[Qgis-developer] Plugin migratin -> mailing list?

[Tim Sutton]

Hi

> On 08 Nov 2015, at 19:05, Paolo Cavallini <cavallini at faunalia.it> wrote:
> 
> Hi all,
> 
> Il 08/11/2015 12:42, Raymond Nijssen ha scritto:
> 
>> For now, let's ask Paolo in what way he is intended to use this.
> 
> I'm surprised of all this discussion. I think the situation is quite clear:
> * we have a problem: major changes in QGIS will break most or all plugins
> * we do not want to leave authors alone in dealing with this
> * the best solution we thought of is to write announcements and tips to
> help them
> * we can send a message to 299 addresses, but it is easier and safer if
> we use software designed for it, and our good old Mailman seems the
> natural solution
> * this allows the list of addresses to be kept secret, and to let people
> unsubscribe (and eventually subscribe) if they wish.
> I frankly do not do not see the issue, sorry.
> All the best.

My 2c:

I think we should revisit the idea of tiered plugins:

* "wild west plugins" - no approval process beyond the basics of it it can pass validation (our current store). It would be accompanied by a cautionary note: ‘Do not enable this unless you are happy with the the fact that QGIS asserts absolutely no confidence in the quality of these plugins.'
* QGIS approved plugins - maintainers *must* to be contactable and willing to receive regular correspondence from QGIS as relates to the management of their plugin. It would be accompanied by a cautionary note: ‘This plugin does not adhere to the coding and style guidelines of QGIS.ORG. Whilst QGIS.ORG provides no guarantee that this plugin is safe to use and fit-for-purpose, it has been developed by a community member who is generally trusted and has made a commitment to maintain the plugin.'
* QGIS recommended plugins - same requirements as in ‘approved plugins’ above, plus the plugins must be compliant with coding standards, HIG (human interface guideline) standards (so they integrate nicely with QGIS look and feel etc.) and the developers are trusted to not insert the odd "rm -rf /“ shell call into their scripts. They would be accompanied by a cautionary note like: ‘This plugin adheres to the coding and style guidelines of QGIS.ORG. Whilst QGIS.ORG provides no guarantee that this plugin is safe to use and fit-for-purpose, it has been developed by a community member who is generally trusted and has made a commitment to maintain the plugin.'

These classifications would be in addition to the experimental / not experimental classification we have (which can apply to any plugin tier above).

Plugin authors should understand that we are placing inherent trust in them and taking an inherent risk by distributing their work and making it available on the computers of hundreds of thousands of users and they absolutely *must* be willing to receive and respond to regular correspondence from us as part of that equation. If we have slipped up on making that clear up till now, we need fix that. To start with we should put all publishers of existing plugins into the ‘wild west’ and they would not be subject to any email distribution so that we keep the status quo. Any author of approved and recommended category plugins absolutely *must* agree to participate in a broadcast email programme intended to inform and advise plugin writers of changes in the QGIS project as pertains to plugin writing and management.

Regards




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-developer/attachments/20151108/135e3db2/attachment-0001.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-developer/attachments/20151108/135e3db2/attachment-0001.tiff>