Any wiki suggestions?
Mar. 16th, 2008 01:34 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png)
Time for the next bit of tech for work: I'm looking for a good wiki. CommYou probably needs a couple of wikis -- one on the internal development server on "how you build this thing", and one on the soon-to-be public server for documentation. (I'm leaning towards a wiki for my documentation, since it allows me to open it up to trusted members of the community to help out. In general, one thing I think LJ's done right is getting the community involved.)
Anyway, there are about six million wiki platforms out there, so I'm curious whether my friends have any suggestions of good ones. My needs include:
Opinions? Eventually I'll probably switch over to using Querki, but I really can't afford the month needed to get that project bootstrapped right now...
Anyway, there are about six million wiki platforms out there, so I'm curious whether my friends have any suggestions of good ones. My needs include:
- I'm more interested in ease of maintenance than most other features: I'd like something that's fairly easy to get configured up and running.
- I don't think I need massive power and complexity -- my needs are pretty straightforward. Some straightforward way to include images would be Very Useful, though. (This is an aspect that many wikis fall down on.)
- It needs to be stylable, but that mostly means that I should be able to apply CSS easily.
- I lean slightly towards something based on JVM/MySQL, since those will certainly be installed and working on the servers, but I'm open to Perl/PHP/whatever, so long as it's a common platform.
- Decent access control is utterly crucial, so I can open editing up to specific members of the community while keeping admin privs locked down. This is *not* going to be editable by the general public, so some kind of group management would be helpful.
- The ability to shove all the wiki framing out of the way would be Very Nice: for purposes of the average reader, I'd prefer that it look a bit more like a conventional web page.
Opinions? Eventually I'll probably switch over to using Querki, but I really can't afford the month needed to get that project bootstrapped right now...
(no subject)
Date: 2008-03-16 06:07 pm (UTC)(no subject)
Date: 2008-03-16 06:16 pm (UTC)(no subject)
Date: 2008-03-16 08:28 pm (UTC)Anyway, I'll give it a look. When I was checking out wikitech last year, MoinMoin was certainly one of the more interesting ones. (Although it does have a mild strike against it that we aren't otherwise using Python, so it adds another server tech that needs to be maintained...)
(no subject)
Date: 2008-03-16 09:14 pm (UTC)(no subject)
Date: 2008-03-16 10:22 pm (UTC)(Doing some research myself, the best-looking system seems to be XWiki, which is JVM-based and does plugins -- but conspicuously talks about the plugins being in Groovy, not Java. This enhances my respect for the developers...)
(no subject)
Date: 2008-03-16 10:39 pm (UTC)I've had a very very limited amount of direct exposure to Java, but the associations that programming Java tends to bring along are enough to act as an active discouragement to learning more about the toolchain. Combine this with relatively poor libre-free JVM support, and the verbosity of the language making the primary way of writing code GUI based (something that I've never been a fan of in my coding: I'm a command line/vim guy), and any interest I have in Java due to some Fancy New Announcement tends to go out the window pretty quickly once reality sets in.
I think that this is just a result of the environment I've typically worked in: Reading your entries has brought a significant level of respect for the Java World from the eyes of someone who is obviously a competent programmer. In general, if someone mentions Java without many other details, over something like say, Python, I have an expectation that the code will be either more basic or more complicated than (or both) a comparable alternative written in a different language. (C gets a 'more complicated and less basic' mark, on the other hand.) There is clear evidence in the posts you've made here that we have similar ideals for workflow of development of a project (even if not being my own boss means that I seldom actually get to follow up on them), despite what I would typically consider a chasm of difference in our toolset.
Hm, this has gone most everywhere. I think that it comes down to "Java has become the entry-level programming language: therefore, lots of programmers in Java only work at the entry-level, and I prefer to work in a different environment than people who work at the entry-level, and therefore my contact with Java has been limited." It's possible that I might like Java but my limited experience suggests that my nature is more Pythonic than... Javaic? Javaish? :)
(I can expand on any of the points if you're interested; I don't know what kind of environment you've typically worked in, so it's possible that this doesn't make any sense. I get the impression that you're older than I am by some number of years, which would probably mean that you were out of the 'entry level' arena long before Java became a defacto standard in it, for example.
(no subject)
Date: 2008-03-17 01:10 am (UTC)I wouldn't actually *recommend* Java at this point, mind. It was a good language in its day, but it stopped being cutting-edge about ten years ago, when "object oriented" stopped being new and cool. Java has become an irritatingly stodgy language, with the result that it's been thoroughly lapped by C#. (Which is by no means perfect, but is at least making a credible effort to keep up with the times.) Java is the great bastion of object-orientedness as a religion, with the result that it has pretty much completely missed the functional revolution. (Much less the declarative/pattern-based one that's gradually coming down the pike.)
But the JVM is another story. That's just a platform, and it's a fairly decent one -- perhaps not quite as good as the CLR, but good enough for most purposes. And there are *lots* of languages on it, some of them just fine. For example, I'm likely to move my active development over to Groovy pretty soon: that looks to be a reasonably modern and sensible language. I mean, it's not that Java is a *bad* language, it's just missing a bunch of important features -- Groovy appears to add those in. The nice thing about the JVM is it has a colossal installed base and a huge number of libraries; pretty much all of those languages can use pretty much all of those libraries. That's why I chose it as the basis for CommYou: I can have my library cake and eat the languages I want.
So I can understand the viewpoint, but I'd caution you to remember to keep the tool and its users separate. It is entirely possible to use it to produce good code -- it's just a bit more effort than it should be. The fact that it has become one of the entry-level languages just demonstrates how much staying power it has developed, and that staying power is almost always because it provides the junior engineers with enough capability to get their work done. That's not a bad thing, per se, so long as you don't let it limit you...
