Wikipedia talk:FAQ/Archive 1: Difference between revisions
Larry Sanger (talk | contribs) No edit summary |
Larry Sanger (talk | contribs) mNo edit summary |
||
| Line 186: | Line 186: | ||
'''Q. Are the casual dropins from the recent publicity drives and recruiting efforts given a misleading impression by the front page? In other words, is any work in progress to adequately define organizational and business models such that "Wikipedia: The Free Enclycopia" can explain what its precise intended committment to Wikipedia.com users and contributers consists of and how it plans to attempt fullfill them?''' |
'''Q. Are the casual dropins from the recent publicity drives and recruiting efforts given a misleading impression by the front page? In other words, is any work in progress to adequately define organizational and business models such that "Wikipedia: The Free Enclycopia" can explain what its precise intended committment to Wikipedia.com users and contributers consists of and how it plans to attempt fullfill them?''' |
||
A. This is a complex issue, and it is not at all well-agreed or well-understood exactly what the problem is in the first place. It requires some discussion to find solutions acceptable to |
A. This is a complex issue, and it is not at all well-agreed or well-understood exactly what the problem is in the first place. It requires some discussion to find solutions acceptable to Wikipedia contributers and users. Some people are under the mistaken impression (or are conveying such an impression, anyway) that the leadership of the project--[[Jimbo Wales]] and [[Larry Sanger]]--regard Wikipedia as a "commercial venture," which is just ''false.'' They have been discussing ways of making the project entirely nonprofit for many months now. Please join us on [[Wikipedia-L]] or on http://meta.wikipedia.com/ if you wish to assist the Wikipedia community with clarifying these questions and articulating well-informed answers. |
||
Revision as of 09:04, 20 February 2002
Discussions not quite ready for the FAQ.
Q. I have just checked the most wanted articles list. both the articles I checked seemed complete Why are they Listed as wanted?
A. Coders are working on this as of 16 Feb 02
Q. How do I get my own wiki? For example for my own web site or business or family discussions or maybe a church group or some other organisation?
A. See the [UseMod:UseModWiki UseModWiki home page] to download the software used to build this site. (Note that the Wikipedia version is slightly different than the official release, but it is mostly the same.) --CliffordAdams
A2. Also, I am probably willing to hook you up with a wiki of your own at http://www.wikicities.com Someday this may be a full service, for now it is ad hoc. --Jimbo Wales
Q. What's the Wikipedia convention on referencing (not quoting from) outside Web and traditional-media sources? What's the line between referencing, "I think _Foobar For Lusers_ is a great source", and shamelessly plugging something? (asked 26 September 2001)
A. I'd say it's a matter of stating the facts from the neutral point of view.
Q. I am asking this again, because I do not feel I received an adequate answer, last time. To write pages on Math, the authors need to be able to include a large number of math symbols, graphs, etc. This is not a question of linking to someone else's images. These are items we need to be able to use freely and in our own original way. It would require, at present, uploading the content to one's own websites, assuming we have them ,and linking to this. For example, to discuss differentiation and integration in Calculus, one needs to have: an Integral sign, a summmation sign, a symbolism for limits, etc. Somehow, writing, the limit of sin(x)/x as x goes to infinity, just doesn't cut it. Already we are all missing the small epsilon for "belongs to," the symbol for "is a subset of," and many more in the Abstract Algebra entries. I think this will impede the addition of much material in the Math area.
A. This is a difficult problem. Part of the problem is that html web pages don't do math symbols very well at all. So almost any solution will be a crutch. Within the next couple of years, browsers should be able to render MathML. This will be a time of much joy. Between now and then, I'm open to suggestions, but I think this is a really hard problem.
(Added note by Zephyr) The best way to put in math would probably be to write it in TeX (or LaTeX) and then use TtH. This will generate an HTML file with all of the math symbols in text form (no images). Here is an example, and here is another. TeX is almost inarguably the standard for mathematical writing.
TtH produces beautiful HTML only equations, but they can't be inserted without being garbled. If this could be fixed, then TtH would definitely be a fabulous answer. -- TedDunning
As much as I admire Knuth, I must disagree totally. MathML should be the ultimate standard, not TeX. For one thing, TeX is more display-oriented and MathML is more semantics-oriented. MathML is also the standard supported by the World Wide Web Consortium, and will therefore likely be supported by browsers (Mozilla and Microsoft are committed to it, for example). TeX is used by a few academics, but they don't represent the real world.
