FANDOM

2,054,160 Pages

MusicBrainz links

Hi Senv,

I recently had an idea how to improve LyricWiki. Usually I'd try to write a bot on my own, but as you're obviously a lot more skilled and since your programming language seems to have a better sourcecode-library for the API I guess it's maybe easier for you?

I thought about using the MusicBrainz Web Service. Step one: Use Category:MusicBrainz/Artist to find the MB-Ids for the artist's albums at LyricWiki. Step two: Use Category:MusicBrainz/Album to find the MB-Ids for those albums' songs. That'd immensely increase the frequency of MB-links on LW-pages, because almost every artist has MB-parameter filled in, but only a few albums and even fewer songs!

Do you think this is doable? - Chris 17:23, July 1, 2010 (UTC)

Nice idea, though not very fresh, - truth to tell, that was one of the initial purposes of Lwt. The same should be said about collecting/filling of metadata from other sorces (amz, amg, dog, acuma & iT). But have you ever tried to emulate such fully automatic (bot) process? Believe me - this seemingly simple task isn't so plain as one may suspect. That's why Lwt never was just a bot, but rather cyborg - I still have no idea how to train Lwt make this job without any human intervention. I hope you've noticed already that even now Lwt is able to make serial updates on all given artist albums/songs metadata (not only mb). But all these updates are made by the cyborg, not a bot, - e.g. I must be an active member/supervisor of this process. I'd be very interested to hear any idea how this process may be fully automated, and your question gives me a hope that you have such idea... If you have - I'm waiting for more details. thx, --Senvaikis (talk) 19:27, July 1, 2010 (UTC)
Hm, tell me what you are currently supervising/checking. What is the step that cannot be done without human help, what cannot be done by fetching XML from both LW's and MB's APIs and comparing string-variables? - Chris 20:24, July 1, 2010 (UTC)
Sorry, Chris, but I desperately need to sleep a little, so till we meet next time - just try, as was suggested earlier, to emulate this process, in consideration of titles/release dates/tracknumbers etc. mismatches, as well as non-one-to-one mapping between LW & mw/it/amg. etc. etc. :) nite, --Senvaikis (talk) 20:36, July 1, 2010 (UTC)
Okay, I'll try. =D Good night. Chris 20:57, July 1, 2010 (UTC) hardly = "nearly not", changed it to desperately = "a lot and now(!)" ;)
 :) - Thanks for a fix - that's not the first (and surely not the last) time when my broken English led to "inverse accuracy" :). Ok, now I hope my mind is virile enough; so - have you tried to find those "bottle-necks", requiring human intervention, or should I describe them? --Senvaikis (talk) 06:54, July 2, 2010 (UTC)
Sure, no problem for the fix :)
Bottle necks I could think of:
  1. 2 albums named identically (f. e. reissues some years later) - solution: log and skip album
  2. One song on multiple albums with different IDs - solution: fill in the first found ID and skip songs which already have IDs
  3. Typos in titles/song cannot be found - solution: skip albums/songs
  4. Multiple artists with the same name - solution: find by comparing the albums' names on matches
  5. Curretly wrong artist-ID in LW - solution: will likely result in a mismatch of all albumnames; error won't spread/reduplicate with every step
I know that skipping a page means to just delay the problem. Always keep in mind that the goal is to insert as many IDs as easily possible, not to definitely insert the ID for absolutely every possible song which has one anywhere in MB.
IMO it's worth a try, because whether a page has no ID because Lwt skipped it or because it's just not there is no difference.

Example for Binärpilot

Method one (MB database triggers LW search)
  1. Binärpilot in Category:MusicBrainz/Artist
  2. Get the MB-ID from LW API Example link
    • Selected ID: 9c6bddde-6228-4d9f-ad0d-03f6fcb19e13
  3. Get the release-groups from MBWS Example link
    • Selected rg-ID: ae601fde-9c36-33f8-8de3-c45c3d50dafc
  4. Get the albums from MBWS Example link
    • Selected album ID: fdfdb3c0-8820-4c22-94e6-6aa136e36c31
  5. Get the album from MBWS Example link
  6. Search the album at LW API Example link
    • Selected album: Defrag (2005)
  7. Get the album from LW API Example link
  8. Compare LW songs (links with colon) with MBWS songs (No matches? Try another rg-ID from MBWS)
    • Insert album ID
    • Insert song IDs
Method two (LW database triggers MB search)
  1. Binärpilot in Category:MusicBrainz/Artist
  2. Get the MB-ID from LW API Example link
    • Selected ID: 9c6bddde-6228-4d9f-ad0d-03f6fcb19e13
  3. Get the albums from LW API Example link
    • Selected album: Defrag (2005)
  4. Search the album at MBWS Example link
    • Selected album ID: fdfdb3c0-8820-4c22-94e6-6aa136e36c31
  5. Get the album from MBWS Example link
  6. Get the album from LW API Example link
  7. Compare LW songs (links with colon) with MBWS songs
    • Insert album ID
    • Insert song IDs

=) Chris 12:51, July 2, 2010 (UTC)

