Wiktionary:Grease pit/2009/October

From Wiktionary, the free dictionary
Jump to navigation Jump to search
This is an archive page that has been kept for historical purposes. The conversations on this page are no longer live.

October 2009

Unicode 5.2

FYI: Some may known already, but Unicode 5.2 was finalized and released earlier today. Of particular interest to lexicographers, it adds seven new scripts as well as many supplemental characters to existing ones for a total of 6,648 new characters. See Unicode 5.2.0 for details. Bendono 23:59, 1 October 2009 (UTC)[reply]

I used Template:Navbox for the new style of Template:index/American Sign Language, but I had to transfer it from the other wiki. As a result, only the base code came, and it relied on classes which don't exist here. As a result, I had to add them manually as element:styles, the result of which is at best an ugly hack. Can we have the relevant classes from w:MediaWiki:Monobook.css copied to our version so that the navbox can use the classes instead (and therefore be customizable, like the rest of our interface, &c, &c.)? —Di gama (t • c • w) 07:18, 2 October 2009 (UTC)[reply]

issues with multiple right-floating things

When more than one right-floating elements (e.g. {{wikipedia}} and an image) are included in an entry, they often leave visually unattractive blank spots (see here). It's sometimes possible to displace it by moving the images around, but that gets harder with every additional image. What can be done about this? —Internoob (TalkCont.) 19:15, 3 October 2009 (UTC)[reply]

Some elements such as the old non-hidden tables {{top}} and {{rfp}} (or {{rfap}}?) don't play nice with rhs floating elements and should be fixed to do so. The sister project boxes can be replaced by the less conspicuous PL items (eg, {{pedialite}}) under "See also". Except possibly for a {{wikipedia}} that take a user to a disambiguation page (serving a function like {{also}}), I don't think sister project boxes should ever be above the first language header. Nor should images usually. DCDuring TALK 19:54, 3 October 2009 (UTC)[reply]
The problem seems to be limited to IE, or at least I don't see it in FF 3.5.3. —Di gama (t • c • w) 03:30, 4 October 2009 (UTC)[reply]
The problem is indeed limited to IE. It isn't a problem in Firefox, Opera, or Safari. --EncycloPetey 15:39, 4 October 2009 (UTC)[reply]
I didn't look at the specific Sandbox demo. In the particular case, replacing the "wikipedia" template with a "pedialite" template shortens the rhs by ~3/4" and lengthens the lhs by 3/8" reducing the imbalance by ~1". Usually we don't have more than one image. The vertical stacking of multiple rhs objects in a short entry can only be definitively addressed by some kind of placement conditional on the length of the page. I wonder if that is something that can be done via Javascript. Conditional rendering seems like a high price to pay for a problem limited to short entries with large stacks of images and sister project boxes, especially since the sister project boxes can readily be placed in "See also", as many here prefer.
I wonder whether we could allow for {{also}} to include a link to a WP page, especially to disambiguation pages there. WP has vastly more complete coverage of proper nouns than we do (and should, for that matter, IMHO) and sometimes user may be looking for them. DCDuring TALK 13:45, 4 October 2009 (UTC)[reply]
That isn't always a solution. Consider the page for (deprecated template usage) needle, which has images of both sewing and medical needles, and which really does need both to be illustrated. The vertical stacking will cause a problem in IE unless Windows 7 has fixed this problem. --EncycloPetey 15:39, 4 October 2009 (UTC)[reply]
For longer entries the appearance problem isn't as ridiculous, at least. I use FF 3.5.3 I have some layout problems with {{top}} which can be addressed by substituting {{rel}}. I have made the substitution and also deployed {{xsee}} instead of {{wikipedia}} at needle. Does that help with IE? DCDuring TALK 18:03, 4 October 2009 (UTC)[reply]
Urgh. That also has problems. For one, it doesn't identify the language of the linked Wikipedia. If a page has more than one language entry on it, that becomes a real problem. Including the language would make the link even messier, though. --EncycloPetey 04:01, 5 October 2009 (UTC)[reply]
I fixed that problem by using {{xsee}} to apply {{pedialite}}. Since it's inline, there aren't any visible rendering problems. I did have to suppress xsee's tendency to bold its parameters, though. This solution allows for other languages too. —Di gama (t • c • w) 04:12, 5 October 2009 (UTC)[reply]
P.S. A problem I just noticed with this method is that I can't remove the terminal full stop, so it would only look good at the end (of the see also list), so more than one WP link wouldn't work. Of course, the full stop could always be removed or made conditional. —Di gama (t • c • w) 04:18, 5 October 2009 (UTC)[reply]
 Zorse on Wikipedia

