Conrad.Irwin

Hello! What does "raising" mean in this text:

SICINIUS

   'Tis this slave;--
   Go whip him, 'fore the people's eyes:--his raising;
   Nothing but his report.

William Shakespeare "Coriolanus" Act IV, Scene 6.

Thank you Igor Skoglund

I have no idea, Shakespeare is far from my literary comfort zone. Conrad.Irwin 21:07, 1 April 2010 (UTC)Reply

diff.​—msh210℠ 00:29, 8 April 2010 (UTC)Reply

That bot is on my TODO list, along with "finish disserting" and "pass exams". Sorry, I can't give an estimate as to when I'll get round to it. Conrad.Irwin 01:10, 8 April 2010 (UTC)Reply

Yeah, a paper and exams might be just a tad more important.  :-)  Good luck with them.​—msh210℠ 15:00, 8 April 2010 (UTC)Reply

I think that if any entry is in Category:Questions about entries for a year or more, and it gets answered, that's better than no answer at all right? If it leads to one problem with one entry being fixed, one is one more than zero, right? Mglovesfun (talk) 22:18, 13 April 2010 (UTC)Reply

You are ignoring the cost of maintaining the categories which is not zero, whether it is -1 or more, I don't know. given that maintenance is something we already have too much of, I'd give it a negative proportional to number of uses and categories. Conrad.Irwin 22:31, 13 April 2010 (UTC)Reply

I find with {{rfv}}, {{rfc}}, {{attention}} and maybe {{rfd}} that they are added more often then they are removed. This seems to be not worse than this. As I say, hopefully we cam add a lot more contributors and then these could be answered very quickly. Mglovesfun (talk) 22:34, 13 April 2010 (UTC)Reply

Indeed, {{question}} is guaranteed to be the same. Out of intereest, what niche does {{question}} fill that is not covered by all those other templates? Conrad.Irwin 22:37, 13 April 2010 (UTC)Reply

Somewhere, a while back, I suggested that template:question simply redirect to template:attention. I still think that's appropriate.​—msh210℠ 15:22, 14 April 2010 (UTC)Reply

The bot enters incorrect anagrams because it ignores the long s.[1] These entries should disappear eventually (the English long s is a typographic style, not a separate letter), but for now... —Michael Z. 2010-04-16 21:39 z

The recent update seems to have cut off all languages which have exactly 10 entries. I'm not sure you intended that, since the description still says "only those with 10 or more entries are shown". -- Prince Kassad 01:33, 19 April 2010 (UTC)Reply

Thanks. I just altered the script to use http://toolserver.org/~enwikt/definitions instead of going through the dump, so if you notice other anomalies let me know. Conrad.Irwin 10:08, 19 April 2010 (UTC)Reply

Last night I dreamed that the mirror in the bathroom had been replaced by a large whiteboard. You were talking to a Wikipedia administrator, who had written a lot of scandalous gossip on the board about some other user. By doing this he had stepped outside Wikipedia's stated goals of civility and openness, but he was relying on the fact that what he had written would be automatically purged from the board, and from history, some time soon. I threatened to take a photograph. Equinox ◑ 21:30, 23 April 2010 (UTC)Reply

Nah, I'm not really into Pokémon.... Conrad.Irwin 23:01, 23 April 2010 (UTC)Reply

Shades of Matsui.​—msh210℠ 15:01, 26 April 2010 (UTC)Reply

Excuse me, but I have every right to request a desysopping of Opiaterein now. Stop reverting my edits. Razorflame 21:35, 29 April 2010 (UTC)Reply

Why do you have that right? This request has been denied the last 4? times you've asked for it - you've asked for it again, and it is (again) being discussed. Conrad.Irwin 21:37, 29 April 2010 (UTC)Reply

Never mind. I've got other things to do. Razorflame 21:46, 29 April 2010 (UTC)Reply

While clicking to add a gloss, the Italian translation was filled into the gloss field. See that? --Bequw → τ 23:57, 29 April 2010 (UTC)Reply

I do indeed. "thinking". Conrad.Irwin 00:06, 30 April 2010 (UTC)Reply

There is a trailing space after trans-top "{{trans-top}} " - I will fix the code to allow for this, maybe we should ask AutoFormat to deal with them. Conrad.Irwin 00:13, 30 April 2010 (UTC)Reply

Now fixed (and it already seems to :p). Conrad.Irwin 00:49, 30 April 2010 (UTC)Reply

Hello Conrad, is it possible or is it wanted to have examples (a German sentence) in the German entries? Thanks for your answer. Nasobema lyricum 15:59, 1 May 2010 (UTC)Reply

Yes please! there's an explanation at Help:Example sentences. Conrad.Irwin 16:00, 1 May 2010 (UTC)Reply

Any idea why it's not showing up? -Atelaes λάλει ἐμοί 21:06, 1 May 2010 (UTC)Reply

document.getElementByID is not a function (you want getElementById) if you haven't already found it, Firefox has a javascript console (Ctrl+shift+J) that tells you these messages. If you need more the "Firebug" FireFox extension [mimicked by Dragonfly for opera or "developer tools" in chrome/IE] provides a fully featured javascript debugging environment. Conrad.Irwin 21:13, 1 May 2010 (UTC)Reply

Yeah, I managed to get it working about the exact same time as you responded. I got Firebug, but haven't yet figured out how to work it. I suppose I should do that sometime, so I'm not doing all this manual debugging. Thanks. -Atelaes λάλει ἐμοί 21:15, 1 May 2010 (UTC)Reply

Not remarkably urgent, but whenever you have some time to waste, I could use lists of pages in Category:Persian nouns, Category:Persian adjectives and Category:Persian verbs that don't link to {{fa-noun}}, {{fa-adj}} and {{fa-verb}}. — [ R·I·C ] opiaterein — 12:22, 3 May 2010 (UTC)Reply

done. Conrad.Irwin 12:47, 3 May 2010 (UTC)Reply

Thanks :) I had another request for Turkish, but I'm not really sure how to handle it yet. The Turkish declension tables have gotten a bit out of hand... — [ R·I·C ] opiaterein — 13:03, 3 May 2010 (UTC)Reply

Sooo I didn't realize quite how many of these there were... and I don't know how much help I can muster, so I have a new favor. :D Most of these are pretty formulaic.

'''{{fa-Arab| `pagename` }}''' ( `transliteration` )

If it's not too messy, could you run conbot to replace that with:

{{fa-`POS`|tr=`transliteration`}}

and then to strip the [[Category:Persian (verbs/nouns/adjectives)]]? That would really really rock. — [ R·I·C ] opiaterein — 22:48, 5 May 2010 (UTC)Reply

Should be running now. Please keep an eye on it - I still haven't worked out how to make spacing correct when removing categories (so that's where it's most likely to go wrong :p). Conrad.Irwin 00:22, 6 May 2010 (UTC)Reply

Famous last words :). Most of the remainder seem to require a head= parameter as they have linked constituents. Conrad.Irwin 00:28, 6 May 2010 (UTC)Reply

I thought about asking you to link the headwords for multi-word entries, but I didn't want to be too demanding :D — [ R·I·C ] opiaterein — 00:59, 6 May 2010 (UTC)Reply

Hell that's what Autoformat's for :D — [ R·I·C ] opiaterein — 00:28, 6 May 2010 (UTC)Reply

Hmmmm Mglovesfun (talk) 22:19, 4 May 2010 (UTC)Reply

Hmmmm indeed. Thanks for pointing it out. Conrad.Irwin 22:38, 4 May 2010 (UTC)Reply

Why did you remove the notice of this from WT:ELE? It is still on the discussion page, but it is not as visible there as on the article itself. BTW I also added a notice of the proposal to WT:BP. Facts707 19:36, 10 May 2010 (UTC)Reply

It was misleading (only one editor that I can see) and badly formatted (such hatnotes, when necessary, are added in indented italics - ala {{also}}). Additionally, it's not really necessary - particularly now it's mentioned on WT:BP too - the name of that page is one of the least serious faults with the page, so advertising more prominently advertised than the /editable version does seems to give the wrong impression. Conrad.Irwin 19:43, 10 May 2010 (UTC)Reply

Ahh, I see the problem. When I first proposed the rename, I suggested simply WT:Entry layout. This was also suggested earlier by another editor and disputed by yet others (I copied those entries into the new request). I later changed my request to WT:Entry layout guidelines, but since that particular name was not suggested by another editor, I should have changed the rename notification to "An editor has proposed that this article be renamed to Entry layout guidelines, along with other suggestions from other editors.", which I have now done (but only on the talk page, not the actual page). But "badly formatted" because something is not in italics sounds a bit harsh. And other requests, such as for deletions, are not in italics. I would have used a template, but apparently there isn't one for moves as there are for deletions {{rfd}} or requests for verfications {{rfv}}. Also I don't see what the placement above the note for the /editable version has to do with it - who is going to look at that version? Facts707 09:35, 11 May 2010 (UTC)Reply

Easiest to use the passive voice, see {{move}} (which I didn't know about until now, but it seems to solve the problem pretty well). As with italics, having things in a box provides clear visual distinction about what is the content of the page, and what is information about the page. Ideally anyone interested in improving the page will look at the /editable version, and not get sidetracked into the much less important naming question. Conrad.Irwin 09:46, 11 May 2010 (UTC)Reply

Hi, Conrad. Due to a quick conversation with Michael at my talk page [2], the template {{lang2sc}} should probably be converted to a hard-redirect to {{langscript}}. Do you agree with this suggestion? Please see its what-links-here and perhaps User:Conrad.Irwin/creation.js/!sc before responding. --Daniel. 20:35, 11 May 2010 (UTC)Reply

So I've recently added the option for showing comparative forms of Persian adjectives in {{fa-adj}}, and I was wondering if it would be possible to have these accelerated with transliteration. This will always be the transliteration of the main word + -tar for comparatives, + -tarin for superlatives. بلندتر is what the output should look like for comparatives, and بلندترین for superlatives. The only differences are the two letters in the tr= and c/s in the fa-adj-form template. Would it be possible for creation.js to read the tr= parameter of {{fa-adj}} to automatically include tr= in the comparative/superlative forms' {{fa-word}}? — [ R·I·C ] opiaterein — 21:18, 14 May 2010 (UTC)Reply

It should be doable, probably easiest using the normal mechanism of adding (yet another) classname <span class="form-of comparative-form-of lang-fa transliteration-{{anchorencode:{{{tr}}}-tar}}"> or something. I'll have a go at programming it in an hour or so (as a reward for doing some revision :s). Conrad.Irwin 21:31, 14 May 2010 (UTC)Reply

Thanks :) Let me know if you need any other info. — [ R·I·C ] opiaterein — 21:38, 14 May 2010 (UTC)Reply

You are welcome. My gratitude for adding required *.--Pierpao 17:01, 15 May 2010 (UTC)Reply

Is there an easy way in MediaWiki:Gadget-vectorTabs.js to hide the "arrow" tab that was hiding the tabs? --Bequw → τ 18:35, 15 May 2010 (UTC)Reply

That should be what "$('#p-actions').addClass('emptyPortlet');" does... Conrad.Irwin 18:37, 15 May 2010 (UTC)Reply

Of course, if I'd taken the time to read what I wrote, I would have noticed the typo in the Wiktionary version that was absent in the local version. Now fixed. Conrad.Irwin 23:02, 18 May 2010 (UTC)Reply

Any chance you could help me out with rfp in hrm? Because it's your favourite interjection, I trust your expertise in the area. --Vahagn Petrosyan 23:39, 16 May 2010 (UTC)Reply

I haz not the IPA at my command. I will try to do you a recording of me reading WT:BP :). Conrad.Irwin 23:41, 16 May 2010 (UTC)Reply

Iz okay? Conrad.Irwin

Sort of :) I think that's /ˈhərəm/ in IPA. Thanks. --Vahagn Petrosyan 00:06, 17 May 2010 (UTC)Reply

Hey, if you're not too tired from hrmmming, do you suppose you could take a look at some of the changes I've made recently. The first draft didn't work on any definition which had an example sentence and a quote at the same time (a rather serious oversight on my part). I've introduced some new code which resolves this, and it works, but it's just so damned clunky....I was wondering if you had any ideas for elegantizing it. Many thanks. -Atelaes λάλει ἐμοί 00:09, 17 May 2010 (UTC)Reply

I am, for today, exhausted. Perhaps I could ask your indulgence while I get the necessary beauty sleep for my lectures tomorrow. After which, I will look at code - shall I move the toggle all to the toolbar at the same time? Conrad.Irwin 00:11, 17 May 2010 (UTC)Reply