Fooey on that. TeX is definitely only used by a few academics, but those few academics happen to include most of the mathematicians in the world. MathML is not usable as an authoring tool (and TtM exists for that anyway). TeX may be display-oriented, but few people actually write in TeX directly. Instead, they use LaTeX or AMSTeX which are both very semantics-oriented. In any case, I happen to live in the present and MathML is definitely in the future and may remain there indefinitely. If we can just figure out how to support the tables that TtH produces then the world be a lovely place (or at least the math pages in wikipedia). -- TedDunning
It really comes down to WYSIWYG (What You See is What You Get) vs WYSIWYM (What You Say is What You Mean). First of all, I would like to say that TeX is used almost exclusively by anyone writing up math. Can you show me any major publication which does not require submissions to be in some variant of TeX? A popular thing about TeX is that (if you so wish) you do not need to worry about formatting, you can just say "this is a section, format it for me". You can still have this flexibility by setting the templates. Have you looked at MathML lately? I was looking at some MathML examples and it seems that TeX is much more straightforward to write. $x^2-4=2$ becomes <math xmlns="http://www.w3.org/1998/Math/MathML"> <mrow> <msup><mi>x</mi><mrow><mn>2</mn></mrow> </msup> <mo>-</mo><mn>4</mn><mo>=</mo><mn>2</mn></mrow></math>. I cannot speak much for this, everybody is entitled to their opinion, but I think that MathML would have a very tough time overthrowing TeX (although it would be good for the TeX to MathML convertor).
Note: There is exactly such a converter from the same guy who wrote TtH -- TedDunning
A minor but important correction to the above: It is LaTeX in what you in principle should not need to care about formatting, just concentrate on the content and structre. Plain TeX, on the other hand, is mostly purely formatting-oriented low-level typesetting language. Unfortunately, there are times in LaTeX when you do have to care about formatting. E.g. there are no tables that would a) work in paragraph mode and b) at the same time not require manually set the table width. But still, TeX math is the best way to write math, and I'd claim LaTeX the best way to write structured documents, despite it's flaws. The usual rule of "least worst" applies. WYSIWYG is never an alternative to me. Lout is an alternative, and has some advantages (e.g. =>, <=) but some major disadvantages: 'x_i^2' (the natural and TeX way to write this) vs. horribly undreadable 'x sub i supp 2'. Anyway, I think this project really needs a way to write math. Current text-emulation is horrible. I'd vote for using TeX math as it is perhaps the most readable and writable format, and most mathematicians know it. Something like LaTeX2HTML could then be used to render the formulas as images (as mathworld did...). When browsers decently support MathML, a converter from TeX to that format can be used. Whatever people might think, MathML is not the ultimate solution, because it is utterly unwritable.
Note that HTML 4.0 contains many math symbols (all greek letters, integrals, summation signs, logical and set theoretic symbols etc.). I have had no problem typesetting math using these, and occasionly a bit of ascii art. So while it would be preferable to have some TeX or MathML mechanism in place, I don't think it's a big issue and the lack of it does not hold up the math pages right now. (If your browser does not support all HTML 4.0 entities, report it as a bug to the manufacturer.) --AxelBoldt
Sorry, Alex, but that's just not helpful advice at all at this point. Most popular browsers in use today fail to support most of the symbols defined in HTML 4, with good reason (mainly the lack of those symbols in the native OS). It's also not a reportable bug (as if it were possible to report bugs in Microsoft products anyway:-), because HTML 4 doesn't require that they be rendered correctly. Some math articles would greatly benefit from typsetting features still not available in HTML 4. It is entirely appropriate for us to investigate other methods, and for the even more important reason that HTML won't allow us to save the semantic information about he formulas, only their appearance. --LDC
- Well, I'm told that the most popular browser, IE (version 5.5) displays all math and greek symbols from HTML 4.0 just fine. (I don't know if one has to download free fonts from Microsoft though.) And free software is usually much better at following the standards; mozilla and konqueror work fine as well. So which ones don't work? Also, I'm all in favor of investigating better math typesetting methods, but I claim that right now the math articles are not held back in any way. --AxelBoldt
How about using EzMath in the pages (http://www.w3.org/People/Raggett/EzMath), and having the wiki translate to MathML or images? It's very readable (and writable) as plain text. Sample expressions: "integral of sin x from 0 to 2pi wrt x", "x^2-4=2", "x squared - 4 = 2".
