:23 PM, Adam Peller wrote:

> We'll probably have to agree to disagree on this, but my take is that
> we should not try to bend another solution to our will; we should just
> find something good and use it.  We need to get out of the business of
> building our own solutions, except where we have to.   Hopefully we
> will find something long-lived and not end up with another jot (that,
> in my mind is the biggest risk, but one that is relatively low with
> Google)

We have never built a solution except for the documentation system, so  
I dont' really know what you mean.  By bend it to our will, I only  
mean that needs to be flexible or complete enough to meet our needs  
and be configured within our environment.   Porting to google is  
minimally less work than any of the other solutions for us.  The only  
thing that we get to do is skip deciding what features we want by  
having that decision forced upon us by an external entity.   For  
giving up control, we do eliminate the need to ever have to make  
decisions about how we'll interact with the community, Google can  
decide that and what the interface to our community looks like.  On  
the other hand, the same could be said of an internal solution.  The  
internal solution is not trying to solve a performance problem as  
people still seem to think from before (it wasn't then either, but  
that was the focus), there should be plenty of packages that can meet  
the requirements (at least to the same extent that google groups does,  
which is itself not an exact match either).


> As for old content, I still recommend we just freeze what we have
> and/or do a one-time export, if that's doable.  Just having it indexed
> somewhere is good enough, in my view.  I don't think the old content
> has to be migrated into the system we end up using.  We need to be
> focused on the future.

Then that solution is perfectly valid for either internal  or external  
migration, so neither solution provides an advantage there, though the  
work still needs to be figured out and completed.

Google buys us already integrated email and thats about it. To me  
thats not worth it.  Remaining -1 on it.

I've already stated my opinion on this and don't really want to argue  
it again, so if there is a vote to go this route and thats what is  
decided then that is fine.  We'll still need people to do the work.   
As I recall, it was by no means clear that people wanted to move to  
google.   It is was split, http://thread.gmane.org/gmane.comp.web.dojo.devel/9466/focus=9504 
  with a slight edge to -1 to google group.

This doesn't mean I think it should stay as it is, it just means that  
the primary reason for this before was eliminated previously, and now  
the other reasons (the interface/usability) are taking the top bill  
and it should be looked at in that light.  We can also compromise in  
other areas than internal/external.  A simple compromise would be on  
two-way email.  If this weren't a requirement, at least at first, then  
package selection is easier.  The google groups solution is already  
itself a compromise on the interface from what people's requirements  
were previously (not to mention the complete loss of ability to decide  
for ourselves where we want to go with it).  From a split community  
perspective, this doesn't do anything more than simply eliminating  
dojo-interest (aside from requiring users to have accounts in both  
systems), so any solution that gets rid of that meets that need.

Anyway, I'm pretty much done stating my opinion on it, I think people  
know where I stand on it so I'll shutup for the rest of this thread  
now.  I'll be glad when is done and out of the way :)

