Encyclopedia > Wikipedia:Feature requests
Wikipedia:Feature requests
No one ever reads this page. (Or at least no one who would actually be adding these features.) Don't request anything here. It will be ignored.
The rest of this page consists of feature requests made under Phase II of the software that are still under consideration but have not been implemented in Phase III (which went into operation on 2002 July 20). Links to old requests are at the bottom.
|
Feature Requests, Enhancements and Suggestions (from Village pump)
- Time-Independent version of Recent Changes: Possibly viewed in terms of number of page views since. Alternatively, Unexamined Changes, those pages which have changed but no one has checked.
- Consider moving from sourceforge to something such as Bugzilla (http://www.bugzilla.org/) or a stripped down version of the sourceforge code. Also considered moving Wikipedia to BerliOS (http://developer.berlios.de), but it would mean moving the implied private user database to a third party.
- Confirmation box for editing an earlier version: Disliked because it makes reverting more troublesome. There is already a bold alert which should be sufficient.
- Add a floating quickbar on the right.
- Add the ability to talk to non-logged in users (IPs).
- What we need is a basic trust metric, a way to distinguish in Recent Changes etc. between an anonymous, unknown user and a well established Wikipedian with a good reputation. Also, being able to talk to anons would make it possible to explain to them why vandalism on Wikipedia is futile (w/ a "Why vandalism doesn't work" page)
Related Changes and self-links
- Wouldn't it be better to change "Related Changes" so that it includes changes of the current page (and maybe also pages that link to the current page)? -Martin
- Yes, I think that would be better. Patrick 13:26 Dec 26, 2002 (UTC)
I'm the guy who removed the self link on list of musical topics. I see why they wanted it now, and I very much agree that Related Changes should include changes made to the current page. Maybe if for some crazy reason people didn't like this, it could be an option in their preferences so that related changes doesn't include the current page. Then, surely we could just automatically strip out self links, as they would then be rendered completely useless, and there would be no possibility of a false positive. Just an idea for the automatically stripping out self-links, there may be some reason against it, but I definitely think related changes should include the current page Smelialichu 17:28 Dec 26, 2002 (UTC)
Wikipedia should be XHTML compliant so that it is easy to convert articles to other formats. See http://www.w3.org/TR/xhtml11/ --TakuyaMurata
They'd prevent a lot of heated arguments and confusion. Lir 02:11 Nov 15, 2002 (UTC)
View linked status on Recent Changes
In order to make sure no new orphans get created - or at least not much - it may be useful to have a note show up in the page's entry in on the Recent Changes page. Maybe a red exlamation mark or so. As soon as the page is linked, it should disappear. Something similar could be done if the article is a stub (I think the definition is an article without a comma?). These tools would make detection of new "malicious" entries easier. June 24 2002 --- jheijmans
It may be useful to be able to see in an article that somebody has voted for it, at least in the case of rewrite/wikification. This may remind an editor that it should check the remarks placed there and - if he thinks it is solved - react and/or remove the article from the queue. This may be tricky though, since we can vote to get an article on that list, but not to get it off. jheijmans
Where have the "votes for NPOV/rewrite/deletion" pages gone? Mswake
Probably wouldn't want it linked on the sidebars, but I think it'd be quite cool to have a Most Linked page that acted like "Most wanted" only it includes existing articles. DanKeshet, Saturday, June 22, 2002
- Yeah, I think that would be cool. Displaying a link to this on the sidebar might be an option for people to select in their preferences (However, I don't see any problem with having this be displayed by default in the sidebar though). Determining popularity in this way is similar to how Google ranks pages and may also be tied into our search function (so long as this wouldn't violate Google's PageRank technology patent). This should also be easy to do: Just write a script to count the number of entries on each articles's "pages that link here" page and then rank the results at [[Special:Most linked to]]. --maveric149
(June 19, 2002)
Would it be possible to not regard links between the user: namespace and the articles? This would clean the "Pages that link here" list for the users that mention articles they edited on their page, and possibly reveal them as orphans. Same thing for the talk: namespace.