Hey, I'm the one asking the favor here. Do it whenever you feel like it. And yeah, feel free to move "all" to the toolbar, if I haven't already (I'm planning on doing it in about ten hours or so). Cheers. -Atelaes λάλει ἐμοί 00:21, 17 May 2010 (UTC)Reply

Toolbox toggle is up and running, so don't worry about that (an hour late, apparently). -Atelaes λάλει ἐμοί 11:39, 17 May 2010 (UTC)Reply

Looks great. As to the duplication of everything for UL/DL, I'd suggest using the toggle as the status indicator - then it doesn't matter which sub-elements are present. Maybe by making its classname dependant on revealed/hidden (but obvious not exactly .hidden) - or just reading out the content of the toggle. Conrad.Irwin 11:53, 17 May 2010 (UTC)Reply

I decided to go with the toggle content, which cleans up the code rather nicely. Thanks for the idea. -Atelaes λάλει ἐμοί 01:52, 18 May 2010 (UTC)Reply

Hi. I don't see the green accelerated links any more. Do you know why that is? I wonder whether it relates to the "new messages" change at the top of pages. Equinox ◑ 12:33, 17 May 2010 (UTC)Reply

They should still be there, but only in the main namespace under a PoS heading, at the request of Mglovesfun. i.e. they still work from my point of view - do you get any error messages, is it certainly enabled? Conrad.Irwin 12:40, 17 May 2010 (UTC)Reply

It started working again. Strange. Equinox ◑ 17:49, 17 May 2010 (UTC)Reply

It was because I'd been messing with JavaScript settings in my browser! Equinox ◑ 14:55, 20 May 2010 (UTC)Reply

:D I've done that before. Conrad.Irwin 15:11, 20 May 2010 (UTC)Reply

Here's a wikipedia article w:Kanji#Readings for a more thourough explaination for the Japanese persepective. I'm unsure about the other languages, but in Japanese, a chinese character usually has more than one pronunciation, or 'reading'. So what the on-reading and kun-reading headings are signifiying is that the chinese characters listed can have the reading of the current page. For example, in なん, which is nan written phonetically, it is saying that the listed characters can take the reading nan. The lists are useful and the current headings are probably the best way to express them as most of the Japanese learing guides I have seen use this terminology. The format for these entries is defined at Wiktionary:About Japanese#Hiragana entries.--balloonguy 17:27, 17 May 2010 (UTC)Reply

Thanks for your explanation, I've added those to the list of headings. Conrad.Irwin 17:46, 17 May 2010 (UTC)Reply

IMO just block the two snazz guys with an expiry time of indefinite. Semper will do it when he logs on anyway. Mglovesfun (talk) 17:38, 19 May 2010 (UTC)Reply

Done. I've added you to rollbackers/patrollers groups again, so that you can automatically patrol the edits you undo. Thanks. Conrad.Irwin 17:42, 19 May 2010 (UTC)Reply

Why am I blocked?Prado ³Prado2

My edits are manual, I do not know how to work with bot. ³Prado2 16:12, 21 May 2010 (UTC)Reply

Remember, please choose here. Sign your post here? And must all of blocked users. QuickGround 04:58, 22 May 2010 (UTC)Reply

Hi Conrad, I've noticed that the vowel pages are empty in the latest index update (a, e, i, o, ö, u, ü). I also noticed that the two-letter links at each header contain the long vowels instead of the short ones as before, e.g. ká, ké, kó instead of ka, ke, ko. Have you made any change in the code? Thanks. --Panda10 19:52, 22 May 2010 (UTC)Reply

Hi, yes - I was implementing the change suggested in Reverse Index above (do you still want that by the way?) to sort the long vowels the same as the short ones (except where the word is otherwise identical) - I've clearly put it the wrong way up in the headings though, sorry. Should be fixable. Conrad.Irwin 20:15, 22 May 2010 (UTC)Reply

Ok. So it seems that while one thing was fixed (short vowel = long vowel in sorting), there is something weird going on with the sorting but I don't understand the pattern. Here is an example:

• Current: odaad, odaát, oda-, oda, óda, odahív, odaliszk, -od, Odera-Frankfurt, -odik, -ódik
• Should be: -od, oda, oda-, óda, odaad, odaát, odahív, odaliszk, Odera-Frankfurt, -odik, -ódik

As for the reverse index, let's wait with it. I tried to figure it out manually, but it's not as easy as I thought. I know how busy you are, so let's postpone it for now. Thanks. --Panda10 21:05, 22 May 2010 (UTC)Reply

It seems to be sorting the end of the word somewhere between a and e, hrm... Conrad.Irwin 21:39, 22 May 2010 (UTC)Reply

My inability never ceases to amaze me - I had perpended a "d" to the sort key (because that was the section header) but inconsistently. Sorry for all the hassle, and please let me know if it is still broken. Conrad.Irwin 21:58, 22 May 2010 (UTC)Reply

Looks good now. :) Thanks! --Panda10 23:20, 22 May 2010 (UTC)Reply

So, you let an "analysis" based on a self-righteous tirade written by a pathological Anglophobe Jcwf and usage options of two corporate programs live, and my accurate evaluation of their inherent worth and relevance die? That's not fair. --Ivan Štambuk 15:12, 23 May 2010 (UTC)Reply

You said nothing that had not already been expounded at length, and you managed to insult people at the same time. Of the other edits I've noticed, people have at least made the pretence of having something to say while doing the insulting. As I said on WT:BP, it's edits like this that cause widespread outrage at the vote, and thus cause people to vote against it - as you don't listen when people tell you to stop being so abrasive (and several people have recently), I figured it was easier to just revert it. Conrad.Irwin 15:22, 23 May 2010 (UTC)Reply

Your second sentence made me ROTFL. My intention was however not to insult anybody but to deliver an accurate estimate of the AugPi's comment in order to prevent others from taking it too seriously. It is very unfortunate that people today are easily "insulted". As if the dismissal of their opinions inflicts some kind of physical harm. OK, I will try to desist or rephrase more thoroughly in the future. --Ivan Štambuk 16:15, 23 May 2010 (UTC)Reply

Thanks for your consideration. Conrad.Irwin 20:16, 23 May 2010 (UTC)Reply

This is "The world". Nothing is fair. Even when it is. O_o

Drinks? — [ R·I·C ] opiaterein — 15:26, 23 May 2010 (UTC)Reply

I'll take the Kool-Aid. --Ivan Štambuk 16:15, 23 May 2010 (UTC)Reply

Is it pimms o'clock? Conrad.Irwin 20:16, 23 May 2010 (UTC)Reply

So... what is it? Mglovesfun (talk) 12:07, 25 May 2010 (UTC)Reply

It should look like a == Heading ==, but without having edit links or appearing in the TOC - I haven't touched it in a while, so it probably doesn't anymore, particularly not if you're in vector. Conrad.Irwin 13:44, 25 May 2010 (UTC)Reply

Admittedly NAMESPACE parameters can be tricky, as you just need one exception to the rules and it doesn't work anymore. Mglovesfun (talk) 11:05, 26 May 2010 (UTC)Reply

For maintenance templates, I use {{{demo|[[Category:...]]}}}. So that they may be used in any place but, for demonstration purposes, you can not have a category. I imagine a similar approach could be used here. Conrad.Irwin 15:14, 26 May 2010 (UTC)Reply

nocat (which I learned on the French Wiktionary) does the same thing, but instead of demo=, it's demo= followed by any characters you like, and it won't make a difference. Have a look at {{delete}}. Mglovesfun (talk) 15:19, 26 May 2010 (UTC)Reply

Right, but that's inconsistent with all the other templates, and requires people to use more text... I'd prefer it to stay as demo=, which is consistent with our other parameters, such as l=, dot= and cap=. As I see you've already decided your way is better, that's fine too. Conrad.Irwin 15:25, 26 May 2010 (UTC)Reply

Not really 'more text', you can just hit any key after nocat= and it works. The thing with l= in language templates, is using nocat would mean you can't substitute it. I admit that's a clever, albeit hacky system. Mglovesfun (talk) 15:36, 26 May 2010 (UTC)Reply

What? nocat=1is 7 characters, demo= is 5 (as the chances of this being used on a page is approximately 1/2000000 it's a moot point). Then you go onto saying "nocat=" would mean you can't substitute it, so that's a disadvantage of your mechanism, or what? (again as these templates are never substituted, it's a moot point). Besides you can make if-statements substable now, see "safesubst:". All in all, I've no real idea what you're talking about; but it doesn't matter, because there is actually no difference between the two. Conrad.Irwin 15:45, 26 May 2010 (UTC)Reply

Yeah the differences are very small. FWIW you can't subst: templates like fr-noun and es-noun-m anyway, which is where you could use this sort of thing. Mglovesfun (talk) 15:52, 26 May 2010 (UTC)Reply

I did say that already :p. {{rfdate}} - I personally think templates like this should copy the style of {{rfd-sense}} et.al. to make it clear they are not part of the content. Conrad.Irwin 15:58, 26 May 2010 (UTC)Reply

I'd like to think I would have done the same thing had I not had a bus to catch. BTW I can't get on IRC using this computer, sorry. Mglovesfun (talk) 21:51, 26 May 2010 (UTC)Reply

Could you possibly update the Ido index now? I've updated my list through the 5th of May. I'll update it through the present tomorrow ;) Razorflame 21:49, 29 May 2010 (UTC)Reply

I'm still waiting on [3] - Wiktionary should be ready in the next few days. Then, yes I will. Conrad.Irwin 22:55, 29 May 2010 (UTC)Reply

Ok, no problems :) Cheers, Razorflame 23:01, 29 May 2010 (UTC)Reply

Is Wiktionary ready for an update in statistics now? Razorflame 13:19, 7 June 2010 (UTC)Reply

Oh, yes! Good point. Conrad.Irwin 13:24, 7 June 2010 (UTC)Reply

Hi. I have been given your name by User:Sven70 as a neutral admin on Wiktionary. I would note that I recently initiated a discussion at the Beer parlour regarding the indef block of Sven70. I am trying to find an admin I can open up a dialogue with, so I might facilitate a discussion regarding (and hopefully involving) Sven70 and the issues surrounding his communication difficulties. I am hoping that you might be willing to investigate that possibility. I shall keep an eye open for any response. Thank you. LessHeard vanU 20:57, 30 May 2010 (UTC)Reply

I am willing to discuss this. I described my opinions in an email I sent to you on the 14th May (23:45). I also think the community has made it fairly clear how they stand - so while I'm happy to negotiate, their wishes must also be respected. Conrad.Irwin 22:19, 30 May 2010 (UTC)Reply

Indeed you did email me, sorry I forgot that. My hope is that by talking directly with Sven70, a lot of the frustration he feels in not being able to respond quickly/comprehensively will dissipate and discussion will be more fruitful. I am not sure that Sven70 wants to "negotiate", but rather discuss in the detail that typing does not permit him the concerns he has. I have advised him that it may not ultimately change the way things are, but he is more concerned that he is properly understood. I have spoken to Sven on voip, and he is easy to understand (providing the line is good) and expresses himself well. My idea is that if I can broker a contact, then Sven and whoever can start a process by which progress to some sort of resolution may occur. If you and Sven were to "chat", is there any rules or issues you would want agreed first? LessHeard vanU 22:43, 30 May 2010 (UTC)Reply

If this could wait until Friday, I'd be much appreciative. My finals start in 36 hours - I should sign off until they're over. Conrad.Irwin 23:06, 30 May 2010 (UTC)Reply

Hey, no problem. Your finals will determine a lot of your life to come; just make sure you are relaxed, and you will do fine. Best wishes. LessHeard vanU 23:15, 30 May 2010 (UTC)Reply

This should be fine, I'll try to create myself a Skype account - not sure when I'll have a suitable period to use it, probably at some point tomorrow. Conrad.Irwin 15:43, 5 June 2010 (UTC)Reply

Hi there Conrad. Can you change the automated creation of Ido plurals that have hyphens in them to not link the inflection line, but for multiple word entries, to link them?

Basically, what I am asking is this:

Say I have a word imprim-proburo. Could you make the automated creation of imprim-proburi to not link imprim and proburi? Say I have a word heleboro negro. Could you make the automated creation of heleboro negro link heleboro and negro?

