User talk:Tpt

From Wikisource
Jump to navigation Jump to search
Grollier's Reading Wheel.jpg

Hello Tpt, welcome to the multilingual Wikisource! Thanks for your interest in the project; we hope you'll enjoy the community and your work here.

This wiki is the original Wikisource wiki, originally hosting works in many languages. The larger collections have been spawned into separate projects, leaving this wiki to serve as a central collaboration point, and as an environment where works without a language subdomain can be started. Refer to our languages list to see which languages still reside on this wiki. You can find a list of the separate language projects on the main page or here and you may want to look at the our coordination page for limitations on placing certain works on the separate language projects.

Most questions and discussions about the community are in the Scriptorium.

The Community Portal lists tasks you can help with if you wish. If you have any questions, feel free to contact me on my talk page!

Zyephyrus 23:12, 23 March 2011 (UTC)

Agréable surprise de te voir arriver ici :)

Admin[edit]

mop : here is the mop.

seau : trouvé ceci en cherchant « seau » sur Commons...

...may be this one is more adequate: seau

Congrats :) --Zyephyrus (talk) 20:12, 22 March 2012 (UTC)

Thanks ! Tpt (talk) 10:06, 24 March 2012 (UTC)

Default header[edit]

Can you put the links on my talk page to the protected page where portal needs to be added?BirgitteSB (talk) 20:40, 14 July 2012 (UTC)

Wikisource:Scriptorium#Needed:_Wikisource:Metadata[edit]

You may be interested in this discussion. I would also love to hear what is the profess with Dublin Core integration? --Piotrus (talk) 00:14, 14 September 2012 (UTC)

Wikidata related[edit]

Hi,

On the test-page http://wikisource-dev.wmflabs.org/w/index.php/Index:Picturesque_New_Guinea.djvu , I notice the thumbnail (and assume all other subsequent useful embedded tidbits) generates the error message:

Error creating thumbnail: Unable to fetch XML for DjVu file

I tried a long time ago to get someone to verify a.) the latest DjVuLibre package is indeed currently installed and b.) if it is, to verify the installed folder paths match as needed by the XML related .exe files.

You see for example, a typicall local default Windows install creates a folder-path containing the needed .dtd file(s) in...

  • ..\DjVuLibre\share\djvu\pubtext\DjVuXML-s.dtd

. . . while the at the same time the executable, DjVutoXML.exe, expects them to be installed in...

  • ..\DjVuLibre\pubtext\DjVuXML-s.dtd

. . . so if you don't create copies of all the .dtd files at the 2nd folder path, an XML will never be generated; thus no "easy" metadata or mapping and so on via XML.

If the same holds true for server installs, then...

  • /usr/local/share/djvu/pubtext/DjVuXML-s.dtd

. . . on our server install probably should be...

  • /usr/local/pubtext/DjVuXML-s.dtd

... instead.

At any rate, I thought I'd better mention it after seeing the error since I never did get anybody to verify my theory years ago. Even more, I recall most of the XML coding was bypassed early on for straight text dumps because it was truely "broken" in the DjVuLibre package at that time - this is no longer the case in the latest build & that is why verifying the latest build is indeed on the server is also helpful regardless of the folder path thing above.

Please respond on my en.WS talk page if you need to. -- George Orwell III (talk) 08:35, 22 September 2012 (UTC)

I think that this bug doesn't append in labs, but there is a bug with instantCommons (the feature that allow not-Wikimedia wiki to use Wikimedia Commons) that break djvu metadata extraction. I've just submitted a patch that solve, I hope, this bug. Tpt (talk) 14:15, 22 September 2012 (UTC)
OK - you would know better than I would.
I was refering to / concerned about these starting points in the code however...
Not that I know what it all means but it seems at one point the XML part was left behind because it wasn't working. That's been fixed & why I wish to know if we have the latest build installed. If it still isn't working - my bet it is because of the folder path mis-match ( Line 296 ). -- George Orwell III (talk) 16:01, 22 September 2012 (UTC)

Status Bar goes html[edit]

Hi,

I noticed you're making some needed changes to the PR extension, some of which will be affecting the table generated for the color coded status bar. The issue, however, is that you folks are still retaining the deprecated syntax in the remake.