Wikipedia

I think langauge-specific links should be kept inside the language headers (not used in {(x)see}). What about using a slimmer version of {{wikipedia}}, such as the one to the right? — This unsigned comment was added by Bequw (talkcontribs).
Now at {{slim-wikipedia}} for people to play with. --Bequw¢τ 17:35, 8 October 2009 (UTC)[reply]
The slimmed one is certainly an improvement, especially for near top-of-the-entry use. Questions remain.
  1. Should any WP template appear above the first language line? I would argue that we should have a link to a en.WP dab page that contained links to Proper nouns that en.wikt does not have (by policy or otherwise). From a user perspective I think such a link is very like the {{also}} or {{xsee}} links which we have above the first language header.
  2. Should we try to minimize the space taken by top-of-the entry material by keeping it to the left of the rhs ToC, images, {{examples-right}}, and the other sister project links. That is, should we use {{xsee}} to contain any top-of-the-page link to WP?
  3. Should we deprecate the large sister project boxes altogether?
  4. Should the See also heading (the home for sister project links if sister project boxes are deprecated) continue to appear after translations. DCDuring TALK 19:44, 7 October 2009 (UTC)[reply]

hav'otha ppl2 probs w/FF353?

its ofn'not respondin'..--219.69.83.119 02:51, 6 October 2009 (UTC)[reply]

When did the problems that you have begin? --EncycloPetey 02:53, 6 October 2009 (UTC)[reply]

sins igotit~somweeks ago-2day4restarts onmy slowlaptop[took 2hrs!,butnow usinOpera2rite this[which dontdo dropdowns+has brite diff-screens,argh!--史凡>voice-MSN/skypeme!RSI>typin=hard! 03:17, 6 October 2009 (UTC)[reply]

That's around the time that MW did a major software update (Sep 16/17). It broke a lot of the bots too. I don't know what solution to suggest (I don't use FF), but it's probably not a problem limited to just you. --EncycloPetey 03:29, 6 October 2009 (UTC)[reply]

u use?btw,i'd/had2re-input'wt;pref'i/Op.-=not browser-INdep?sory4spelin,ta4repls!

Which one do you use? By the way, I had to re-input WT:PREFS in Opera - such not browser-independent? Sorry for my spelling, thanks for replies! Also, this FireFox version also doesn't reestablish my tabs on restart > problems not just with Wiktionary, but of general nature...
L☺g☺maniac chat? 13:34, 6 October 2009 (UTC) +--史凡>voice-MSN/skypeme!RSI>typin=hard! 14:25, 6 October 2009 (UTC)[reply]

Wiktionary mobile

Wiktionary on mobile is shocking or doesn't load at all, I tried on different mobile phones. Logging in requires an extra page load to get to the page you want originally. Any plans to make it friendlier on mobile phones? Not that one could contribute much on mobile phones but reading answers or checking watchlists would be good to have, more importantly - simply using for users. --Anatoli 03:44, 6 October 2009 (UTC)[reply]

Is this a Wiktionary-only problem, or does it happen with Wikipedia as well? --EncycloPetey 03:47, 6 October 2009 (UTC)[reply]
Wikipedia is OK but I only looked at articles, not special pages. On my Nokia 6700, the main frame gets really narrow, breaking short words in half, so that my watchlist (with many suppressions) looks incredibly long (from top to bottom). Anatoli 04:03, 6 October 2009 (UTC)[reply]

This whole lot needs to go into the correct category here. If anyone thinks they can correct the related templates (starting with nan) please do, then these whole lot will be in the right category. For the subcategories like nan-tw:Conjunctions, as dialects, these are wrongly named too, right? Mglovesfun (talk) 17:15, 8 October 2009 (UTC)[reply]

