Grants talk:IEG/Pronunciation Recording (Finish incomplete GSoC project)

Add topic
From Meta, a Wikimedia project coordination wiki
(Redirected from Grants talk:IEG/Finish Pronunciation Recording)
Latest comment: 6 years ago by Nemo bis in topic Wikimedia France

Takeover proposal[edit]

Today, I sent the following message:

Dear Mr. Flaschen, Hi Rahul,
since the Pronunciation Extension was not fully completed and thus won't be installed by sysadmins on the WMF cluster, I
would like to invite you to share your opinion about the IEG-Proposal to provide the extension as a gadget[1] or even
join that proposal! I have experience with gadgets (I coded and administrated the Wikivoyage logo election voting
system[2] using a script that makes use of modules hosted at Commons[3], for example). The same approach can be used
here to temporarily provide PronunciationRecording as a gadget. This option is also much more flexible than development
in Git/Gerrit and it is closer to the community; changes can be deployed faster. Note that Commons already serves some
scripts[4][5] to other wikis thus proving experience and stability despite all the criticism about Commons and as a
prove of concept.

What do I have in mind with Pronunciation-recorder's i18n?
Well, if the extension (or only its i18n-strings) are installed at Commons, this can be used. Otherwise I will write a
maintenance-JavaScript that fetches the translations from and copies them at Commons (so they can be
used by RL). I know that the latter sounds awkward to a professional but this is nothing you will have to care about.
Once you deploy the extension, the local messages can be deleted by a script at Commons. Nothing really time- or

Why do I think it is a great idea?
Because you will be able to get feedback before something is deployed; issues can be discussed on-wiki what the
Community is used to do all the time and they can be addressed more quickly and local Wiktionary administrators decide
whether they want to install the software and in which fashion.

Thanks for the great work you have already done.

Hoping for your feedback
Kind regards
Rainer Rillke
Community Administrator at Wikimedia Commons

[4] (to Wikidata)
[5] (to Meta)
First, it's really great that you also want to push this code forward. Rahul has also told me he plans to continue working on it. It is true that the full scope of the GSOC has not been completed yet. However, the full scope is not necessarily required to get it on the WMF cluster. The people reviewing for deployment will probably be concerned with code quality and conventions (e.g. bugzilla:54724 and bugzilla:54907), and possibly design to some extent (though it may not be a blocker). Although I do think the Wiktionary integration is important, I don't see why it should be a blocker to getting it installed on Commons (which would already be useful). In summary, deployment of the existing extension may not be that far away.
That brings us to the extension v. gadget issue. I just saw the IEG note, "Extensions or software features requiring code review and integration cannot be funded". I wasn't aware of this before. Although I think I understand the motivation, it's somewhat unfortunate, because it draws a line between employees and the community. It's also important to note that WMF employees are not the only people who have +2 (i.e. the ability to merge). Given that this is already an extension, perhaps it makes sense to make an exception to this rule.
There are various advantages to keeping it as an extension:
  1. It's already one, so people can just keep working without repackaging.
  2. Extensions have internationalization (i18n)/translation support, while gadgets do not (yet). As someone who has first-hand experience developing both extensions and gadgets, I know the annoyance of needing i18n workarounds for gadgets. Rainer's workaround idea (or something similar) would probably work, but it requires spending time on a secondary task.
  3. More flexible technically. For example, an extension has the ability to add its modules only to certain pages (e.g. only the main namespace and Special:PronunciationRecording). A gadget is on for every page, or off (unless the special withJS parameter is used).
  4. We're using a MIT library, so it needs investigation whether that can be legally imported on-wiki (or we could ask the authors to agree to CC-BY-SA 3.0).
As for the upload code/FormData, I was thinking that initially the UploadWizard code could be cleaned up so it was more usable from third-party code. After that, it should be moved to core, so other extensions don't need to depend on UploadWizard.
However, the most important thing is just that everyone who wants to work on the code is able. We don't want to have duplicate work, or for people to step on each other's toes. If Rahul is okay with this approach, and the licensing works out, I can support this as a workaround due to IEG's funding rules (though I would still prefer if they can make an exception). However, I think after the IEG, it should be moved back to the extension (note, that would require a notice that anyone working on the gadget must agree to the extension's license, GPLv2). Superm401 | Talk 03:38, 3 October 2013 (UTC)Reply
Thanks for replying.
As for the licensing question, I believe it is possible to use MIT works under CC-By-SA 3.0 with the short MIT-disclaimer as an attribution requirement for the CC-By-SA-use. This is, BTW, a general issue with on-wiki scripts; Although licenses like GPL would be compatible with MediaWiki's license, they aren't with CC-By-SA. I hope that we will get the possibility using real software licenses for scripts and Lua modules in future suppressing the text that the above content would be licensed under CC-By-SA as scripts are no real "wiki content" but software. -- Rillke (talk) 08:12, 3 October 2013 (UTC)Reply
Additionally, the content-model for js-pages, like gadgets are, is javascript. Hence, it is not text but can be considered as software or a file. Media-Files at Wikimedia Commons can be licensed under various free licenses. Cf. commons:COM:L. I don't see why this should not apply to software files. -- Rillke (talk) 08:45, 9 October 2013 (UTC)Reply