Nice job, Chris! I'm really impressed - you've listed almost all mb metadata search-related methods, used by Lwt :). But the steps, described by you in the most details, practically don't need any human intervention and are used by Lwt automatically without any problems for a long time, while such 'simple' actions, mentioned by you almost without any comments, as 'compare strings', selecting matching albums or (especially) songs, - aren't so simple. That's exactly the place where Lwt is usually allowed only to select required information and make initial matching procedures - then it waits for my approval to update (often preceeded by some manual rematching). Example you have used wasn't the best choice, if you wanted to disclose mentioned bottle-necks, when one LW album has multiple mathes on mb, or mb album is only the part of Lw album. All data should be taken into account - release dates, album tracks count, different interpretation of multialbum releases in mb and LW, song titles matching etc. etc. No need to list all possible reasons of song titles mismatching - believe me, there are a lot of them. This action is the most complicated in this task, at least for me (sorry, - for Lwt). And I haven't got yet any universal algorithm for it's 'full automation'.
I agree with you - we could simplify the task, taking only exact string matches. But keep in mind that such simplification would lead not only to narrowed list of matches. There's one side effect of such simplification which I'd not like much more - let's name it 'data mixing', when songs of some release, having real (easy-human-readable) matching in mb, may get their mb data from different mb releases just due to some negligible (again - for human, not bot) string mismatches. What do you think about that?
Anyway, thank you for your nice efforts to waken me from some kind of lethargy - may be I should think about this task more... having a time... if you help me... :) --Senvaikis (talk) 20:16, July 2, 2010 (UTC)
I know what you mean by those nasty variations: I recently had a case with Nagelfar and the album Srontgorrth.
Track one: Kapitel Eins. Der Frühling: Als Die Tore Sich Öffnen... (Freiheit Oder Untergang?)
Red = variable
  1. First word "Kapitel" or "Srontgorrth - Kapitel"
  2. "Eins" or "1"
  3. "." or "," or ":"
  4. Umlaut "ü" written as "ue"
  5. ":" or songtitle ends or the folling in brackets
  6. (as always) some mismatches on "Die" (LW:PN) vs. "die" (original)
  7. "..." yes or no or as this one coherent letter
  8. "(Freiheit Oder Untergang?)" appended or not?
That with 5 songs in various combinations, all of them appearing on other albums as well before included into that one album. A nightmare! This is why I (as an exception) deleted all the pages not matching the naming scheme I use instead of redirecting them. I just didn't want to have different 30 messy names which no one would ever guess again.
And yeah, I understand what you mean by "data mixing". The song Oceans of Time by Axel Rudi Pell f. e. appears on Oceans Of Time (1998), The Ballads II (1999) and The Wizard's Chosen Few (2001). Would be "bad" if at MB's page to "Oceans Of Time (1998)" album there was a typo like "Ocean Of Time" and therefore the ID was the one from "The Ballads II (1999)". Hm...
I could help you in that way:
It currently appears to me that you discarded/delayed the idea of inserting the meta-data as cyborg because you do not have the time or aren't interested in supervising the process all the time. Well, I actually am, I am used to it from my own Greasemonkey scripts... That is: I am willing to run and supervise such a cyborg.
  1. Show me your current interface. If it's good, ignore the next two steps and simply "cut" it out into a new program (you wouldn't want me to get a full copy of Lwt, would you?)
  2. Your interface had a bigger potential to optimize it. I'll create an optimized one, adapted to my needs and demands on such a tool
  3. You adapt the current sourcecode to the new interface
  4. I create a new account "MusicBrainzBot" (like "GracenoteBot") which is the cyborg's (then bot-)account here
  5. You give me access to the program or hand it over completely
  6. Finally, I'll run and supervise it
You see what I mean by "I'd try to write a bot on my own", meaning I'd also try to write a cyborg on my own. The problem is: I simply can't. It's too difficult in the programming language I use. It's not made for XML parsing I guess, because it's always been very difficult to do easy tasks, f. e. for login (before logintoken was needed):
function Login(Name, Password: String): TStringStream;
var Parameters: TStringList; Response: TStringStream;
begin
  Response:=TStringStream.Create('');
  Parameters:=TStringList.Create;
  Parameters.Add('action=login');
  Parameters.Add('lgname='+Name);
  Parameters.Add('lgpassword='+Password);
  Parameters.Add('format=xml');
  try
    HTTP:=TIdHTTP.Create;
    HTTP.Post('http://lyrics.wikia.com/api.php', Parameters, Response);
    result:=Response;
  except
   on E: Exception do
     begin
       result:=nil;
     end;
  end;
end;
And navigating through the XML output is even trickier. I guess it's easier for you, isn't it? Looking forward to hearing from you! - Chris 22:11, July 2, 2010 (UTC)
Glad to hear you are ready to help me, - that's nice, thanks. Unfortunatelly, can't say the same about me - currently I haven't free time enough to meet such help requirements. You are programmer, so I have no need to explain you what does it mean - "cut module from main programm and adjust it to your requirements, as well as other, independand from you, requirements, such as making it user/hardware/proxy/firewall independand, etc. etc. so, so, - maybe later :). And why don't you want a full Lwt?
Btw, about interface - though this Lwt module has been slightly changed, but actually more than year-old screenshots (A-D) still may be used as quite suitable illustration of current functionality & interface of this module (tab "Tracks"). cu, --Senvaikis (talk) 20:58, July 3, 2010 (UTC)
Hm, ok, I'll then try to do it on my own. You grow with your tasks.
Why I do not want my own copy of Lwt? First, because I don't think I'm a good user for Lwt. As you developed it, you know how it works, what it does when you click there and there. It is your program. It's only easy for you to use it. I'm afraid I cannot understand the program to the level you do, and I guess I couldn't improve, extend or fix anything there, because it's written in .NET.
Second I don't feel comfortable with the idea of me "stealing" you the tasks your Lwt currently is able to do. Too easy you could get the impression of "I'm not needed anyway, Chris does what I did, my time has come to search for some new task" and so on. And that wouldn't be true, you are needed here, and it was a very great loss if you left! Seriously, I mean it. Together with me you are the only really skilled programmer still very active here as far as I can see. teknomunk has temporarily left, and Will hasn't had too much time on his hands in the recent past, either. 6 is a template guru, but I haven't ever seen anything programmed for PC by him.
Truth to be told, I'm afraid of the prospect being the only active (= doing edits here) active (= currently developing something) programmer here. I don't know if I'm forgetting about anyone here atm, but I think I felt alone if you left...
And third it's not my style of program I find comforting, very packed GUI (which is the logical consequence for a program with that amount of functionality). This is why LWS has this module-building and the 3-step GUI. You do not see what you currently do not need.
Okay, you grow along with your tasks, so we'll see where this challange gets me. The biggest problem will definitely be the work with the APIs and their XML. As said, Delphi isn't good at working with XML. - Chris 23:05, July 3, 2010 (UTC)

