[postgis-devel] is PostGIS ready for git ?

[Paragon Corporation]

Strk,

I'm going to top post cause I don't want to loose anything you said by accident.

1) One thing I don't like about PostgreSQL setup is they don't accept pull requests from github.  I think that's a big non-starter for us.
So if we can avoid that that would be great.
2) In large scheme of things $120,000 is not much 9barely enough to secure a pre-payment for FOSS4G).  Especially since we can't even send our core developers and maintainers to conferences for free with it. Which I always found a huge pity.
3) Helping Martin is great, and yes part of the reason I took on doing Jenkins, configuring postgis.net  and all that stuff is that it was a learning experience for me to get comfortable working on Linux
Within the context of a project I love. So yes there is a personal feeling of double  accomplishment there.

If what you propose works that would be great.  Though I'm still unclear of how much of a burden NOT USING githubs's issue tracker would be but still accepting pull requests from it.
So whatever solution we come up with still has to have a component of github in it.


-----Original Message-----
From: postgis-devel [mailto:postgis-devel-bounces at lists.osgeo.org] On Behalf Of Sandro Santilli
Sent: Wednesday, October 14, 2015 2:22 AM
To: PostGIS Development Discussion <postgis-devel at lists.osgeo.org>
Subject: Re: [postgis-devel] is PostGIS ready for git ?

On Wed, Oct 14, 2015 at 01:30:44AM -0400, Paragon Corporation wrote:
>  
> 
> > What about PostgreSQLs git repositories? They don't have "a special place" for extensions?
> 
> They have their own which is not github and they have a mirror on github.
> 
> http://git.postgresql.org/gitweb/

Pretty much what I'm proposing for ourselves (having our own).

> The problem with relying on OSGeo infrastructure too much is that we 
> have unpaid volunteers.  And trust me maintaining bots and all that 
> stuff is a big burnout job. Maintenance is way more work than cost of servers.

[...]

> Rather than asking for things we should insure our hard-working people 
> get paid or are better rewarded before we start demanding things.

I fully understand both points, and surely can feel the problem (git/trac was requested 4 years ago...).

If OSGeo cannot pay for the infrastructure maintainance we'll need another option. But it's a pity, looking at the surplus of the last
3 years:

 ~47k USD in 2013 - http://wiki.osgeo.org/wiki/OSGeo_Budget_2013
 ~40k USD in 2014 - http://wiki.osgeo.org/wiki/OSGeo_Budget_2014
 ~75k USD in 2015 - http://wiki.osgeo.org/wiki/OSGeo_Budget_2015

First point in the OSGeo Goals list is "to provide resources for foundation projects", third is "promote the use of open source software in the geospatial industry":

 http://www.osgeo.org/content/foundation/about.html

That said, I don't like "asking" either so I'm actually trying to help the good Martin Spott on the task of allowing hosting git repositories on OSGeo machines, and I don't think we're far.

What we're up to is a system in which you'd be able to maintain the current trac instance and no need to migrate any existing ticket.
We'll be able to register multiple repositories with trac and reference changesets in each from the tickets, so that existing tickets will keep pointing to the old repo, while new tickets will possibly refer to new repos.

Is this work not worth the trouble, you're suggesting ?

I'm learning as I go, which is one of the basic freedoms Free Software provides. Something that would not be possible by just moving the code to github. I'm learning to setup my own trac/git instance, shall github fail, OSGeo stop providing infrastructure or whatever.

So, to me, it's time spent much better than on migrating tickets to github, even if no Open Source company offerend fundings to do that (yet).

--strk(learning_to_fish);
_______________________________________________
postgis-devel mailing list
postgis-devel at lists.osgeo.org
http://lists.osgeo.org/cgi-bin/mailman/listinfo/postgis-devel