User talk:Tpt

Berlin hackathon invitation

I invite you to the yearly Berlin hackathon, 1-3 June in Berlin, Germany. Registration is now open. If you need financial assistance or help with visa or hotel, then please register by May 1st and mention it in the registration form to request subsidy.

This is the premier event for the MediaWiki and Wikimedia technical community. We'll be hacking, designing, teaching, and socialising, primarily talking about ResourceLoader and Gadgets (extending functionality with JavaScript), the upcoming launch of Lua for templates, Wikidata, and Wikimedia Labs.

We want to bring 100-150 people together, including lots of people who have not attended such events before. User scripts, gadgets, API use, Toolserver, Wikimedia Labs, mobile, structured data, templates -- if you are into any of these things, we want you to come!

I also thought you might want to know about other upcoming events where you can learn more about MediaWiki customization and development, how to best use the web API for bots, and various upcoming features and changes. We'd love to have power users, bot maintainers and writers, and template makers at these events so we can all learn from each other and chat about what needs doing.

Check out the the developers' days preceding Wikimania in July in Washington, DC and our other events.

Best wishes! - Sumana Harihareswara, Wikimedia Foundation's Volunteer Development Coordinator. Please reply on my talk page at mediawiki.org. Sumana Harihareswara, Wikimedia Foundation Volunteer Development Coordinator (talk) 01:50, 20 April 2012 (UTC)

GSoC / OPW IRC AllHands this week

Hi, you are invited to the GSoC / OPW IRC AllHands meeting on Wednesday, June 26, 2013 at 15:00 UTC (8:30pm IST, 8am PDT). We have done our best finding a time that works decently in as many timezones as possibles. Please confirm at qgilwikimedia.org so I can add you to the calendar invitation and I have your preferred email for other occasions. If you can't make it's fine, but let me know as well. We have left a notice to all students, but your help is welcome making sure yours is aware about it. Thank you!--Qgil (talk) 18:08, 24 June 2013 (UTC)

Status of your semantic genealogy extension

Dear Thomas, are you still working on your semantic genealogy extension? Looks like this is what I was looking for, but I don't want to use an experimental extension in my wiki. Do you know when this extension will be in beta status or even stable? Many thanks! --84.154.114.148 19:53, 25 October 2013 (UTC)

This extension is in a very experimental status so I believe that you shouldn't use it in your wiki. Tpt (talk) 06:23, 26 October 2013 (UTC)

Hi Thomas, are there some written lines about the usage of this extension? If you say very experimental, what could it do in the worst case? - Greets from Hunsrück, Andy

Hi! No, there isn't. I've abandoned the development of this extension before having a working product. So, this extension does nothing currently. Tpt (talk) 16:30, 14 February 2014 (UTC)

Hi, thanks for that fast reply. Are there other possibilities to create a familytree on my wiki? I found a template:chart on wikipedia :-) But no way to get it on my wiki * Andy

I think It's possible (and pretty easy) to reuse Wikipedia templates if you have required extensions (ParserFunctions and maybe Scribunto) installed. If you don't want to use Scribunto, you should have a look to the French version of the chart template that is basically the same template but without Scribunto dependances. Tpt (talk) 17:12, 14 February 2014 (UTC)

O.k. That sounds good. Maybe generally not so bad to know about how to use or reuse templates :-) For now thank you very much. It will take some time till I have space to try * So long & L.L.R.R. - Andy

Google Books > Internet Archive > Commons upload cycle

Hello Tpt, GSoC begins soon. Yann volunteered to help with this project, would you be available as co-mentor for the MediaWiki (and JavaScript?) side of things? I'm not even sure if that would be part of a MediaWiki extension or just a tool on Labs, but I think your technical advice would be useful to make decisions such as this, if applications from interested users arrive. :) --Nemo 18:15, 20 February 2014 (UTC)

Added you

Hi Tpt,

I added you to the new list of useful or interesting things that volunteer devs have done at How to become a MediaWiki hacker/Volunteer achievements. I hope that's okay. If you can expand or correct this new list, please help me. Whatamidoing (WMF) (talk) 23:41, 19 June 2014 (UTC)

Urgent request for new sidebar: for Commons links, please link to *both* gallery *and* category if possible

Urgent request: for Commons, please link to *both* gallery *and* category if possible

This can (presumably?) be achieved by using Commons category (P373) and Commons gallery (P935)

Otherwise people are getting concerned that "this mean we are going to need to create hundreds of thousands of basically redundant galleries replicating categories just so that links will work? That seems insane." -- from thread at Commons:Village_pump#Interwiki_links_on_categories

(Posted on the talk page, but cross-posting here in the hope that you will see it).