I like having these links show up. However, I think it perfectly justifiable to make this an option set in user preferences, with the default (which is what shows up to visitors) to not display these links. Alternatively, we could have things listed in different groups. That is, "The following encylopedia articles link to ..." and "The following other Wikipedia pages link to ...". As for detecting orphans, I agree with jheijmans that links from other namespaces should not count. — Toby Bartels, Wednesday, June 19, 2002
It should be cool to be able to put a latex environment to edit equations. It's a personal project that I'm working on (WikiLatex) but I should be great to have it here also.
Fix "Stub articles" sort order
The current sort order for stub articles is nearly useless. What's needed is to pick a threshold for what constitutes a stub article (e.g. 1000 characters), and then sort that list by how many pages link to the given stub. This would make the "Stub articles" list work similarly to the "Most wanted" list, the latter of which is very effective. RobLa -- April 13, 2002.
Would be nice to have:
- Serif font (?)
- Footnoted URLs should be listed as footnotes! Currently you just see the [1], [2], and that's all... (In fact, all external links should have their URL's listed at the bottom as footnotes. AxelBoldt)
- Some ability to link higher-resolution images for diagrams, maps, etc.
At some point the number of users statistic is becoming meaningless due to accidental log ins which create multiple users or users who are just no longer active. Would it be possible to take users who have not been active for a long period of time (3 months?, 6 months?) out of the system or out of the statistical listing - then the statistic could be number of users active in the last x months? Trelvis Mar 12
The most popular page should have an explanation of what the number is. How long a time period are those numbers collected under? Could it be reset every week so it shows this weeks most popular, rather than the most popular all time? Trelvis
Could the search list be ranked either based on traffic or like Google by the number of pages linked to it, so that the most popular pages come up first. This would prevent some of the wading through a large list of obscure pages and redirects which are less popular?
When you get a search result could we add an option to do a Google search on the subject? (I found this useful in the old software)
Also when the search results come up any wiki links in the example text which have a searched phrase do not work because the html bold tags are added into the link - so far I haven't seen many accidentally formed pages with the bold tags built in, but this should be fixed. It might be on the bugs page already.
I know I have seen a request for advanced search capabilities somewhere, but I will restate that request here for compactness. Trelvis
I would like to have redirected articles and "Complete list of Encyclopedia topics" pages omitted from search results. AxelBoldt
- I agree that it shouldn't search the body of a #REDIRECT, but it should still search the title; otherwise, the failed searches page become unnecessarily longer when users type in a misspelled word for which there is a redirect but which is not used in any regular article.
Looking at mis-spelled search requests for 'Circumsision', 'Circumsicion', 'Circumsission' and 'Lamberghini', we should probably use something like Soundex or Metaphone to search for a sound-alike article if a literal search fails. Note that recent versions of PHP have a metaphone() function built in.
See http://www.zend.com/manual/function.metaphone.php for more details.
In fact, I would go further: search engines should
- match exact (case sensitive!) matches first, and then
- match (case insensitive) 'exact' matches second, and then
- I disagree strongly; we should never be case-sensitive. --BV
- fall back to ignoring diacritics (as if ASCII-only), and then
- fall back to Metaphone or other similar scheme (just replacing all vowel sequences by '*' normalises to a sufficient extent for many purposes)
- Agreed. --Brion VIBBER
Just looking at the failed searches shows that about half of them would succeed given some very simple normalisation. Many of the others would work if a combination of guess-the-spaces and stemming was used. Wikipedia is small compared to the Web, and so techniques like this will improve recall without deluging the reader in dross, providing that exact matches and article-title matches take priority. The Anome
The failed search page could use some work. When you type in a single letter (for example, some poor chap who wants to find the entry on the letter J) it tells you that you probably entered a word with less than three characters, this should read less than TWO characters, as two character searches work. I also suggest adding the following text to the failed search page
If you are looking for Wikipedia articles on letters in the Latin alphabet, please use these links: A, B, C, D, E, F, G, H, I, J, K, L, M, N, O, P, Q, R, S, T, U, V, W, X, Y, Z
That would make it MUCH easier to find the letters from the search bar.
