2,054,106 Pages


Mobile notes from Turkey

Nice weather, perfect service, beautiful Turkish people - it's almost heaven. The only one disadvantage, and it was a big surprise for me - LW is hardly accessible for mobile users. On the other hand, maybe it's very good for such lw-addicted persons like me...Senvaikis (talk) 15:58, October 1, 2009 (UTC)

Haa haa don't want to be blocked now do you ;) Enjoy!  Яєdxx Actions Words 21:44, October 1, 2009 (UTC)
Strange, but today I'm reading Red's note in much mobile/human adapted form. Was that result of my note or just I was too drunk yesterday? 16:29, October 2, 2009 (UTC)
If you can't even remember whether you were too drunk the answer is very likely yes ;-) As for accessibility… I suspect you wouldn't be able to stay away from LW even if the only access was via smoke signals. — 6×9 (Talk) 18:00, October 2, 2009 (UTC)
Rare case then 6's conclusion was wrong - if he looked more careful into above msgs, he'd noticed that last one is signed as anonymous, and that's the main reason, why lw pages look more readable - they aren't affected by senv's skin ;) 18:35, October 2, 2009 (UTC)(senv)

Spam block still a problem for LWT?

I'm finally getting back into the swing of things & clearing out my talk page. I noticed this convo mentions that your URL was spamblocked but at the time you didn't think it mattered since LWT looked like it might never be able to run again (but now it IS fortunately :D)... is the URL still being blocked or is everything okay?
-Sean Colombo 23:42, October 15, 2009 (UTC)

Glad to see you again in my room after such a long pause :). Regarding spamblocked address: maybe my broken English was the reason of some misunderstanding, because I never wanted to say that LWT operatibility is somehow dependent on this address accesibility in LW. It was used to publish online LWT One-Touch App, if someone would like to use it. DaKing was the only one it's user (for some time), so when even he stopped using it, I decided to terminate vain efforts supporting published LWT app. That's why I told that blocking this address shouldn't be a big problem for me or anyone else. Btw, I saw KyleH's comment's that now we have a possibility to overwrite some LW external blacklist items by editing local whitelist... Anyway thanks for solicitude , cheers --Senvaikis (talk) 12:06, October 16, 2009 (UTC)
Tried to use KyleH's suggestion. Works! Now blacklisting of Ankhara official site is overwritten. Thanks, KyleH, --Senvaikis (talk) 10:16, October 18, 2009 (UTC)

Who tells me when it'll be finished?

(moved to AP on 6's request).--Senvaikis (talk) 13:11, October 23, 2009 (UTC)

Spam whitelist

That's no problem. Feel free to whitelist any sites you need here. Angela@fandom (talk) 22:41, October 23, 2009 (UTC)

Soular:Tomorrow Never Comes

Hello :) I was looking for the lyrics Tomorrow never comes by Soular..and I've found only these...but the lyrics of this song are not full...please...can you send me full lyrics of this song on my e-mail I really need them but I can't find them anywhere :( Please, help me, I'll be very grateful :) --IP: 16:21, 4 November 2009

I'd like to help you, but unfortunatelly I haven't any lyrics sources but those available for you as well :). After a short googling I may only state that you was right - all this song lyrics pages, available on inet, are just a clones of the same incomplete lyrics. The only one proper solution in such case - just write missing lyrics yourself, while listening to the song. I'd gladly make this job for you, if I were native Englishman ;). So, I'd better recommend you to ask another our admin, User:Redxx, for this favour - she simple adores such challenges :). Btw, - I liked this song too; good luck, --Senvaikis (talk) 17:45, November 4, 2009 (UTC)
Green check 6 isn't the only one with a wand :-)  Яєdxx Actions Words 05:28, November 6, 2009 (UTC)
Well Done!, thx --Senvaikis (talk) 08:57, November 6, 2009 (UTC)