Thanks, ~~~~

IA upload tool is not working for me again

This is the message I got from IA upload robot. I was on the IA website and logged in to IA as ineuw, and I am also automatically logged into the commons as well. I thought that I should let you know.

Internal error

Message/link copied from top of the error page Ineuw talk 16:39, 28 November 2014 (UTC)

ia-upload is down?

Hi, I'm trying to upload from Archive.org to Commons with your great tool ia-upload but it's down right now. It is something temporarily or permenent(I hope not)? Thanks! --Dvdgmz (talk) 12:38, 29 January 2015 (UTC)

I've restarted the webserver. Thank you for te report. Tpt (talk) 13:03, 29 January 2015 (UTC)

Now works fine. Thanks a lot! --Dvdgmz (talk) 16:05, 29 January 2015 (UTC)

Hi, I'm afraid it is down again. Sorry. --Dvdgmz (talk) 18:20, 29 January 2015 (UTC)

It looks to work. Tpt (talk) 06:54, 30 January 2015 (UTC)

Changements de style de police, la taille et la hauteur de la ligne?

Je cherche des informations sur les changements de style de police, la taille et la hauteur de la ligne, des dernières mises en œuvre sur Wikisource anglais (Page et Main namespaces). Pourriez-vous s'il vous plaît me dire où je peux voir ces changements? Ineuw talk 02:37, 18 May 2015 (UTC)

S'il vous plaît oublie ça. En attendant, wikisource:User:George Orwell III m'a fourni avec l'information, et j'ai aussi trouvé un Firefox add-on (très utile) qui m'aide avec les calculs. Merci. Ineuw talk 05:42, 19 May 2015 (UTC)

No user in pagequality tag

Hi Tpt, as the developer of the Extension:Proofread Page i'd like to ask you about something stange at dewikisource. With this search i found a lot of pages that have an empty user parameter in the pagequality tag. I do not know why this happened but most pages seem to be rather old. The "little" problem with this is that when make an edit without changes anything the user parameter is set to the user that made this "edit". Although i am not sure what this parameter is used for i'd like to repair the effected pages. What user would be the appropriate one? Thank you, --Aschroet (talk) 15:00, 9 August 2016 (UTC)

Hi! It seems that for these pages the Proofreading level has been set to validated without any user name by ThomasBot (see this diff for example when it was still possible to do anything on the proofreading status using a bot. A simple way to repair the pages would be to make ThomasBot to edit again the pages. But as the "user" parameter is currently only used to disallow the proofreader to validate the page I don't think it's important to bother about this problem. Tpt (talk) 09:11, 17 August 2016 (UTC)

Ok, thank you for the background information. Since it was not very complicated to derive the user that set the status of the page use my bot to modify the parameter accordingly. Just to be clean at this. --Aschroet (talk) 11:38, 18 August 2016 (UTC)

Killed python3 process running on tools-login

Hello! I've killed a python3 process you were running on tools-login since it was taking up too much CPU and making the instance unresponsive for other users. In the future, please run long running processes on the grid. Thank you! YuviPanda (talk) 03:54, 17 December 2016 (UTC)

Arabic Alphabet

Dear Tpt. Can the OCR work on Arabic and Persian Texts? --Yoosef Pooranvary (talk) 09:39, 3 June 2017 (UTC)

@Yoosef Pooranvary: I assume you are talking about the "OCR" button that is used by a lot of Wikisource, aren't you? It uses Tesseract (an open source OCR system) that has models for Arabic (ar) and Persian (fa). We could definitely make sure that our Tesseract version works with Arabic and Farsi and deploy the OCR button on ar and fa Wikisources. If the results are bad there is the option to use the Google OCR that allows a limited number of requests but gives sometime better results. Tpt (talk) 14:36, 3 June 2017 (UTC)

The problem is that I don't see that button on the Fa wikisource! --Yoosef Pooranvary (talk) 15:45, 3 June 2017 (UTC)

