Bug 15218

Created attachment 15491 [details]
GPG Public Key

Created attachment 15492 [details]
SSH Public Key

On Wed, Mar 26, 2008 at 04:35:27PM -0700, bugzilla-daemon@freedesktop.org wrote:
> I am a KDE contributor. I am going to start to work on a freedesktop
> specification for what we call JobViewServer. I would love to have hosted my
> proposal (and further proposals I would love to contribute to, or work on) on
> the freedesktop site (working and fixing details on people.freedesktop.org).

Have you had much interest from non-KDE projects (e.g. GNOME, XFce,
Mozilla) in this at all?

Actually, I am a convinced defender of the Freedesktop project itself. I always bet for standards, and I am working on a specification of a standard for Freedesktop (JobViewServer). This will benefit all desktops. Of course I am interested in the rest of the desktops, in what making as much things as standard as possible so we can give better experience to the free desktop users.

On Thu, Mar 27, 2008 at 04:40:49AM -0700, bugzilla-daemon@freedesktop.org wrote:
> Actually, I am a convinced defender of the Freedesktop project itself. I always
> bet for standards, and I am working on a specification of a standard for
> Freedesktop (JobViewServer). This will benefit all desktops. Of course I am
> interested in the rest of the desktops, in what making as much things as
> standard as possible so we can give better experience to the free desktop
> users.

Sure, but there's a difference between a specification on fd.o, and a
standard.  A standard has support from all the major interested parties,
whereas a specification is developed by one group in isolation, and
possibly published on a site such as this, the Open Group's, whatever.
So I'd really like you to discuss this with some GNOME/XFce/whoever
developers and attract some more support, lest you design a
specification that turns out to be unworkable for other people for some
reason.

Cheers,
Daniel

> Sure, but there's a difference between a specification on fd.o, and a
> standard.  A standard has support from all the major interested parties,
> whereas a specification is developed by one group in isolation, and
> possibly published on a site such as this, the Open Group's, whatever.
> So I'd really like you to discuss this with some GNOME/XFce/whoever
> developers and attract some more support, lest you design a
> specification that turns out to be unworkable for other people for some
> reason.

Of course. We only wanted to make sure that this became stable and is trustable (so there is an implementation already, ours) before asking it to become a standard.

After having it implemented correctly on KDE (we already have) I wanted to write a paper about the spec itself, and then start to get feedback from the rest of the desktops, if there are problems, or if it is OK.

Probably I didn't give enough information. The spec is about http://dot.kde.org/1169588301, and of course the spec itself is based on D-Bus interfaces.

As you want, I can work on it and directly post the request on the xdg mailing list, but I'd love to do such changes in a place in freedesktop, so it is cleaner to change things (if needed or requested by others) rather than in a mailing list.

Tomorrow (or should I say later today?) I will write a more detailed and expanded description of the specification we designed with everybody in mind. As said previously, we finished our implementation (after having designed the specification in a very general and cross-desktop way), and it works smoothly.

From my understanding, the point where we should start to develop this specification (that can later become a standard, or not) should be here, at freedesktop, since this specification benefits all desktops, absolutely all of them, so that's a reason to take in consideration for working on this specification. All desktops are interested in showing somehow time consuming operations and inform the user about them.

So, all requirements have been obeyed from my point of view:

- Has been written about this issue in xdg. As I said before, I will do in a very higher verbose mode tomorrow.

- I have spoken with the mathusalem developer, and we were interested in merging efforts in this way.

- The reason of working in freedesktop is of course, for having an objective point of view of the issue, and accept really valuable input, what is always really good for any possible specification.

Daniel:

I appreciate the rigor you are attempting to apply to the fd.o process. It's a refreshing change to some of the wild west style of years past which has resulted in things like D-Bus interfaces registered under "org.freedesktop" which never went through any such gate-keeeping.

Of course, it becomes a chicken-and-egg problem if a relevant specification can not be at least drafted at freedesktop.org. For something that is of obvious use and utility to all involved but which has not been addressed, the relevancy should be without doubt.

Whether it transitions from "proposed spec" to "fd.o adopted spec" to "de facto free software standard" is something that can only be answered through a process of engagement. freedesktop.org is the place for that engagement, and this particular specification will be based not just on a theory but a working implementation; in turn that working implementation was built on a design that had cross-desktop utility in mind from the very beginning. In the end, the stakeholders involved will have a chance to weigh in on its potential status as an fd.o spec; that is not possible, of course, if the spec is not given a venue in which to thrive.

Rafael is simply asking for such a venue so he can continue the process he has already begun. He has an implementation (in production!) based on an internal spec generated after consultation both with KDE and with other projects (e.g. Mathusalem). He has also given notice of his efforts on xdg@.

To quote the freedesktop.org front page, "Unlike a standards organization, freedesktop.org is a "collaboration zone" where ideas and code are tossed around, and de facto specifications are encouraged." Rafael's efforts are precisely in line with that ethos.

Let's get him his account so he can start working. We could always do as others have in the past and host the drafting of these specs elsewhere, but that would seem to be in direct conflict with the stated position and goals of freedesktop.org. It would also, I think, send the right message to Rafael and others that would follow him.

Cheers, Aaron.

Hi,

On Thu, Apr 03, 2008 at 06:01:35PM -0700, bugzilla-daemon@freedesktop.org wrote:
> I appreciate the rigor you are attempting to apply to the fd.o process. It's a
> refreshing change to some of the wild west style of years past which has
> resulted in things like D-Bus interfaces registered under "org.freedesktop"
> which never went through any such gate-keeeping.

Yes, it's somewhat overdue by this stage ...

> Of course, it becomes a chicken-and-egg problem if a relevant specification can
> not be at least drafted at freedesktop.org. For something that is of obvious
> use and utility to all involved but which has not been addressed, the relevancy
> should be without doubt.
> 
> Whether it transitions from "proposed spec" to "fd.o adopted spec" to "de facto
> free software standard" is something that can only be answered through a
> process of engagement. freedesktop.org is the place for that engagement, and
> this particular specification will be based not just on a theory but a working
> implementation; in turn that working implementation was built on a design that
> had cross-desktop utility in mind from the very beginning. In the end, the
> stakeholders involved will have a chance to weigh in on its potential status as
> an fd.o spec; that is not possible, of course, if the spec is not given a venue
> in which to thrive.

Indeed, I fully agree with everything you've written (including the rest
which I didn't quote).  I'm trying to just get some time (hopefully the
weekend includes some, but we'll see) to write everything down and get
it sorted.

Hi  Rafael, how'd you go writing up a detailed version of the spec?

Hi Benjamin,

I have done an extense introduction on xdg mailing list. Here it is the reference:
http://lists.freedesktop.org/archives/xdg/2008-April/009410.html

Hi all,

I wonder how the account creation or whatever is going, so we can start working on this freely on freedesktop itself.

You can already check the previously pointed out mailing list thread on xdg to see others interest in this solution.

I don't think a complete, accepted standards project should be required to get an account, so I went ahead and added it.  It may take up to 8 minutes for LDAP to propagate.