New iTunes format: regex'es with samples

  1. Artist: http://itunes\.apple\.com/((?<story>[a-z]{2})/)?artist/((?<artist>.*?)/)?id(?<artistID>\d+)
  2. Album: http://itunes\.apple\.com/((?<story>[a-z]{2})/)?album/((?<album>.*?)/)?id(?<albumID>\d+)
  3. Song: http://itunes\.apple\.com/((?<story>[a-z]{2})/)?album/((?<album>.*?)/)?id(?<albumID>\d+)?i=(?<songID>\d+)
Notes about optional groups
  • If iTunes story country is ommited, "us" is supposed
  • Artist or album title groups haven't any influence to the link - they may be substituted by any text or simply ommited (same as @akuma, not like amz)

So, new-style *f.iTunes might look like:

  • |iTunes = 139775 (not sure, but seems that artistIDs are the same for all countries; if so - no need for ?s=)
  • |iTunes = 78681 (for USA story; &s=143441)
  • |iTunes = 250453528?s=gb (for UK story; &s=143444)
  • |iTunes = 78681?i=78671 (for USA story)
  • |iTunes = 250453528?i=250453535?s=gb (for UK story)

--Senvaikis (talk) 14:07, November 14, 2009 (UTC)posted in a weak hope to liven up new iT stuff solution

I think the letters "id" should be left in in templates for the ease of copy/pasting (double click highlights a whole word) and to make it easier on newcomers because a number by it self is more ambiguous. LYRIC-Humbug wordsdeeds 06:43, January 7, 2010 (UTC)

A job for LWT?

Hello Senv,

have a look at the page of Soda Stereo - using RegExp and Word and Excel all together I managed to change every link to each word's first letter uppercase. Is LWT now able to go through all pages and change the "album" part of their {{Song}} templates according to the page? Until now the pages either link to albums that are wrongly spelt, to completely wrong albums that even never existed or to no album at all. Could this be a LWT job? - Chris 21:22, December 6, 2009 (UTC)