Tesseract data seems not there for the Person language (far). @Phe: could you add them to the Tesseract instance running on labs? Tpt (talk) 15:56, 3 June 2017 (UTC)

Visual editor -- quality flag

While proof reading in page namespace in Tamil Wikisource using visual Editor there is no option to change the quality flag from "pink" to "yellow" or "green" or other colours. Hence it is posing difficulty in proof reading. Regards -- Balaji (talk) 19:09, 12 December 2017 (UTC)

Hello! If you click on the proofreading quality banner you should get an "edit" button that opens a selector allowing to pick the quality. Tpt (talk) 21:48, 13 December 2017 (UTC)

@Tpt: there is an edit button on the banner and I can change the colour in the selector but the "done" button stays grey so can't validate the change :/ @Balajijagadesh: for information. Cdlt, VIGNERON (talk) 15:21, 18 December 2017 (UTC)

@VIGNERON: Thanks for the detailed effort. Thanks @Tpt: for the reply. But as Vigneron said I am also unable to save the status after changing it. The issue is there atleast in ta.ws. Thanks -- Balaji (talk) 15:29, 18 December 2017 (UTC)

Hello, Tpt. If you happen to have a Phab task for this, then please feel free to add a link to the top of Topic:U380pzytya6t2m2h (just in case anyone else has the same question in the future). Whatamidoing (WMF) (talk) 23:04, 27 January 2018 (UTC)

MW2SPARQL

I'd like to play with that tool, but none of examples from the page above doesn't seems to be working and [1] return "No webservice". --Ghuron (talk) 08:07, 4 May 2018 (UTC)

Sadly this tool is broken because of a big change in the database replicas on tools labs. I have not yet found a good way to fix it. Sorry. Tpt (talk) 09:13, 4 May 2018 (UTC)

I was working on toolforge on the queries, that combines sparql and categories, but accidentally discovered your tool. It looks like it make sense to fix your tool rather than reinventing the wheel. May be I can help here? --Ghuron (talk) 09:26, 4 May 2018 (UTC)

It would be great! The blocker problem is described here: phabricator:T188506. An solving option would be to do an API call when mw2sparql service is started to retrieve all namespaces and then generate speicfic rules in this file for each namespace. But it is probably going to affect performances. Tpt (talk) 09:32, 4 May 2018 (UTC)

Is there a way how we can chat more conveniently? Some messenger or irc? --Ghuron (talk) 09:46, 4 May 2018 (UTC)

I'm Tpt[m] on IRC. Tpt (talk) 09:47, 4 May 2018 (UTC)

I was able to compile and (somewhat) run it locally, and now I can see why this is so difficult. I cannot just hardcode namespaces in .java or .config, because we need namespace prefixes in mapping.ttl I cannot easily hardcode them in mapping.ttl as well, because ontop sql parser in somewhat limited. So basically we need one rule per namespace per language which seems to be overkill. Let me think if there is another way --Ghuron (talk) 13:24, 7 May 2018 (UTC)

Hello Ghuron. Thank you for having stated to work on it. During your thinking, you should probably also take care of the efficiency of SQL queries, the Wikipedia tables are huge so we need to make sure that the queries generated by ontop hit the good indexes. Tpt (talk) 12:04, 8 May 2018 (UTC)

Can't catch you on IRC :) It looks like there are no convenient way to get namespace prefixes in DB layer. Instead of that we can:

1. Modify .ttl rules, so they will return urls with "fake" prefixes like https://en.wikipedia.org/wiki/ns:14:Articles_needing_additional_references_from_March_2013
2. Add in SPARQLActions.evaluateTupleQuery() results post-processing that will replace fake prefixes with real ones e.g. ns:14: -> Category:

It's not the most elegant solution in my career, but it will return this tool to operational mode. What do you think? --Ghuron (talk) 13:47, 8 May 2018 (UTC)

It looks indeed like a possible good solution. But we need also a first step: preprocess the SPARQL query to do the Category: -> ns:14: conversion if there are some constant URIs in the query. The other possibility I see is to write some code that generates the R2RML code for each namespace. Tpt (talk) 14:22, 8 May 2018 (UTC)

Please review [2] --Ghuron (talk) 16:12, 10 May 2018 (UTC)

