devicemap-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Werner Keil <>
Subject Re: [PROPOSE] Release DeviceMap-data and other dependencies (with connection URL etc. fixed)
Date Wed, 07 Sep 2016 15:41:03 GMT
Similar to e.g. Tamaya, Karaf has everything under a single codebase. is what gets deployed
shows, there's also versioning of these artifacts.

So since everything is "code" also in this case (sources like .asciidoc get
interpreted into .html or other formats) then why can't similar sources
also be made available to the DeviceMap site?
It should not matter if we use SVN or Git either.

Code in both SVN or Git repositories are useless to the URL loaders and even
won't work, because the SVN browser turns it into a special form like
And the URL loader can't handle these dynamic URLs either.

Eberhard way back tried to use the sources from SVN and it failed, only
afer I pointed it to the plain URLs on the VM the .NET clients work since

Instead of Karaf, let's look at
It contains 1.1.0/ <> and other
folders for releases of this artifact.

Not signed ZIP archives like etc.
Users would have to download first and use as local update sites.

So if this alternate (more convenient for some) distribution of some parts
of Sling works, why should it not work for DeviceMap?

Are some projects "more equal"?;-|


On Wed, Sep 7, 2016 at 5:02 PM, Werner Keil <> wrote:

> Then why can Karaf publish something DeviceMap can't ?
> Is their "tutorial/latest" simply part of the "site" which does not fall
> under the release constraints?
> Again, as all the clients stand, especially the .NET ones (but without
> significant improvement to add a "Remote JAR" option, also the Java Client)
> a ZIP or JAR file on a remote server is worthless to all DeviceMap clients.
> If it's not possible to make the "data" folder and URL available in a
> proper way ( will likely continue to exist,
> only with the "retired" warning if it gets archived, so
> hypothetically also would, but nobody is
> even able to start or maintain the VM after the PMC stops) then I don't
> think any further release is worth the effort.
> Werner
> Hi,
> On Wed, Sep 7, 2016 at 1:42 PM, Werner Keil <> wrote:
> > ...Can we turn
> >
> > into
> > ...
> No, because that data must be published via the Apache release process
> as per and then
> distributed at least via and possibly
> other common channels.
> Once again, for this to happen you just need to create an archive of
> what needs to be released, sign it and make it available so that this
> PMC can vote on it - while it still exists.
> -Bertrand

  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message