The reason for this is because if it links proburi and imprim (and imprim itself is not an Ido word), it makes more work for me because I have to unlink them manually. Could you possibly make it so that it doesn't do that for just the words with hyphens? Thanks, Razorflame 17:36, 1 June 2010 (UTC)Reply

Is it always the case in Ido that hyphens will be connecting two non-words? Conrad.Irwin 17:48, 1 June 2010 (UTC)Reply

proburi is a word (plural of proburo), but imprim is not. No, it will not always be the case, but in most instances, it will be. Maybe if you could just remove the option to automatically create hyphenated Ido plurals? I can make them manually pretty easily ;). Razorflame 17:51, 1 June 2010 (UTC)Reply

Should be done. Conrad.Irwin 18:03, 1 June 2010 (UTC)Reply

Ok. Thanks for the help! Razorflame 18:19, 1 June 2010 (UTC)Reply

I noticed. Thanks for reverting, and noticing me =) --Diego Grez 17:54, 1 June 2010 (UTC)Reply

Hi. I wonder if you could have a look at [[User:Msh210/h1.js]] and tell me whether you think this will work, and whether you think the benefit in visibility is worth the resource drain. Thanks. Tweaks also most welcome (just edit the page).​—msh210℠ 19:19, 2 June 2010 (UTC)Reply

I am not sure who benefits from things like this - is it just IE6 users or do other windowsy folk need them too? The code looks fine, and compared to the vast swathes of javascript we already force on people, it's not huge (we could even make it only load for those browsers that need it). There are a few minor javascript errors, but the general approach is fine: src() isn't defined where you use it first, so put "function src()" at the top. Then wrap the whole thing in addOnloadHook(function(){ ... }); so that it doesn't try and find the heading. Also declare "var i", and use .className += " " + scr() rather than .setAttribute("class"). One other thing that would seem to fail is for titles with spaces and punctuation - which presumably shouldn't count towards which script it is in - may cause it to fail completely; not sure the best way to fix that, maybe just sum all the src() occurrence and chose the highest. I would be quite excited if we could make this work as an alternative to our current Xyzy approach too (either in Javascript or PHP). Conrad.Irwin 19:47, 2 June 2010 (UTC)Reply

Thanks much for the suggestions. I've tweaked it to allow for punctuation, I think. I have no idea who needs this and who not. Is it only IE(≤6)? Also, I'm not sure what the table below is. They're not the most common characters in each script (or are they? 41 is uppercase), so what makes these characters switchworthy?​—msh210℠ 20:08, 2 June 2010 (UTC)Reply

Prince Kassad may know. They're the first character with that script after a run of characters with a previous script, 0x41 = A, based on the unicode character database. Conrad.Irwin 20:21, 2 June 2010 (UTC)Reply

/0/python/lib/unicodescript/unicodescript$ cat scripts.txt | grep -v 'Common' | grep -v 'Zzzz' | grep -v 'Inherited' | uniq -f1 
00041 Latin
00370 Greek
003E2 Coptic
003F0 Greek
00400 Cyrillic
00531 Armenian
00591 Hebrew
00606 Arabic
00700 Syriac
00750 Arabic
00780 Thaana
007C0 Nko
00800 Samaritan
00900 Devanagari
00981 Bengali
00A01 Gurmukhi
00A81 Gujarati
00B01 Oriya
00B82 Tamil
00C01 Telugu
00C82 Kannada
00D02 Malayalam
00D82 Sinhala
00E01 Thai
00E81 Lao
00F00 Tibetan
01000 Myanmar
010A0 Georgian
01100 Hangul
01200 Ethiopic
013A0 Cherokee
01400 Canadian Aboriginal
01680 Ogham
016A0 Runic
01700 Tagalog
01720 Hanunoo
01740 Buhid
01760 Tagbanwa
01780 Khmer
01800 Mongolian
018B0 Canadian Aboriginal
01900 Limbu
01950 Tai Le
01980 New Tai_Lue
019E0 Khmer
01A00 Buginese
01A20 Tai Tham
01B00 Balinese
01B80 Sundanese
01C00 Lepcha
01C50 Ol Chiki
01D00 Latin
01D26 Greek
01D2B Cyrillic
01D2C Latin
01D5D Greek
01D62 Latin
01D66 Greek
01D6B Latin
01D78 Cyrillic
01D79 Latin
01DBF Greek
01E00 Latin
01F00 Greek
02071 Latin
02126 Greek
0212A Latin
02800 Braille
02C00 Glagolitic
02C60 Latin
02C80 Coptic
02D00 Georgian
02D30 Tifinagh
02D80 Ethiopic
02DE0 Cyrillic
02E80 Han
03041 Hiragana
030A1 Katakana
03105 Bopomofo
03131 Hangul
031A0 Bopomofo
031F0 Katakana
03200 Hangul
032D0 Katakana
03400 Han
0A000 Yi
0A4D0 Lisu
0A500 Vai
0A640 Cyrillic
0A6A0 Bamum
0A722 Latin
0A800 Syloti Nagri
0A840 Phags Pa
0A880 Saurashtra
0A8E0 Devanagari
0A900 Kayah Li
0A930 Rejang
0A960 Hangul
0A980 Javanese
0AA00 Cham
0AA60 Myanmar
0AA80 Tai Viet
0ABC0 Meetei Mayek
0AC00 Hangul
0F900 Han
0FB00 Latin
0FB13 Armenian
0FB1D Hebrew
0FB50 Arabic
0FF21 Latin
0FF66 Katakana
0FFA0 Hangul
10000 Linear B
10140 Greek
10280 Lycian
102A0 Carian
10300 Old Italic
10330 Gothic
10380 Ugaritic
103A0 Old Persian
10400 Deseret
10450 Shavian
10480 Osmanya
10800 Cypriot
10840 Imperial Aramaic
10900 Phoenician
10920 Lydian
10A00 Kharoshthi
10A60 Old South_Arabian
10B00 Avestan
10B40 Inscriptional Parthian
10B60 Inscriptional Pahlavi
10C00 Old Turkic
10E60 Arabic
11080 Kaithi
12000 Cuneiform
13000 Egyptian Hieroglyphs
1D200 Greek
1F200 Hiragana
20000 Han

If you're still working on finals (or still celebrating their completion), please feel free to not respond to or even read this until a more convenient time. I was thinking that it should be possible to grab the nearby pages from the indices, at least those which are auto-generated. This would have a few advantages, I think. First, I would expect it to be faster, as, from what I can tell, hippietrail's approach actually munches data every time, instead of simply gathering it. Also, this would keep all the code here, on the one project (as opposed to having to call toolserver). Finally, this would allow any custom ordering concerns which go into the indices to be represented in the nearby pages. My skills, while, I think, developing nicely, are still nowhere near up to the task of implementing such a thing, so all I can do is nag you to do it for me. So, if you have better things to do, obviously feel free to do them instead. I just thought I'd throw it by you, in case you hadn't already thought of it. Oh, and when you do have more time, expect me to ambush you sometime on irc and force you to explain some of the stickier points of the translations editor. You've made some excellent documentation, and so I'm making slow progress on my own, but I expect I'll still need some hand-holding on a few bits. Cheers. -Atelaes λάλει ἐμοί 14:31, 5 June 2010 (UTC)Reply

Hey. I'm FINISSSHED! so will have time for this stuff as soon as I work out how not to sit around grinning like a maniac all day. I'm on IRC on-and-off depending on what else is up, so should be able to come across each other at some point. The idea behind hippietrail's script is to get all the information at once - and in every language - loading all the indices on pages like a will take a long time (and you'd still need to guess for languages with no indexes). I did talk to him about re-using the sorting code, but he couldn't find an easy way to integrate it all together, so tries to use the inbuilt locales. (He also does a lot of clever things to try and optimise retrieval of results in a single SQL query, if I remember correctly). I think that it should stay on the toolserver, but that we should probably move it to the wiktionary shared account so that when it has problems more people (currently just hippietrail and I) can try to fix them (want to join in?). Integrating sorting (and, more importantly I think, inflected-forms removal) from the indexing is definitely possible, but... :). (For what it's worth, the current problem is just that the pages are ISO 8859, but served with a utf-8 content type, I'm not sure whether that was caused by toolserver "upgrading" things or a mistake on hippietrail's part, but I can't edit his files so I can't fix it - I may try creating a clone at /~enwikt/, but he tends to rely on lots of perl modules, and I haven't worked out how to install them properly yet. Conrad.Irwin 15:41, 5 June 2010 (UTC)Reply

Hey there Conrad. At the very bottom of the page, it is listing 21 pages as belonging to the language "english" (uncapitalized). Is there any way that you can sequester those pages so that they can be fixed? Thanks, Razorflame 14:35, 7 June 2010 (UTC)Reply

Thanks for pointing it out - they're now fixed. Another "little joke" from our friend Rising Sun :). Conrad.Irwin 14:45, 7 June 2010 (UTC)Reply

Hi Conrad,

Please have a look. User_talk:Conrad.Irwin/editor.js#zh_translation. Please reply there if you do. --Anatoli 03:19, 8 June 2010 (UTC)Reply

I would love your input at "Proposal for (toned) pinyin words" at BP. So far I've gotten very little response. Thanks. ---> Tooironic 08:34, 8 June 2010 (UTC)Reply

Hello Conrad, I think your bot is malfunctioning slightly. It is removing loads of good anagrams. Take for example this edit or this edit. I suspect that it might have something to do with the fact that these anagram links don't have a #langname piped, because along with these removals I see it adding piped anagrams, but I could be wrong. --JorisvS 13:47, 11 June 2010 (UTC)Reply

Gah. It's because it wasn't supposed to be doing Finnish - but I configured it wrong. Thanks for pointing it out. Should it do anagrams for Finnish in the future - if so how should it deal with ä and Ž etc.? I'll re-set it going with the problem solved, and then find the list of pages to revert. Conrad.Irwin 13:58, 11 June 2010 (UTC)Reply

Well, if it can, wouldn't it improve Wiktionary? I don't know how it then should handle these letters. I don't speak Finnish and just stumbled upon your bot in action. I have read that French ç is treated as a regular c, but then again French doesn't consider this a separate letter, while Finnish considers ä a separate letter but š and ž variants of s and z respectively (see: w:Finnish alphabet). --JorisvS 14:12, 11 June 2010 (UTC)Reply

I doubt it bothers anyone at all if a bot removes these. Mglovesfun (talk) 14:28, 11 June 2010 (UTC)Reply

It still shouldn't :p. Would you like it do add them all instead? Conrad.Irwin 14:28, 11 June 2010 (UTC)Reply

Hello, Conrad. I noticed that, while mentioning that "board game piece" is not a word, you removed it from a Wikisaurus list.

As you can see, the page WS:board game piece lists the hyponyms draught and stone, among other words, which are also hyponyms of WS:board game and WS:game. Since I feel that all these WS pages would be rather incomplete without listing such types of game board pieces, I came to at least one possible conclusion: to list "board game piece" as an hyponym of "board game", thus logically resolving my concerns on incompleteness, by having draught as an hyponym of board game piece, which is an hyponym of board game, which is an hyponym of game.

I've had one or more recent conversations on this subject. [4] As a result, similarly, "beautiful woman" and "ugly woman", despite not being words, are listed as hyponyms of WS:woman.

While I understand that "board game piece" is not a word, I think nonetheless that it should be listed as an hyponym of "board game". --Daniel. 18:24, 14 June 2010 (UTC)Reply

It seems good to link to the Wikisaurus entry (truth be told, I didn't notice the [Ws] link when I made my edit), but not to link to the dictionary - is there a template for that? I've just hacked it in at the moment, but it might be nicer just to use the link. I'm not sure in what way a "draught" is a hyponym of a "board game", a meronym of the hyponym ("draughts") seems more likely. I was also not sure whether to list pieces or piece, I went for pieces as they are often taken as a unit, on reflection I've updated it to be piece so that the dictionary link is more useful. Conrad.Irwin 19:21, 14 June 2010 (UTC)Reply

Hmm, looks like I've mistyped or misunderstood something in my message; let me correct myself and organize my thoughts:

• You're right, draught (the piece) is indeed a meronym of draughts (the game). In addition, draught is a hyponym of board game piece (which names a WS page), which is a meronym of board game. Therefore, unless I'm mistaken once again, draught is also a meronym of board game.

Then, my suggestion of linking WS:board game piece to WS:board game remains. Thanks for readding that link. I agree with you on "piece" being better than "pieces". I believe that a template to link to the WS page but not to the respective dictionary entry does not exist, but may of course be created; although, I'm not sure if we want to standardize the format "piece     [Ws]", comprised of a link to the dictionary entry which is different from the link to the WS page, because it is apparently not usable to remove the eternal red links from WS:woman. --Daniel. 20:09, 14 June 2010 (UTC)Reply

I think you used hyponym in place of meronym in your first paragraph, it's clear that you meant what seems correct to me. Surely you could just pick one of the synonyms to use? Otherwise you could word the link as Ws:ugly woman? Conrad.Irwin 20:18, 14 June 2010 (UTC)Reply

On the suggestion of just picking one of the synonyms of WS:ugly woman to name that page, I have yet another concern: "ugly woman" seems very natural and self-explanatory, unlike WS:beldame, WS:brown bagger, WS:hag, WS:harridan, WS:moose, WS:munter and WS:Sybil. Therefore, according to my standards, I couldn't find any synonym good enough to name the page.

