[asterisk-dev] Named ACLs

[Kevin P. Fleming]

On 06/13/2012 11:00 AM, Olle E. Johansson wrote:
>
> 13 jun 2012 kl. 17:28 skrev Kevin P. Fleming:
>
>> On 06/13/2012 09:46 AM, Olle E. Johansson wrote:
>>
>>> In response to the review:
>>>
>>> I think you've totally missed why John Todd and I came up with the ideas around Named ACLs. They have to be MANAGEABLE. That's why you will find a lot of work with trying to find out a way to manage NACLs in AMI and the CLI in my code. I would suggest that you take a look at this part too.
>>
>> And I think you've missed a very important point: the code that has been posted so far is a *framework*, clearly marked as not yet finished, that was put up for review in order to ensure that the framework was sound and was headed in the right direction to support what was desired. It is not done, not even close to done, and no statement to that effect has been made.
> So should I take it that you don't want feedback on framework reviews? You make me confused. I say that I think that the code in this review has totally missed the point of the whole project as outlined at one point by John and myself. You say I missed the point - which point? Don't you want' feedback?

I have not seen any feedback on the code itself from you, did I miss it? 
All I've seen is comments that we have 'missed the point'. Since the 
code is not done, you must believe that the framework that is posted for 
review is unable to support your desired usage modes. If that is the 
case, can you describe why that is the case, so that we can better 
understand what the framework must be able to do?

'missed the point' is not a constructive criticism, it does not help the 
code developer improve the code in any way towards your goals.

As I understand it at this point, the framework that has been posted 
*will* be able to support your desired usage (dynamic changes of the 
named ACLs with nearly immediate updates in the consumers of them), but 
with the addition of persistent storage and other aspects.

>>
>> As it turns out, now that we've had the opportunity to spend more time thinking about how the end result needs to work, we've tasked Johnathan with stepping back and documenting the desired use cases and proposed design on a wiki page so we can ensure that the project is headed in the right direction. This should have been done originally, but was not, and for that we'll take the blame.
>
> I am still confused on why Jonathan was not tasked with getting back to me and John to figure out why we did this in the first place and what state the code is in. But I guess I'm missing your point again and apologize for that.
>
> Sorry for disturbing the conversation.

Can we please attempt to keep these discussions about the technical 
merits of proposed solutions?

We have already apologized twice for mishandling this project, and we 
have tasked Johnathan (and Matt Jordan) with discussing this with you, 
but not until there is a wiki page that represents our understanding of 
what the project is supposed at accomplish. There's no point in having 
such a discussion without a common basis for all parties to read and 
provide criticism of. In fact, other than responses to code review 
feedback, that is the *only* thing of substance that has been worked on 
in this project since late Friday (our time), when this whole uproar began.

I ask that you treat us with the same patience and respect that you feel 
you deserve (something about a 'golden rule' comes to mind here), and 
that you also understand that we are all human, we all make mistakes, 
and the best way we can deal with them is to understand what went wrong 
and take actions to correct the situation. This will not happen 
overnight, but it will happen, and you'll see the evidence of it here on 
the mailing list, since this is the proper forum for this discussion.

-- 
Kevin P. Fleming
Digium, Inc. | Director of Software Technologies
Jabber: kfleming at digium.com | SIP: kpfleming at digium.com | Skype: kpfleming
445 Jan Davis Drive NW - Huntsville, AL 35806 - USA
Check us out at www.digium.com & www.asterisk.org