Things like border=0, cellpadding=0, cellspacing=0 & align=center are all outdated not to mention non-compliant under HTML5. I understand why width=$1 must be kept (for the time being) but we could do without the rest can't we?

Before I saw the commits today, I had come up with...

		$output = "
<table class='pr-quality' style='border-collapse:separate; border-spacing:0; border-width:1px; border-style:outset; line-height:40%; margin: -0.45em 0.00em 0.00em 0.01em;'>
<tr style='text-align:center;'>
<td class='qualityf' style='padding:0; width:0;'>&#x3000;</td>
<td class='quality4' style='padding:0;' width=\"$q4\"></td>
<td class='quality3' style='padding:0;' width=\"$q3\"></td>
<td class='quality2' style='padding:0;' width=\"$q2\"></td>
<td class='quality1' style='padding:0;' width=\"$q1\"></td>
<td class='quality0' style='padding:0;' width=\"$q0\"></td>
$void_cell
</tr>
</table>";

... and the $void_cell would just be something like....

		"<td class='qualitye' style='padding:0;' width=\"{$qe}\"></td>"

.... producing {basically)

 

Doesn't have to be exactly the same, but would you please consider making some of the same changes to lower the amount of stuff marked problematic by W3's validation tool at least? TIA. George Orwell III (talk) 19:25, 7 May 2013 (UTC)

Thanks for the report! I'll have a look to it. But, I think you should rather reports bugs on bugzilla that allows a better traking of issues. Can you open a bug here please? I'm automatically mailed at the opening of a new bug, so am aware of the requests here as if they were done on my talk page. Tpt (talk) 20:11, 7 May 2013 (UTC)

1.23wmf10 question[edit]

In the new.....

includes/page/ProofreadPageDifferenceEngine.php


... Line 67, isn't there an extra letter "p" in 'pproofreadpage_page_status' ?


I'm pretty sure that is why the difference heading for status changes looks like this [1] instead of this: <proofreadpage_page_status> . -- George Orwell III (talk) 04:39, 10 January 2014 (UTC)

Thanks! I've just fixed this issue in the master branch. I'll try to make it backported to the version that will be deployed on Wikisources Tuesday. Tpt (talk) 12:05, 10 January 2014 (UTC)
Very good :)

Please also see my very last outdent (Jan. 11) over on the en.WS Scriptorium. -- George Orwell III (talk) 00:05, 11 January 2014 (UTC)

FYI... "my very last" linked above on Jan. 11 was made into a formal bug # 61220 . -- George Orwell III (talk) 02:24, 15 February 2014 (UTC)

plwikisource import[edit]

Thanks for the imported pages. Please note, that I did not ask for import all pages, but only the listed ones. My intention is to split the multilingual index into three parts: Polish will remain in plwikisource, Wymysorys to be moved here and German to be moved to dewikisource (still in TODO). Ankry (talk) 22:41, 14 March 2014 (UTC)

I have made a steward request. Ankry (talk) 22:55, 15 March 2014 (UTC)

window[edit]

What is this for? Helder 00:21, 30 August 2014 (UTC)

Help with harassment by cs.wikisource admins[edit]

Hi, I see you are admin of this wiki. On the local Village pump there's written: "This is the central location for discussing issues with Wikisource, and the place to ask for help." So I posted a plea for help there but noone replied. Would you please have a look? I'm being harassed by cs.wikisource admins, I was banned indef but I never did anything wrong on the wiki at all. Thanks in advance. --Auvajs (talk) 17:40, 20 January 2016 (UTC)

wsExport[edit]

Bonjour, j'ai une petite question qui concerne une différence observée avec la version francophone et qui se retrouve dans les documents exportés.

Ici, dans l'espace principal, par exemple ce texte, dans le code HTML, on a la div ayant pour identifiant mw-parser-output, à l'intérieur de laquelle il y a, juste après la div ayant pour id headertemplate, une autre div n'ayant pas d'identifiant mais dont l'attribut style a pour valeurs width:36em; margin: 0px auto; text-align: justify; text-indent: 2em; (le gras est la partie problématique). Si j'exporte mon document EPUB, tant que la largeur de mon lecteur EBUB est >= à 36em tout va bien, mais lorsqu'elle est inférieure, les éléments contenus dans un bloc ayant pour propriété CSS width:100% (relative aux 36em, donc), comme le modèle ct par défaut, ne vont s'afficher que partiellement (Voir image).