Developing of MusicBrainz Cyborg

Current status is available for download. Not too much functionality (basically just login, logout and artist-search so far). Needs a new order of elements, I think I'll order artist/album/songs vertically and only show one of them while the other two slide closed again. - Chris 03:20, July 4, 2010 (UTC)

Next steps done:
  1. Rudimentary implementing of XML displaying was improved a lot! This is important because that way I can use the lines for getting content
  2. Layout improved
Still needs a lot of work... Next step is graphical displaying of the XML results. I'll use records for artists and display the information in a string grid. I hope I'll get the cyborg completely done in the first weeks of my summer vacations. - Chris 01:45, July 5, 2010 (UTC)
Glad to hear about your progress; let me know if you have some problems. Btw, from a quick googling on XMLDocument, SelectSingleNode, XPath together with Delphi I decided that you shouldn't have any problems here too. Good luck, --Senvaikis (talk) 10:40, July 5, 2010 (UTC)
No problems so far. Just one other question: could you do me a favor and batch-update the second {{Song}} parameter (artist)? I batchmoved two artists and had to replace it everywhere. Artists are "HELLYEAH" (moved from "Hellyeah") and "Matchbox Twenty" (moved from "Matchbox 20"). Would be cool because that are many pages. - Chris 13:20, July 5, 2010 (UTC)
DoneGreen, --Senvaikis (talk) 14:10, July 5, 2010 (UTC)
Greetings from Sir Thanksalot ;D - Chris 17:59, July 5, 2010 (UTC)
Can you do that for *NSYNC, too? Need "*NSYNC" in there. - 00:37, July 11, 2010 (UTC) DoneGreen
Thank you =) - Chris 14:30, July 11, 2010 (UTC)
Still thinking about how to do it. I've recreated the program once more, because what started good turned out to become unhandy. I'll do without animation now, just need too much work. Furthermore I'll try to use own units for each source (LyricWiki, MusicBrainz, AllMusic, ...), so the source-code of the main-program is disentangled and clear again.
And I have one question: I noticed that the LW-API-XML is quite bad (unfortunately). As you certainly know, it's
<getArtistResponse>
    <artist></artist>
    <albums>
        <album></album>
        <year></year>
        <amazonLink></amazonLink>
        <songs>
            <item></item>
            <item></item>
            ...
        </songs>
        <album></album>
        ...
    </albums>
</getArtistResponse>
where it should be
<getArtistResponse>
    <artist></artist>
    <albums>
        <album>
            <title></title>
            <year></year>
            <amazonLink></amazonLink>
            <songs>
                <item></item>
                <item></item>
                ...
            </songs>
        </album>
        ...
    </albums>
</getArtistResponse>
How do you parse that, or do you use the HTML-response? Or don't you use "action=lyrics&func=getArtist" but one of the "usual" queries like the one I listed above? - Chris 20:51, July 9, 2010 (UTC)
 :) That's a twice-told tale, Chris - just take a look at this discussion - it's starting point and result for a start. Too late now for details, so all the rest comments - tomorrow, nite, --Senvaikis (talk) 22:08, July 9, 2010 (UTC)
Thanks, that fixXML is the only thing I needed to know. No need to maltreat your fingers any more. - Chris 09:49, July 10, 2010 (UTC)

Stevie Wonder - Have a talk with God

There are people who have left the problems of today

Lead them to conclude that for them life is not the way

But every problem has an answer and if yours you cannot find


... why writing more if i can't edit the lyric? - Francescortiz 13:53, 6 July 2010

Hi, Francescortiz, hope you've found already the reason of truncated/protected lyrics. Believe me, most of us as well are unhappy about Gracenote licencing restrictions. But if your interest is limited only in mentioned song lyrics, just take a look at one of the last it's editions, before restrictions were applied ;). And - please, sign your messages. --Senvaikis (talk) 19:53, July 6, 2010 (UTC)

New userbox

{{Userboxes/Cyborg Editor}} - Originally only intended for me, I think you might maybe like it, too? =) - Chris 01:20, July 17, 2010 (UTC)

Weird LWT behaviour

Hi Senv, I just thought I'd inform you about this - Chris 22:27, July 17, 2010 (UTC)--Senvaikis (talk) 14:39, July 18, 2010 (UTC)

Thanks, Chris, though imho, one and the only weird error, happened to occur while tuning Lwt.Google.lang module, harldly may be qualified as any behaviour (manner of behaving or acting). cheers, --Senvaikis (talk) 14:39, July 18, 2010 (UTC)
Hehe, yeah, I thought that this was a quite unique error, but rather tell you something you already know than maybe keeping an unintended error alive. =) - Chris 15:37, July 18, 2010 (UTC)

LyricWiki:Job Exchange

Hey Senv,

it's me again, your shade with those annoying requests and questions ;) Once again I want to ask you for joining a project of mine. This time however, it's not my tasks but our users' tasks that will occasionally pop up at LyricWiki:Job Exchange. The sence behind it is really simple and explained there.