Yes, these were misnamed, but we never sorted them out. The issus with (e.g.) "nan-tw" is that it is not a dialect, but a script difference (Min Nan in traditional script) and we didn't sort out naming of these. (We = me, A-cai, and a few others.) Does need sorting, these are, of course, not topic categories. Robert Ullmann 01:06, 10 October 2009 (UTC)[reply]
Subcategories moved into Category:Min Nan language. There's still much to cleanup, but at least it's all in the same place. --Bequw¢τ 04:02, 13 October 2009 (UTC)[reply]

tbot, and various stuff

I've been doing work on Tbot (which you may have noticed). And fixing some other things. I had a flaming power strip yesterday morning, you might have noticed some service interruption ;-)

Tbot is eliding script parameters when not needed (and mostly when doing other things). I've also added a bit to the {{t-sect}} optimization. (Which, as a reminder, should only be used with {{t}}.)

Tell me if you see any anomalies. If I'm not apparently around, send SMS +254 722 929 463.

Should be okay :-) Robert Ullmann 01:03, 10 October 2009 (UTC)[reply]

Templates

Is there any guide somewhere to whatever the language is that templates are written in? I have never really understood all the #ifs and #eqs, and those seemingly endless iterations of curly brackets. But I would rather not have to ask someone every time I want to create a new one of any level of complexity. Ƿidsiþ 12:22, 10 October 2009 (UTC)[reply]

http://www.mediawiki.org/wiki/Templates is a good starting point, and you'll also want to read up on http://www.mediawiki.org/wiki/Help:Parser_functions_in_templates and http://www.mediawiki.org/wiki/Help:Extension:ParserFunctions, for even more stuff there's http://www.mediawiki.org/wiki/Help:Magic_words . Conrad.Irwin 12:27, 10 October 2009 (UTC)[reply]

Thanks. That's helped a lot. Ƿidsiþ 13:05, 10 October 2009 (UTC)[reply]

Trennbare Verben

I have added the template {{de-trennbar}}, to allow a prettier representation and categorization of trennbare Verben, such as anmelden.

I'd like to hear your opinions. Thanks, Peleg 10:22, 14 October 2009 (UTC)[reply]

I've added {{de-verb-strong}} to allow this nicely for strong verbs (such as ansehen) as well. Peleg 12:13, 14 October 2009 (UTC)[reply]

etw. jmdn. etc.

I've added the templates {{etw}} and {{jmdn}} (as examples) to use it in a proper way with the <abbr> tag. I find it nice and useful. Look here: interessieren and tell me what you think. Thanks! Peleg 12:15, 14 October 2009 (UTC)[reply]

"WT:" become equivalent to "Wiktionary:"?

The English Wikipedia (and maybe other Wikipedias too, I don't really know) has an interesting feature which makes typing "WP:" in the search bar be identical to typing "Wikipedia:", meaning that typing in "WP:Administrators" into the bar brings you straight to "Wikipedia:Administrators" and even has "Wikipedia:Administrators" appear in the dropdown list as it's being typed. Is there any way this could be implemented on Wiktionary, so that "WT:" appears as "Wiktionary:"? I know that most pages in the Wiktionary namespace have shortcuts, but quite a few don't and those that do aren't the easiest things to remember... --Yair rand 00:03, 15 October 2009 (UTC)[reply]

But you can already use Special:AllPages for that. I find the search window tool more useful as it currently stands, allowing me to quickly determine whether there is a page beginning with that sequence of letters. I can understand the utility on an encyclopedic site like Wikipedia, but on Wiktionary, where every word should eventually have an entry, that kind of redirect could lead to problems, especially if there are abbreviations beginning with "WT:" in any language. --EncycloPetey 01:48, 15 October 2009 (UTC)[reply]
But we already have wt: as a dedicated namespace, so we already can't really use it for entries, in that it won't show up in main-space dumps and therefore won't show up on mirrors, won't get handled properly by interwiki-bots, etc. I agree with Yair rand that there's no reason it should be its own namespace, rather than an alias for the project namespace, now that the software supports that sort of aliasing. (Dunno what we have to do to make that work, though.) (But it looks like we'd need to have a developer make the change: mw:Manual:$wgNamespaceAliases.)RuakhTALK 02:20, 15 October 2009 (UTC) edited 02:25, 15 October 2009 (UTC)[reply]
I did not realize that WT: was its own namespace, in which case my concerns are moot. --EncycloPetey 02:22, 15 October 2009 (UTC)[reply]
Yeah, it's kind of the "bastard child" of namespaces. —RuakhTALK 02:25, 15 October 2009 (UTC)[reply]
So how does one request that a developer make a change like that? --Yair rand 12:36, 16 October 2009 (UTC)[reply]
And if we do change that, we might as well also change WS: to be equivalent to Wikisaurus: as we have no simple way to go to those pages. --Yair rand 23:01, 20 October 2009 (UTC)[reply]

