[Qgis-psc] Chat unification

Sun Nov 8 10:54:11 PST 2015

Hi (with apologies for the rather long post below)

> On 08 Nov 2015, at 17:06, Sandro Santilli <strk at keybit.net> wrote:
> 
> On Sun, Nov 08, 2015 at 02:25:17PM +0000, Anita Graser wrote:
>> On Nov 8, 2015 10:29 AM, "Sandro Santilli" <strk at keybit.net> wrote:
>>> 
>>> On Sun, Nov 08, 2015 at 09:35:21AM +0000, Nathan Woodrow wrote:
>>>> Oh we can just host something like rocketchat which is open but not old
>>>> school like IRC
>>> 
>>> Sorry but... what's does "old school" mean exactly ?
>> 
>> I think what feels old school today is the lack of a friendly web chat
>> client (I.e. no local installation necessary) which non technical people
>> would be comfortable using.
> 
> Had you tried https://kiwiirc.com/client
> and/or http://webchat.freenode.net/ ?
> 
> (How) do they fall short on "friendliness” ?

I think this and the previous thread on chat clients conflates two separate discussions, so let’s speak to these separately:

Discussion 1) Do we find it acceptable to use non-FOSS software as part of our project offering?
=========================================================================

Lets first do a little stock take:

* When we decided to drop our forums some years ago we were happy to adopt to using stack exchange (also a proprietary platform) because it provided a better platform with good critical mass, and I don’t really see how this discussion is substantially different. StackExchange and the contributions of people like Anita and Nathan on it are a huge asset to the QGIS project in terms of helping our community, and I believe this far outweighs the fact that the platform is proprietary.

* We have always hosted our code on a proprietary platform - we moved our source code from SourceForge (which was FOSS until 2001 and thereafter proprietary) to GitHub (which is proprietary). Funnily enough we are still listed as ‘beta' software on source forge: http://sourceforge.net/projects/qgis/ :-)

* We run our domain on go daddy.com which is proprietary.

* We build our windows binaries using proprietary compiler (in the old days Radim I think used to build under mingw but that had issues with linking to binaries created on other compilers)

* Our OSX build is built using clang which I guess is FOSS, but I think much of the glue that makes it usable is proprietary OSX stuff

* We host our services on Hetzner which uses a proprietary web panel for administering restarts, tickets etc.

* We even make it accessible for people to use proprietary databases like Oracle, SQL Server, 

* We have an official twitter account somewhere again on a proprietary platform

* Even in OSGEO, our parent project it looks like they are using slack (and probably other proprietary stuff that I am not aware of) - http://osgeo-org.1560.x6.nabble.com/Slack-and-OSGeo-td5220636.html

I could go on, there are many more places where, if we were Richard Stallman clones, we would have cried ‘foul!'. We need to decide as a community what our values and objectives are. Personally my objective in joining the QGIS project back in 2002 is still the same as it is today: to democratise access to GIS so that we can take better care of our earth. I love Free Software and it is the perfect way to achieve that goal, but it is not enough to be Free Software. We also have to be better, more accessible (in cost, language, availability of help and all other senses of the word) and provide our users the tools they need to actually do their work - which is precisely why we have Oracle, SQL Server, MrSID, ECW support and have made other ‘compromises' beyond that which Richard Stallman would find unforgivable. 

The reason I liked gitter is because it provides a richer, more accessible way to interface with our users. I wasn’t promoting it because it is proprietary, I was promoting it because it provides a great experience and it is free (small ‘f’). If we as a community decide that being Free Software in all aspects of the project is a fundamental, uncompromisable premise then I am all on board with that, but lets apply it uniformly through the project and not just in this one case. It seems to me though, based on the above, that our approach to date has been rather more pragmatic: “Lets promote the virtues of FOSS, while doing what we can to make it something that is accessible and useful to as many people as possible.” For me the place to be uncompromising is in the adherence to publishing QGIS as FOSS / GPL, our documentation as open content, and doing everything we can to foster an open and collaborative community. Beyond that I think being pragmatic makes sense because we have limited time and resources and we want to make the most of that time in the creation of an awesome Free GIS.

Discussion 2) What do we wish to offer our users for an online realtime chat experience?
===================================================================

Here are my criteria for a good chat system (aka reasons why I like gitter):

* works on all platforms nicely including mobile and web browser
* sends me a notification by email if I miss a direct message in the chat because I am away
* supports pasting of screenshots and other media which I can view inline with the chat
* integration of notifications from github when someone comments on an issue, PR, travis fails, someone makes a commit etc. (I like the way gitter does this to the side panel so that it is not cluttering up the main chat conversation)
* automatic referencing of issues and PR’s just by typing #1234 (or whatever the issue number is) in your message
* maintains session state as you move online / offline so that you don’t miss anything in the chat history
* supports for rich markup so you can use e.g. code blocks, bold, headings etc.
* its extremely simple to use and approachable

Here are some things gitter in particular misses:

* not FOSS
* mobile version misses GH notification sidebar
* no support for moderation

This last item is my main concern with  gitter, but all of these features provide a very friendly and useful platform which IRC does not offer out of the box. I did try out kiwi irc - it seems nice but I don’t think it address enough of the things above to make it a credible alternative to something like gitter. I put up the gitter channel as an experiment to see if others like it and the more modern experience it provides. I agree that having one chat platform would be good, though for me IRC is not the best choice any more. 

https://rocket.chat looks really nice and ticks the FOSS box. It seems like the mobile client can’t be configured to use your own server, but maybe that is ok for now. I set up a #qgis channel on https://demo.rocket.chat/channel/qgis - come and try it out if you like. I don’t know what they offer in terms of reliable service, but if it seems  good and it ticks a lot of the boxes above, I would be more than happy to go with rocket as a more modern alternative to IRC.

For me its -1 to use IRC as our official platform, having used ’next generation’ chat platforms, its hard to go back and I don’t think it is a good experience for new users, but I am more than happy to spend some (more) time and effort in finding a good FOSS alternative if that is the preference of the rest of the community.

Regards

Tim

> 
> --strk;
> _______________________________________________
> Qgis-psc mailing list
> Qgis-psc at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/qgis-psc

Tim Sutton
QGIS Project Steering Committee Member
tim at qgis.org

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/qgis-psc/attachments/20151108/c32d3774/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: PastedGraphic-1.tiff
Type: image/tiff
Size: 9882 bytes
Desc: not available
URL: <http://lists.osgeo.org/pipermail/qgis-psc/attachments/20151108/c32d3774/attachment.tiff>