Depending on your decision there will be more or less tasks to be offered. One of the tasks could be serial update of SH.artist after batchmoves. Looking forward to hearing from you. - Chris 00:35, July 25, 2010 (UTC)

Nice idea, Chris. I'll join to this project gladly, the rather that sometimes Lwt keeps it's hands in pockets for a very simple reason - just I haven't any job for it ;).
There are more typical possible tasks which may be addressed to me/Lwt via JE:
  • Creating album pages for all artist albums, (desirable - having non-red tracks & {{Album Art}}s on the artist page)
  • Resolving all redirects on the artist page (this should be done before previous task)
  • Filling the most common metadata (wp, iTunes, amz, discogs, mb, goear) for all given artist albums & songs (required: all albums & songs of the artist to be filled, e.g. metadata will be collected and filled only for albums and songs, having pages, backlinked to the artist). That may be one of final stages in artist/albums/songs editing, before making them golden.
The set of typical tasks for Lwt surely may be much wider, - just now I'm on vacation, sitting on the shore of the lake, and using extremely slow mobile connection (EDGE), so not all hard-traffic-tasks are suitable in such circs...
Once more - good initiative! --Senvaikis (talk) 12:51, July 25, 2010 (UTC)
Okay, thanks =D This is great because the more tasks we offer, the better it is going to be. There are some tasks missing anyway, which I'll add in the next days.
Yeah, I know that slow web is a pain. I just wanted to notify you about the new JE. The tasks there are just examples so far. Please note how I did it with your comments on "Edit Page": I moved them to your own section. Thanks to the new extension "Loops" we can add as many numbered parameters as we want, so a chain of contributors per template is no problem. Each admin has it's own section everytime he participates. That's the idea behind, to track who's done what (f. e. look at "Batch Move"). I tried to create them in a realistic way.
And now: enjoy your holidays! - Chris 16:25, July 25, 2010 (UTC)

de-URLing

Since you're apparently begging for more work for LWT… :-) Here's something to do for you. Just to make sure prefix the country code if the link is to a non-English WP. Replacing the underscores and percent codes might be a good idea as well. — 6×9 (Talk) 14:57, July 29, 2010 (UTC)

DoneGreen --Senvaikis (talk) 16:25, July 29, 2010 (UTC)
Quick work! Despite mangled sentence above – somehow a "to" wandered left Thanks! — 6×9 (Talk) 16:40, July 29, 2010 (UTC)

Albumcovers from Lwt

It's nice to see that Lwt uploads the albumcovers from Discogs (saw it on Nylon Beat). But there are some points I want to mention:

  1. 500×500 pixels is our maximum for Fair use. Could Lwt resize the pictures before uploading them? Slightly bigger images are "tolerated", but resize for dimensions bigger than 520×520, then to 500×500 of course.
  2. File covers' names do not have to comply to LW:PN, so the file to 12 apinaa should be File:Nylon Beat - 12 apinaa.jpg, and the {{Albumcover}} template should have the real title as second parameter.
  3. When inserting the cover, does it automatically update the albumpage? Because it looks like it forgot Nylon Beat:Satasen Laina (1997). Maybe you want to look into the code.
  4. If it updates the pages with covers, it could also insert the DC-parameter for them. If DC has got the cover, it likely has a master/release page for that album ;)

What do you think? - Chris 07:52, July 31, 2010 (UTC)


  1. I always treated dog as one of the most reliable sources for a legal info. Yes, remarkable part of cover images@dog have ~600 px resolution, but I never thought (and still don't think) that such oversizing may be treated as an essential crime, especially when image source is provided in {{Album Art}}. You are the first, telling me that dog's resolution is inadmissible. Take a look at images, uploaded by other admins and you'll find that I'm not lonly ;) Yes, when amz have required cover image with tolerable quality, - usually I'm using it; otherwise dog is being searched as possible image source, then - all the rest available sources.
  2. We have been talking about that already, but I may repeat - Lwt isn't smart enough for such 'proper' casing. Yes, LW:PN isn't obligatory for files, but it isn't forbidden too, so Lwt is going to comply LW:PN for a very simple reason - that's simpler ;)
  3. You know the answer, I suppose :). Yes, I'm updating/touching albums after cover images upload; and yes, I'm updating/touching images after album creation. But not automatically, - that's the reason you've found forgotten one.
  4. Yes, that's possible (in current module - with additional requests). But why only dog? Why not to add asin param, if image is available on amz too? Why not to add all asins to all songs of the album (in most cases amz cover images are going from MP3 department, so songs asins are available also)? One of my favorite Voltaire proverbs says: "The archenemy of the Good is the Better". I'm using this proverb when being too lazy to modificate/improve something in my program (or when someone asks me to make some of my programs to do almost all possible tasks at one single click) :)
--Senvaikis (talk) 13:31, July 31, 2010 (UTC)
  1. I know about big covers uploaded by admins (esp WillMak is well-known for it^^), but with covers it's the same as with the lyrics: they are objects of copyright; the only reason why we may collect and display them is the wonderful Fair use law, allowing to present copyrighted material for documentary purposes. But the first point (read the template) is: "Though this image is subject to copyright, I feel its use is covered by the U.S. fair use laws because it is a low resolution copy of an album cover." Dimensions over 500 pixels per side can hardly be considered "low resolution", if you keep in mind that good printers will print a really good-looking cover out of a picture with only 300px per side. Therefore, if I find big covers uploaded by users, I download them, resize them, and re-upload them. This is a very timeconsuming activity though, as I have to do it all by hand. I don't know, there is no maximum resolution by law, but when I was new here yes, I've been a newbie one year ago ;) I asked what dimensions would be considered "low resolution" and they told me "500 pixels per side as a maximum, in some rare cases we tolerate slightly bigger images". Since then I always keep an eye on it. Furthermore there is no reason for higher resolutions, as the images are never displayed any larger 200px than per side (on the main page as AOTW). Well, I don't know, if resizing is too complex... I thought it was easy with .NET as the graphics-editor I use is written in .NET (Paint.NET).
  2. Okay, if you apply LW:PN to filenames it's okay with me, but then please do not change the second parameter in {{Album Art}} if it only differs in capitalization. =)
  3. ...But not automatically... okay, that was what I needed to know. Otherwise it'd have been a malfunction of your code =)
  4. ...in current module - with additional requests... oh, I thought you'd get them by a request including the master/release code anyway. I've not looked at the DC API too closely yet. I'm currently fighting our own API for my cyborg (which is pretty good looking so far btw.) :D
Chris 09:27, August 1, 2010 (UTC)
Yes, image resizing before reuploading really is a snap for .NET - it'd take a single line in Lwt code. But image, resized by standard .Net methods, not always will have a smaller size (in bytes), in addition, sometimes quality of reduced image also may be affected. So, I'm still not sure that it'd be a very good idea to resize all covers from dog. On the other hand, if community decided that it must be done, - Lwt'd take this task :)
Glad to hear about progress & nice looks of your cyborg, --Senvaikis (talk) 17:47, August 1, 2010 (UTC)