Template that uses Special:Index

What's that tidy template that allows you to find all your subpages and uses the page above as a base? Cheers. Mglovesfun (talk) 14:27, 16 October 2009 (UTC)[reply]

Well, there's Special:PrefixIndex. Is that what you were looking for? L☺g☺maniac chat? 15:03, 16 October 2009 (UTC)[reply]
It is, but I'm talking about a template I have seen before. Mglovesfun (talk) 15:19, 16 October 2009 (UTC)[reply]
oh. Well that I don't know. :) L☺g☺maniac chat? 15:25, 16 October 2009 (UTC)[reply]
{{subpages}} Conrad.Irwin 17:42, 16 October 2009 (UTC)[reply]

Insertable templates not working?

The insertable items in the "templates" section when editing an entry seem to be broken. (I'm using Opera 9, and always was.) Since a few weeks ago, each space-delimited word is separately clickable and inserts itself alone, e.g. {{alternative spelling of|}} is three separate links that just insert their own text, rather than being a single link that is inserted as a whole template. Why has this changed? Equinox 20:12, 17 October 2009 (UTC)[reply]

They're working for me in Safari on a MacBook. It may be a Preferences issue, or else Opera-specific. --EncycloPetey 01:58, 20 October 2009 (UTC)[reply]
The odd thing is that I haven't upgraded my Web browser or (as far as I recall) changed my preferences. Equinox 03:49, 20 October 2009 (UTC)[reply]
*sigh* because browsers hate me? The culprit is (presumably) my edits to MediaWiki:Edittools. It seems that there are more differences than I expected among various browsers in handling of non-breaking spaces. (I've already had to make changes to the code to address discrepancies between Safari and Firefox.)
It should work again now. I'll have to decide whether I want to (1) download Opera to figure out exactly what its issue is, (2) take a different approach entirely, or (3) give up.
RuakhTALK 03:49, 20 October 2009 (UTC)[reply]
Yes, it works again. Thanks. As far as I know, Opera is generally very standards-compliant and should have a decent JavaScript/ECMAScript implementation. But if you would like me to try any particular mini-script to diagnose this issue, feel free to contact me. Equinox 03:51, 20 October 2009 (UTC)[reply]

Schönfliess

I was putting in some crystallographic term at nl.wikt like nl:draaispiegelas. When doing that you easily get into the notations for point and space symmetry groups like S4 or I41/amd or so. Although I am not looking forward to putting in all 230 space groups or so, I do think an article explaining what a symbols like S4 stands for would be nice. Problem is: how do you get a superscript into an article name? Is there a way to do that? Jcwf 20:48, 17 October 2009 (UTC)[reply]

We seem to have used the characters ₂,₃,₄ on occasion (e.g. (deprecated template usage) H₂O), but it doesn't seem ideal and I don't know whether they go up to 9. Equinox 20:58, 17 October 2009 (UTC)[reply]
Thanks Equinox, actually if there is a six that would pretty much do for most crystallography, although some symbols have h, v or d Jcwf 22:03, 17 October 2009 (UTC)[reply]
They do go up to nine. Appendix:Unicode/Superscripts and Subscripts There's also a subscript v, but none of the other letters you mentioned. (However, they will be encoded in Unicode 6.0 which will be released in 2011, so you might want to wait until then) -- Prince Kassad 22:42, 17 October 2009 (UTC)[reply]

Shorthand for "capitalise first letter of link"

Considering how often it comes up — any time you want to link the first word of a definition — I think it would be good to have a shorthand for a link whose first letter is to be capitalised in the display, e.g. [[\sulphuric acid]] might produce [[sulphuric acid|Sulphuric acid]]. Is it practical, desirable? Equinox 19:14, 19 October 2009 (UTC)[reply]