Namespaces
I'd like a way to list all pages in a namespace. For example, http://www.wikipedia/wiki/wikipedia%3A should list all pages in the Wikipedia namespace that the current user can access, and http://www.wikipedia/wiki/special%3A should do the same as special:Special pages. --Damian Yerrick
It would be nice to have a history table (and an auto-generated graph as well if possible), showing the trends, day by day, week by week, etc. for each of the monitored variables. This would give an instant overview of what's going on.
See MRTG for an example of this sort of thing (for network traffic in this case), or the trends graphs at seti@home for another. -- The Anome
See also the plots at http://www.distributed.net/statistics/ --Damian Yerrick
A very minor issue: I don't want my minor edits showing up on my contributions page. I don't consider myself to have contributed the article on Agatha Christie, for instance, and my contribution to it (a typo correction, IIRC) was so minor as not to deserve notice. I would not, however, mind have the page list articles I instigated, e.g. Dziga Vertov and Dave Brubeck--those, in my mind, are more properly contributions. Best, Koyaanis Qatsi
- It seems this has been changed, but I don't like it at all. Previously I've been able to set my preferences so that the meaningless distinction between major and minor edits effectively disappeared. No longer is this the case. --Zundark, 2002 Feb 2
Listing New Articles on User page
It would be nice if the ten newest articles that someone started would appear at the bottom of their user page. --Chuck Smith
A source:namespace that only can be edited by sysops or trusted longtime users and be called upon by individual lines in an article. For example, typing [source:Origin of Species/Chapter 1{1-15}] within the edit box of an article causes the display lines 1 through 15 of Chapter 1 of Charles Darwin's Origin of Species in that article (working in a similar way as placing a url of an uploaded image in the same edit box -- except the result will be selectable text). This way, one could call upon any set of lines within the Origin of Species and annotate until they are blue in the face without changing what Darwin said.
www.seedwiki.com has a great idea in its page editing: popup menus with all Wiki Shortcuts they use. When one Shortcut is selected it should be inserted in the text. On the bottom where Wikipedia writes DO NOT USE .... it has a nice summary of Wiki text formatting. Would be helpful. -GillianAnderson
List the number of pages which are redirects. Calling redirects junk pages is not very accurate. --Chuck Smith
This is a rather esoteric feature, so bear with me. It would be very cool to be able to enter machine readable metadata about pages, and specifically, to allow RDF triples to be entered.
A crude prototype of how this might work is running on the RDF wiki (http://swag.webns.net/rdfwiki). The basic concept is to have three URIs where the second URI describes the relationship between the first and the third (e.g. <http:...Mona+Lisa> <http:...Resides+In> <http:...Louvre>)
The way that I could see this working is that one namespace exists for relationships (the second URI), and that the first and the third URIs are normal Wikipedia articles. The relationship lists would reside on a special page associated with the first URI.
-- user:RobLa
I think there is a need for some sort of metadata system on Wikipedia.
XML/RDF data source would make it much easier to accomplish. But for the
time being with are stuck with raw text.
Is any metadata namespace or something along the lines the previous feature
request in the works ?