Statistics for oversized images in AC

Current AC contains 5594 oversized images (treating as oversized all images with resolution, exceeding 500*500 = 250.000 px2).
Here's more detailed statistics for their distribution by size:
Size is grater then, px2 Count
2500005594
3600002056
4000001808
5000001520
1000000461
500000023
So, while we are not determined about dog res (<=360000), Lwt could resize at least those 2056 images, oversized even according to my 'liberal mindset' :)
Does anybody have some objections/suggestions?
--Senvaikis (talk) 13:32, August 2, 2010 (UTC)

DoneGreen - resized 1608 images - the most oversized part of AC (with size > 490000 px2). --Senvaikis (talk) 21:12, August 3, 2010 (UTC)

Current situation: there are 4254 images in AC, exceeding 500*500 resolution limit; 464 of them are still sized over 600*600 px. --Senvaikis (talk) 09:26, August 4, 2010 (UTC)
Did I already mention I like your tables? =D I'll ask Sean if it still is true that we prefer 500x500 - Chris 11:16, August 6, 2010 (UTC)
There's no need to ask Sean if we prefer 500*500 - the confirmative answer is predetermined. The question should be: what is the size limit, making the image inadmissible.  :) --Senvaikis (talk) 11:57, August 6, 2010 (UTC)

/* Just a Question? */

hi there

if you have the time, would you take a look at the pages "Knuffelrock" and "De Pre Historie" ?

Just wondering: is this ok, or should i add the album on that page? for Knuffelrock it can be done, but i would really not do it for De Pre Historie.

Thanks for the trouble

LyricWyverex

Hi, Wyverex, nice to hear you again after such a long pause. But all this time I was watching your progress and now I'm glad to say - you became a steady, substantive and prolific editor. Thanks for your remarkable contribution.
Regarding your question, I'm not sure if I've got the core of your question, but I'd recommend you to borrow an idea of organizing such albums-lists form Bravo Hits pages. If you wanted to ask something different, - please, feel free to reask me in more details (you know - I'm very bad in the finer points of English) :)
nite - --Senvaikis (talk) 21:23, August 9, 2010 (UTC)


Mass-editing required

All (or at least most) of this user's creations lack the closing lyric tag & SF… — 6×9 (Talk) 20:14, August 15, 2010 (UTC)

And all pages containing a ' need to be moved to the pages where the letter after the ' isn't the capital letter but the LC one ((s)he wrote It'S instead of It's). I checked all contributions. - Chris 23:39, August 17, 2010 (UTC)
Okay, forget about that, I now moved, joined or redirected the pages myself. - Chris 22:52, August 29, 2010 (UTC)

2 Artists with same names

Got just one question: What if there are 2 different artists with the same name? (e.g. Maxim, there is also a german reggae artist called Maxim) How should I name the second artist (like "Maxim (German)" or how?) I'm looking forward to hearing from you - FPJ 20:29, August 17, 2010 (UTC)

There's a disambiguation template {{Disambig}}, made specially for such cases. It's being used for listing all homonymous artists, as I've made in Maxim (Disambiguation). Then another template {{WrongPage}} should be used at the very begining of each listed artist page (as done with Maxim & МакSим). I left the page of Maxim (GE) untouched intentionally - feel free to fill it or ask me, if my instructions weren't clear enough :).--Senvaikis (talk) 12:59, August 18, 2010 (UTC)
Thank you very much, I think I got it :) FPJ 14:42, August 18, 2010 (UTC)

1 album 2 artists

I wonder, Is there any problem if I redirect to same album from two different artists?

I did it in this album:

Split EP (2004) from Agalloch

And

Split Ep (2004) from Nest

Thanks in Advance. -Schakalopoulos 03:26, 19 August 2010 (signed 4u by senv)

Yes, it would be ok if you created one single album page for this album, having it's representation on both artist pages, but actually you've created two pages Agalloch:Split EP (2004) and Nest:Split EP (2004), e.g. - two separate albums in terms of LW. I'd recommend to make one collaboration album page for album "Agalloch & Nest:Split EP (2004)", redirect both current albums to this newly created page and include it's representation into both artist pages.
Hth, and happy editing.
Oh, and please - sign your messages with "~~~~"
--Senvaikis (talk) 08:01, August 19, 2010 (UTC)

If Lwt needs work

then I got something for it. Thanks Smile - Chris 22:31, August 28, 2010 (UTC) Sorry, but I had to change the format again.

