Re: [hudson-dev] [Hudson-Dev] Re: Specifying the sections where your plu

[Winston Prakash <winston.prakash@xxxxxxxxx>]

On 1/5/12 12:57 PM, Henrik Lynggaard Hansen wrote:
> 2012/1/4 Winston Prakash<winston.prakash@xxxxxxxxx>:
> > > Hi
> > >
> > > One thing that is not documented is the values which can be used in that
> > > tag.
> >
> > Yes, we should document it. Unfortunately no solid list was ever
> > consolidated.
> > Let us work on that.
> sure, let me start by making the initial suggestion
>
> - scm: for all scm plugins (as today)
> - scm-related: for scm helper plugins (as today except git,label column)
> - trigger: for buils triggers (as today)
> - builder/buildstep: for plugins that extend Builder. As today except
> a few to clean out like project health, build pipeline, post build
> task and the such
> - slave: anything related to managing slaves (as today except system load avg)
> - notification: for plugins that notify after (as today except copy to
> slave, hudsontracker?)
> - report: build reports
> - publish/upload: any artifact uploader/publisher
> - security: for security and user management
> - fun: for none serious plugins such as cigame etc.
>
> I know I am missing some categories, but since many of them are
>
> changes from the current topic list
> - suggest slaves and cloud plugins are one category
> - drop the maven category
> - add a library category for things like static-analysis-core, token-macro etc.
> - drop the misc category and just use uncategorised

Sounds like a good list. Could you start a wiki page to list the Plugin 
Categories, so that others can chime in.

> Maybe we should build an alternative topic page while we figure this out ?

Agree

> > > also should we specify the tier in the pom ?
> >
> > I'm not sure if it should be part of the POM. The tier is decided based on
> > popularity of the plugin (number of downloads etc)
> Speaking of which, did we ever get the download/install stats for
> plugins fixed,
If you mean fixing the Public/private key issue to unscramble the 
collected data, if so I fixed it in 2.2.0.
It has a unique server key, so the collected data will be more accurate/

The download data prior to 2.2.0 could be interpreted in terms of unique 
IP. But may not be accurate.
> and is the data publically available ?
The data is collected as part of apache log rotation. I have to parse 
and create the statistics

What I think for the long run is to build an application that receives 
the download request which saves the stat directly in a database. The 
advantage of this approach would be; direct display of download stats by 
that application on demand.

- Winston

> Best regards
> Henrik
> _______________________________________________
> hudson-dev mailing list
> hudson-dev@xxxxxxxxxxx
> http://dev.eclipse.org/mailman/listinfo/hudson-dev