-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Rob Church wrote: This is as much just to isolate it as anything else, though. More people have SSH access for committing (though with a restricted shell), and the machine automatically runs the updated parser tests, so I'm leery of putting that in our core cluster. If a MediaWiki committer (or someone who gets into one of their computers or SSH keys) abuses that, they can only really use it try to hack into my blog. In theory we could isolate some of that better and go ahead and host it in the main cluster, or move it elsewhere. There was some controversy over this, primarily in that it was a surprise to those of us who worked on the software at the time. The position I took then was that Wikimedia is MediaWiki's number one _customer_, but not necessarily _owner_. We never really got around to figuring out a compromise or alternate ownership of the name that I know of, and now Tim and I are on the foundation's payroll, so I don't know whether or not we care about making a distinction like that anymore. - -- brion vibber (brion @ pobox.com / brion @ wikimedia.org)

On 07/01/07, Matthew Flaschen <matthew.flaschen(a)gatech.edu> wrote: We do; we use svn+ssh as our protocol. However, to stop us running riot and storing p0rn on brion's server, we're a little bit locked down. :) Rob Church

Brion Vibber wrote: Thanks. Your explanation made sense. I never really thought about what svn+ssh really meant... Matthew Flaschen

On 07/01/07, GerardM <gerard.meijssen(a)gmail.com> wrote: I agree that it is unfair that Brion's stack gets piled high (I should point out that Tim isn't exactly stuck for things to do most of the time, either) with projects that someone "official" promises at Wikimania, but then bitches about when their arbitrary deadline is not met. Agreed; if Wikimedia is really serious about pushing wiki, then it should continue to support MediaWiki in other environments, not just its own. It is reasonable for us to keep the focus of the software, of course - MediaWiki is a wiki engine - but it would be unreasonable to turn around and deprive existing third parties of much of the support and general community that's sprung up around MediaWiki itself. I don't think, however, there's any danger of this not happening. Absolutely! New developers come in all very excited about the thing - they've submitted a few patches, we've generally used those to help mould them into shape, and we've got a feel for them - but they're still very "green" - I certainly was. It would make excellent sense if a more experienced developer was assigned to "mentor" newbies, and I don't mean that in the sense of being up their noses - I mean that in a sense of giving their changes a quick eyeball, and being available to answer the odd email about how something should be done. Of course, we don't want to abandon eyeballing everyone else's work, either... Rob Church

Phil Boswell schreef: Hoi, There are two ways in which you can read a proposal for change. You can read it as a criticism and you can read it as a possible road forward that intents to allow us to do even better in the future. With a person that would be tasked to support people new to MediaWiki, it is not at all a good idea to have this person be the only person that these new people are exposed to. Development is done best when there is a community of people working together. This will not change, what would change is the scale of the influx of people. You may already notice that the amount of people working on the MediaWiki code is increasing. You may have noticed that some aspects of the MediaWiki code are actively developed but are of no relevance to other groups that do not have an interest. This means that we now have several sub communities of developers all involved in MediaWiki all being part of the MediaWiki ecology. My feeling is that as the use of MediaWiki is increasing, the scope of what people work on will increase at the same time. This is to be expected to have a direct impact on the scope developers work on as well. Consequently, it will be of an increasing importance to both maintain the architectural integrity of MediaWiki and allow for the growth of the developer community at the same time. Thanks, GerardM

Rob Church schreef: Hoi, Not necessarily anything. I do agree with Phil that a good job is being done. I also think that as the amount of developers grow, as there are more Wikis deployed, it becomes increasingly important to maintain coherence. We need unity in diversity. Thanks, GerardM

On 10/01/07, Erik Moeller <erik(a)wikimedia.org> wrote: OK - what does officially putting MediaWiki under the umbrella of the WMF do for MediaWiki that is an improvement on the present situation? List with + and -. - d.