Thanks for merging it! Unfortunately it doesn't work very well. Run the first query from MW2SPARQL and you'll see https://en.wikipedia.org/wiki/TemplateAmbox instead of https://en.wikipedia.org/wiki/Template:Ambox I will look into that tomorrow. --Ghuron (talk) 18:10, 14 May 2018 (UTC)

Can I ask you to redeploy application, since I fixed today a few bugs? --Ghuron (talk) 13:55, 16 May 2018 (UTC)

It should be done. Do you have a Wikimedia Tools Labs account? If you have one I could grant you rights on the project and show you how to do the deployment (doing it while we are both on IRC is probably the most efficient way). Tpt (talk) 16:10, 16 May 2018 (UTC)

"Ghuron" is my wm-tools account. Can we do it tomorrow? If yes, what is the preferred timeslot? --Ghuron (talk) 16:25, 16 May 2018 (UTC)

Sure! I'll be in the train between 10am and 4pm CEST. I should have a good network connection at least during the first 3 hours. So, pick anytime in it. Tpt (talk) 17:29, 16 May 2018 (UTC)

WSexport down, again...

Salut !

https://tools.wmflabs.org/wsexport/tool/book.php

Sais-tu ce qui se passe avec les tools ? ils sont très fréquemment en panne, ces temps ci... --Hsarrazin (talk) 09:47, 3 October 2018 (UTC)

C'est redémarré. Assez souvent les tools sont plantés car le système de fichiers partagées entre les serveurs qui est utilisé sature vite quand les tools font trop d'écritures simultanées dessus. C'est bien pénible. Tpt (talk) 16:35, 3 October 2018 (UTC)

Interviewing you before I present on ProofreadPage for a WikiConference workshop on using Wikisource

First of all, thank you very much for maintaining this tool. ProofreadPage has been the goto collaborative transcription and proofreading tool that I've been using for my work for the last decade or so. Later on this month, I'll be facilitating a workshop on using ProofreadPage in Wikisource at the WikiConference North America in Columbus, Ohio. Before I do so, I'd like to ask you a bunch of questions about the history of ProofreadPage, it's maintenance, your vision for it, the coding community around it, and any UI/UX decisions made vis a vis other online collaborative transcription tools (both open and closed source). Could we connect to chat, and if so, what would be your preferred means of communication? My email is aharon at opensiddur dot org. Thank you! Aharonium (talk) 22:54, 9 October 2018 (UTC)

page navigation tabs

Tracked in Phabricator
Task T225749

I don't know why the order of page navigation tabs on Persian Wikisource is so messed up. It's order should exactly be the opposite of the English Wikisource. Would you please tell me how can I change it? --Yoosef Pooranvary (talk) 18:59, 12 June 2019 (UTC)

Hi! Yoosef Pooranvary. I believe you would like to swap the other of the previous page and the next page links? The order is actually exactly the opposite of the one on English Wikisource. The file creating the links is this one. Feel free to submit a change on Gerrit, I would be very happy to review it. Tpt (talk) 10:11, 13 June 2019 (UTC)

Since I'm not a pro like you, I decided to use Photoshop instead!

This is how the tabs should look like. I'd be so thankful if you change the order like the above picture. --Yoosef Pooranvary (talk) 15:46, 13 June 2019 (UTC)

Thank you! I have created a task about it and a change that implements it. Tpt (talk) 19:12, 13 June 2019 (UTC)

Thanks a million! I noticed the change. You're great.--Yoosef Pooranvary (talk) 11:13, 27 June 2019 (UTC)

needed info on index pages

Hi. I need below variables to make a bunch of templates on Wikisource based on index pages. What I mean is that when I include an index page name as a parameter, it sould return these variables:

• Total number of pages
• Validated pages
• Proodread pages
• Problematic pages
• Notproofread pages
• Textless pages

--Yoosef Pooranvary (talk) 07:40, 19 July 2019 (UTC)

Hey! At my knowledge, it's not possible for now to access these data from templates. I plan to provide a Lua API for that, but it's not done yet. If you are familiar with Scribunto internals, feel free to try a first version of it. Tpt (talk) 08:38, 19 July 2019 (UTC)

Adding as a co-maintainer

I would like to know whether you can add me as a co-maintainer for the tool Phetools. I saw some issues related to the tool and I would like to try whether I can solve it or not. Adithyak1997 (talk) 11:17, 30 November 2019 (UTC)

