couchdb-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Damien Katz <>
Subject Re: Transactional _bulk_docs
Date Fri, 06 Feb 2009 18:44:03 GMT

On Feb 6, 2009, at 1:20 PM, Ted Leung wrote:

> On Feb 6, 2009, at 8:04 AM, Damien Katz wrote:
>>> This may hint at one of my biggest concerns here, the  
>>> balkanization of the PMC from the rest of the community.  I don't  
>>> think I've ever seen a project where the dividing line between the  
>>> PMC and the rest of the community was so often and brightly drawn.
>> Please tell me more about this brightly drawn line. We should get  
>> rid of it.
> I don't know if this is what Geir means, but there is a lot of use  
> of the letters PMC in messages on the list.   Much more than usual,  
> in my experience.    That may sound like a small thing, but that  
> kind of small thing, repeated over time,  can lead to a highly  
> stratified community, which is undesirable.
>>> There's a comment later in this thread where Chris describes his  
>>> role as ... well, for lack of better words, your handler, to  
>>> shield you from the community.  Things like that feed my  
>>> perception.  More on that later.
>>> I have an idea :
>>> 1) Stop coding.
>> No. The changes I am making need to be made regardless. They  
>> encompass much more than just the transaction change.
> Is there a way to commit some of the unrelated changes over time?    
> I find that it's easier to understand a series of  small patches as  
> opposed to large ones.
>> Then why are you being so combative and insulting to our team?
> Damien, I am sorry that this has ended up at such a contentious place.
>> The only complaint is that I didn't run some stuff over the mailing  
>> list before writing code.
> I think that part of why we are having contention is wrapped up in  
> the sentence above.   When I read this sentence, I feel that you are  
> being dismissive of the use of mailing lists.   When I originally  
> chimed in on this thread it was in response to your statement that  
> the decision was made in IRC:
> On Feb 4, 2009, at 11:11 PM, Damien Katz wrote:
>> Ideally yes, but real time communication with everyone together is  
>> damn useful.
>> -Damien
>> On Feb 5, 2009, at 2:07 AM, Ted Leung wrote:
>>> Uh, project decisions are supposed to be made in the public  
>>> mailing lists...
>>> Ted
>>> On Feb 4, 2009, at 6:51 PM, Damien Katz wrote:
>>>> This decision was discussed and made on IRC.
>>>> -Damien
> Honestly, I felt that your response was dismissive of my concern.

I'm sorry Ted. There are lots of things going on here, I didn't pay  
enough attention to your concerns and I worded my response poorly. I  
really have been heads down focused on code, and when I am the words  
don't come so well.

But I have to ask, why does it take nearly 2 days to respond with that  
concern? I'm asking rhetorically, not as a knock on you. This is an  
issue with mailing lists, it doesn't allow very well for incomplete  
thoughts. Misunderstandings tend to fester longer than they should.

Now, the real problem I think, is the word decision. The only  
"decision" that was made on IRC was a plan of action for me to  
proceed. But others have taken this to mean we make major decisions on  

This wasn't a formal decision of the the final state of the code.  
Really, it wasn't a decision at all, it was me bouncing some ideas off  
project members so that I feel comfortable moving ahead in my own  
private branch.

Basically, the "decision" was that we need an alternative transaction  
model for partitioned databases as the current one is unworkable. Also  
it was decided that the code would preserve the current capability, as  
a just in case we must have both models.

These decisions are entirely appropriate for IRC, IMO. They are  
guidance so I can proceed with some confidence down a path. These  
aren't the final decisions that I want everyone to be a part of.

Again Ted, sorry I seemed dismissive your concerns.


> The ASF has minimal requirements of projects but one of the key  
> requirements is that decisions be made on the public mailing lists,  
> in a way that invites participate of an ever growing community.    
> There is probably no quicker way for a project to invite scrutiny  
> than to have something seem not quite right in this area.   Time and  
> again problems in this area have proved to lead to problems with a  
> project.    I was, and am, responding to something that experience  
> tells me is a warning sign.   I am not trying to disparage you or  
> the project by doing this.    The reason I am writing this now is  
> that I want to be sure that I am conveying how important this issue  
> is, not only to, but to the the foundation as a whole.
>> I haven't checked in anything yet, and I won't until we've had more  
>> discussion, review and testing. This problem is a non-issue,
> I look forward to seeing the code, and the ensuing discussion.
> Ted

View raw message