[Wikipedia-l] No new wikis

[Daniel Mayer]

Erik wrote:
>I am a bit concerned about the pace at 
>which new Wikipedia spinoff projects are 
>created. 

Isn't spin-offs the whole point of the Wikimedia idea?

>Wiktionary was a good idea, because it filled 
>a gap that was there -- we received lots of 
>dictionary entries, so it seemed like the logical 
>conclusion to start a wiki-dictionary.

We also get a lot of very long articles that are more suited for a textbook. 
Oh wait...

>But now we have a new "Textbook-Wiki" which 
>was started without much discussion - 

Not much discussion? There was so much discussion that we very quickly got our 
own mailing list. 

>possibly a good idea, but also possibly too 
>specific - 

Too specific? Have you browsed any online book seller for textbooks? There are 
thousands of different types - if anything it is a far more ambitious project 
that creating just one encyclopedia. We plan on writing hundreds of different 
textbooks. Specific indeed.

>and shortly afterwards, a "Wiki-Quote" project was 
>created. 

That one blindsided me too - where was the discussion for that?

>Now people are talking about creating a 
>"Wiki-Piki" for pictures.

Terrible names - sounds like a fast nose picking. But having a centralized 
database for images that all Wikimedia projects can share is a very good 
thing. But that is a longer term goal and (as you say below) there are more 
important immediate needs to take care of.

>This is all nice and good, but haven't we learned 
>anything from the Wiktionary experience? Wiktionary 
>was set up without much thought as to how the wiki 
>process could be applied to a dictionary; it took 
>months to formulate some kind of standard template,

Huh? The template and the processes you speak of were worked out the WikiWay 
on live data. Wikipedia was similarly chaotic during its first few months. To 
work this stuff out you have to work with real data - mock-ups on meta are 
little help here (and having them "clog" meta's RC would probably be frowed 
upon). 

So we have learned something from the Wiktionary experience; we started with a 
project with the vague goal of having it be a dictionary. We experimented on 
different ways to make that happen with a great variety of different data. So 
different ways to organize those data were explored. Ambiguity problems arose 
between different meanings of words so we reorganized articles. Somebody had 
the idea of having translations in entries - similar ambiguity problems arose 
and were taken care of. Now instead of just a dictionary we have a dictionary 
to end all dictionaries whose entries will be all words of all languages.

It is difficult to imagine all that being worked out on meta.

>and we still don't have Wiktionaries in other languages. 

In time. Wikitionary has the goal to define every word in every language so in 
a sense it is already somewhat internationalized (of course the descriptions 
are in English). This, IMO, brings down the demand for other language 
versions (they are still needed though).

>Wiktionary could have benefitted a lot  
>from better planning before it was set up. 

Perhaps. But do read the above.

>I'm not sure I like the Wiki-Quote idea at all, 
>as it intersects a lot with Project Sourceberg, 

That has been my reaction too and I unsuccessfully tried to expand the focus 
of the project to include really big quatations (such as books). 

>is not very wiki-like (a quote is a quote) and not 
>very compatible with the open content idea. 

Giving the source of the quotations and giving commentary about them  is 
wiki-like. 

>Wiki-Quote was only very briefly discussed.

I don't remember the discussion at all...

>Furthermore, it's not exactly like we have lots 
>of free resources. Our database server, pliny, is 
>down on its knees, the full text search on the  
>English wiki is now permanently disabled, we 
>have only a couple of active server administrators,
> and hardly enough developers to address 
>problems in the software.

These are all very valid points.

>I propose that
>1) we do not start any new Wiki spin-off 
>projects until our current resources have 
>been substantially expanded;

I agree - that may have to wait until well after Wikimedia has a bank account 
set-up and is able to accept donations/grants. But that should encourage much 
discussion on how Wikimedia will function and on how to organize additional 
projects. 

>2) we formalize a process for starting such 
>projects, e.g. a planning period of at least 3 
>months on Meta with exact specifications as 
>to what is to be placed there. 

I wouldn't put such a time period on it since many ideas never are discussed 
much. I would set milestones (such as figuring out a name, the parameters of 
what just should and should not be in such a project etc.). 

>After this period, users on the Meta wiki should  
>vote on whether the new wiki should be set up 
>or not. 3 months may seem long, but if interest 
>can't be kept up that long, the idea may not be 
>so great after all.

Hm. I wonder if there would have been a Wikipedia at all if a similar process 
was vetted at Nupedia... Sometimes the best thing to do is jump in head 
first. But dead wikis are bad so something is needed to keep the still born 
rate low. 

>Otherwise I see the danger that we'll end up with 
>lots of nice ideas that all go nowhere, like the sep11.
>wikipedia.org (which IMHO should never have been 
>set up in the first place).

Call me Cassandra but I predicted that. I suggested several times that we 
should have a general tribute wiki and that we shouldn't limit such a project 
to a single event. I still think that that would be a viable project.

So yeah - until our load/server/software problems are fixed the non-developers 
should concentrate their efforts on the already existing Wikimedia projects 
and subprojects and on devising processes on how to best form new ones. In 
short; no new Wikimedia projects. 

--- Daniel Mayer (aka mav)