August 16, 2002
The most wanted feature is great, but it doesn't give enough results. Right now, most of the most wanted are things like '11th century BC' Olof
- It probably would look better without that kind of time line article cluttering the listing, (Currently 26 out of 50) Most of them would only be templates anyway. Perhaps if we each took one on they could be cleared up quickly. Even so increasing the number to the 100 most wanted would make sense. That would give each contributor a greater chance to find something in his own filelds of interest. Eclecticology
- 21 out of 50 (April 15, 2002). But if we put stubs into the Year in Review articles, linked to other Years in Review, we would link to other Year in Review articles and bring them into WantedPages. I'd say if the first character of an article's title is a digit, demote it by 3 links in WantedPages. --Damian Yerrick
Plain text list of all articles
Hi! First off, I'd like to say that I think the new Wikipedia scripts are wicked cool. So, here's my little itty bitty feature request: you know the special page that lists all articles? There should be a way to fetch a text/plain listing. Just straight up one-article-name-per-line text. (The reason I'd like this is so I could fetch such a listing via cron, say, nightly, for tab completion in my Wikipedia emacs thingy.)
Maybe a Least wanted page, akin to the Most wanted one, is useful. It gives a list of those articles that nobody visits, and these pages may needs some visitors to edit them, or maybe to put up links at other/better places.
Since it is essentially no different than the most wanted page (just a small change in the query) I suppose this shouldn't be difficult to make. jheijmans
I agree, but let's call it the "least popular pages" or "least visited pages" instead. What should be the definition? We could sort them by
- total number of visits ever, or by
- total visits since a given date, or by
- least recent visit timestamp, or
- (my choice) use an exponentially decaying estimate of visit rate, with a decay half-life of about (say) 1 month.
Note that essentially all the pages will get visited by robots roughly daily, so pages not visited by people may get lost in the noise. Wikipedia visits for the most wanted pages seem to follow Zipfs law, and probably so do the least popular ones. The Anome
- I agree, the name least wanted is very well chosen. As for the definition, I think any will do. The most wanted uses the second option. I figure that bot-visits do not really matter, assuming all pages approximately get the same number of bot-visits. One disadvantage, however, is that new pages will automatically show up there as well, so maybe the third option is better, or, your preference (though that requires more programming of course, so my 'it's easy' argument no longer holds. jheijmans
Links to redirects to nonexistent pages
If I spontaneously link to a nonexistent page, I sometimes worry that the page already exists under another name. Well, if this bothers me, then that's what search engines are for (at least assuming that I'm looking for words with more than 4 letters in them!). But I realise that not everybody else is so thoughtful, so often I'll create orphan redirects to pages that I'm working on, using names that I think are likely spontaneous links.
What's missing is an option when both occasions arise. I make a spontaneous link to a nonexistent article, and while I know that it doesn't exist under another name, I worry that other people may spontaneously link to the same topic under another name. I would like to redirect alternative likely names to my name — I mean, to the one most in line with Wikipedia:naming conventions. Unfortunately, if I do this, then when such a spontaneous link appears, people will think (from looking at the link) that the article has already been written, when it hasn't. This is quite unsatisfactory — indeed, I've voted such redirects for deletion to avoid just this problem.
My proposed solution: When creating a page, each link is followed enough to see if it exists. I say, see if it's a redirect too, and if it is, then follow it further to see if the page that it redirects to exists. Then if it's a redirect to a nonexistent page, format the link as if it were itself a nonexistent page.
Toby Bartels, Friday, June 28, 2002
I was just editing a page, and I accidentally pressed return when the scope of the cursor when somehow connected to the submit form buttons but not inside a specific text window.
This caused the browser to react as if I had pressed the Save button, which I wasn't ready to do yet.
No real harm done — I only wanted to Preview once to check for typos, and it turned out that there were no typos, so the only thing that was missing was that I didn't get a chance to check the minor edit box.
However, I think that it would be safer to set things up so that the default button is Preview, rather than Save.
I'm not sure what in the HTML code makes the Save button the default; perhaps the browser (IE) is just guessing, and all that we have to do to make it guess correctly is to put the Preview button first???
That would be a good idea anyway, to make newbies think about Preview before they think about Save.
Also, while I'm on the subject, it would look keener if the Cancel link were a button as well, even though that's completely unnecessary, since it has more in common with the Save and Preview buttons than with the Editing Help link.