[Note added by Serge Winitzki] Here is a compromise: you type math in TeX and see HTML and images (http://www.mathcircle.org/cgi-bin/mathwiki.pl?) Unfortunately, the (free) software for this "mathwiki" implementation is not in production shape right now. Please contact me if you would like to get in touch about this. (http://www.geocities.com/CapeCanaveral/Lab/5735/1/email_me.html) I intend to work on improving mathwiki code later this year (I am not the principal author). -- S.W.
This is a very good solution since it is fast and completely portable among browsers. Getting the generated image font sizes to match up with the text would make this an excellent choice. The image description could use TeX so that Lynx users would find it useful as well. --Jonathan
Jonathan, first, welcome to Wikipedia, and second, please stay on top of this. CliffordAdams is the main programmer for UseModWiki, but I'm not sure he's interested in including functionality of the sort you describe. But if he doesn't do it, I suspect we'll be able to interest other programmers to work on his code and do it. Anyway, our math section is very active and competent and there's definitely a crying need for this sort of feature. --LMS
There is a patch for ZWiki called LaTeXWiki. It lets you type in LaTeX and ultimately renders it as a png grafic. The LaTeXWiki page is http://latexwiki.rootnode.com/
Q. Why don't we simply use html?
A. Good question. Wiki software does use some html, but not all of it. One reason is for simplicity when typing--for instance this text is indented because I typed a : character at the start of the line. In order to do that in HTML, I would have to type much more. The wiki syntax also prevents many common errors, like not closing a bold-tag and making all the following text be bold.
Another reason for not using raw html is security--several old versions of popular browsers have security problems that can be triggered by arbitrary "malicious HTML". Even without considering security, malicious users could do things like Javascript popup windows or page redirects if they had full HTML ability. Several "experimental" sites that allowed full-HTML editing have suffered such attacks, including a couple other wikis that allowed arbitrary HTML.
There is a "raw-HTML" option in the wiki code, but it is defaulted to "off" for the public release. I strongly recommend leaving it off for public wikis. (If that option was enabled, you could type "<html>
Header
</html>" to create a header.) See the Meatball Wiki RawHtmlWiki page for earlier discussion of these issues.
The HTML tags directly supported in the current version are i, em, b, strong, tt, and pre. There is also a pseudo-tag "nowiki" which you can use to prevent the wiki from formatting your text (like NotALink--edit to see the tags). The "code" tag acts like a combination of the tt and nowiki tags. The latest release (0.91) adds a new wiki-syntax for h1-h6 (h1 is =text=, h2 is ==text==, etc...).
The Wikipedia may have different needs than the typical wiki. The formatting rules are easy to change, and I am willing to try new things for this wiki if the maintainers of this site want changes. --CliffordAdams (Later, the Wikipedia maintainers have added support for several extra tags like h1, font, sub, sup, font, table, and a few others. Eventually the documentation will be updated.)
Q. What about non-ascii characters? We could write latin1, or (I think this would be better) use the HTML Codes (i.e. ü for "u). This would also fix the math-symbol-problem, cause W3C-HTML 4.0 supports many of them, e.g. Α for a capitalized Alpha, α for a low one, ∫ for a integral symbol ect. Konqueror supports this (and I think IE too, but I don't know.)
A. Just because the codes are defined in HTML4 doesn't mean they actually work in any common browser. See the Wiki special characters page for a detailed discussion of what is generally safe and what isn't. Of course, this will change over time as more browsers come to support more features.
Q. Perhaps this question is stupid or perhaps this is not the right place to ask it, but I want to be absolutely sure before committing any blunder that could hurt Wikipedia: I would like to write an article about Julian the Apostate (Roman emperor), which I would like to base on an essay I prepared for a high school course some two years ago. This essay relies heavily on a number of books. It is not cut-and-paste work and it contains just a few verbatim quotations, but nevertheless I "used copyrighted work" to write it, because I read the books and got most of the factual information from them. I guess it is not a problem to publish this essay on Wikipedia (including a bibliography, of course), or is it?