@Adithyak1997: Thank you for agreeing to step in. phetools is a collection of script from user:Phe without much documentation so getting into the code is quite difficult and I have never managed to do it fully. We could try to make OCR work again but we should take care not to disrupt the other running things. The proper solution to the OCR problem is creating a real OCR service for Wikisource. @Phe: Are you ok with granting access right to Adithyak1997? Tpt (talk) 21:23, 30 November 2019 (UTC)

@Tpt: Thanks for replying. Just for your information: If I am granted the right, I will be trying to do two things. Firstly, restarting the service. This needs to be done since many times Phe was able to solve the issues related to OCR through restarting. Secondly, to check whether the html file is freezing due to the any content related policy issues or not. I can wait until a reply comes from Phe. But I guess he is inactive for around 1 year. Adithyak1997 (talk) 22:20, 1 December 2019 (UTC)

Hello. I have the same request. User:Phe has been inactive for several years. The Match_and_split of https://admin.toolforge.org/tool/phetools worked with errors before. And now it stopped working at all in the Russian Wikisource and I think in others too. As I understand it, this is the script code, which is run from the gadget. A year ago, I left a message on the gadget's TalkPage that it was necessary to optimize the code and remove the incorrect prefix names that depending on the utf-8 encoding; none of the engineers are active on TalkPage either. But the problem is in the back-end code. Apparently the code stopped working due to the fact that it was written in Python v2.7, the support of which ended on 2020-01-01, and one of its problems was precisely that it did not support utf-8 without complex "acrobatics" ... Plus, it uses pywikibot, which has a built-in "site.proofread_page_ns" method with which help optimize it. Also other errors must be fixed. And it is desirable to rewrite it to the current Python v3. Also in the Phe repository there is a error fix pull with no response for a year.
In general, it seems Phe left the project, his scripts have accumulated critical errors and are not supported. But some of them are needed by user from many countries. Can someone maintain the tool and GitHub repository? --Vladis13 (talk) 12:38, 27 May 2021 (UTC)

Hi Vladis13. Sorry for the answer delay. It would be great indeed if someone can take over phetools, maybe splitting the tools into more manageable pieces. phetools have been mostly migrated into python3 but the currently running version has diverged from the github version so it is not reflected in the github repository (only phe as commit access to the github repository). @Xover, Candalua, and Phe: are you ok with giving access to phetools to Vladis13 or does one of you want to have a look at the Match&Split bugs? Tpt (talk) 05:55, 29 May 2021 (UTC)

Is there some Git repository with the newest version? Because I am currently editing code from this GitHub. It have "common" scripts that are in Python 2 there. Updating them to Python 3 will require update of probably all depending phetools scripts. I don't feel like reinventing a wheel that's already debugged. Vladis13 (talk) 07:02, 29 May 2021 (UTC)

Sadly not yet. I just generated a diff file with the changes in the local clone with respect to phe github version. A clean thing to do is maybe fork phe github project and apply the changes there. The major changes there are python 3 migration and the use of svg instead of png for ProofreadPage statistics. Tpt (talk) 07:15, 29 May 2021 (UTC)

@Vladis13 (CC Candalua) I can certainly take a look at the issue. I wasn't aware that there were any major ongoing problems with phetools after we fixed the OCR problem. Match&Split works fine on enWS so it's not a universal problem. Could you give me a description of what fails, what the symptoms are, etc.? Xover (talk) 07:09, 29 May 2021 (UTC)

I found several bugs there, the fixes for which I don't see in this diff, I'm not sure if this should work at all. It will be long and unproductive to copy the code here. In addition, I found a bit of code that can be optimized and formatted according to the PEP8 standard. I also added PDF support; the current version only extracts text from DJVU. I did not yet applied the diff, so will see it some late. I made a fork of repository, and think to add my code edits to there. Vladis13 (talk) 09:02, 29 May 2021 (UTC)

I think if your aim is to rewrite all of it in Python3 and reformat it to some other style guide (which makes tracking changes very challenging) then it would probably be better to set up your own tool based on your fork of Phe's code. After the OCR fixes last year, all indications are that the Phetools are working well and stably for most projects (the Match & Split problem on ruWS is news to me). Starting a massive rewrite in situ, with the inevitable small bugs and downtime along the way, is probably not a good tradeoff for most projects (depending on the severity of the M&S problem for you, that calculation could look different for ruWS, of course). Xover (talk) 06:08, 30 May 2021 (UTC)