On the other hand, your suggestion of placing a direct link to Wikisaurus seems good enough to me. Then, I've implemented it in WS:beautiful woman, WS:woman and WS:board game. I've chosen the final format "Wikisaurus:ugly woman" to imitate other similar inter-WS links, that exist mainly in See also sections. --Daniel. 22:09, 14 June 2010 (UTC)Reply

I like it. Conrad.Irwin 22:13, 14 June 2010 (UTC)Reply

Hi Conrad. I was reading the more recent comments made in WT:RFDO#Template talk:q. Michael Z. made a good point (in his post timestamped 2010-04-12 17:46; ¶5): "The use of empty {{q}} is very bad practice. Then when I view the source wikitext the most important word or phrase is omitted. The wikitext breaks silently and permanently if I copy it anywhere else. Doesn't this break Mediawiki's search?" To rectify this, would it be possible to edit the template so that, when the editor writes {{q}} (without defining the first parameter), the template automatically adds the {{PAGENAME}} (as {{q|PAGENAME}})? Would prefixing the {{PAGENAME}} with subst: (as {{subst:PAGENAME}}) achieve that?  — Raifʻhār Doremítzwr ~ (U · T · C) ~ 22:12, 14 June 2010 (UTC)Reply

No, I don't think that's possible. We could write a bot to do it every so often - or ask autoformat nicely. I'm currently thinking that any kind of additional mark-up in these quotes is bad, and we should just drop the emphasis altogether, but that may be a heretical view. Conrad.Irwin 22:15, 14 June 2010 (UTC)Reply

Yes, I saw that opinion expressed. I personally disagree, but there's no reason we can't accommodate those who think that. WT:PREFS would allow the formatting that {{q}} applies to be changed however one wished, be it a box, emboldenment, or nothing at all. In the meantime, I'll go ask AF for a favour.  — Raifʻhār Doremítzwr ~ (U · T · C) ~ 22:36, 14 June 2010 (UTC)Reply

If you need a momentary distraction, here is a special request. Can you create a definitions file from Finnish Wikipedia with definitions of English words? http://download.wikimedia.org/fiwiktionary/20100605? It's for an offline dictionary on my blackberry. ~ heyzeuss 14:39, 15 June 2010 (UTC)Reply

Most certainly :). http://toolserver.org/~conrad/finnish-english.tsv (1.8MB) Conrad.Irwin 14:50, 15 June 2010 (UTC)Reply

Thank you! ~ heyzeuss 15:04, 15 June 2010 (UTC)Reply

You have adjusted {{ws}} so that "{{ws|ugly woman|-}}" renders an item without a hyperlink. This solution is okay, yet imperfect. The second parameter of "ws" is normally used for a gloss, as {{ws|self-evident|Obviously true, and requiring no proof, argument or explanation}}; with your adjustment, the second parameter is overloaded with an additional meaning. It seems to me that "{{ws|ugly woman|nohyperlink=1}}" would be better. What do you think of it? --Dan Polansky 08:23, 16 June 2010 (UTC)Reply

I think that, in the case there is no link, there should never be a need for a gloss. The solution is similar to how a lot of our other templates work to "get rid of a link", so it's easy to remember. The problem with nohyperlink=1 is that it's not used anywhere else (that I know of), and it's not nolink= nopagelink= or any host of other variants (sure you could make other variants work too, but then the wikitext will be harder to work with automatically). If you're concerned about additional meaning, think of it as an optional second-positional boolean parameter that changes the behaviour of the template to hide the link and gloss, as the gloss would never be "-", it's easy to tell which is intended. I strongly prefer my solution, which is why I did it that way in spite of what you suggested originally, because it is easier to use, understand, and remember. Conrad.Irwin 09:55, 16 June 2010 (UTC)Reply

I strongly dislike your solution. Positional parameters should not be overloaded with meaning. A parameter should not take on a new meaning in a situation in which its main meaning is inapplicable: each meaning should have a dedicated parameter. There is nothing wrong with "nohyperlink" not being used by other templates given that other templates do not have a parameter with this meaning. Your solution is faster to type, but hardly easier to use, understand, and remember. I don't think at all it is easier to understand: it looks as if the intention were to have no gloss, but the effect is to have no hyperlink. --Dan Polansky 12:54, 16 June 2010 (UTC)Reply

Your assertion that "positional parameters should not be overloaded with meaning" is demonstrably ignored throughout Wiktionary - other than that, it's a matter of personal taste; I happen to prefer solutions that are easier to remember, use and process over those that are perhaps technically more sound, but I can see where you are coming from. A more tasteful solution might be to create {{ws-only}} for this case, but I don't really see the need. Conrad.Irwin 14:22, 16 June 2010 (UTC)Reply

I did not know that my assertion was ignored throughout Wiktionary; that is a pity anyway. I would go for "{{ws|ugly woman|nohyperlink=1}}" while creating {{ws only}} as a shortcut for that, given that a shortcut seems required by you, and possibly also by Daniel. ("ws only" instead of "ws-only", on the model of "ws sense" and other Wikisaurus templates.) I really find it hard to accept needless overloading of parameters. --Dan Polansky 14:52, 16 June 2010 (UTC)Reply

You've used {{en-noun}}, surely? Not to mention {{context}}, perhaps the most widespread "offender". Wikitext isn't a programming language, so conventions from there really don't make so much sense. A shortcut is definitely required, or I'll go back to using the much easier to read plain link format that you seem to dislike. Conrad.Irwin 14:58, 16 June 2010 (UTC)Reply

I disagree with the following quote from Dan Polansky: "it looks as if the intention were to have no gloss, but the effect is to have no hyperlink".

Apparently, the intention of {{ws|ugly woman|-}} is to have both no gloss and no hyperlink. Namely, there are four possibilities:

• {{ws|ugly woman|-}} (no hyperlink and no gloss)
• {{ws|ugly woman}} (hyperlink and no gloss)
• {{ws|ugly woman|abc}} (hyperlink and gloss)
• (then, a gloss and no hyperlink is inapplicable)

That's very easy and intuitive, in my opinion, in addition to self-explanatorily appearing instantaneously in any WS page where they are used. --Daniel. 15:07, 16 June 2010 (UTC)Reply

Well, I don't think that this is intuitive at all; where does it follow that {{ws|ugly woman|-}} means no hyperlink? What is it apparent from? Not apparent at all, if you ask me. I admit that these sorts of claims about what is intuitive and what not are rather subjective. Anyway, you are two for one solution against me one for another solution, so let it be. --Dan Polansky 15:32, 16 June 2010 (UTC)Reply

Can you please archive your talk page? It takes a few minutes to load on my PC, unless I use "tricks" to read new sections, such as typing |&action=edit&section=x somewhere or comparing revisions. --Daniel. 14:35, 16 June 2010 (UTC)Reply

He's probably still trying to find another version of "1".  :-P -Atelaes λάλει ἐμοί 14:39, 16 June 2010 (UTC)Reply

It's actually because I still need to do that Mapudungun thing right at the top - maybe it's time to admit I never will. Conrad.Irwin 14:44, 16 June 2010 (UTC)Reply

Well, you could replace "Create more flexible and robust hiding mechanisms, to keep Atelaes from crying." on your todo list with it. I got sick of waiting on that one, cried a little, and then started doing it myself. -Atelaes λάλει ἐμοί 14:49, 16 June 2010 (UTC)Reply

But then I'd "never" do it :p. Conrad.Irwin 14:51, 16 June 2010 (UTC)Reply

That's much better. Thanks for archiving it. --Daniel. 15:10, 16 June 2010 (UTC)Reply

Hi Conrad, there is something going on with the vowel pages (except A). Index:Hungarian/e, Index:Hungarian/i, Index:Hungarian/o, Index:Hungarian/ö, Index:Hungarian/u, Index:Hungarian/ü. It seems that the first 1-3 sections are out of place. Can you please take a look when your time allows? Thanks. --Panda10 20:50, 16 June 2010 (UTC)Reply

Should now be fixing itself (will take about half an hour to run). Sorry, again. Conrad.Irwin 15:39, 17 June 2010 (UTC)Reply

Thank you! :) --Panda10 20:54, 17 June 2010 (UTC)Reply

A lot of the recent improvements I've made to the targeted translations script were based on code I'd seen in your editor script. I just thought it'd be appropriate to note that, and say "thanks". -Atelaes λάλει ἐμοί 12:04, 17 June 2010 (UTC)Reply

Ay, C-dawg, wassup ma nigga? Do a favor for a brother: I need two wikilinked lists of absolutely ALL (including form-of) {{xcl}} and {{hy}} entries. I intend to check and expand them all. Peace out. --Vahagn Petrosyan 14:20, 17 June 2010 (UTC)Reply

All is good, big V. http://toolserver.org/~conrad/requests/ contains three files for you, "armenian-old-armenian.tsv" is a list of all armenian and old armenian words with definitions, armenian.list and old_armenian.list are the wiki-format lists. Enjoy. Conrad.Irwin 14:47, 17 June 2010 (UTC)Reply

Thanks. Now looking forward to translation table creator. --Vahagn Petrosyan 15:02, 17 June 2010 (UTC)Reply

I have a few bug fixes for exsiting stuff to do first, but one day :). Conrad.Irwin 15:05, 17 June 2010 (UTC)Reply

I figure you are very busy these days, but can you update http://toolserver.org/~conrad/requests/armenian.list when you have a minute? Only that list, not the others. --Vahag 16:32, 8 December 2010 (UTC)Reply

Would you be willing to look over the new code at User:Atelaes/QuoteHide.js, just to make sure nothing's wrong with it. I've checked and double-checked, and checked on all my browsers, and I'm almost positive it works fine. However, I'd very much appreciate another set of eyes before I push it to everyone on the interblags. Many thanks. -Atelaes λάλει ἐμοί 11:49, 18 June 2010 (UTC)Reply

Overall it looks great - modulo some bizarre indentation :p, though it seems to fail on pages with nested definitions for two reasons, firstly because each nested definition with a quotation gets two buttons, and secondly because the outer definition's toggle affects the inner definition's quotes. I'll have a go at fixing it. Conrad.Irwin 13:53, 18 June 2010 (UTC)Reply

I knew I couldn't restructure the whole thing without a few issues popping up. Thanks for the fix. What's wrong with my indentation? -Atelaes λάλει ἐμοί 14:01, 18 June 2010 (UTC)Reply

Nothing serious, it's just not like mine :). Are we ready to go then? Conrad.Irwin 14:04, 18 June 2010 (UTC)Reply

By ready to go, do you mean ready to put them into common.js? If so, yes, I think so. With your new hiding interface, there's no need for a PREFS option to turn them off, as Wiktionary just remembers the user's preference by magic. All we need is to kill the WT:PREFS option to turn them on, add the code to common.js, and maybe protect the code itself (I think we'd be kicking ourselves if some vandal inserted some malicious code to all our users that way. :p). -Atelaes λάλει ἐμοί 14:09, 18 June 2010 (UTC)Reply

Thank you for spotting the problem. The first letter (N) was the Greek one (echo Νoun | od -c gave: 0000000 316 235 o u n) --flyax 21:44, 18 June 2010 (UTC)Reply

Just came across these: [5], [6]. Don't know if you've already fixed this. --Bequw → τ 06:50, 21 June 2010 (UTC)Reply

It is caused by pressing save before the page has loaded, it's not really fixable. Conrad.Irwin 09:10, 21 June 2010 (UTC)Reply

It's not a "bug", it's people not reading the entry before they hit save. It's more like misuse than a bug. Mglovesfun (talk) 09:45, 21 June 2010 (UTC)Reply

So I ran my crapbot to shift Maltese nouns off {{infl}}, but for whatever reason, whoever was doing Maltese didn't use the template to specify plurals very often, so there's a lot of raw text like {{mt-noun|m}} (''plural'' '''[[ x plural ]]'''). If you have time, could you hunt down stuff that links to {{mt-noun}} and replace that stuff so it looks like {{mt-noun|m| x plural}} or {{mt-noun|f| y plural}}? (It's possible that some of them won't have that m or f, so it'll be important to switch {{mt-noun}} with {{mt-noun|| x plural }} or the gender will come up as a question mark.)