Dustin

> -Adam
>
> On Mon, May 11, 2009 at 3:31 PM, Dustin Machi  
> <dmachi at dojotoolkit.org> wrote:
>> I've stated my opinion on this previously and it remains  
>> unchanged.  I
>> do not think that moving this stuff outside of our infrastructure  
>> will
>> do anything to ease the maintenance of our community.  I've stated in
>> the previous thread that the ongoing maint is not the primary issue,
>> it is migration where the time is needed and this exists in either  
>> the
>> case of external software or internal software. In the case of
>> external software there is either a need to import current data into
>> the new system or continue to maintain the existing system, which  
>> both
>> complicates searching of the forums as well as eliminates the ongoing
>> maintenance argument.
>>
>> The primary impetus of the previous discussion was due to performance
>> problems that existed at the time and those problems were solved.
>> While the conversation devolved into everyone's wishlist of features
>> for forums, it was secondary to the initial problem.  I agree that we
>> should migrate to something else since noone is willing to put effort
>> into it, but i fail to see why it needs to be done externally.  I
>> don't see how it can be though of as simplifying maint by moving it
>> externally.  It simply makes it more difficult to integrate and bend
>> to our will.  The migration and maint tasks exist in either case and
>> community maint is made more difficult.
>>
>> Futhermore, it is a complete misinterpretation to think that it was
>> previously implied that 'django would solve all our problems'.  We
>> simply said we would rather have standalone packages for the
>> individual components and I continue to believe this should be the
>> approach that we take.  The reference to django was only preference  
>> to
>> solutions that were written with or used django would be weighted as
>> we have active developers and contributors who are familiar with it
>> and could make modifications and updates as necessary.  We are of
>> course free to choose a forum (or any service solutions for that
>> matter) that is composed of something else, but the trade off comes
>> from the number of people willing to learn and work on it...the
>> primary drupal problem.
>>
>> So while I think its the wrong decision to move everything offsite  
>> and
>> I think we will regret doing so in the future, we can proceed with
>> that option, but will still need to find people to do the migrations
>> and maint.  There should be at least one primary Ops person per
>> foundation project who is the primary ops/support person for that
>> project. I'll then support these people from the foundation  
>> perspective.
>>
>> -1 offsite apps, +1 getting rid of dojo-interest,
>> Dustin
>>
>> On May 11, 2009, at 1:26 AM, James Burke wrote:
>>
>>> On Sun, May 10, 2009 at 11:56 AM, Peter E Higgins <dante at dojotoolkit.org
>>>> wrote:
>>>> And not that I want to rehash this topic, but I am getting
>>>> increasingly
>>>> disgusted by the separation. Half the questions that _could_ be
>>>> easily
>>>> fielded by the committers are going to the dojo-interest group,  
>>>> which
>>>> seems VERY FEW of the committers read or respond to. The other half
>>>> go
>>>> unanswered in the forums. I only assume few committers read the
>>>> forums
>>>> because of the horrible horrible usability of the forum software. I
>>>> assume because I've done nothing but erase spam messages from
>>>> topics for
>>>> the last two weeks, I've had to plain turn off commenting on some
>>>> of the
>>>> current front-page blogs because I have been unable to keep up, and
>>>> no
>>>> one seems to be helping. Even worse, Drupal's crap usability
>>>> prevents me
>>>> from erasing spam comments _while_ comments are disabled or "read
>>>> only",
>>>> so in some cases new spam is coming in while I'm attempting to  
>>>> erase.
>>>> I've yet to find a way to bulk erase comments.
>>>
>>> If it is any consolation, I also spent a good amount of time  
>>> starting
>>> last weekend removing spam and marking some posts as read only.
>>> Annoying it has gotten so bad lately.
>>>
>>> I favor using Google Groups or stackoverflow.com for the forum-like
>>> help services. stackoverflow.com might be better at dealing with the
>>> spam vs Google Groups.
>>>
>>> It seems best to move as much infrastructure to services that are
>>> better staffed/focused to deal with these issues. As Peter points  
>>> out,
>>> no one seems to have an interest/time in touching the drupal config
>>> (excluding Dustin's efforts to just keep it running). I am also  
>>> bummed
>>> about the planet issue.
>>>
>>> James
>>> _______________________________________________
>>> dojo-contributors mailing list
>>> dojo-contributors at mail.dojotoolkit.org
>>> http://mail.dojotoolkit.org/mailman/listinfo/dojo-contributors
>>
>> _______________________________________________
>> dojo-contributors mailing list
>> dojo-contributors at mail.dojotoolkit.org
>> http://mail.dojotoolkit.org/mailman/listinfo/dojo-contributors
>>
> _______________________________________________
> dojo-contributors mailing list
> dojo-contributors at mail.dojotoolkit.org
> http://mail.dojotoolkit.org/mailman/listinfo/dojo-contributors



More information about the dojo-contributors mailing list