Sure, it'd be an easy job for Lwt, but I don't think it's a job you'd want to start from. Wrong albums in a song header is a bare malady with all this can of worms named "Soda Stereo". Take a look at this list to understand what I mean: most of SS songs have from 2 to 5 ghosts. I don't think it would be a good idea - filling/fixing all metadata for all those ghosts before deleting/redirecting them. So, you may want to clean this garbaged list at first, before running this job :). cheers, --Senvaikis (talk) 23:11, December 6, 2009 (UTC)
Added pseudo-OS list to SS page for you - just to make your job easier :). Pseudo- because not all these songs are really orphaned. Some of them have backlinks to album pages, but album pages aren't synched to artist page content. So, that's a perfect sample, how making all artist page from scratch may be easier than existing garbage can editing... Good luck :), --Senvaikis (talk) 08:56, December 7, 2009 (UTC)
...Changed my mind and reduced mentioned OS to list of real orphans - don't wanna confuse other editors. You may compare them going through revisions though.
Btw, this 'pearl' researching disclosed once more some old LW problems we were talking about so many times:
  • Doubling info on artist & albums pages. Doubling efforts for info entering & editing isn't a main problem here. Absence of any mechanizm for synchronization of info in both containers is far more serious problem. I'm afraid you'd be blindsided if tried to make this synchronization site-wide check. And what's more interesting - that leads not only to double or ambiguous info. That may lead to virtually lost info also. Example: some song isn't included in the artist page and in any album, listed on that page. So, it must be orphaned. But isn't, because it's backlinked to some (the same artist) album, not listed in artist page, or some differently named album 'ghost' (or to Gracenote page - see below). So, artist page may look perfect, all albums - too; orphaned list may be empty, but so named Oblivion - not.
  • WLH. It's by all means one of most important tools for managing our site. Sad, but it still fails to give us 100% reliable info. If you wanna make some research like this, take my tip - touch all pages you wanna work with before starting.
While these two problems aren't solved, managing such SS-like mess isn't easy task even for Lwt, let alone ordinary editor.--Senvaikis (talk) 11:26, December 7, 2009 (UTC)
...and forgot to mention one more 'goody', making our life even 'more interesting' - Gracenote. Now it has totally swamped Orphaned pages list, making it absolutelly useless; to crown it all, it serves also as very good 'mask' from orphans searching, providing them backlinks and such way condemning to real Oblivion. --Senvaikis (talk) 13:41, December 7, 2009 (UTC)
Hmm, okay, let's see *reads it again and again*
I think you try to tell me, that there are several songs that exist as content pages, although they should be redirects, f. e. because they are wrongly named such as Soda Stereo:Angel electrico which should be a redirect to Soda Stereo:Angel Electrico. If so, I'm sorry I started the topic too early. When I hit the page first (that was when I was categorizing the pages in Category:Split) I noticed the "Request for Edit" at the top. The pages were all capitalised like Soda Stereo:Angel electrico which was the reason for the request. So I bookmarked the page to take care of it afterwards (like I did with Pete Francis which was seemingly missing only a few edits - which wasn't the truth, but that's another story...). Well, as I said I bookmarked it to move the pages later.
But: when I was ready with Category:Split and was about to start moving, I found out the first 4 pages or so that I opened all linked to correctly set redirects. Several other randomly chosen links did as well, so what I thought was that someone had moved all pages earlier and then just forgot about the Request-template at the artistpage. So I just edited the links to skip all those redirects.
Seeing that only 3 of them were redlinks afterwards I once again felt confirmed in that all the wrongly spelt pages were redirected already. So when I made that post I thought all wrong pages had been taken care of so far. Now that I see they're not I'll do that of course before you start.
As I'm here once, I'd like to ask you a question: when I was splitting Category:Split I saw several pages with a certain "OS-scetion" characteristic:
  • Pages not properly spelt by the rules of LW:PN were still in the alphabetically ordered list, but all of them were redlinks
  • There were links to the properly spelt counterparts of all redlink-pages
  • These links were in the same order like the redlinks
  • They were listed in one block exactly before the first "LW:PN-caused redlink"
  • These characteristic could only be found in the OS-section, but not in the albums' sections
I forgot to check some of these pages, but is that type of edits made by LWT?