Sorry I'm always coming to you with silly simple stuff like this :D But for my future reference, so I can try to do it myself, how do you make a wildcard in mw python files? — [ R·I·C ] opiaterein — 23:37, 21 June 2010 (UTC)Reply

I'll have a look at this tomorrow. What kind of wildcard do you want? Conrad.Irwin 23:21, 29 June 2010 (UTC)Reply

Something so that the text around a certain word/group of words will be edited, like the (''plural'' '''[[ and ]]''') but not what's between the pipes. — [ R·I·C ] opiaterein — 23:31, 29 June 2010 (UTC)Reply

Sorry for the delay. What you want to do is use capturing groups - I think. The example I used to do most of this list was quite convoluted (because I am checking for an optional "|m" and using "|" if that is not present) but the idea should be clear - if not there are plenty of tutorials, or just ask for clarification. Conrad.Irwin 09:33, 18 July 2010 (UTC)Reply

newtext = re.sub(r"\{\{mt-noun(\|[mf])?\}\} *\('*plural'* *'*\[\[(.*)\]\]'*\)", lambda m: "{{mt-noun%s|%s}}" % (m.group(1) if m.group(1) else "|", m.group(2)), text)

Three cases left - your turn :p. Conrad.Irwin 09:33, 18 July 2010 (UTC)Reply

• bank: {{mt-noun}} (''plural'' '''[[banek]] (bank), '''[[bankijiet]] (bench)''')
• dada: {{mt-noun}} (''plural'' '''[[dadi]]''')
• passat: {{mt-noun|m}}, '''[[passata]]''' {{f}}

Regarding {{ang-conj}} and {{ang-verb}}, the switch for the extra categories doesn't seem to be working. Ifnothing else, {{#ifeq:{{{type|}}|strong... would work, but if a switch works for the head word, why wouldn't it work for the categories too? Mglovesfun (talk) 18:24, 25 June 2010 (UTC)Reply

I assume this was fixed by Ruakh? Conrad.Irwin 23:18, 29 June 2010 (UTC)Reply

Yep. Mglovesfun (talk) 13:43, 9 July 2010 (UTC)Reply

Hiya. Hope you're well. Is this template still wanted or to be deleted?​—msh210℠ (talk) 16:59, 29 June 2010 (UTC)Reply

Ya, all is well I just snuck off for a choir tour for a few days without telling anyone :D. I've deleted the template. Conrad.Irwin 23:17, 29 June 2010 (UTC)Reply

Hey, I'm trying to allow toggling of the TOC in my tabbed languages bit, but the right hand gadget/pref seems to prevent any such toggling. You appear to be the creator of this capability. Anyway, I was wondering what the reasoning behind the 'display:inline' portion was, as this is the problematic part. Many thanks. -Atelaes λάλει ἐμοί 19:17, 4 July 2010 (UTC)Reply

I don't know, and think it's very peculiar, it was probably copied from somewhere else so feel free to change it. Conrad.Irwin 07:00, 6 July 2010 (UTC)Reply

Can this be updated? We're nearly three months now since the last update. --EncycloPetey 17:14, 7 July 2010 (UTC)Reply

Now done - though it was more recent than that. I've taken to updating Wiktionary:Statistics/generated to avoid colliding with human edits. Conrad.Irwin 09:00, 8 July 2010 (UTC)Reply

Please can this be updated. I worked with entries T-Z, so there are now much less. --Volants 13:41, 9 July 2010 (UTC)Reply

I think I did about the first 2000 but nearly went insane after that. Now I just fix them as I find them - with all the parts of speech and all the languages, we may be talking tens of thousands of entries. Mglovesfun (talk) 13:43, 9 July 2010 (UTC)Reply

Yes, maybe it will take one year to clean up them all. Now I'm injured, however, I have much more time to work on all of this. --Volants 14:00, 9 July 2010 (UTC)Reply

Only 5710 to go :). Thanks a lot! Conrad.Irwin 06:59, 10 July 2010 (UTC)Reply

How hard would it be to scan generally for entries without correct PoS category (not just English & "noun")? There'd be lots of lists, but we could link to them from language-specific cleanup areas. --Bequw → τ 13:18, 14 July 2010 (UTC)Reply

A more-than-bitesize portion of the entries in there are sumo terms. A gulp-size portion of them are RFV candidates. They were created by SemperBlotto, though. --Volants 18:37, 15 July 2010 (UTC)Reply

Really not hard at all, just add things to User:Conrad.Bot/Missing_categories and I'll try and hook up the upload code to upload them all when it creates them. Conrad.Irwin 08:55, 18 July 2010 (UTC)Reply

Would you be able to bot replace {{cite book}} with {{reference-book}}? I made the first redirect to the second, but I eventually want to get rid of it entirely to avoid confusion with {{cite-book}}. Cheers. --Bequw → τ 13:20, 14 July 2010 (UTC)Reply

Yes, of course - were you planning to do the same for the other cite_ templates?. Sorry for the delay in replying, I seem to have accidentally landed myself an 80-hour-week job :). Conrad.Irwin 08:38, 18 July 2010 (UTC)Reply

Thanks. I was planning on merging the code for the cite-* and quote-* templates. Probably having one set call the other with a different default prefix. Does this mean our done with your dissertation? --Bequw → τ 15:00, 18 July 2010 (UTC)Reply

Yes - I'm now graduated and everything, which I was told meant I'd get more free time. Working as first employee for a startup [7] seems to be even more intense - maybe I just did university wrong :). Conrad.Irwin 16:45, 18 July 2010 (UTC)Reply

Hey again. I was asking Atelaes back in january how to make a index and he mentioned a slow way of doing it and certain magical ways of creating one with the help of a great wizard called Conrad =). I have added alot of words since january and it would be fun to have an index, If you find the time to help me it would be awesome. -Edelstam 17:11, 25 March 2009 (UTC)Reply

Hey Conrad, been a while, thanks for updating the mapudungun index again. I can't get on the irc anymore because got this TOR program on my computer and I can't get it to work with some IRC servers. So I'll have to ask you here. I have been working some with trying to start up the mapudungun wikipedia and there they are working with some separate solutions for the different orthographies(what I mistakenly have called alphabets) within one wikipedia. I looked some at the serbian wikipedia where they use two different alphabets for example. I got some comments about the mapudungun language category here on english wiktionary, that it was confusing to have words from two different orthographies in one place. So I was thinking if we could possibly adopt some solutions the serbian language category has here. They have for example a index with the two different alphabets, look here. Since we only have words in two orthographies it would mean a similar solution to the one the serbian category uses. What do you think about this, is it alot of work? -Edelstam 20:14, 26 April 2009 (UTC)Reply

Hi Edelstam, the main problem is automatically telling the two apart. If there is some indication in the page text, or even a way of testing the page titles, that will tell me which alphabet which entry belongs to, then it shouldn't be that hard. I can obviously look for some letters that only appear in one of the alphabets, but I don't know enough about Mapudungun to know if I will always be able to tell the difference. Conrad.Irwin 20:19, 26 April 2009 (UTC)Reply

Yeah I was thinking about how to separate the words since they don't look different like the Serbian Cyrillic- and Latin alphabet words. Right now all words written in the raguileo orthography have this in the article tho: (using Raguileo Alphabet), look at cew for example. Some words are the same in both orthographies and in those cases they look like this word: awka. Words written in Unified alphabet look like this: chafod. Can you use that for the bot? -Edelstam 20:47, 26 April 2009 (UTC)Reply

Yes, that's enough. I'll try to change the indexing scrit at some point. Conrad.Irwin 08:19, 28 April 2009 (UTC)Reply

Hi again, it's been a while since I worked on the mapudungun but I just got some time again to add words. I asked you back then in 2009 if you could possibly change the index so that it looks like Index:Serbian. Right now there are words in the current index written in two different orthographies which are the two top ones in this page Appendix:Mapudungun alphabet. People has also added words which has not been catogoriesd, so it would possibly be good to have an Index like the serbian index but with link to a page with uncategorised words aswell. What do you think of all this? -Edelstam 10:41, 15 July 2010 (UTC)Reply

Sorry. I'll try to fit this in. Conrad.Irwin 09:15, 18 July 2010 (UTC)Reply

It now splits the page into three, those that are written in Raguileo (the majority) go on one, Unified on the next, and unspecified on a third. Conrad.Irwin 16:26, 15 August 2010 (UTC)Reply

Thank you very much! -Edelstam 09:36, 17 August 2010 (UTC)Reply

I am trying to hire someone to do the same thing you did for the English Wiktionary to extract all the definitions, but for the Chinese Wiktionary (and maybe later the Japanese one). I've looked through the code and it didn't seem too complicated. I was wondering how long it took you to write and how tough it might be for someone else (unfamiliar with Wiktionary) to adapt it to the Chinese Wiktionary. Thanks! Swiftly 19:08, 15 July 2010 (UTC)Reply

It's not really possible to say how long it took to write, the actual code took a few hours - but I wouldn't have been able to do it so quickly without knowing quite a lot about how the English Wiktionary works and a week or so of finding problems on wiktionary and fixing them. I just had a quick scan through the Chinese Wiktionary, and frankly it looks very scrappy - though the Japanese one may be better. I would imagine you would need a reasonably significant amount of effort to understand the various formats you can find, but the actual extraction of words shouldn't be too hard. I'd be happy to lend advice as you need it, but I'm suddenly much busier than I thought I would be, so I probably can't do the coding (though please use anything I've already written if it helps!). When you get something working, I'd be happy to host it next to the English definition lists (if you want). Conrad.Irwin 08:49, 18 July 2010 (UTC)Reply

Okay, thanks for taking a look. I hope my rented coders can handle it--I may try to pass some specific questions on to you if we get stuck. Yeah, if it starts working, I think it'd be useful to host it next to the English definitions. Thanks! Swiftly 16:25, 19 July 2010 (UTC)Reply

On second try, the rentacoder got something working. Turns out there aren't all that many definitions of Chinese words there; we were only able to get about 8000 definitions, half in Chinese, half in other languages. I can send you the script that does the parsing if you're still interested in putting it up there. How should I send it to you?--Swiftly 14:39, 2 November 2010 (UTC)Reply

Cool! I'm <wiktionary username>@gmail.com Conrad.Irwin 15:34, 2 November 2010 (UTC)Reply

At WT:RFM#Category:Radio, there may be a dialog (now mostly a monolog) about problems in mapping contexts to categories, based on the instance Category:Radio. Even if there is a good ad hoc solution for the specific case, isn't the general problem one that is approaching the point where it needs to be addressed? DCDuring TALK 11:08, 16 July 2010 (UTC)Reply

Hi, could Conrad.Bot make an Index:Anglo-Norman? Also could it overwrite Index:Old English, that hasn't been updated since 2007, and it's almost entirely composed of red links, some of them look to me like invalid page names. Cheers, Mglovesfun (talk) 08:18, 31 July 2010 (UTC)Reply

I moved its contents to Wiktionary:Requested entries (Old English)/list, and have populated Anglo-Norman and Old English. (Treating the ƿ like a form of w). Conrad.Irwin 17:31, 15 August 2010 (UTC)Reply

Hi Conrad.Irwin. I wanted to get in touch with you and a few other members of the Wiktionary community about the upcoming rollout of the Usability Initiative features (Vector and enhanced editing features). I was referred to you by Ariel who thought you might be a good person to reach out to. We’ve already rolled out the changes to approximately projects and are currently planning on rolling out the new features the remaining Wikimedia projects (including Wiktionary) in late August (probably the 4th week).

As I’m sure you know, some gadgets and customizations may not be compatible with Vector. Would it be possible for you to help us identify the major incompatibilities? We’re using Bugzilla to track issues (please file under "Usability Initiative"). Also, our FAQ page has info on how to test Vector.

