Month: July 2005

Hey John, I fully agree with your point of view about how fd.o projects should be.

I more or less used xdg-list to get a huge amount of opinions. I tried filtering some of them into a wiki page. I decided to go public with that wiki-page and let interested people add their needs. It turned out to be more or less a small success. The wiki-page did, indeed, grow. And most additions where not being added by idiots (unlike the flame wars on xdg-list). The idiots decided not to waste their time with it. Once done with that, I decided to shut up about it. And I started to search for people who are interested, I started a mailing list and etcetera. So just like you suggest.

I’d like to make sure, however, people understand that this wiki page isn’t “dconf”. It’s not. Not at all. It’s just the opinion of a lot people about desktop configuration management. So Colin, you’re blog-title is wrong! That image isn’t about “DConf”. It’s unjustified to use that drawing to taint peoples opinions about “dconf”. Please don’t.

DConf is a research project. It’s about “seeing what happens if you do this or that”. The exercise of telling people about a shared configuration management infrastructure is part of that research. Viewing our communities fail to agree on this stuff is a lesson that we are learning today. For me, it’s also part of the research: The results are miserable. I decided to stop that experiment.

I learned GNOME nor KDE are ready to cooperate on this level today. And it’s not only the cooperation between the two teams. I’m sure, John, you are learning how difficult it is with “D-BUS”. In my humble opinion shouldn’t cooperation be that difficult. So my plan is to keep this project a research one. I don’t think it’s a good idea to be very “public” about “dconf”. Yet you guys forced me to defend this ghost. I hope this illustrates why I wrote my previous reaction, and why I said “I don’t (yet) like doing that (blogging about dconf)”.

I hope the nonsense reactions (that I removed) from my blog and nonsense blogs and comments about “dconf” that suddenly pop-up on various places also illustrate why I think it’s better not to be very public about “dconf”. Some are really hilarious!

Hey John and Ross, now you guys forced me to blog about “dconf” :p. And trust me: I don’t (yet) like doing that. Simply because “dconf” is at this moment “only an idea”. It’s a “concept”. It’s not yet code. You can’t yet touch it. Please understand this!

And please understand that we are not about flamewars on xdg-list. We are actually thinking about and investigating the concept. We are doing meetings. Etcetera. But we aren’t promising anything! It’s unfair of you to start saying “dconf is crack” when we haven’t even started it. Sure you meant “that E-Mail is crack”, but a lot people might interprete your blog as: “dconf is crack”. That’s unfair and will kill the project. In my humble opinion shouldn’t the free software community behave like that.

I’d like to make sure people understand that the proposal of Jamie isn’t really about DConf. It’s about a project that he’ll be doing himself

So far the only (more or less official) conceptual proposal, coming from me, about DConf, is this one. And this drawing will give you a visual overview of the different (proposed) components and their responsibilities.

I’m glad you guys read the mailing list. But please don’t take things out of the context. Please make sure people will read the conversations, not just one E-mail of anybody who’s subscribed to the mailing list.

Also note that storing the configuration data in a CVS server isn’t addressed in any of the current proposals. So DConf, it’s current proposal, isn’t about that at all. Again, please keep things in context. Only one person requested this on xdg-list a very long time ago. None of the people who are interested in dconf have ever acknowledged that this is an important feature request.

I also like to repeat that DConf is (in my humble opinion) a research project. It’s not something that will be usable soon. Please do not start thinking that we’ll have a solution for the next major GNOME or KDE release. It’s not that simple!

Oh and by the way: Yes, you’re right, we do need a realistic-minded project leader.

I decided to take a look at this bug about eog. It’s a feature request to have printing support for eog. It used to be available but got removed for some reason. So I reimplemented it using libgnomeprint and libgnomeprintui. It’s my first time I used gnome-print so please review before approving. There’s still a little issue with the paper-margins for example.

By the way, why isn’t the gnome-print API available on developer.gnome.org? Is it a secret API? Or did people just forgot to add it? At Foundation level we are talking about “Certification for GNOME apps”, but at documentation level we aren’t telling people how to do simple things like printing a document. IMHO there should also be more samples. Simple ones like: “How to print a GdkPixbuf using gnome-print and get scaling it on the paper right”. I actually had to start searching koders.com to get basic answers about how to use gnome-print. That shouldn’t be, there should be samples and real documentation.

By the way Tim, if it’s not a problem for you it’s more easy for me to commit patches like this myself (after approval of course). That way my tree doesn’t get a huge CVS conflict. However, if you prefer to commit it yourself, it’s not a real problem for me.

You can find some photos of the Evolution hackers (and contributors) at GUADEC this year here.

ps. Veerapuram Varadhan made these photos