A. First, see fair use. Excerpts for comment and criticism generally fall under those guidelines. The fact that your paper was done for school adds additional protection in that context that doesn't apply to the Wikipedia context, so it is possible that excerpts that would be considered fair use for a school paper might be more problematical when placed in Wikipedia, even though we might be able to claim that this is an "educational" project. But if they are small excerpts clearly identified as such, collected and used for comment and criticism, I don't think there will be a problem. The worst thing that could happen is that one of the authors requests us to remove it; we either capitulate and suffer no consequences, or else we argue fair use if we think using the material is important enough. --LDC
Q. What constitutes a minor edit?
A. Well, obviously, simply fixing a few spelling errors is a minor edit, as (I imagine) would simple rearranging one or two sentences. A major edit is basically something that makes the entry worth relooking at--either through substantial additions or reorganisation--or fixes a major error.
Q. How can I find a list of all pages linked from a given page?
A. At the top of each page there is a button, Links to This Page which provides two lists. The first is a list of links to the page but to which the page does not have a link. The second is pages with which the current page has a mutual link arrangement. The links embedded in the page are not provided ..... probably because viewed as redundant.
Wow, what a huge improvement to the FAQ! Thanks a lot, Hornlo! --LMS
Q: What is the meaning of the ? links?
A. The ? link indicates that an article with that name has not yet been started. So you can click on that link and start a page with that name. But be careful--there may already be articles on similar topics, or an article on the same topic under a different name. It's pretty important to hunt around for similar topics first.
Q. Why are some links preceded by a slash?
A. The slash indicates that the linked page is or will be a subpage of another. If there is no target specified, the page will be a subpage of the page you are currently on, unless you are on a subpage already. (There can only be one level of subpages. For instance, a link to /Newpage from Algeria/Economy would make the page Algeria/Newpage, not Algeria/Economy/Newpage.) Users can also specify which page the link should be a subpage of, for instance: Poker/High-low split.
There is an ongoing debate on Wikipedia about the usefulness and limits of subpages, with the debate centering around the contextualization and implied hierarchy of them.. See Larry Sanger/Why I am suspicious of subpages and Larry Sanger/The case against subpages for most (all?) of the debate. But it seems everyone agrees that subpages are useful for commentary, debate, and strictly personal pages such as To-Do lists.
- ----
- Yeah, the very case insensitivity whose lack is causing your contributors huge headaches (and causing untold numbers of duplication of effort and redirections) will cause problems when it is implemented. Lots of articles will need renaming to the new standards.
- On the other hand, its the type of problem that gets worse the longer it is ignored. Since you must (probably) go down for a piece while the upgrade is done, anyway, some additional time will be required to automatically convert all article names to the canonical form and detect and somehow handle those article names whose canonical names are the same. The longer you wait, the greater will be this time and the more manual intervention will be required.
- If we had some data on the basics of how Wikipedia was set up (platform, how the articles are stored, etc) someone (perhaps I) might be able to offer some help in this matter. And no, I am not about to download and study the source code to find out. The payoff (for me) is not high enough.
- --Buz Cory
Q: There is a problem with linking to books. Some pages link to Amazon. You probably all know why you shouldn't do it. What should be done with such links ? Should we choose some "official" books eshop ISBNs link to ?
A: Whatever you do, don't restrict users from things
subQ: If they want to, they still can, but we shouldn't promote this.
A: Why link to a bookstore anyway? There's a wonderful system called interlibrary loan in place. In a matter of days, you can get pretty much any book ever printed delivered to your local library. Why not simply link to the entry in the Library of Congress catalog: it has all the information you need about the book, should you want to buy or borrow it. --AxelBoldt
subQ: Does it also work in Europe ?
Q: What is the ruling on cross-referencing entries? I wrote an entry for Tasmania, but then created a separate entry for Van Diemen's Land. This seemed natural, the old term still occurs in historical documents and folk songs prior to the 1850s, and Brittanica has a similar entry. But then I wondered if it was strictly necessary, due to the search engine. Have a look and give me some stylistic pointers.
A: It also occurs as late as the mid 80s on a U2 album. :-)
- MHO is that if the two entries are substantially different then write an entry on both, but if they are just two ways of expressing the same thing then the less common one deserves a redirect to the more common one. This is only my opinion, and may not be shared by others. You can make a #REDIRECT page by typing #REDIRECT at the far left (no leading space, all caps) and then following it with [[the name of the page you'd like to link to]]. --KQ
Search:
Q. Why doesn't Wikipedia's search look at titles of pages? Sometimes, I want to go directly to a specific page title.
Q. How long does it take for new pages to show up in Wikipedia's index?
Augh! My eyes! :-) Please change it back to the way it was, or make the headings much smaller. --LMS
- (ensuing discussion moved to Ben Finney/Wikipedia markup. I think this issue is important to resolve for Wikipedia; if you do too, please continue discussing there. -- Bignose)
Q. Should Wikipedia include topics which are very, very specialized? (I'm thinking about an article on my home town, population 200; or an article about my late grandmother, who could not have been called "famous"; or an article about the history of an old one-room schoolhouse near my house.)
