I've been looking for a way to give back to this community (and the GNU/Linux community as a whole) for months now, and recently, find myself with a more flexible schedule. Well, I'm not one who has trouble putting two and two together... This is my chance! But what can I do?
I figured my skills would best be put to use contributing to the ArchWiki, and spent the last few days thoroughly exploring the wiki and related forum posts. It seems a major effort began back in 2006 to dramatically re-structure the wiki, but activity seems to have died off recently, as far as I can tell.
I'm left with a few questions and ideas I'd like to put forth before really "diving in" and getting dirty.
* There was a fair bit of discussion about moving from MediaWiki to another system--is this still in contention?
* Which structure is the wiki aiming for:
a) Monolithic: Massive articles covering everything and everything else about a topic (e.g.: the Beginner's Guide, the Xorg article, etc.)
b) Fragmented: Smaller articles aimed at covering one very specific topic and covering it well (borrowing from the UNIX philosophy)
Personally, I think the 'fragmented' approach leads to a much more maintainable wiki. Furthermore, it tends to reduce duplication of effort (there are probably a dozen different articles all containing information on "configuring pacman," for example).
I think right now the wiki is in need of a "benevolent dictator" to take charge and make some major direction-changing decisions (wiki structure, templates, conventions) before we can really work towards a real "professional"-looking and feeling wiki.
Yes, the alternative argument has merit: "It's a community effort! We should all decide how to shape the wiki, etc." Unfortunately, this decision-making process has been going on nigh three years now. We need definitive action!
I'm willing to volunteer a significant amount of my time towards improving the ArchWiki, but I think the focus needs to shift to the underlying framework first, and enough of this "Just fix up spelling, etc. while we decide how to improve the wiki" nonsense.
M*cr*s*ft: Who needs quality when you have marketing?
Monolithic certainly works well for the installation guides. Compare our guides to this convoluted, unreadable mess of links.
Consider that link to be the anti-archwiki.
More important than the format of the wiki, is the quality of the articles. We have been striving for high-quality entries that are easy to follow, provide smooth reading, and are complete.
If you want to contribute, do so in whatever style you excel at, just be sure that the quality is there.
In my experience, fragmented articles often do not give enough frame of reference and are too easy to botch up by not providing a thick enough base to ever expand upon. Large articles are easier to maintain in my view, because it is easier to trim out unnecessary bloat by making an additional page, if required.
You mention the plethora of redundant articles (you used configuring pacman as an example)... If you want a place to start, perhaps simply cleaning up what we already have (thus making any potential future migration easier). You could start by condensing such redundancies, cleaning up other pages, or filling in obvious blanks in the wiki (I can't think of any off the top of my head, but I know I've stumbled across a few of them).
Cleanup is always welcome, whether that involves putting some of the template tags on existing articles (deletion, stub, unclear, etc.) or sitting down and filling out some of the articles. Just reading through the wiki you will find things that could use some TLC.
I'm trying to change the current attitude I perceive as "just affix another band-aid to that man's gaping wound." Work with me!
Misfit318, you found a good example of disorganized documentation, yes. As long as we're finding messes of links, why not take a gander at the Table of Contents for the Official Install Guide or Beginner's Guide articles?
I imagine people use the wiki primarily in one of two ways:
a) Search for a specific article. In this case, I think smaller, focused articles yield more relevant results. Currently, for example, any search for Xorg takes me to a page with WAY too much information for most cases--unless I'm installing Xorg for the first time.
b) Navigating from the main page, following links through each category. I think the wiki currently does an acceptable job here... and appreciate how difficult it is to create an effective scheme.
Ghost1227 and toofishes, I want to address the wiki "framework" first. The wiki needs a well-written style guide/doctrine. It needs templates for new articles (HOWTOs, software installation/configuration, hardware, etc.) so articles don't need cleanup in the first place! There's a hole in your boat: fix the hole before focusing on pumping the water out!
M*cr*s*ft: Who needs quality when you have marketing?
calm down... I agree, though, the wiki needs another makeover.
I don't think it's a good idea to use the beginners guide as reference. I vote for exclusion of that article from the discussion of
specialized vs. well.. uh the word was monolithic
because it's a guide from start to end about something special (installing and setting up an arch linux system). it should be huge, it should be covering almost everything about almost everything relevant in that stage of the installation/configuration process. It's not helpful for new users if you provide a list of articles which might or might not be exactly what they are looking for (which is an understandable guide for setting something up for the first time)
furthermore specialized does not exclude monolithic. I don't think the xorg article is too long. what would you throw out, pointone? It's about the configuration of Xorg, thus it contains exactly what it claims to contain and only that. the fact that Xorg is a beast to configure changes nothing.
The goal is not to have small articles, but to have articles that cover exactly what they claim to.
as far as i can tell, you, pointone try to suggest a generalized structure with this thread. when you say you want more focused articles, what do you mean. I think we all agree that losing information is not part of a successful restructuration, which means that you need to get all the information about Xorg into smaller articles which are then spread over the wiki (preferably in a category Xorg). How is that helpful? you already have this. the Table of content does exactly this. it just catenates your focused pages and provides links to the relevant section instead of a new page.
I agree though on the fixing the hole before pumping water out
I will see that i can use some of my term break to fix articles in the wiki which are sorely out of date (I noted some of them somewhere, if only i could find my notebook (a real one, like with dead trees in it and stuff )