That's just for looks though, not important.
— Toby Bartels, Wednesday, July 3, 2002
- I have to disagree about the default behavior with return/save. I for one hit enter all the time to save, and this is the most common behavior of webforms like edit windows. --maveric149
It certainly is the most common behaviour, but I think that we should encourage uncommon behaviour.
Too often I see edits where somebody corrects minor typos that they just made — I did this often myself when I first started out.
If people Preview when they expected to Save, this causes no real harm; if people Save when they expected to Preview, then things can get messed up, even if only temporarily.
But I do agree that my feature request should be implemented only if writers agree with me in this respect rather than with you; this will merit discussion.
— Toby Bartels, Wednesday, July 3, 2002
There is no good solution to such a black/white issue. Perhaps the problem is in viewing it as an either/or situation. My ideal solution would impose a "default default" behavior for newcomers (Preview by default, in this case), but can easily be customized by more advanced users to do the more 'natural' thing (Save by default). Just as the login info can be saved on the local machine, so would be such customization preferences. I'm sure you can think of other such useful preferences! David 12:38 Aug 14, 2002 (PDT)
For myself, I'd be quite happy if pressing return did nothing. It does nothing inside the big text field, after all, so why should it do something inside the summary? I've no objection to a user setting that changes this, of course, except clutter. (BTW, I doubt that very many people are reading this page. I'm trying to figure out how to bring these requests back to life. I intend to move some of them to SourceForge.) — Toby 00:16 Aug 15, 2002 (PDT)
There are a lot of articles that follow a naming convention that is useful for disambiguation, but a bit too long for casual linking, for instance C programming language (which is in 90% of the places linked as "C"). What I thought was, to give each article an optional field to specify an "overlay name", that would be used instead of the "ordinary name". Although not directly intuitive, this could be quite convenient for many sorts of articles, so in my opinion the pros outweigh the cons. --Uri 02:32 Aug 16, 2002 (PDT)
To some extent, we have this. In an article named B (programming language)[?], you can make a link [[|C]], and it will be rendered as [[C (programming language)|C]]. This is a situation where using "natural disambiguation" hinders rather than helps! — Toby 02:45 Aug 16, 2002 (PDT)
Here's something I've been thinking about which might be pretty easy to implement. How about a method for automatically generating nice, neat, numbered footnotes? You could embed something like [[note: the North American Land Giraffe is an example of a species so rare no members have ever existed.]] into the text of an article, and then when it gets converted to HTML for display the tag gets replaced with a number that's linked to an anchor down at the bottom of the page with the text of the note in it. Wikipedia's hyperlinks between articles make footnotes not as important as they would be for a paper encyclopedia, but there are still situations where they're very handy; annotating tables of data, for example, where there isn't room to include the text and it doesn't warrant a whole separate article. You could even get fancy and have [[note:]] tags within tables get placed immediately below the table itself, associating them more closely. Bryan 07:04 Oct 6, 2002 (UTC)
Suggestions for additional book sources - see Wikipedia:ISBN
When searching, if a redirect comes up, we should provide the title of the page it redirects to, not the title of the redirecting page. For example, if you search for "nine eleven", then September 11, 2001 Terrorist Attack should come up in the list of "Article title matches", rather than Nine eleven.
Also, if two redirects to the same page come up (eg nine-eleven and nine eleven), then only one should be displayed, not both. Similarly, when searching for "list christians", the redirect at list of christians shouldn't be in the list because the target of that redirect (list of Christians) does appear. Martin
Numbering and Referencing Equations, Figures, and Tables
It would be nice if Wikipedia could provide syntax for automatically numbering figures, tables, and equations, and referring to them in an article.
It is more professional and scalable to have an article say "This is shown in Fig.1" than "This is shown in the figure below." However, figure numbering is impractical in the Wikipedia as it stands, because someone else may add a figure before that point, and all the figures would have to be manually renumbered.