And another one: look at Buddy; Ltw picked exactly the wrong of two artists from MB (the second), although the first result was the correct. Why? Akuma was wrong, too. Please don't understand me wrong, I love Lwt for filling in such stuff, but especially the case of MB makes me wonder what went wrong? Usually I'd expect Lwt to take the first link, but as it didn't there seems to be some criteria, that however misled it... - Chris 22:47, December 7, 2009 (UTC)
  • Regarding your set of questions about OS section: I can only say that normally Lwt shouldn't edit OS section at all. Sometimes it may be done by senv, too lenient to jump into another window unher his own account ;). But I'm not sure I have understood a problem you were talking about - if you wanna know who made those edits - take a look at edits history; otherwise - give some example, sharing more light on the issue.
  • Regarding Buddy: it's quite possible, even more - I'm sure that's not the only one Lwt error ;). Unfortunatelly haven't time enough to look into details right now to detect if that was a systematic bug or some sporadic error. I promise to make this check later. Thanks for noticing that, and let me know if you find some Lwt misbehavior in the future.--Senvaikis (talk) 15:00, December 8, 2009 (UTC)
Checked Buddy (It wasn't mentioned, but discogs link was wrong also). News aren't very good. It wasn't a sporadic error. And what is worse, it isn't even a bug. It's a result of incautious solution of lenient developer.
Situation'll be coherenter if you take a look at Lwt jobs journal, (2009-11-02 - 2009-11-03). Now I can start explaining. While updating artists akuma, I noticed, that akuma artist page contains mb, dogs, wp & ms links also. So, the idea to grab this ready-for-use info was natural and attractive. The question was only one - what's the reliability of this info? I walked through most of known (for me) artists akuma pages and find info quite reliable, with some exceptions though (see notes in journal). But number of these exeptions seemed to me negligible enough to be not treated as serious counterargument to use this info. So, I decided to save my time and efforts and use this akuma info if available instead of making special procedures for batch mb, dogs, wp & ms metadata searching & updating. Truth to tell, some of them (mb, dog) were awailable (and much more reliable) in Lwt, but akuma approach was much quicker :). Now let's return to poor Buddy to see why I think now it was a wrong solution. If you try to make a search for "Buggy" on akuma and click on first link in artist name matches list, you'll get exactly the same result as Lwt did: p707298. So, we have one of those 'negligible' exceptions case here. The problem is that this error is multiplied then by number of other links in this page, and that can't be treated as negligible error anymore, 'cause that's already systematic error. Still don't know a real extenth of this negligence results. Number of pages, touched in this process is ~18000. It's quite possible Lwt'll need to go once more through all these pages checking metadata reliability and removing erroneous if found. Yep, that may be a real job for it...--Senvaikis (talk) 21:14, December 8, 2009 (UTC)
Maybe it'd be a security feature to fetch the WP arcticle for an artist, check reliablity by f. e. checking the hometown, any album's title or something. Senv, don't be too concerned about these wrong links; if you multiply the wrongly added links, you also have to multiply the correctly added links - and that will keep the levels at the same points like they were when you checked the reliability. 18000 pages, maybe 500 of them wrong. You cannot multiply the 500 by 4 and compare 17500 to 2000, but you'd have to compare 70000 to 2000 links then, which again looks very good, doesn't it? =) Keep up the great work! - Chris 22:15, December 8, 2009 (UTC) P. S.: You could also use this feature once again after you finished the link verifications: MB also has got relationships like Akuma, especially for DC, WP. Furthermore, album pages have got the ASIN on it, which could be filled in by Lwt going through the MB album category and checking the pages for ASINs.
Checked all ~18000 pages with akuma-based metadata. Musicbrainz ID was taken as touchstone of testing. Result - two lists with pages, having mismatches between akuma-based mbID and MB-search result: MB ID Mismatches and MB Titles Mismatches. Initial research shows that most of these pages should be checked manually that's not my favorit kind of job, as you know ;)--Senvaikis (talk) 14:01, December 13, 2009 (UTC)
I checked some links, and well, let me tell it's usually not LWT's fault something went wrong. 411 for example should rather be The 411 (referred to as "all girl band" at MB). There are 2 artists named 411, one is the current id, one is the search id, and meant is a totally different group... - Chris 11:11, December 31, 2009 (UTC)
Surely the list of such slips is substantially longer, but - please don't be mad at Lwt - we both think that we've done already our best here and aren't going to return to that ;). Happy New Year, --Senvaikis (talk) 11:46, December 31, 2009 (UTC)
I'm not mad at all, far from it! What I said that LWT made a perfect job for the parameters it got, but these were the wrong ones because some users made mistakes earlier^^ - Chris 13:22, December 31, 2009 (UTC)