I look forward to your feedback! Howief 23:37, 6 August 2010 (UTC)Reply

You may have noticed that Conrad has not been around since July 18 or so. I am not sure who else is in any position to help. You might try RobertUllman, Bequw, and, possibly Hippietrail or Ruakh. They may be able to help you find others. I hope you can delay implementation for some time as we probably have lots of little gadgets that we need to make wikipedia-oriented software do a job it is not intended to do. I don't believe we have ever received and lately sought much technical help from Wikimedia, though I have regularly heard about the bad consequences of unannounced "minor" changes. DCDuring TALK 00:02, 7 August 2010 (UTC)Reply

I use the Vector skin currently and can't think of any problems. There's no real code change, per se, just a change in the default preference. It *should* be relatively smooth. --Bequw → τ 03:00, 7 August 2010 (UTC)Reply

I'd feel better if we had our full complement of adepts around when the cut-over happens. I hope that my sense of dread will be an object of humor afterwards. DCDuring TALK 03:53, 7 August 2010 (UTC)Reply

Thanks for the suggestions. I will leave a message for RobertUllman and others. Hopefully we can get some folks to help identify issues with Vector. Howief 21:57, 11 August 2010 (UTC)Reply

Hi, the main "problem" I can think of is trying to get our edit-tools into the new editor. I haven't looked into it for a while though so maybe that's already solvable. Most of the other gadgets and tools seemed to "just work" last time I checked. Conrad.Irwin 09:44, 15 August 2010 (UTC)Reply

Thanks for responding Conrad.Irwin. When you get the chance, would you mind taking a look at the edit-tools to make sure they're Vector compatible? This would be really helpful. Howief 17:45, 18 August 2010 (UTC)Reply

Hello, I was looking at the entry sp and noticed that p**s had been listed by your bot as an anagram (and vicey versey). I don't think that's right, so I removed the anagrams. Is there a way to change the bot to not do that? ~ lexicógrafo | háblame ~ 01:33, 14 August 2010 (UTC)Reply

Thanks. I'll update the bot. Conrad.Irwin 09:45, 15 August 2010 (UTC)Reply

Hi Conrad.Irwin,

My name is Robert and I'm an administrator at the Romanian Wiktionary. I don't know if you're the right person to ask, but is there a possibility to use the same function in another project like the Romanian Wiktionary? In other words, can I import this function? Since I don't have any programming skills to be proud of, I can't really do it on my one.

Best Regards,

--Robbie SWE 12:51, 20 August 2010 (UTC)Reply

Hi Robbie, The answer is that you can use most of it, but you'll need a javascript coder to fix up the wiki-specific stuff (unless you happen to use exactly the same templates as we do). Conrad.Irwin 21:00, 1 September 2010 (UTC)Reply

a lot!

No worries :) — all it really needs now is some abstractions around lists, if you're feeling like coding ;) Conrad.Irwin 19:54, 7 September 2010 (UTC)Reply

Hi Conrad,

I was wondering — now that we have your cool safesubst: thing, that means we can make language templates do more sophisticated things now, right? IIRC, the main barrier to complicated language templates was that {{subst:nah}} needs to result in bare [[Nahuatl]], and it used to be impossible to make complicated things disappear during subst-ing but still work during regular transclusion.

This raises the possibility that, for example, {{Xyzy}} could be made obsolete by language templates that know what their corresponding default script templates should be.

What do you think? Is that a reasonable path to go down?

(But now that I think about it, I realize that we could always have done something like this, by having parallel templates; for example, we could have had {{sc:he}} redirect to {{Hebr}}. So maybe it was intentional that we kept language-templates so limited?)

—Ruakh_TALK 14:30, 10 September 2010 (UTC)Reply

I have not weighed this up in much detail — it sounds like you have a pretty good understanding of what's going on. It's a trade-off between template load-time and template parse-time, making the language templates clever increases parse time on pages that don't want the cleverness. Xyzy tries quite hard to avoid loading extra templates at the expense of increasing parse-times — so making each language template know which script template would increase load-time for invocations that use scripts (each language and script template is needed), and increase parse-time for pages that don't; all to save some parse-time on pages that do — it sounds like it's not going to be a winner overall. Hope that's useful — I'll try and spend some more wiktionary time soon. Conrad.Irwin 23:19, 23 September 2010 (UTC)Reply

Hi Conrad, at one point I think you (or someone else?) mentioned that it would be possible to amend the custom.js code to anonymously log the number of viewers using each preferences. That would be quite useful information for WT:GP#reorganizing PREFS, or really anytime in the future. Let me know if you think it's doable. Cheers. --Bequw → τ 03:22, 13 September 2010 (UTC)Reply

I've implemented a collector for the raw data, parsing and filtering it should be very easy. I've deliberately not included timestamps or user identifiers for privacy purposes, though if you think dates would be useful, we could do those. Conrad.Irwin 23:08, 23 September 2010 (UTC)Reply

Thanks. Feel free to send me the raw data to analyze. --Bequw → τ 02:47, 24 September 2010 (UTC)Reply

It's at http://toolserver.org/~enwikt/prefs_stats/ Conrad.Irwin 08:45, 24 September 2010 (UTC)Reply

I'm getting a "not found" image at the bottom left corner of every page. Is this connected to the change? --Yair rand (talk) 01:10, 26 September 2010 (UTC)Reply

Hi. Someone has inquired whether this tool is available for Arabic. If not, would your codebase be able to work on Arabic Wiktionary with a bit of massaging. (I'm not asking you to make it work; I'm just wondering if it is worthwhile telling them about the code). John Vandenberg 09:00, 16 September 2010 (UTC)Reply

Yes, it probably would. Please feel free to email me (or to pass on my email address <wiktionary username>@gmail.com), if you want more information than is there. I can hopefully set it up to run automatically like the English ones if that would be good. I don't know any Arabic at all, but it looks like arwikt-defs-latest.tsv contains something useful, so wouldn't be too hard to tweak. Conrad.Irwin 22:17, 23 September 2010 (UTC)Reply

Could you please update this tool to work on the Vector skin? Thanks, —Internoob (Disc•Cont) 20:51, 2 October 2010 (UTC)Reply

Switched it to work in Vector. Though it could be prettier. --Bequw → τ 22:39, 2 October 2010 (UTC)Reply

Thanks. Though by the way, why does its expand/collapse arrow go the opposite way of the other boxes? —Internoob (Disc•Cont) 19:40, 3 October 2010 (UTC)Reply

Could you make your Accelerator work with Esperanto verb forms? All your need fill in the edit box with is a subst: call to new eo form, like {{subst:new eo form|ekzempl|as}}. (It would be ideal if it did that for noun and adjective forms, too.)

If it could read {{eo-part-sublemma|ekzempl|anto}} and fill in edit boxes with {{subst:new eo form|ekzempl|antojn|unc=yes}} as appropriate, that would also save a lot of work. − Robin 21:27, 4 October 2010 (UTC)Reply

This isn't necessary as I have a bot that can handle verb forms, and include pronunciations at that. I don't run it frequently as people just kept adding to the list and it takes a while to set up. So if you have verbs that you'd like the forms of uploaded in mass, add it to User:Opiaterein Inflectobot/nomnom. — [ R·I·C ] opiaterein — 02:08, 5 October 2010 (UTC)Reply

See [8]: Teχ is not an anagram of et, but bot does not identify χ. --Hydrox 12:28, 9 October 2010 (UTC)Reply

I think he's already aware of the problem. Ideally, the bot should just discount all words/terms with non Latin characters, including 0123456789. Also, he's on wikibreak right now due to work. Mglovesfun (talk) 12:32, 9 October 2010 (UTC)Reply

It would be great if you could produce a list of words the bot has added to its database containing non-latin characters for control. --Hydrox 14:35, 9 October 2010 (UTC)Reply

Alright so, I was thinking I might like to have a list of pages in Category:Russian verbs that don't link to {{ru-verb}}. I'm pretty sure I got everything off {{infl}}, so everything will either be linked to ru-verb or raw '''идти''' (idti) [[Category:Russian verbs]] style stuff...

However, I have a feeling that this list might be pretty long, though... So I was thinking as an alternative to that, maybe you could write me an mw python file I can run on opibot to replace that raw inflection-line stuff with {{ru-verb}} code? There's a considerable amount of variation so if this seems like somethin you could help me with when you've got the time, I'll get more info on what would be required. Either way, thanks in advance as always :) — [ R·I·C ] opiaterein — 02:35, 21 October 2010 (UTC)Reply

Just wanted to let you know that an erroneous anagram was added by your bot to prone: perſon. The long s seems not to register as a letter, but perhaps you're already aware of this problem? When you do fix it, I think it would be a good idea to treat it as the same letter as s (as é, ê, è, ë, etc., is to e in English). – Krun 15:05, 21 October 2010 (UTC)Reply

Same thing's happened with fuþorc/Corfu, just so you know. ;) Haipa Doragon 03:59, 2 January 2011 (UTC)

I thought the bot wasn't going to remove human-added anagrams? See diff.​—msh210℠ (talk) 17:10, 25 October 2010 (UTC)Reply

That was indeed the intention. I'm sorry about all the mess this particular bot has caused, and will try and fix it when I next have the time. Conrad.Irwin 13:28, 26 October 2010 (UTC)Reply

Hi, have you had a look at this? The feature seems not to be working after the introduction of the vector skin. --Harald Khan Ճ 14:07, 26 October 2010 (UTC)Reply

Interesting, it seems to work for me — are you getting any javascript errors? Conrad.Irwin 21:18, 26 October 2010 (UTC)Reply

No, the extra tools are simply gone. They return, both in IE and FF, when the vector skin is "turned off". --Harald Khan Ճ 21:40, 26 October 2010 (UTC)Reply

Hi, I want to get a list of definitions from Catalan wiktionary for a free (as in freedom) project. I've found your listdefn project at toolserver, but I'm unable to achieve your results, create.sh calls a missing latest_dump.sh file. I'm not a power user, but I know something about regex. Can you help me, please? I want to know how to parse Catalan wiktionary, because I want to be able to parse monthly, or yearly, if needed. Thanks in advance. --Unjoanqualsevol 13:39, 18 November 2010 (UTC)Reply

Hi, latest_dump.sh should just print out the file name of a dump file. You can make it as simple as:

#!/bin/bash
echo /tmp/cawiktionary-20100912-pages-articles.xml

Let me know if that helps. (Obviously you'll have problems at the moment because http://download.wikimedia.org/ is broken). If you do get something working, I'm happy to try and host it alongside the en.wiktionary ones. Conrad.Irwin 23:21, 18 November 2010 (UTC)Reply

Thanks for your hints. Yeah, I don't have cawikitionary dump now, :(. If any result is achieved, I will post here. Regards, --Unjoanqualsevol 10:27, 19 November 2010 (UTC)Reply

The RAE recently stopped considering ch and ll as separate letters (this goes beyond removing them for collation puposes). Could you do the same for Index:Spanish? Thanks. --Bequw → τ 16:39, 28 November 2010 (UTC)Reply

Done. Conrad.Irwin 19:18, 13 February 2011 (UTC)Reply

I tried to report this bug using Bugzilla, but the MetaKeywords extension isn't listed. The signature of the wfMetaKeywordClearCache function is incorrect and using the extension with PHP 5.3.5 throws the following error:

Detected bug in an extension! Hook wfMetaKeywordClearCache failed to return a value; should return true to continue hook processing or false to abort.

Backtrace:

1. 0 /czdata/ec/phase3/includes/Article.php(2138): wfRunHooks('ArticleSaveComp...', Array)
2. 1 /czdata/ec/phase3/includes/Article.php(1818): Article->doEdit('right ?[...', 'beginning to re...', 98)
3. 2 /czdata/ec/phase3/includes/EditPage.php(1058): Article->updateArticle('right ?[...', 'beginning to re...', false, false, false, , false)
4. 3 /czdata/ec/phase3/includes/EditPage.php(2588): EditPage->internalAttemptSave(false, false)
5. 4 /czdata/ec/phase3/includes/EditPage.php(391): EditPage->attemptSave()
6. 5 /czdata/ec/phase3/includes/EditPage.php(273): EditPage->edit()
7. 6 /czdata/ec/phase3/includes/Wiki.php(553): EditPage->submit()
8. 7 /czdata/ec/phase3/includes/Wiki.php(70): MediaWiki->performAction(Object(OutputPage), Object(Article), Object(Title), Object(User), Object(WebRequest))
9. 8 /czdata/ec/phase3/index.php(117): MediaWiki->performRequestForTitle(Object(Title), Object(Article), Object(OutputPage), Object(User), Object(WebRequest))
10. 9 {main}

The fix is very simple. Here is the patch:

--- MetaKeywords.php	(revision 80574)
+++ MetaKeywords.php	(working copy)
@@ -117,7 +117,7 @@
 }
 
 //Updates the cache if [[MediaWiki:Metakeywords]] or [[MediaWiki:Metadescription]] has been edited
-function wfMetaKeywordClearCache( &$article, &$wgUser, &$text ) {
+function wfMetaKeywordClearCache( &$article, &$wgUser, $text ) {
 	global $wgMemc;
 		$title = $article->mTitle;

I don't work on Wiktionary and don't monitor its talk pages. So, if you want to contact me, try Citizendium, user Dan Nessett Dnessett 23:41, 20 January 2011 (UTC)Reply

I'm looking for a way to parse out translations from an xml dump. Is there an existing tool for that? ~ heyzeuss 10:57, 5 February 2011 (UTC)Reply

You can try out this script, that I have written to generate bilingual off-line dictionaries from the translations sections of the dump. Matthias Buchmeier 09:18, 7 February 2011 (UTC)Reply