In LaTex, there is a facility to automatically generate such numbers. For example,
Equation (\ref{newton2}) is Newton's second law.
\begin{equation}
F = ma
\label{newton2}
\end{equation}
However, in Wikipedia we do not have specialized syntax for making equations, unlike \begin{equation} in LaTeX. An unofficial convention seems to be to put equations on a standalone line with ':'
- like this
but this may be used for other purposes, such as quoted text. Similarly for figures. -- CYD
Format of moved pages in Recent Changes
When you move a page, the comment attached to the move appears to be "Moved to new_article_name", which is obvious from the name of the article. Wouldn't it be better to have the comment as "Moved from old_article_name"? -- SGBailey 00:04 Feb 7, 2003 (UTC)
- It appears on the line of the old article name, and in the history of that, "moved to" makes therefore more sense than "moved from". It would be better though, if the name change were also in the history of the article with the new name (here of course "moved from"). - Patrick 12:09 Feb 7, 2003 (UTC)
- Also, the "N" for "new article" appears on the line of the old article name. I think it's far more logical to have this "N" appearing together with the new article name. D.D. 11:54 Feb 7, 2003 (UTC)
- But the article at the now article name isn't a new article: it's the old article, just in a new place. It's the old location that has the new content.
- --Paul A 07:59 Feb 11, 2003 (UTC)
- An example (open quote...
- (diff) (hist) . .N Dependent areas; 09:57 . . Dhum Dhum (Talk) (moved to "Non-independent_areas")
- (diff) (hist) . . Non-independent areas; 09:56 . . Dhum Dhum (Talk) (dependent areas / other non-independent areas)
- ... close quote) There was an article in Dependent areas(DA), there was no article called Non-independent areas (NIA). There is now an article called NIA which is new (it didn't exist before) so it needs an N whilst the old article (DA) doesbn't need an N since it was just an edit to replace the entire content by a redirect. It doesn't matter that that isn't how it is handled within the database, what matters is how it is perceived by the general punter. This should have read (open quote...
- (diff) (hist) . . Dependent areas; 09:57 . . Dhum Dhum (Talk) (moved to "Non-independent_areas", now #REDIRECT)
- (diff) (hist) . .N Non-independent areas; 09:56 . . Dhum Dhum (Talk) (moved from Dependent areas) (dependent areas / other non-independent areas)
- ... close quote). IMO anyway -- SGBailey 10:18 Feb 16, 2003 (UTC)
Is it possible to add to the Wiki Markup to HTML conversion automatic conversion double dashes to proper em-dashes? Wikipedia is now totally sprinkled by double dashes that ends up looking a bit odd (or perhaps I'm just picky?) -- Egil 11:26 Apr 10, 2003 (UTC)
The last time that this came up, there were still a lot of readers using Netscape 4, which can't handle proper em dashes. (If anybody wants to see how their own browser does it, here it is, inside some quotation marks: "—".) Also, I'd be wary of an automatic conversion until we're sure that double hyphens are never needed for anything common (<nowiki> will suffice for the very uncommon). I myself used to use em dashes until I realised that some people couldn't see them. -- Toby 02:53 Apr 13, 2003 (UTC)
I often found my login session expired by the time I finish typing an article. The article was then listed by my IP address instead of my login name. It should be nice to have an option to set the session to last longer. -- User:Kowloonese Apr 28, 2003 (PDT)
There are many articles that share the same format with similar inclusion other than the main content. One example is the articles for the cneturies, decades and years. It would be nice to factorize the repeating parts in a form of macro that can be included to ensure uniformity across all similar page.
- Wikipedia:Phase II feature requests: features requested under phase II of the software
- A gloassry for babylon[?] with all the entries of wikipedia.
All Wikipedia text is available under the terms of the GNU Free Documentation License
|
Search Encyclopedia
|
Featured Article
|