Deadend Pages

I'd like to work together with you: I'm tidying up Special:DeadendPages, and my helping script compiles a list of pages that have to be moved due to LW:PN. Could LWT import these lists and move the pages, that will be listed on any interface-page created for that purposes? As the list is generated by me, you may even choose the format, f. e.:

Single plaintext items like
  • AFI:Kiss and Control ///// AFI:Kiss And Control
  • A Red Season Shade:Last Light on our Walls ///// A Red Season Shade:Last Light On Our Walls
  • A Red Season Shade:Shades of Truth ///// A Red Season Shade:Shades Of Truth
  • Aaliyah:More Than a Woman ///// Aaliyah:More Than A Woman
  • ...

You know, I'd do it myself, but it's hard for me since I may move only 2 pages per minute (unless there's some possibility to move more pages without being admin^^). What do you think? - Chris 13:22, December 31, 2009 (UTC) P.S.: Of, of course, yeah, there are pages that do not have to be moved. Well, what to do with them? Store them and merge them later, I guess. LWT cannot judge which lyric version is better^^ - Chris 13:36, December 31, 2009 (UTC)

Sure, it may be done, with some stipulations though. One you've mentioned already; the second one is the fact that Lwt account isn't an admin too (therefore - can't detete pages), so I should run it under my own account for lists containing non-empty destination pages (provided that someone have made mentioned "judging" already ;)). Format of the lists doesn't matter for me - Lwt is able to digest all the most popular formats used in such cases (but favorites are xml & csv).--Senvaikis (talk) 14:00, December 31, 2009 (UTC)
So that is ok with you? =) - Chris 15:43, December 31, 2009 (UTC)
Oh, and you shouldn't overwrite pages, if you want, I can give you your own ftp Account to my webspace so LWT can upload a html file with the links it couldn't move because the target already existed. - Chris 16:29, December 31, 2009 (UTC)
(edit after edit conflict)
Sure, with two notes:
  • http part is both redundant and useless - Lwt will cut it away anyway - it needs pure pageTitle
  • If we wanna work with single list for both types of moves (empty/non-empty destinations) - the third node, pointing to the destination type would be desirable (f.e. - boolean "Empty") . Lwt would be able to process all nodes with empty="true", me - the rest. Alternative - generate two sepparate lists: for_Lwt.xml and for_Senv.xml
Now - sorry, but just now I'm gonna engage in more serious things - a New Year meeting ;) HNY! --Senvaikis (talk) 16:32, December 31, 2009 (UTC)
It cost me a lot of time, but I finally made it, using a LW-API query to find out whether a page exists or not. Look at the results now. I added IDs just in case you want to log LWT's actions. If you don't need them, I'll drop them for the first real file to work on. Later, I'll create a new xml-file whenever it seems sensible, and (with the ID^^) LWT could find out whether it's a new log or not.
However, thank you Senv, you're a good pal. =) Happy New Year to you =D - Chris 20:55, December 31, 2009 (UTC)


Hey, could you touch the pages you moved afterwards again so they disappear from the category? And could you please touch these? Thanks, 18:05, January 4, 2010 (UTC)

DoneGreen, but not sure if that helps--Senvaikis (talk) 19:41, January 4, 2010 (UTC)
P.S. Hope you've noticed that pages with non-empty destination weren't moved. It seemed for me that some of them should be redirected instead of moving. --Senvaikis (talk) 19:56, January 4, 2010 (UTC)
Hi, Chris. Noticed changes in your movables list. Processed all nodes with empty destinations. Once again - list items with non-empty destination weren't moved - they contained pages, which should be redirected, not moved (see this f.e.). We've been talking about that already - cyborg just can't be that judge, deciding which page version should be recognized as being 'lawful'. So, if you wanna put this process on some more regular backbone, we must describe it more strictly.
BTW - this post saving was blocked by spamfilter, so I was forced to whitelist your hosted site address ;) --Senvaikis (talk) 09:03, January 11, 2010 (UTC)

The Wave (and iTunes links)

Hey Senv, I saw you mention google wave on Seans talk page, and seeing as I've got a Google Wave account too, I've created a public wave for us lyricwikians to 'meet'. Search for "with:public lyricwiki" then through your lyricwiki friends in "The LyricWiki Wave".

I've linked to your iTunes Links rundown from the Community Portal, in a hope to get some talk about what should be done. Cheers, LYRIC-Humbug wordsdeeds 05:08, December 8, 2009 (UTC)

Thanks, Humbug; sadly today I'm 'on the race', so - till tomorroww (or late night) :)--Senvaikis (talk) 05:42, December 8, 2009 (UTC)

Melting SNLI

Just can't believe that real SNLI contains only 40 songs. I suspect this number must be ~80. So, now I'm going to make an experiment, touching all pages in archived SNLI list. If my inkling appears to be wrong, - I'll remove this message, posted just for fixing the starting point of experiment (SNILI.count=40). --Senvaikis (talk) 14:14, December 13, 2009 (UTC)

I'm surprised, but snli.count has grown only by 1 in result of touching. Does that mean that melting has stopped? Anyway, - leaving this msg undeleted - just for history. --Senvaikis (talk) 14:47, December 13, 2009 (UTC)

Merry Christmas!


--Senvaikis (talk) 11:13, December 24, 2009 (UTC)

Community content is available under Copyright unless otherwise noted.