(no subject)
Date: 2008-03-17 01:15 am (UTC)you were out of the 'entry level' arena long before Java became a defacto standard in it, for example.
Oh, sure. I mean, I started doing really hard-core Java programming back in, what?, I guess 1996. (During the Netscape/IE 3 generation.) At the time, I was writing what may well have been the most sophisticated applet in the world: a self-installing, self-modifying, cross-browser multimedia client applet for an educational MUD we were building for the DoD. Fun stuff, and at the time it *was* quite cutting-edge -- indeed, the idea of an object-heavy language like Java was still controversial, and there was a lot of resistance to the idea that you could do "real" work in it. But that was a while ago now...
(no subject)
Date: 2008-03-16 06:18 pm (UTC)Now, this may or may not meet all your needs. It's not currently being supported so far as I know: WYSIWYG. The example Wikis are no longer up, for instance.
But if you want a Wiki that gets out of the way and lets you work, with password protection, it's pretty good.
(no subject)
Date: 2008-03-16 08:29 pm (UTC)(no subject)
Date: 2008-03-16 07:22 pm (UTC)(no subject)
Date: 2008-03-16 08:31 pm (UTC)Out of curiosity, which problems of wikis are making you unhappy?
(no subject)
Date: 2008-03-17 01:03 am (UTC)(no subject)
Date: 2008-03-17 02:06 pm (UTC)(no subject)
Date: 2008-03-16 08:38 pm (UTC)(It also looks like it's significantly higher-powered than I really need, though, so that's okay. Not quite the right tool for a project that's still this small...)
(no subject)
Date: 2008-03-17 01:07 am (UTC)(no subject)
Date: 2008-03-17 02:10 pm (UTC)But anyway: I could probably swing $1200, but I don't love the match in terms of the license scheme. In particular, my intent is to open the thing up to members of the community, and I have no idea how many people that's going to be. Confluence looks to be designed more for in-house enterprise use, but that isn't exactly what I'm doing.
So I'll probably go with the freeware (having come across XWiki, which is probably a better match for my needs anyway). But thanks for the pointer...
(no subject)
Date: 2008-03-17 01:09 am (UTC)(no subject)
Date: 2008-03-17 03:21 am (UTC)(no subject)
Date: 2008-03-17 03:22 am (UTC)(no subject)
Date: 2008-03-17 04:25 pm (UTC)(no subject)
Date: 2008-03-17 04:31 pm (UTC)(no subject)
Date: 2008-03-17 04:52 pm (UTC)(Actually, it's really very ambiguous about personal use. It's kinda-sorta free for educational use, but I don't see *anything* talking about personal use licenses. I suspect that, if it's just one or two people, they just don't care much...)
(no subject)
Date: 2008-03-17 04:24 pm (UTC)(no subject)
Date: 2008-03-17 04:35 pm (UTC)Doku does look quite decent, though. Syntax is a little idiosyncratic, but that's not unusual, and it's richer than many. Image support is quite powerful in some ways, although it doesn't appear at first glance to do automatic management of them. (Which I would really love -- it appalls me how few wikis do this.) I don't love PHP, but it's probably good enough for this job.
Useful suggestion -- I'll keep that in mind as I evaluate my options. Thanks!
(no subject)
Date: 2008-03-17 05:13 pm (UTC)(no subject)
Date: 2008-03-17 07:24 pm (UTC)(And FYI, it has a Windows self-installer that's pretty slick: took all of five minutes to get the thing up and running, and I'm playing around with it now. *Damned* impressive stuff -- indeed, powerful enough under the hood that I'm going to have to seriously consider changing course on Querki, and maybe make it an XWiki enhancement instead. It's the first wiki I've come across that already has most of the DB infrastructure that Querki needs. I may play around in my spare time, and figure out how long it would take to get the LARP-writing module adapted to it...)
(no subject)
Date: 2008-03-17 07:34 pm (UTC)(no subject)
Date: 2008-03-17 07:36 pm (UTC)(no subject)
Date: 2008-03-17 07:41 pm (UTC)(no subject)
Date: 2008-03-17 05:54 pm (UTC)I think that's a feature, rather than a problem, if you ever intend to let the public see the thing. Wikis are working space, web pages aren't. Trying to make one look like the other will tend to confuse the audience.
(no subject)
Date: 2008-03-17 07:39 pm (UTC)*MediaWiki* is designed as a massively co-operative enterprise, with social controls taking the place of formal ones. That's one valid and common use case for wikis, but far from the only one; indeed, arguably not even the dominant one. It's pretty hard to do successfully -- if you don't have a massive user base, keeping the spammers at bay is difficult.
I suspect that the majority of wikis aren't even remotely that publically-editable (a very large fraction at least require some kind of approval for editing), and many don't look quite so much *like* a wiki. For example, it's quite clear that LJ's public face is more or less entirely built in a wiki: they just choose (correctly, I believe) to not put the wiki-ish features front and center. They're there under the hood, I'm sure, but the average user doesn't see things like edit links.
So it's not so much a matter of matching the tool to the audience, as the look-and-feel you apply to that tool. I want a wiki for my infrastructure, for all the usual reasons: it's easy to edit, I can easily provide editing control to trusted users, it deals with version control, and stuff like that. But to the typical CommYou user, I want it to look more like a web page, because that's what the typical user cares about. They don't care about how this site works, or how one edits it -- they just want to find the answer to their question efficiently, with as little hassle as possible.
(One of the things that's appealing to me about XWiki, the current leading contender, is that it provides immense customization of the look-and-feel. You can have one skin for the administrators and editors, and another for the normal end users. Very sensible, and a feature I might make use of...)