Lwt doesn't need work, but it's always ready to do some work, if someone asks :).
DoneGreen, with 2 notes:
--Senvaikis (talk) 05:46, August 29, 2010 (UTC)
Hello Senv, thank you very much, and:
  1. This is weird. The only reason I can think of is that they were put into multiple categories by hand.
  2. The program only suggests the absolutely neccessary titles, not the aesthetically best-looking (LWS tries that, however...). For example, for "Rose:Yes We Did (alternative version)" it suggests "Rose:Yes We Did (alternative Version)" and not "Rose:Yes We Did (Alternative Version)".
Maybe I should change the rule, that any character behind a non-alphanumerical character is uppercased, if the non-alphanumerical is the first character of a word or the title. That way "Xy (z)" would turn into "Xy (Z)", and "(x Y) Z" will be changed to "(X Y) Z", but "Xy(z)" or "X-y-z" stay as they are and are not changed to "Xy(Z)" or "X-Y-Z". - Chris 11:03, August 29, 2010 (UTC)
(e.c.) You positivelly know my sligtly infidel position regarding all these capitalisation rules. If I(/lwt) had my way, LW'd be case-insensitive. Ok, let's treat that as a bad joke ;).
Anyhow, while rules are such as they are, I'm trying to be as loyal to them as possible (within the pale of common sense though). So, if lw:pn says that all pagetitles must be inicaped, - ok, let it be so. But why then should we complicate our life even more than it's required according to lw:pn, titling song as Rose:Yes We Did (alternative version) or Rose:Yes We Did (alternative Version) which, truth to tell, seems even worse compared to Rose:Yes We Did (Alternative version) imo), instead of simple 'dogmatic' Rose:Yes We Did (Alternative Version)? Moreover - look once more at the Rose's page and try to find at least two songs with title "Rose:Yes We Did...", having different lyrics. Yes, there are two songs: Yes We Did & Yes We Did (alternative version) with 'different' lyrics. Difference - 2 inicaps!. Shortly - in terms of LW (lyrics! site) they both represent the same one song, ergo - should be represented by one page. That's why (until someone ever will find an alternative version of song with really different lyrics) I'd name this song as "Yes We Did (alternative version)".
lol, I know, you wasn't asking about that, I know :). Coming down to earth - what do you think about 'dogmatic' "(We Are) Performance"?
--Senvaikis (talk) 13:00, August 29, 2010 (UTC)
I'm currently waiting if Wikia is able to reset Botanic's password, because I cannot find it anywhere on my PC anymore... Could you touch the pages in Category:Unlicensed Lyrics for me please? - Chris 12:49, August 29, 2010 (UTC)
OK, but only 2-3h later. Now - Turkey, basketball: Lithuania-Canada, and nothing else egsists! Keep fingers crossed for Lt! --Senvaikis (talk) 13:00, August 29, 2010 (UTC)
Oh, you haven't told about 18,432 cat-members... But - Lt:Ca=70:68, and Nothing Else Matters!. Job's runnin'.--Senvaikis (talk) 15:36, August 29, 2010 (UTC)
DoneGreen. --Senvaikis (talk) 16:57, August 29, 2010 (UTC)
Oh boy... I once again wrote a comment and did not save it but left it open in another tab. This is what I wanted to write some hours ago:
Fingers are crossed for Lithuania! Smile
  • what do you think about 'dogmatic' "(We Are) Performance"?
Agreed! Grin
And now: thanks for touching them. - Chris 17:37, August 29, 2010 (UTC) Yeehaw it worked!

Disambiguations

How come that songs by June (FR) were moved to titles that start with "June (FR):"? I'm pretty sure the rules are exactly the other way round: only add the country code if there's an identically named page by the artist without countrycode. Maybe you confused the scripts "Change artist" (move pages) with "Update Artist"? Wink - Chris 17:37, August 29, 2010 (UTC)

Something wrong with June (Disambiguation) or June? --Senvaikis (talk) 17:48, August 29, 2010 (UTC)
Seems I've got what you wanted to say - songs of both June's may have preffix June, if their titles are different? (i'm still slow-thinking after basketball(& beer) :) --Senvaikis (talk) 17:57, August 29, 2010 (UTC)
I positivelly should reduce count of mugs per game - must confess, I still don't understand how may song, performed by June (FR), have a preffix "June:", even if June has never performed this song. Maximum what Lwt was going to do with all June (FR)'s songs was intention to batch-change their sh.artist to 'June (FR)' and add alias 'June'. And I still don't understand what rules where broken here.
Please tell me, what advantage do you see in your approach, when song, performed by LILITH (BOL), is titled as Lilith:Under The Lilith's Spell. Open this song page and click on the 'performed by' link - oops, - and suddenly you find yourself in the page ot Lilith, but from Poland!
Anyway, seems it's time to go to bed for me,- nite, --Senvaikis (talk) 20:19, August 29, 2010 (UTC)
  • songs of both June's may have preffix June, if their titles are different
Yes, that was what I intended to say. A song by June (FR) should have the prefix "June" for several reasons:
  1. I still don't understand how may song, performed by June (FR), have a preffix "June:", even if June has never performed this song.
    That is true and wrong at the same time. It's the linking which makes the prefix seemingly wrong. But, as a matter of fact, the artist who sung the song is named "June", not "June (FR)". The name "June (FR)" is a compromise that LyricWiki came up with to solve a naming conflict between two artists with exactly the same name. As long as it's doable, a less-than-ideal should be avoided. Why should we make a compromise if we can have all songs at their usual pagenames? It is true, that the songs are not by June, but they are by June, not by June (FR). I think you get what I mean.
  2. If you search for a certain song by Muse here in LyricWiki, you expect it to be at "Muse:Song", don't you? Now where would someone expect a song by June?
  3. MediaWiki makes up the description and the keywords by the the pagetitle. The keywords for June (FR):Innocence: Innocence lyrics,June (FR) Innocence lyrics,Innocence by June (FR) lyrics,lyrics,LyricWiki,lyricwiki,June (FR):Innocence,June (FR):Innocence,June (FR):Innocence,June. The description and keywords are what Google and other search enginges use.
This is why I think the rule is absolutely sensible. About sh.artist: We've got the possibility to link to another page than the displaytext; we can use it where necessary and sensible. If a link to an album was wrong, the link had to be changed, not the pagetitle of the page that links to the album.
But tomorrow I've got to go to school again. My vacations are over and I won't be here as often now as in the past weeks, at least not in the morning. However I guess you (meaning all of LW) won't notice that too much, because during the last weeks I used to be still asleep at that time - ergo: no difference. Good night, Chris 22:38, August 29, 2010 (UTC)
May you take a look at June(fr) now and tell, if current layout correspond to your guidelines? Some things are still not quite evident for me. First of all - why then most of disambiguated pages don't correspond to this layout? And how then should be formated non-latin-titled pages with romanised part in the title? tia, --Senvaikis (talk) 14:06, August 30, 2010 (UTC)
If everything is like with Un peu de temps, then everything is like it should, yes.
Romanized is a bit different: the parts are inserted to keep it readable for those who cannot read Chinese, Cyrillic, etc. There the romanization in the title again helps to increase the search engine exposure, because the romanization is in the keywords again. Furthermore, the romanization can be entered as search term and find the pages then. - Chris 15:46, August 31, 2010 (UTC)
Thanks for clarification, seems now I've got, and agree about romanization. Still, what about bolded question part, (or particularly - Liliths, managed by yourself as anexample for some user)? --Senvaikis (talk) 15:57, August 31, 2010 (UTC)
Oh yes, I forgot about that part. Well, the disambiguated pages are only created if the name was already taken by another artist here in LyricWiki: Let us assume a user wants to create the page for a French artist named "A Band" and finds out that the page is occupied by another artist/band. (S)He will disambiguate the name then and create a page at "A Band (FR)". Our Wiki software is not able to detect such cases and pre-creates the content with items like this: # '''[[A Band (FR):SONG TITLE|SONG TITLE]]'''. This structure is used by the creator to fill in the song- and album-links. Then (s)he saves the page and clicks on the redlinks to create one songpage after another. The result is a complete disambig-coded discography. (Side note: this is also a common reason how not LW:PN conform titled pages are created: the album was created with the wrong links and someone uses theses links to create the pages afterwards).
With all the "Lilith"s it's a very special and tricky case, because there are so many bands: "Lilith" from Poznań, Poland, "LILITH" from Radom, Poland, another "LILITH" from Bolivia, and again another "Lilith" from Spain. It is true that I helped with the pages, but that work was more about moving and splitting the mixed information to the correct pages, as well as help with a split album. Look at the corresponding section on my talkpage. - Chris 16:53, August 31, 2010 (UTC)
All that's clear, but the question is: what part of disambigued sets are organized according to current June layout? Don't you think, that some clarification should be done in lw:help, if most of these sets don't satisfy your guidelines? --Senvaikis (talk) 17:19, August 31, 2010 (UTC)
Sorry, but could you rephrase your question "what part of disambigued sets are organized according to current June layout?" please? I don't get what you mean...
Do you mean, I should write a coherent text that explains the complete rule so we can publish it? Unsure
It really bothers me that I cannot find where I read all these "rules". I'm pretty sure I didn't think out them on my own; I either asked someone, saw any summary of edits/mass movements, or read it anywhere in the help. Maybe I saw a readly disambiguated pageset, or maybe even Sean himself told me how to do it in IRC one and a half year ago? Sceptical hmm...
I'll search around a bit more, until now I only found the "how to disambiguate the artist page" part on LW:PN... - Chris 18:19, August 31, 2010 (UTC)
Yep... :) Yes, I just wanted to state that even if these guidelines/rules exist, they are likely unknown for most users. It's suffice to look at our (two admins! ;)) discussion to come to a conclusion, that some additional clarification should be published (or moved into more noticeable place). In the meantime - thanks for instructing at least one heavy-minded user :). nite, --Senvaikis (talk) 19:33, August 31, 2010 (UTC)