A. I can think of no reason why not. However, if your grandmother's name was Margaret Thatcher or Whitney Houston, don't be shocked if someone moves your homage to make room for what most readers would be looking for.
One of the great things about Wikipedia is that it doesn't really matter how obscure a topic is. Maybe no one will ever read your article. Or, perhaps they will. It doesn't matter. Britannica can't afford the time, money, or page space to have an article about your hometown. Wikipedia can, if you want to write it.
A2. One reason why not is that minutiae would clutter up the search engine. This was the old objection to doing this; now that the search engine looks at article titles rather than complete texts this may not be as significant.
Q. A stupid question, probably. I saw a link *Download Data* in the *Help* page and clicked it to download all the data to my computer. But there was no permission to access the download folder. Is it available only to priviledged users?
A. There is some discussion in the mailing list archives (links above and on main page footer I think) of maintaining a very flat "Democratic" hiearchy consisting of sysops, knowledgable community users, and new users. Basically the new users would have a couple of minor restrictions to protect the site from initial errors by novices learning the Wiki(pedia) Way or local ropes. I notice that my download button has also gone away so perhaps this mechanism has been added or the download was loading the system to badly or nonfunctional. It would seem reasonable to avoid accidental massive downloads by beginners and casual users so perhaps it has been relocated.
Q. A related question, is a business plan or organizational charter for Wikipedia.com intended to be published at some point such that its contributors can be confident it will always be available free to the public at large?
A1 This is under discussion at http://meta.wikipedia.com/ which is an associated space setup for people interested in working on the site and community infrastructure. This preserves Wikipedia.com scholarly atmosphere and focus on improving and adding content to Wikipedia.com articles.
Q. Is the wikipedia code going to be released under a free or open license? The reason I ask is that it seems to me that the information in the wiki entries is greatly diminished if one cannot access the related wiki links and track to more detailed or general information on the subject of interest.
A. From browsing email list archives it appears there has long been an informal understanding among the wikipedia community of developers that while wikipedia.com is a commercial venture the source to operate the site would be open or free. Browsing the wikitech-L archives reveals that this has been now formally documented and finalized by inclusion of the GPL text into the CVS tree on the 14 Feb 2002.
Q. I am concerned about the backing up of data, however and would like to know more about that. How about wiki-mirrors? I wonder how many giga-bytes just the current versions of all the articles take up. --maveric149
A. One developer on the wikitech-L list stated that for testing purposes he had a "old" download snapshot of 21000 plus articles that was 30 MB and provided a link for developers interested in stress testing. I am unclear whether this included previous versions and attribution data as seen on the history page. user:mirwin
Q. Is the complete database (all data required to efficiently setup a mirror or free fork of the site)download functionality currently available or will it be extended at some point in the future for free (as in beer) or for fee?"
A. ?
Q. Specifically is the attribution data necessary to comply with the specified FSF license contributors submit under by clicking submit on change or article submission part of the primary schema of the content database or a separate shema processed by the GPL source code that runs the site.
A. ?
Q. Are the casual dropins from the recent publicity drives and recruiting efforts given a misleading impression by the front page? In other words, is any work in progress to adequately define organizational and business models such that "Wikipedia: The Free Enclycopia" can explain what its precise intended committment to Wikipedia.com users and contributers consists of and how it plans to attempt fullfill them?
A. This is a complex issue, and it is not at all well-agreed or well-understood exactly what the problem is in the first place. It requires some discussion to find solutions acceptable to Wikipedia contributers and users. Some people are under the mistaken impression (or are conveying such an impression, anyway) that the leadership of the project--Jimbo Wales and Larry Sanger--regard Wikipedia as a "commercial venture," which is just false. They have been discussing ways of making the project entirely nonprofit for many months now. Please join us on Wikipedia-L or on http://meta.wikipedia.com/ if you wish to assist the Wikipedia community with clarifying these questions and articulating well-informed answers.