There is the code I use to output translations to add into the Indexes (it used to be in awk, but I rewrote it in Python :p). Unfortunately it's in a terrible-terrible mess (as it is completely tied to index creation, so does other similar tasks like French synonyms), but you can get it from http://jelzo.com/stuff/get_trans.py (the xmlreader it needs is from Pywikipedia). I want to tidy this up and make it generally usable, one day. Conrad.Irwin 09:08, 8 February 2011 (UTC)Reply

The indexes haven't been updated in quite a while. Could you run Conrad.Bot to update them again? --Yair rand (talk) 08:12, 7 February 2011 (UTC)Reply

Yes — I tried doing this the day before you commented. Unfortunately it seems that my copy of Pywikipedia is out-of-date, and in order to update it I need to update my SVN (in order to update that I need...) etc. I'll try and get it done (including Bequw's Spanish request) this week. Conrad.Irwin 09:04, 8 February 2011 (UTC)Reply

Done. Conrad.Irwin 19:18, 13 February 2011 (UTC)Reply

I would like to know your preference as regards the use of "<" vs "from" in the formatting of etymologies in Wiktionary, whatever that preference is. Even explicit statement of indifference would be nice. You can state your preference in the currently running poll: WT:BP#Poll: Etymology and the use of less-than symbol. I am sending you this notification, as you took part on some of the recent votes, so chances are you could be interested in the poll. The poll benefits from having as many participants as possible, to be as representative as possible. Feel free to ignore this notification. --Dan Polansky 12:15, 14 February 2011 (UTC)Reply

Would it be possible to populate Index:Swedish? The alphabetic order is exactly as on the table of contents there, so it should not be too hard. -- Prince Kassad 15:31, 14 February 2011 (UTC)Reply

I'll give it a go; see also User talk:Yair rand Conrad.Irwin 20:02, 14 February 2011 (UTC)Reply

In User:Mglovesfun/vector.js I made this change (ignore the sv bit). Remind me again how I would apply it to other templates like {{past of}}, presumably $1 and $2 would become $2 and $3, for example. PS you can reply here or see me on IRC in the enxt few days if you prefer. Mglovesfun (talk) 15:16, 21 February 2011 (UTC)Reply

Note that KassadBot does the reverse.​—msh210℠ (talk) 18:17, 21 February 2011 (UTC)Reply

Sort of; it adds a link only when none is present to avoid using {{count page}}. This has been brought up on User talk:Prince Kassad (I think, anyway). Mglovesfun (talk) 18:47, 21 February 2011 (UTC)Reply

You can do this without changing the dollars by changing it to /\{\{(?:plural of|past of)\|.../. (?:A|B) says "match either A or B, but don't capture it into a dollar variable". You might find [9] and [10] helpful. Conrad.Irwin 20:14, 21 February 2011 (UTC)Reply

I get it now; note for msh210 that it will only remove non-piped links, not [[fan|fān]]. Mglovesfun (talk) 00:37, 22 February 2011 (UTC)Reply

This has been blank for a long time - will it ever be used again? Mglovesfun (talk) 10:49, 1 March 2011 (UTC) (delayed signature, mea culpa)Reply

I doubt it. I really need to fix the anagrams bot — it's arguably the worst mess I've ever been the cause of... Conrad.Irwin 06:39, 3 March 2011 (UTC)Reply

I've marked this as inactive as it's hilariously empty. If you have time to generate it (or anyone else with the code) please do, thanks. Mglovesfun (talk) 18:41, 25 February 2011 (UTC)Reply

How different is Yiddish alphabet from Hebrew? Conrad.Irwin 06:40, 3 March 2011 (UTC)Reply

Ohhh shoot, thanks for telling me! I didn't know it was actually saving it as a different page revision each time! :| GreenGenie 06:59, 27 February 2011 (UTC)Reply

Doesn't matter :). Keep up the good work. Conrad.Irwin 07:02, 27 February 2011 (UTC)Reply

Thank you for the cool update of Index:Czech :). --Dan Polansky 12:19, 28 February 2011 (UTC)Reply

No worries — if you have other languages for me to consider, I'll happily consider them. Conrad.Irwin 06:41, 3 March 2011 (UTC)Reply

There is the vote Wiktionary:Votes/pl-2011-02/Deprecating less-than symbol in etymologies, which would benefit from your participation, even if only in the role of an abstainer. Right now, the results of the vote do not quite mirror the results of the poll that has preceded the vote. There is a chance that the vote will not pass. The vote, which I thought would be a mere formality, has turned out to be a real issue. You have taken part on the poll that preceded the vote, which is why I have sent you this notification. --Dan Polansky 08:24, 10 March 2011 (UTC)Reply

If possible and not too much trouble, could you populate Index:Danish in the same way as several other languages?--Leo Laursen – (talk · contribs) 11:40, 16 March 2011 (UTC)Reply

I've cleared out Wiktionary:Finnish inflection types/nouns as best as I could, finally. Can you check to see if any of them are still being used? ~ heyzeuss 23:22, 21 March 2011 (UTC)Reply

Can you give an advice about applying your script to ru-wiktionary? We have language sections under "h1" not "h2". What should I correct in the code? --Infovarius 16:08, 22 March 2011 (UTC)Reply

There are a number of Hebrew letter-vowel combinations whose transliterations are ambiguous, so a map for Hebrew would transliterate only a partial repertoire. I had thought that that's no problem: the rest will yield an error. But actually (if I'm not mistaken) the extension will pass a character through untransliterated if there's no rule for the character. That means that either (a) some words will wind up mistransliterated (i.e., including some transliteration and some Hebrew) or (b) one would need to include a dummy target like the null string to transliterate ambiguous things to. Neither looks usable AFAICT. Would it be possible (and I know this is very late in the game) to modify the extension so that it can, as an option (either a parameter in the parser-function tag or, probably better, a flag at the top of the map), yield no transliteration or some specified error when a substring to be transliterated is not in the map?​—msh210℠ (talk) 17:26, 24 March 2011 (UTC)Reply

This is not easy to change, unfortunately. How long is the blacklist of characters that can't be transliterated? If it's of specifiable length, then it may be possible to add an extra feature that first checks for any blacklisted characters and generates an error. Would that work? Conrad.Irwin 20:07, 31 March 2011 (UTC)Reply

For Hebrew, yes. (Well, as all alphabets are finite and all words have finite length, it will work for any language, actually.) Thanks for considering it.​—msh210℠ (talk) 20:15, 31 March 2011 (UTC)Reply

Hi,

Can I request indexes for Chinese (Mandarin), Japanese and Arabic, please! Also, Korean, Thai, Persian, Vietnamese and Hindi would be nice to have. The current Arabic index is a joke. The Japanese is very small, perhaps should be moved to give space the new one. Perhaps it's fine just to index Chinese and Japanese by the first character, otherwise it may be hard.

Can you do the Korean by the jamo symbol? Hint: the Hangeul formula is: [{(initial)×588}+{(medial)×28}+(final)]+44032. E.g. 한 is made of three jamo: ㅎ + ㅏ + ㄴ and it's Unicode value is {(18 × 588) + (0 × 28) + 4} + 44032 = 54620. --Anatoli 04:50, 26 April 2011 (UTC)Reply

Hi there. If you ever have time, could you regenerate User:Conrad.Irwin/oldanonpages with pages older than about 18 months? Cheers. SemperBlotto 21:17, 30 April 2011 (UTC)Reply

Done. Conrad.Irwin 08:33, 3 May 2011 (UTC)Reply

Hi there. What kind of words are marked with "<partlynew>"?

[11] Why line with 'Lichtenstein' entry was changed from # [[Liechtenstein#Polish|Liechtenstein]] ''proper'' {{audio-list|Pl-Lichtenstein.ogg}} [[Liechtenstein|*]] to # <span class="partlynew">[[Liechtenstein]]</span> [[Liechtenstein|*]]? Maro 20:41, 7 May 2011 (UTC)Reply

Partlynew means that the index script thinks that the page doesn't exist in that language, or the page is not a lemma. Because the definition reads '{{l|en|Lichtenstein}}', the indexer treats that as a non-lemma entry (assuming that all definitions made up of one template are form-ofs). I will try to change that for {{l}} before I update the index next time. Conrad.Irwin 23:16, 7 May 2011 (UTC)Reply

So this one is something I've been putting off for a long time and will probably continue to try to put off, so you can take your time with it, but I do need a list of any entry that uses more than one template in the category Category:Turkish declension-table templates. (I'm going to be consolidating them so that one template can do the job of up to 3 so they don't all need to be called individually.) Thanks in advance. — [ R·I·C ] Laurent — 19:16, 8 May 2011 (UTC)Reply

Hi there. Would you consider automagically populating Index:Venetian in the same way that you do for Italian? At the moment (certainly from the latest dump) there are very few Venetian words - but I have started on a project to add them, and have added a few hundred already. Cheers. SemperBlotto 16:48, 4 June 2011 (UTC) p.s. The index has been set up, but is currently totally empty.Reply

It's currently only on one page (at Index:Venetian), once it crosses a thousand words or so, it will be split into separate pages (I can force the latter behaviour if you really want). Conrad.Irwin 16:24, 6 June 2011 (UTC)Reply

Well done. p.s. Where do the red links come from? SemperBlotto 16:26, 6 June 2011 (UTC)Reply

Translation tables (you can click the asterisks to find them). Conrad.Irwin 17:43, 6 June 2011 (UTC)Reply

I've rolled back the latest bad update to Wiktionary:Statistics/generated. SemperBlotto 16:01, 18 June 2011 (UTC)Reply

• I rolled back too far. Prince Kassad has done it properly. SemperBlotto 16:35, 18 June 2011 (UTC)Reply

Thanks — I'll see if I can make the uploader a bit less trusting. Conrad.Irwin 19:37, 18 June 2011 (UTC)Reply

What is our best estimate of the number of lemma entries in a language? To me it looks like English might have only 220,000, which is a bit disappointing.

The number of gloss definitions in English doesn't seem to follow an 80-20 pattern, suggesting that we don't reflect the full polysemy of many polysemic words or over-represent terms with single senses. DCDuring TALK 16:54, 18 June 2011 (UTC)Reply

Using the same criterion as the gloss definitions vs. form-of definitions, there are 225352 pages with an English lemma definition. It's entirely possible that the numbers are very bogus — I haven't updated the heuristics in a long time — https://github.com/ConradIrwin/wiktionary/blob/master/wiktionary/__init__.py#L348 .

Some English entries may have both gloss and form-of definitions, eg, many -ing and -ed forms of verbs are also nouns and/or adjectives. Also, some plurals, and terms defined using only(?) {{non-gloss definition}} that are lemmas, though these would be a small number.

I'm not much help on reading Python, though I'll try, let alone correcting or improving it.

The late Mr. Ullman had code that counted headings. Is there code that counts L3 and lower headings by language? I am only interested in English and have worked assiduously to eliminate spurious ones, especially in English.

What prospects do we have for getting any more of Ullman's code working? You don't seem to have had much time for en.wikt lately, but any help or advice is appreciated. DCDuring TALK 20:26, 18 June 2011 (UTC)Reply

I have code that finds all the sections with definitions in them and checks against WT:POS for the missing ones, which I've been fixing myself — this is a vague project to make http://toolserver.org/~enwikt/definitions/ as good as possible. I don't yet have counting for all the other headings but I could try and get something working. What exactly do you want? Conrad.Irwin 01:21, 27 June 2011 (UTC)Reply

Hi, cirwin. I notice that the bot doesn't remove deleted entries from Index:Old Armenian. E.g. this and this are still in the index. --Vahag 18:19, 19 June 2011 (UTC)Reply

Hey Vahag, the Old Armenian index was deemed to be "too big" to fit on a page by the indexer, so it now updates Index:Old Armenian/ա, etc. instead. Conrad.Irwin 18:53, 19 June 2011 (UTC)Reply

I see. Thanks, C-bear. --Vahag 19:21, 19 June 2011 (UTC)Reply

Hi, Conrad. The bot isn't updating Index:Armenian/ու. --Vahag 06:09, 18 July 2011 (UTC)Reply

Hi Conrad,

I believe you mentioned once that the language indices don't include entries whose only definitions are entirely template-generated. Would something like this:

# {{form of|blah blah|foo}}.

(where the period is outside the template) end up in a language-index? If so, would you be willing to change that?

I ask because people are (once again) discussing standardizing the presentation of form-of templates, and one idea that's been floated is to simplify the dot/no-dot thing by removing the dots entirely, and having them just be tacked on at the end.