how

how u can create badges? if u can --just me sistemx 18:35, September 5, 2010 (UTC) i mean how achievemnts can be created

Sorry, Sistemx, but I never used to create any badges, so you'd better look for someone else help. I'd ask Chris, f.e. ;). --Senvaikis (talk) 18:45, September 5, 2010 (UTC)

Parodies

I asked Trainman. he said parodies are fine, -Mace170 01:31, September 9, 2010 (UTC)

Yes, song parodies are fine indeed, but I was speaking about other kind of parodies - editing parodies, with a slite hope that my joke will turn your attention to the fact, that all your editions break allmost all possible pagenaming/editing rules and require additional fixing, without any sign of progress. If you don't like reading LW:Help, try at least learning on examples, studying revisions, made on your 'parodies', to avoid the same errors in the future. Sorry if my joke wasn't evident for you. And please, - sign your messages with "~~~~".
--Senvaikis (talk) 05:09, September 9, 2010 (UTC)

Strange 'Duplicates'

Lwt just has finished cleaning duplicated images. All dups are cleaned atm, but several odd images - some mistreated by api as dups, some - non-existing at all.

Strange images, treated by API as duplicates
Image Duplicate Notes
File:Chicago - Stone Of Sisyphus.jpg File:Zac_efron_in_2008.jpg Image info page doesn't show duplicate
File:Jello Biafra & Mojo Nixon With The Toadliquors - Prairie Home Invasion.jpg [[:File::ACDC_COVER_BANDA_FOUR_BROTHERS_.wmv]] Image info page doesn't show duplicate
File:Weezer - ‎The Lion And The Witch.jpg [[:File::ACDC_COVER_BANDA_FOUR_BROTHERS_.wmv]] Image title starts from hex '00 0E'
File:Weezer - ‎Weezer - The Red Album.jpg [[:File::ACDC_COVER_BANDA_FOUR_BROTHERS_.wmv]] Image title starts from hex '00 0E'
File:‎Mandragora Scream - Madhouse.jpg [[:File::ACDC_COVER_BANDA_FOUR_BROTHERS_.wmv]] Image title starts from hex '00 0E'
File:‎The Academy Is... - The Academy EP.jpg [[:File::ACDC_COVER_BANDA_FOUR_BROTHERS_.wmv]] Image title starts from hex '00 0E'
File:‎ The All-American Rejects - The All-American Rejects.jpg [[:File::ACDC_COVER_BANDA_FOUR_BROTHERS_.wmv]] Image info page doesn't show duplicate; starts from hex '00 0E'
File:倖田來未 (Koda Kumi) - MOON.jpg File:‎Cover_MOON_(CD).jpg Image info page shows duplicate as existing image; Dup title starts from hex '00 0E'

--Senvaikis (talk) 10:26, September 13, 2010 (UTC)

