incubator-tashi-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Michael Stroucken <>
Subject Fwd: Re: Incubator to Tashi community:
Date Mon, 12 Aug 2013 20:08:42 GMT

On 12.08.13 14:58, Craig L Russell wrote:
> Hi Richard,
> Tashi looks like it suffers from "good enough" syndrome.
> It works well enough for its intended purpose and doesn't aspire to greatness, so there
is little development activity.
That it does, but it seems that the VM management community has migrated
to the more well known alternative offerings, eg. Apache Cloudstack.
What is the Tashi project still able to contribute?
> In order for the project to be successful at Apache, it should have a community of users
and developers who can collaborate and evolve the project. At the least, it must have three
or more developers who are active and can prepare and vote on releases.

> With about 30 subscribers to dev and 30 subscribers to user, it sounds like there is
interest in the project.
It's been a long time since we heard from any of them though. Our
original sponsoring organizations have either disappeared or thrown
official support behind a commercial alternative.

I am not involved with that group anymore, but I see Tashi is still
popular with the students. But it is good enough for them, and I don't
think that they would be too eager to update the environment with any
improvements that we may come up with.

However, if they generate improvements, I would include that quickly
into Tashi.
> Are any of the subscribers to dev and/or user willing to take more of an active role
in developing and sustaining the project?
Tashi is very dependent on underlying tools - last time I was very
active on it, KVM functionality and formats were changing weekly. So
work probably needs to be done to support improvements in that area.
Some users have approached me with interface improvement suggestions,
but I can't currently think of any new functionality I would want to
add. I have more of an interest in replacing the Python/RPyC setup with
Java for maintainability and type safety.

For the hardware management side, I don't have a physical cluster
anymore, so I don't think I can contribute much there.

So given that the project may still be continually improved at a slow
pace, not receive much public interest, not attract additional
developers, probably never reach a community size that graduated
projects have - would it be best for Apache and Tashi if we moved to GitHub?


View raw message