Thanks in advance!
—Ruakh_TALK
12:19, 26 June 2011 (UTC)Reply

Done :). Conrad.Irwin 01:11, 27 June 2011 (UTC)Reply

Wow, fast! Thanks! —Ruakh_TALK 01:38, 27 June 2011 (UTC)Reply

Hi Conrad, I was told at the helpdesk that you may be able to help me get a plain-text dump of English words in Wiktionary. Actually, I have since figured out that I can save the HTML from the pages at Index:English, and then strip out just the words. That would be perfect, but now I've noticed that the Index:English pages do not seem to include noun plurals, verb inflections and so on, which I would like to capture if at all possible. Do you know if there's some other list available that does include those things? Thanks for any help you can give. 86.181.202.192 01:35, 30 June 2011 (UTC)Reply

Sure, there are TSV files at http://toolserver.org/~enwikt/definitions/ — is that any good for you? Just to get the words out should be a simple case of extracting the 2nd column and removing duplicates. Conrad.Irwin 05:53, 1 July 2011 (UTC)Reply

Hi, Thanks very much for your reply. That looks great, but I unfortunately I have no way of reading those "gz" files. I don't suppose there's a plain-text (or .zip?) version available anywhere? 86.148.152.95 11:23, 5 July 2011 (UTC)Reply

Do you have a script that can archive all the closed discussions on WT:RFD? The page is now very long (again), but archiving by hand is a royal pain. --EncycloPetey 16:48, 6 July 2011 (UTC)Reply

I'd like to invite you to come to the New Orleans Hackathon 2011. We're getting together folks like you -- template, script, tool, extension, and gadget writers -- to participate, give feedback, test, and hack with us.

At the event, MediaWiki developers and Wikimedia operations engineers will be working on Wikimedia's gadgets/extensions/tools support, authorization/authentication strategy, dev-ops virtualization, and general training and hacking. And we'll improve and discuss the Wikimedia Labs projects infrastructure and other stuff that makes it easier for anyone to supercharge Wikimedia with awesomeness.

The event is open to anyone who wants to come and contribute, and is an opportunity to spend time with senior MediaWiki developers & ops engineers, write beautiful code, and learn about the latest developments. We'll write code together, discuss the software, and hold little workshops.

If you can make it to New Orleans, Louisiana, USA, 14-16 October 2011, we'd love to have you. Please add your name to the attendees list.

(You additionally might be interested in the Brighton hackathon this fall.)

Thanks! mw:User:Sumanah (Volunteer Development Coordinator, Wikimedia Foundation)

Hi Conrad,

The toolserver definitions dump is a great resource, but one thing I've noticed is that the order of the definitions is different to how they appear in Wiktionary. For example, the first two Wiktionary definitions for the noun cat are A domesticated subspecies… and Any similar animal of the family Felidae… whereas in the dump they are A catamaran and A catfish. It looks like the dump definitions may have been sorted alphabetically.

Is it possible to get a dump where the definitions are unsorted?

Kind regards, Mike 86.129.80.33 16:54, 25 August 2011 (UTC)Reply

I've tried changing the sorting criteria, so they're now sorted by only language name word; though this had the side-effect of making it also sort case-sensitively. Is that better? Conrad.Irwin 20:20, 25 August 2011 (UTC)Reply

This is great! Now definitions are in the order that they appear in Wiktionary and hence the most obvious definition appears first.

I've only been looking at the English dump but I can't see any issues with the sorting. I think I'm correct in saying that words have always been sorted case-sensitively (uppercase first) and - to me - this isn't an issue.

Many thanks for your time and effort doing this. A brilliant resource just got even better! Mike 2.31.8.190 21:44, 25 August 2011 (UTC)Reply

Hi there. To me, your quote looks like the use of (deprecated template usage) after + (deprecated template usage) spring line. SemperBlotto 07:30, 28 August 2011 (UTC)Reply

Reading the various quotes again, I think that is always the case. I've deleted my version for now. Conrad.Irwin 17:43, 19 September 2011 (UTC)Reply

Mzajac pointed out in a post on the talk page of the definition editor that editor.js's redo button is active even when there's nothing to redo, which could be confusing. It would probably be helpful if the redo button was disabled=disabled when the redostack is empty, and the same for undo and save changed when the editstack is empty. (Also: +poke for looking at the transliterator issues :) ) --Yair rand 18:00, 31 August 2011 (UTC)Reply

I originally did this, but the re-enabling was buggy so it would sometimes be *disabled* even when you needed it. Feel free to implement it right :). Conrad.Irwin 17:43, 19 September 2011 (UTC)Reply

Hi Conrad,

Can I kindly request Mandarin and Japanese DB dump, similar to Index:Russian, please? Not sure what challenges it may cause, though, and I have no idea how hard it it. I mentioned the idea for the Japanese index in BP. --Anatoli 05:03, 19 September 2011 (UTC)Reply

Hi Anatoli, Yes — in theory. I have quite a backlog of things to do though. There are some more requests for indexes further up this page as well. I do plan to do them, but I'm not sure when. The more specific you can be about what you want, the easier it will be when I do have time (or for anyone else to pick this up).

Thanks, Conrad. I guess, I would agree to the simplest possible extract, where entries are indexed by their first character, be it kana or kanji symbol, words in Romaji (Roman letters) could be excluded, they are duplications of other words, anyway. In case of Mandarin, only hanzi (Chinese chracters) words, please. When you have time to do, please let me know if there are some possible variants or if you have questions. --Anatoli 22:28, 19 September 2011 (UTC)Reply

Hi there. I see that the automagically maintained indexes haven't been updated for some time (e.g. Italian not since June). Is there any chance of running Italian, Venetian and French? SemperBlotto 08:29, 29 September 2011 (UTC)Reply

p.s. Someone has been trying to create a Telugu one manually - could you do that as well??

download at http://toolserver.org/~enwikt/definitions/ is dropping all downloads at approx. 3.8M point. don't know if bandwidth-related. sorry if message constitutes inappropriate use user talk... Guest

I can't replicate this, does it still happen? Conrad.Irwin 23:25, 29 October 2011 (UTC)Reply

Something's wrong here. Look at the history and what your bot did for the latest dump. Somehow over two thirds of Wiktionary magically disappeared. -- Liliana • 11:26, 26 October 2011 (UTC)Reply

Now fixed. Sorry about that, the definitions extractor crashed half-way through. It worked when I just ran it again... Conrad.Irwin 23:25, 29 October 2011 (UTC)Reply

The same issue, or a similar one, seems to have affected some of the language indices. —Ruakh_TALK 04:20, 30 October 2011 (UTC)Reply

Hi there, could you please share your view here? I would like to understand, from a technical perspective, what can be done to fully automate anagram list generation.

Also, I believe there is a problem with {{alphagram}}, see discussion. Thanks. 124.147.76.165 05:32, 3 November 2011 (UTC)Reply

Just a reminder of the upcoming Brighton developers' event. Sumanah 13:47, 6 November 2011 (UTC)Reply

hey could you move it to lowercase for me, I cannot, thanks.Westernstag 03:25, 7 November 2011 (UTC)Reply

What would be needed to allow other languages to be indexed as well? Could indexing be added for Old Norse, Old and Middle High German, Old Saxon, Middle Low German, Old and Middle Dutch, Old Frisian and Gothic (in Gothic script with romanization)? I don't know if this is hard to do, and whether the work is worth it for these languages that don't have many entries yet. —CodeCat 17:23, 13 November 2011 (UTC)Reply

In theory its as simple as adding alphabets to the bot script — if you could make the copy-pasteable, I'm happy to give that a go.

It's also useful to know which letters are only considered different at the second-degree (i.e. a < ä, but äb < ac). (And whether or not the language counts ä as a different letter, or just a modified a). Conrad.Irwin 05:36, 16 November 2011 (UTC)Reply

All of those languages except for Gothic and Old Norse used the medieval Latin alphabet, consisting of Aa Bb Cc Dd Ee Ff Gg Hh Ii/Jj Kk Ll Mm Nn Oo Pp Qq Rr Ss Tt Uu/Vv/Ww Xx Yy Zz. I don't think I and J were considered separate letters, or U, V and W, but for modern purposes it's probably best to treat them separately. In Middle Dutch, J was written instead of I when it appeared after another I, so II was written IJ. The other languages probably did similarly as well, although I don't know for sure. Vowel length is written with macrons on Wiktionary but it's not part of the entry names.

For Old Norse, the alphabet is similar to that of modern Icelandic: Aa Áá Bb Dd Ðð Ee Éé Ff Gg Hh Ii Íí Jj Kk Ll Mm Nn Oo Óó Pp Rr Ss Tt Uu Úú Vv Xx Yy Ýý Þþ Ææ Œœ Øø Ǫǫ, but this is the standard 'normalised' alphabet, in the original writing J was not distinguished from I, nor was V from U, KV was often written QU as in Latin, and I don't think vowel length was marked either. I'm not sure what the proper ordering of the last four letters should be either. The letter Ęę also appears sometimes but I don't think it's part of the regular normalisation, it's often replaced with Ee.

Gothic used its own alphabet, which is detailed at Wiktionary:Gothic transliteration. A vote has decided to allow romanized entries in Gothic, so it would probably be good to have two indexes for Gothic: one in the Gothic alphabet (with its ordering) and one with the Latin alphabet (Aa Bb Dd Ee Ff Gg Hh Ii Jj Kk Ll Mm Nn Oo Pp Qq Rr Ss Tt Uu Ww Xx Zz Þþ Ƕƕ). Like in the other languages, macrons are used to indicate long vowels but these are not part of the entry names.

I hope that helps? —CodeCat 11:49, 16 November 2011 (UTC)Reply

Hi, is working on the issues with the transliterator still on your todo list? I'd just like to know if it's still something that's going towards getting done at some point, or what... Thanks. --Yair rand 04:35, 28 November 2011 (UTC)Reply

Vaguely, though I'm struggling for time. I hoped to do this last time I had a break from work, but other things overtook me (as I no longer have a PHP development environment set up, there's a bit of grunge work to do before I can even start) — I'll try to have another go over Christmas unless someone beats me to it. Conrad.Irwin 02:08, 30 November 2011 (UTC)Reply

Hi Conrad,

Starting with last month's run, the Hebrew index no longer includes words from ====Translations==== sections, only words from ==Hebrew== sections. The same seems to be the case with other language indices. Was this change intentional?

(I'm not particularly objecting, if it was; I just want to make sure.)

Thanks in advance!
—Ruakh_TALK
03:44, 29 November 2011 (UTC)Reply

That was not intentional, I'll see if I can find out what's wrong. Conrad.Irwin 02:08, 30 November 2011 (UTC)Reply

Thanks! —Ruakh_TALK 05:36, 30 November 2011 (UTC)Reply

Hi. Could you also make the bot update Index:Armenian/ու? Words starting with that letter are instead being added to Index:Armenian/ո. --Vahag 12:37, 30 November 2011 (UTC)Reply

Hi Conrad. Can User:Conrad.Irwin/English nouns without categories be updated again? There's probably not many left to do now. --Simplus2 11:51, 28 December 2011 (UTC)Reply

After finding the subsections in Wiktionary Common.js, I had to use it too, I wanted to understand it, so I rewrote it a bit. As you can find, I prefer using jQuery where I can, made it ready for some MediaWiki where globals disappears and JSHint'ed it. Thanks for this script :) --TiCPU 05:41, 4 January 2012 (UTC)Reply

$j(document).ready(function () {
	// ResourceLoader vars
	var wgNS = mw.config.get('wgNamespaceNumber');
	var wgTitle = mw.config.get('wgTitle');
	var wgAction = mw.config.get('wgAction');
	var wgScript = mw.config.get('wgScript');
	var wgPageName = mw.config.get('wgPageName');

	if ((wgNS === 2 || wgNS === 8) && wgTitle.lastIndexOf('.js') !== -1 && wgAction === 'view') {
		var pres = $j('#bodyContent pre');
		if (pres.length === 1) {
			var spans = $j('span.coMULTI', pres);
			var section = 0;

			spans.each(function (i) {
				var res = null;
				var text = $(this).text();
				if (text.length > 0) {
					// TODO \n+ doesn't seem to work here
					res = text.match(/\/\*\s*<\/pre>[\r\n\s]+(==+)(.*?)(==+)[\r\n\s]+<pre>\s*\*\//,"gi");
				}
				if (res !== null && res.length !== 0) {
					section++;
					var l = res[1].length < res[3].length ? res[1].length : res[3].length;
					var h = document.createElement('h' + l);

					h.innerHTML = 
						'<span class="editsection">[<a href="' +
						wgScript + '?title=' + wgPageName + '&amp;action=edit&amp;section=' + section +
						'">edit</a>]</span><span class="mw-headline">' + res[2] + '</span>';
					pres[0].replaceChild(h, this);
				}
			});
		}
	}
});