Re: [pgsql-www] New Mailing List : pgsql-gui-dev@postgresql.org ?

Le 29.11.2017 15:37, Stephen Frost a écrit :
> Magnus, all,
>
> * Magnus Hagander (magnus(at)hagander(dot)net) wrote:
>> On Wed, Nov 29, 2017 at 3:24 PM, Simon Riggs <simon(at)2ndquadrant(dot)com>
>> wrote:
>> > On 29 November 2017 at 20:29, Damien Clochard <damien(at)dalibo(dot)info> wrote:
>> >
>> > > I'm sorry to insist but my original message was a month ago and I have no
>> > > idea how to proceed...
>> >
>> > Back in 2004, there was huge interest in setting up a special list to
>> > discuss PITR. We did, then everyone waited for me to write the
>> > feature.
>> >
>> > As a result we said "no more special purpose lists".
>> >
>> > If anybody is actually planning to write emails to this list, why not
>> > start writing them now? Why wait?
>> >
>> > Just use the prefix "GUI". If we get enough people writing emails and
>> > code, then maybe a special list is worth having.
>>
>> I think that the core difference is that pitr was basically a subset
>> of
>> -hackers. (as was the win32-hackers list, which I believe was around
>> the
>> same time)
>>
>> My understanding at least is that the target audience for a "gui list"
>> has
>> very little overlap with the target audience for the hackers list,
>> right?
>
> I'm a bit concerned about adding yet-another-mailing-list too, and I'll
> point to 'pgsql-cluster-hackers' as an even more recent case where we
> added a list and it ended up dying off.
>
> Now, it's a good point that cluster-hackers is also just another subset
> of -hackers, and perhaps gui-dev is enough different that it won't
> suffer from this issue, but I have to ask this:
>
> What of pgadmin-hackers ? That list certainly gets traffic on it
> regularly as it's being actively developed, but are people expecting
> that traffic to move to this new list? If not, why not? If it's
> because there's not enough commonality between pgadmin and whatever the
> list is for, then I'm confused as to the point of the list. If the
> idea
> is to move the pgadmin-hackers traffic over to this new list, then that
> would settle my concerns that it ends up being a dead list.
>
> If there's other active projects that need lists to facilitate the
> development of the project, then that also seems more likely to be used
> moving forward, but having one general list for what seems like a bunch
> of independet projects strikes me as a bit odd, especially when the one
> I'm most familiar with (pgadmin) is going to continue to be
> independent (if that's the case, which I'm guessing it is..).
>
>

Hi !

If we had only a handful of graphic tools for Postgres, I would
definitely agree with you. But so far we have identified 60 developers
actively working on 35 different tools. Diversity is a chance for open
source projects but at this level of fragmentation it begins to hurt
everyone : developpers are wasting time and code on the same issues,
same needs.

Our goal is not to unify every projects into one but we need a space to
discuss about common APIs, compare charting libraries, frameworks,
common libraries, standard JSON outputs, common design principles,
shared experiences, etc.

You are right that this list could end up being a dead list. I cannot
estimate how much traffic it will receive. But I am positive that it is
clearly not a subset of psql-hackers or any other list, or at least most
of the 60 developers we have identified are not active on any other
mailing lists of the community.

--
Damien Clochard