Btw, there are at least 70 web-unavailable files like :ACDC_COVER_BANDA_FOUR_BROTHERS_.wmv, which all are 'uncategorized images' according to API. --Senvaikis (talk) 13:10, September 13, 2010 (UTC)

Spam Filter

Hey, I'm getting a false positive on the spam filter when attempting to edit [[Heaven_Shall_Burn:R%C2%ADsandi_Von_(Outro)]], due to the bad characters in the page name. It's a duplicate of Heaven Shall Burn:Rìsandi Von (Outro). --LYRIC-DarthNemesis 16:38, September 13, 2010 (UTC)

Editing (and therefore redirecting) wasn't possible, but moving was, thus leaving a redirect behind. — 6×9 (Talk) 17:47, September 13, 2010 (UTC)
Thx, 6. --Senvaikis (talk) 17:49, September 13, 2010 (UTC)

amg 2 akuma (&2iT)

Not all artists with an allmusic page also have an akuma page, despite existing ID (like this example). Apparently, making things simple for us is pretty far down on akuma's priority list :-) — 6×9 (Talk) 19:14, September 19, 2010 (UTC)

Yes, agree with both your remarks, especially with second one, just my intentions weren't so simple, as may seem at first glance :).
Akumas filling is just subsidiary, preparatory stage of more important job - remarkable update of artists iTunes. I discovered that iTunes also may be searched by AMGID. That's why now I'm running through all artists in Category:Allmusic/Artist, getting their AMGID. Then iTunes of all these artists could be rechecked using this amgId-iT relation. Akuma's filling was made just for one simple reason - there's one-to-one relation between amgId & akuma. The fact, that some miserly part of them are pageless, is unessential imo - anyway, if this artist will ever have his page @akuma (that'll happen when amg page itself gets some content), - this link will be working. Any other akuma-search will never give you some different, more accurate akuma value ;).
nite,--Senvaikis (talk) 21:41, September 19, 2010 (UTC)
DoneGreen, ~4650 artists iT-uned (en route - 180 wrong amgs detected). Let me know if you notice some errors. Cheers, --Senvaikis (talk) 08:05, September 20, 2010 (UTC)
Those I checked were all OK. Only oddity is Colosseum II – instead of the usual artist page I see the "We are unable to find iTunes on your computer" message. ID is the same as I get with their link maker though, so it's no mistake at our end. — 6×9 (Talk) 13:24, September 20, 2010 (UTC)
What a pity! Is that possible, that while akuma, iTunes (and many others, I suppose) are amgID-related and searchable by amgID, it can't be used as search parameter in it's own native DB - allmusic.com (e.g. it's impossible to get amg link, knowing only amgID). Just can't believe, - I'm positvely missing something here. Any thoughts? --Senvaikis (talk) 16:56, September 20, 2010 (UTC)
Well… not quite impossible ;-)
Actually, I couldn't quite believe that it's impossible to search AMG directly, so I just stupidly incremented the opt1 parameter until I hit Gold. Sometimes, brute force is the way to go :-) — 6×9 (Talk) 17:40, September 20, 2010 (UTC)
Well Done! :), thanks!, --Senvaikis (talk) 18:39, September 20, 2010 (UTC)

bot syntax explanation?

just was working on a page and what i assume was your bot came through and capitalized. cool stuff. but it left this cryptic edit message: "frmt, lang: English (Gc:64%)". any explanation of what that means? especially the gc:64%, i was curious about it. thanks, --Betarepeating 05:44, September 21, 2010 (UTC)

Hi,
usually these revision notes are more usefull for Lwt (strictly speaking, - not a bot, more 'cyborg') or me, but surelly you should know their meaning also :)
  • Starting part just says that lyrics was formated (usually that means sentence-case capitalizing, some accents replacing, html-decoding etc.)
  • Second part is self-explained, imo: English language of the lyrics was detected and according SongFooter attribute set to the new value (|language=English).
  • Final part is just additional information about reliability of language detection. Lwt uses Google language detection api for that purpose. One of parameters this api returns together with the detected language, is "Google Confidence" - value from 0 to 1, proportional to the detection reliability. So in our case language was detected with .64, or 64% reliability.
hth, --Senvaikis (talk) 06:43, September 21, 2010 (UTC)
thanks! --Betarepeating 16:45, September 21, 2010 (UTC)

TimeOut

Timeout

--Senvaikis (talk) 17:01, September 22, 2010 (UTC)
Lwt informs about going on two-weeks strike. --Senvaikis (talk) 21:12, September 23, 2010 (UTC)

Serbo-Croatian --> Bosnian

Why you are change Indexi:Svet u kome zivim, from Serbo-Croatian to Bosnian. There is no Bosnian in this time, and for sure there wasn't in that time. Gojkok 13:53, 2010 September 26

My change was caused by very simple reason: language Serbo-Croatian is still not included into LW languages category, so it was displayed as non-existing language (red link). Group Indexi is from Bosnia and Herzegovina according to Indexi page, so I looked at wp and found a Bosnian language as the main official language in that country. And if you noticed, this language, contrary to Serbo-Croatian, is displayed as blue-link, e.g. existing language. So, wp doesn't know Serbo-Croatian language, Lw doesn't too, but both do know Bosnian language - that's why I decided that this information is sufficient for a change I made. That doen't mean that this change can't be reverted - if you are from Bosnia and Herzegovina, you may know better :).
One more note: please, sign your messages by "~~~~"; - this time I've done that for you.
--Senvaikis (talk) 14:43, September 26, 2010 (UTC)

Language

I'm really sorry, I always forget to specify the language when it comes about the lyrics. Thank you for modifying my posts, I'll try to pay more attention in the future. -188.26.198.204, 12:35, 2010 September 28 (UTC)

That's ok - You are always welcome :). Wouldn't you care for creating an account? --Senvaikis (talk) 13:56, September 28, 2010 (UTC)
Community content is available under Copyright unless otherwise noted.