I definitely agree that a big rewrite of phetools is maybe not the best idea. If we want to improve them significantly I believe it would be nicer to split phetools into different tools, each one with a specific goal in mind (match&split, OCR, statistics). It would allow each of them to move at their own speed and have their own set of maintainers. Updating an URL in on-wiki JS gadgets is not a big deal. Tpt (talk) 06:12, 30 May 2021 (UTC)

I made an updates this week. See the Pull from my fork here. You can add the conversion to PEP3 in the first commit completely freely, I did it automatically. For other updates. Official support for Python 2 is over, debugging and adding something on it is extremely problematic, it's a matter of time before its support is removed from Pytwikibot and WMFTools. With utf-8 there is of course a terrible trouble, perhaps it work correctly only with the Latin alphabet. In addition, there are a few more fixes. See the description in the Pull. Vladis13 (talk) 14:47, 7 June 2021 (UTC)

Regarding the division into separate instruments. Are you suggesting that we split this whole mass of directories into separate tools? This would be a good option, because when I started to change something in the code there, I had to redo all, since there are a lot of interconnections. Match & Split yes can be freely separated. But the question is different, I spent a week rewriting this code. Does it make sense to invest in it, if this running tangle on outdated Python2, without support by a long-inactive user Phe, and with forbidden access to the script folder on the server and missing `ws_namespaces` files in the repository... are everyone afraid to touch and update? Even if, for example, I highlight this tool and statistics tool, will it replace the current Phetools? Or will everything remain as it is and I will just waste a lot of time and distract you? Vladis13 (talk) 14:47, 7 June 2021 (UTC)

Proofread Page Index attributes

How do I get the attributes from Proofread Page's Index namespace? I need to display them in corresponding Page namespace. --Ans (talk) 20:05, 3 September 2021 (UTC)

Sadly it's not easy. If you have an attribute named "Header" it is automatically used to pre-fill the header of the page during its creation and similarly for the "Footer" attribute. Outside of these (and the CSS stylesheet), you can fetch the content of the Index page using a Lua module (the getContent method of the mw.title object). There is a work in progress Lua API for ProofreadPage that should make it much easier. Tpt (talk) 12:23, 4 September 2021 (UTC)

May we add parser function like {{INDEX:author}}, {{INDEX:title}}, {{INDEX:volume}}, {{PAGENUM}}? It would be much easier to be implemented than Lua API, and also simpler to be used by editor.

--Ans (talk) 02:40, 8 September 2021 (UTC)

We definitely can do that but it is also easy to implement a similar template {{INDEX|author}} in a few line of Lua code. I can write it for you when the Lua API is going to be deployed hopefully very soon. Tpt (talk) 05:45, 8 September 2021 (UTC)

IA-upload is not working?

Hi, I tried uploading 3 files from IA, and this is the result:

An error occurred: Code: badinteger Message: Invalid value "" for integer parameter "filesize". Result: {"error":{"code":"badinteger","info":"Invalid value \"\" for integer parameter \"filesize\".","*":"See https:\/\/commons.wikimedia.org\/w\/api.php for API usage. Subscribe to the mediawiki-api-announce mailing list at <https:\/\/lists.wikimedia.org\/postorius\/lists\/mediawiki-api-announce.lists.wikimedia.org\/> for notice of API deprecations and breaking changes."},"servedby":"mw1382"}

Could you please take a look? Thanks! Bennylin (talk) 18:20, 18 February 2023 (UTC)

Hi Bennylin, sorry for that. I don't see the error in the log. To help me reproduce the problem more easily, which IA files were you trying to upload? And under which name in Commons? 2A01:CB14:AF5:8C00:4AE6:4E46:EAB8:7B8E 18:58, 19 February 2023 (UTC)

Hi ,it was 686222, 686223, and 517479. Bennylin (talk)

Phetools wikisource statistics

Hello, what is wrong with the statistics page for wikisource: https://phetools.toolforge.org/statistics.php ? Can you or some other technician repair it? --Havang(nl) (talk) 21:02, 14 March 2024 (UTC)