I can't imagine the developers doing anything to wikisyntax at this point that would cause existing good links to go bad just to add a shortcut, so it probably would have to be a parser function which wouldn't be that much of a shortcut. The only alternative I can think of would be to give a purpose to the redundant [[|sulphuric acid]] which currently gets saved as [[sulphuric acid]], so it wouldn't affect direct coding, tho it might affect a substitutable template that made use of that redundancy. — Carolina wren discussió 00:27, 20 October 2009 (UTC)[reply]
It ought to be possible to set up a simply-named template like: {{1|F|f|irst}}. That is, if a template name can begin with a numeral, we can set up a template that swaps the first letter out for the link/text. We need the template rather than automation. Relying on automation could send the user to a German noun by mistake. --EncycloPetey 01:56, 20 October 2009 (UTC)[reply]
If we go with a template, there wouldn't be any need to specify both the lower and upper case letters as separate parameters. [[{{{1}}}|{{ucfirst:{{{1}}}}}]] would be the simplest code version of this, sans any bells and whistles such as non-Latin script or foreign language entry support, and needs only one parameter, the entry to be linked to. — Carolina wren discussió 05:28, 20 October 2009 (UTC)[reply]
If a template is created, it should always be used with subst: (otherwise, it would make the contents less readable and loading the page would be slightly longer, without any good reason). Lmaltier 19:31, 20 October 2009 (UTC)[reply]
Template:1's been created. Please tweak as needed.​—msh210 14:24, 29 October 2009 (UTC)[reply]

Creating pages on ar.wiktionary

I have no trouble creating or editing pages on ar.wikipedia and no trouble editing pages on ar.wiktionary but every time I try to create a new page on ar.wiktionary, I end up downloading some index.php . There are other Arabic Wiktionarians who do not have this problem. I have tried different browsers and computer systems but I still get the same results. Can anyone explain what is going on? :)--Thecurran 09:21, 20 October 2009 (UTC)[reply]

Does anyone on English Wiktionary also edit on Arabic Wiktionary? :)--Thecurran 15:40, 25 October 2009 (UTC)[reply]
You should provide a link to an example of this. Nobody knows what you mean. —Stephen 22:42, 31 October 2009 (UTC)[reply]
Could your browser be misconfigured? It should be displaying the page content of index.php rather than prompting you to save it to disk. Equinox 22:52, 31 October 2009 (UTC)[reply]

In Arabic, the creation link for ألبانيا is http://ar.wiktionary.org/w/index.php?title=ألبانيا&action=edit or without diacritics, البانيا is http://ar.wiktionary.org/w/index.php?title=البانيا&action=edit . In English, Albania- is http://en.wiktionary.org/w/index.php?title=Albania-&action=edit . In Spanish, Albania- is http://es.wiktionary.org/w/index.php?title=Albania-&action=edit . In French, Albanie- is http://fr.wiktionary.org/w/index.php?title=Albanie-&action=edit . In Russian, Албания- is http://ru.wiktionary.org/w/index.php?title=Албания-&action=edit . In Chinese, 阿尔巴尼亚 is http://zh.wiktionary.org/w/index.php?title=阿尔巴尼亚&action=edit or in Traditional, 阿爾巴尼亞 is http://zh.wiktionary.org/w/index.php?title=阿爾巴尼亞&action=edit . I thought I would be able to create all of these pages if I was logged in. Of course, I would not create the ones that end in hyphens or nine. Can you two or anyone else trigger the Arabic links appropriately once you're logged in to the Arabic Wiktionary? :)--Thecurran 13:36, 1 November 2009 (UTC)[reply]

In Arabic Wikipedia, ألبانيا9 is http://ar.wikipedia.org/w/index.php?title=ألبانيا9&action=edit . In English, Albania- is http://en.wikipedia.org/w/index.php?title=Albania-&action=edit . In Spanish, Albania-http://es.wikipedia.org/w/index.php?title=Albania-&action=edit . In French, Albanie- is http://fr.wikipedia.org/w/index.php?title=Albanie-&action=edit . In Russian, Албания- is http://ru.wikipedia.org/w/index.php?title=Албания-&action=edit . In Chinese, 阿尔巴尼亚- is http://zh.wikipedia.org/w/index.php?title=阿尔巴尼亚-&action=edit . :)--Thecurran 14:32, 1 November 2009 (UTC)[reply]