On 19/01/07, Anthere <Anthere9(a)yahoo.com> wrote: This being the case, we generally develop functionality with a semi-defined project scope in mind for MediaWiki; it's a wiki engine, mostly driven by the Wikimedia interpretation of a wiki. A *lot* of what goes into MediaWiki (aside from bug fixes) is new functionality developed *for* use on Wikimedia wikis; a lot of extensions are developed to facilitate extras for Wikimedia. Having said that, there *are* things going into development which Wikimedia might never make use of, but which fall under the general scope of what MediaWiki is supposed to function as; a lot of additional extensions and plugins, for instance - stuff like LDAP authentication and other corporate/academic-used features. I don't see any real problem with the current interpretation of things. MediaWiki is as open as Wikipedia - the code is licenced under the GNU GPL and anyone who really feels strongly about it can fork it quite happily at any time. A huge number of organisations can and do take a vested interest in developing the software to its fullest potential. The superb effort going into things like Semantic MediaWiki and WiktionaryZ may never have been on the agenda in 2003, and may not be completely adopted in all Wikimedia projects, but they're very much in tune with what MediaWiki is. What is MediaWiki? It's a free form wiki engine which has a purpose. That purpose is to make sharing of ideas, information and content easy for anyone. How is that counter to any Wikimedia ideology? It's not. I think MediaWiki development works better when everyone feels they're in a comfortable environment where they're free to work on their own portions that interest them. Sure, they have to put up with me swearing every so often...sure, they have to be careful. But I personally, and I suspect others, have a nice comfortable feeling of not having to be accountable to the Wikimedia Board of Trustees. I like that; I can view all my users as equals. In case it's not clear; I support the current status quo. Rob Church

On 08/01/07, Brion Vibber <brion(a)pobox.com> wrote: That's just begging for a "somebody set us up the bomb" type response. Rob Church

On 1/7/07, Simetrical <Simetrical+wikitech(a)gmail.com> wrote: Commit now. For great justice. -- -george william herbert george.herbert(a)gmail.com

Jay R. Ashworth schreef: And to "us" Thanks, GerardM

Hoi, The notion that the MediaWiki software is almost entirely developed for one organisation is seriously wrong. OmegaWiki, formerly WiktionaryZ has realised a lot of functionality already and this is just one other project. There are many more projects that have developed on MediaWiki and much of this development, like OmegaWiki, is as Free as the WMF developed functionality is. As to distributed MediaWiki, as you may remember the Vrije Universiteit of Amsterdam (Andrew Tannenbaum's department) is actively working on a distributed Mediawiki and is particularly interested in the distributed network functionality that is required. This includes things like making sure the content is near to where it is requested. The WMF could provide a meeting place for organisations that use MediaWiki stimulate cooperation. The WMF can provide a developer that has as his task to mentor new developers, particularly students from Universities that want to be technically involved in MediaWiki projects. My expectation is that it will be possible to do some 50 projects in a half year (only the MW support .. not supporting the content part of the project) this is likely to lead to a retention of developers of in between five to ten percent and will as a consequence be a good investment. The organisations that develop MediaWiki may also need support to build extensions so that their code can be part of the main MediaWiki code. This can be a paid for service. When it is not the WMF who does this, another organisation may be willing to provide this service ... I am sure there are more things that can be done when MediaWiki has its organisational part developed it may even generate money for the Foundation. Thanks, GerardM David Gerard schreef:

On 08/01/07, Gerard Meijssen <gerard.meijssen(a)gmail.com> wrote: I stand corrected :-) Is as much of this as possible in the main line of development, or regularly merged? Excellent! That's the sort of thing I mentioned. Trouble being that would take an experienced MW coder who wants to do that and can do that as well as code. Yep. I wonder if there's any money in MediaWiki programming consultancy as yet. - d.

On 09/01/07, Gerard Meijssen <gerard.meijssen(a)gmail.com> wrote: Mmm. There's talk of my company selling my services to a customer as a MediaWiki sysadmin, as they have several installs with *no sysadmin responsible* ... and I'm going to have to learn to code soon whether I like it or not ... - d.

On 09/01/07, Rob Church <robchur(a)gmail.com> wrote: NOOOO YOU BASTARD AIEEEEEEEEE /me runs away - d.

On 10/01/07, Matthew Flaschen <matthew.flaschen(a)gatech.edu> wrote: I was talking as in Unix sysadmin; I suspect Gerard meant programming as well. - d.

On 10/01/07, Gerard Meijssen <gerard.meijssen(a)gmail.com> wrote: Christ, keep yer hair on! Seriously, though, this sounds rather similar to the job that Wikia employed a few people (Sannse and Mindspillage) to perform some time ago, and it'll probably be a growing trend as wiki culture refines itself. Rob Church