Ce problème n'existe pas sur la version francophone (je n'ai pas testé les autres) parce que le conteneur n'est pas limité à 36em. On remarque également sur l'image d'autres problèmes, tels que l'apparition des numéros de page, ou encore l'indentation des paragraphes générés avec <poem>, particulièrement embêtant lorsqu'un paragraphe s'étend sur plusieurs pages.

Penses-tu pouvoir résoudre tous ces petits problèmes ? :-s Reptilien.19831209BE1 (talk) 06:33, 7 May 2018 (UTC)

C'est corrigé pour le problème des 36em. J'ai remplacé width:36em; par max-width:36em;. C'est corrigé aussi pour les numéros de pages. Pour les poèmes il faut que je regarde cela. Tpt (talk) 08:41, 7 May 2018 (UTC)
Me voilà embêté parce que je n'ai plus de contenu dans le document exporté. J'ai bien la page de titre, ainsi que la page à propos, mais aucune donnée provenant de la transclusion. J'ai essayé avec un autre texte qui n'est pas de moi et qui utilise également <pages />, idem. Par contre les autres textes semblent fonctionner correctement. Reptilien.19831209BE1 (talk) 13:10, 7 May 2018 (UTC)
C'est réparé. Désolé pour le problème. Tpt (talk) 12:39, 8 May 2018 (UTC)
Merci, je confirme que ça fonctionne. Désolé à mon tour, j'avais entre-temps renommé le livre ce qui avait ajouté un problème de redirection. Reptilien.19831209BE1 (talk) 13:47, 8 May 2018 (UTC)

Traductions libres et ProofreadPage[edit]

Bonjour, je suis confronté à un petit problème technique et j'aurais voulu avoir tes bons conseils sur la meilleure marche à suivre.

Le texte Index:Émile Gérard - On diner à conte-cour, 1885 (in BSLLW s.2-t.7 p.263-265).djvu est écrit en wallon liégeois dans un vieux système orthographique (avant Feller 1900), et on aimerait avoir la possibilité de le retranscrire dans d'autres systèmes orthographiques (Feller ou Rfondou) en vis-à-vis du document original, dans l'espace page, ce qui est beaucoup plus commode. Comment faudrait-il procéder ? J'ai pensé à ce qui suit : en considérant que la sous-page Page:Émile Gérard - On diner à conte-cour, 1885 (in BSLLW s.2-t.7 p.263-265).djvu/Rifondou/1 fait apparaître l'image de la page 1 du FS, j'ai envisagé qu'en créant un sous-index Index:Émile Gérard - On diner à conte-cour, 1885 (in BSLLW s.2-t.7 p.263-265).djvu/Rifondou on pourrait profiter des fonctionnalités de <pages/>, comme j'ai tenté de le faire sur la version francophone, et transclure aisément sans faire appel à de multiples {{:Page:...}}. Mais ça ne fonctionne pas comme je l'espérais. Aurais-tu une idée ? Ceci concerne presque tous les textes en wallon. D'où la nécessité, selon nous, d'avoir un système plus pratique qu'une traduction faite directement dans l'espace principal. Reptilien.19831209BE1 (talk) 09:29, 10 May 2018 (UTC)

Hum, le fait que Page:Émile Gérard - On diner à conte-cour, 1885 (in BSLLW s.2-t.7 p.263-265).djvu/Rifondou/1 fonctionne pour l'affichage du fac-similé n'est pas une fonctionnalité "attendu" de ProofreadPage. C'est pour cela que le reste ne marche pas (navigation entre pages, <pages/>...). Je n'ai pas très envie de modifier ProofreadPage pour faire marcher tout le reste car ce mode de fonctionnement ressemble un peu à un hack et si on l'utilise il faudra que ProofreadPage fonctionne correctement pour ce cas là pour X années à venir. Je recommandait donc de soit ne pas faire la conversion dans l'espace page, soit avoir dans chaque page deux <section>, l'une pour le texte original et l'autre pour la transcription et ensuite d'utiliser pour la transclusion le paramètre onlysection de la balise <pages/> (il fonctionne comme fromsection mais s'applique à toutes les pages). Tpt (talk) 16:59, 11 May 2018 (UTC)