Is there anything that bars editors from creating articles in a wiki before doing other types of edits there? :)--Thecurran 14:36, 1 November 2009 (UTC)[reply]

Wikisaurus

I have long wondered why certain Wikisaurus entries (like Wikisaurus:machine) have odd blue dotted lines halfway across the page from all the words. Recently, while using the Firefox browser instead of my usual Google Chrome, I noticed that the lines actually appear under the words in that browser. Anyone know why it messes up in Chrome and how this can be fixed? --Yair rand 19:21, 20 October 2009 (UTC)[reply]

Seems to be an obvious css bug on Chrome's part, though I do not know how we can fix it. For the record, the dotted line indicates the presence of a tooltip giving a short definition. Circeus 04:55, 20 November 2009 (UTC)[reply]

User:Visviva's preloadText into site JS

Now that preloadText is being used by creation.js (only for esperanto at the moment, but probably for more languages as time goes by) and by various other projects we have - would it make sense to put a version of it into Common.js? Conrad.Irwin 00:26, 23 October 2009 (UTC)[reply]

Polish ł

Is it possible, for the Polish letter ł to appear in the search results when I type l? It would be cheaper, than to buy me a Polish keyboard. --Volants 13:30, 23 October 2009 (UTC)[reply]

Buy a Polish keyboard. (In other words, no, it's impossible) -- Prince Kassad 13:35, 23 October 2009 (UTC)[reply]
Shame. Maybe I can petition the WMF who can buy one for me, as a scolarship of something. They do that sort of thing, don't they? Or get some free software, I'll look for it... --Volants 13:41, 23 October 2009 (UTC)[reply]
I'm wondering why you can't use the Polish keyboard layout with your US (?) keyboard. -- Prince Kassad 13:43, 23 October 2009 (UTC)[reply]
Try WT:PREFS "show a special character input (like the one beneath the edit field) for the search field). Conrad.Irwin 14:35, 23 October 2009 (UTC)[reply]
You can install Polish Keyboard layout in Control Panel if you are using Windows. Maro 22:16, 23 October 2009 (UTC)[reply]
I've chosen Conrad.Irwin's suggestion. Thanks. This money I've saved will now be spent on something better. Maybe for a microphone. --Volants 12:14, 26 October 2009 (UTC)[reply]

Edittools

Currently, the Latin/Roman section is sorted by diacritic. While this is pretty nice, I think it would be much more intuitive and easier to search through if all the letters were simply sorted alphabetically. That way you don't have to look through the entire section to find your character, but just snap to where the base character appears and look in there. Any comments? -- Prince Kassad 18:33, 24 October 2009 (UTC)[reply]

The sorting by diacritic serves two purposes: it helps you find the character you're looking for, and it lets you know what a given character is. Your suggestion might help the first purpose, but I think it could significantly harm the second one. (I don't think I could reliably distinguish ǎ from ă, for example, at normal font sizes.) —RuakhTALK 06:54, 25 October 2009 (UTC)[reply]
Hm, this is an interesting point. The problem is that currently nobody uses the Latin/Roman section simply because it's so counter-intuitive, so it should be redone in some way. If differentiation is an issue, maybe we should make the section a table. -- Prince Kassad 14:20, 25 October 2009 (UTC)[reply]
Or in a larger font.​—msh210 14:15, 29 October 2009 (UTC)[reply]
That would look horribly ugly. -- Prince Kassad 18:39, 29 October 2009 (UTC)[reply]

Could someone please add an f= and fp= parameter for this (with an #if) for adding feminine singular and plural form. See this edit if you're unsure of what I'm on about. Mglovesfun (talk) 11:36, 27 October 2009 (UTC)[reply]

Huh? You mean like adding {{#if:{{{f|}}}|, ''feminine singular'' '''[[{{{f}}}]]'''}}{{#if:{{{fp|}}}|, ''feminine plural'' '''[[{{{fp}}}]]'''}} before the end bracket? --Yair rand 15:48, 27 October 2009 (UTC)[reply]
I'm not exactly sure if that's what you meant, but I added it to the template. Should I also add m= and mp= parameters to it? I notice that on amye you just added (m: amy) at the end of line. --Yair rand 16:22, 27 October 2009 (UTC)[reply]
I'd say no; we don't do that with other templates like fr-noun. Mglovesfun (talk) 17:53, 27 October 2009 (UTC)[reply]

Python bot broken?

Hello. I'm planning to rerun User:Dawnraybot, but the bot isn't working. According to Metawiki all bots are Wikimedia bots are broken since September. But I've seen other bots working here. What to do? --Rising Sun 09:54, 29 October 2009 (UTC)[reply]

  • SemperBlottoBot is working normally. What are the symptoms? SemperBlotto 09:56, 29 October 2009 (UTC)[reply]
    • I try to login using login.py, and it flashes with somehting like "runtime error. Bad magic number in .pyc file" (but it is hard to read, because it flashes for a millisecond). My updated user-config.py is:
use_api = True
mylang='en'
family='wiktionary'
usernames['wiktionary']['en']='Dawnraybot'
console_encoding = 'utf-8'
put_throttle = 0

Also, when I try to run the frfromfile.py, error message is

Traceback (most recent call last):
  File "pywikipedia\frfromfile.py", line 37, in <module>
    import wikipedia, config
ImportError: Bad magic number in G:\WT\pywikipedia\pywikipedia\wikipedia.pyc

The question is, what is a magic number, and how do I make it a good magic number? --Rising Sun 10:04, 29 October 2009 (UTC)[reply]

Traceback (most recent call last):
  File "G:\WT\pywikipedia\pywikipedia\user-config.py", line 3, in <module>
    usernames['wiktionary']['en']='Dawnraybot'
NameError: name 'usernames' is not defined

I remember this one from the first time I tried to set up - but I don't remember what I did when that happened before.--Rising Sun 10:16, 29 October 2009 (UTC)[reply]

You probably need:
usernames = {}
usernames['wiktionary'] = {}
usernames['wiktionary']['en'] = 'Dawnraybot'
or equivalently:
usernames = {'wiktionary': {'en': 'Dawnraybot'}}
Conrad.Irwin 11:24, 29 October 2009 (UTC)[reply]
Tried those, same error. I might just reinstall everything from scratch instead. --Rising Sun 12:28, 29 October 2009 (UTC)[reply]
Can be fixed somehow; I'm using stuff; let me sleep any try to catch you in my daytime ;-) Robert Ullmann 01:01, 30 October 2009 (UTC)[reply]

Bad magic number means the .pyc file wasn't compiled with the same version of Python (or was corrupted somehow). Normally Python will automatically recompile if the magic number is an older version. Did you downgrade the Python version? You can always just delete the offending .pyc file and let it rebuild it.

Re-syncing the pybot framework from SVN is probably a good idea. The "use_api" setting shouldn't (can't) be messing up the "usernames" (is just somehow masking it with an earlier error).

You certainly don't need to create the usernames array; config.py does that, and then execs user-config.py. You might try excuting config by itself, as a main program it should dump out all the settings, including what is in user-config. If it faults, that will be interesting. Robert Ullmann 16:33, 30 October 2009 (UTC)[reply]

I see you've gotten it working ;-) Robert Ullmann 12:24, 31 October 2009 (UTC)[reply]

Service interruption

For some unknown (and previously unseen) problem service is down here; this affects AF, Tbot, Interwicket. All of which recover normally from problems of less than 48 hours, and eventually from problems of any duration.

My apologies. It is 4AM here, I will chase this in a few hours when everyone (else) is awake! Robert Ullmann 00:56, 30 October 2009 (UTC)[reply]

Should be normal now. Robert Ullmann 16:17, 30 October 2009 (UTC)[reply]

Create a Template:wrong script (or not)

I was thinking of alter kaker which is "Yiddish", apart from Yiddish doesn't use the Latin script. Why not create a {{wrong script/box}} per {{nolanguage/box}}. Either that, or it's possible to merge the three ({{notenglish/box}}) but I can't think of any decent reason to do so. Mglovesfun (talk) 16:26, 31 October 2009 (UTC)[reply]