Hello Rainer Rilke,

I am glad that you have showed interest in our extension, but sticking to the proposal I would like to mention that Integrating to Wiktionary is important I feel because, this project idea is basically a feature request . It would not be right on our part to not complete what we have started. I am interested in knowing more about your plans. I am a little busy with my own college work and will be off for a little outing soon, so do give me a detailed description about your idea.--Rahul21 (talk) 20:25, 8 October 2013 (UTC)Reply

Rahul21, thanks for replying here; can you estimate how long it will take to complete your work on the extension? If it is less than one year, I am going to withdraw this proposal as it wouldn't be worth the efforts; if it is more than one year, providing your code+ some amendments (including in-page-integration like your initial mockup and uploading to Commons from within Wiktionary) as a gadget will enable you to get feedback early so you can see what works and what needs improvement or must be re-considered. Communies often have surprising demands for developers and all kind of bugs appear when one executes code client side like we do here so it would be a good trial for low cost. -- Rillke (talk) 08:45, 9 October 2013 (UTC)Reply

Eligibility confirmed, round 2 2013[edit]

This Individual Engagement Grant proposal is under review!

We've confirmed your proposal is eligible for round 2 review. Please feel free to ask questions here on the talk page and make changes to your proposal as discussions continue during this community comments period.

The committee's formal review for round 2 begins on 23 October 2013, and grants will be announced in December. See the schedule for more details.

Questions? Contact us.

Siko (WMF) (talk) 05:43, 4 October 2013 (UTC)Reply

Community involvement[edit]

Dear Rillke: Since you said in your proposal that you were planning to check with the community to see of your work was useful, I came to this page to ask if you had asked for any opinions before putting in the proposal. I see by reading further up the page that at least one person has asked for this feature. Have you asked for input at any Wikiprojects or other discussion pages to see if others agree? There doesn't seem to be much on the project page in the community involvement section. Anne Delong (talk) 21:21, 11 October 2013 (UTC)Reply

Hi Anne, before I can ask the community whether they want this, I have to know whether it would be beneficial/required at all. I'll e-Mail Rahul and ask about the timeframe in which he would like to complete extension work (see conversation above).
Since the extension was developed with co-operation of the Wikimedia Foundation, I am assuming that the need for the general functionality was elaborated carefully enough. If this proposal doesn't pass (due to shortcomings in time for community consultation), it's not a big deal for me; I just wanted to give it a try. -- Rillke (talk) 12:37, 12 October 2013 (UTC)Reply
Hi Rillke - I'd encourage you to not give up on this before you've even begun :) We do need to see that the community being impacted has been notified as part of the process, even if they choose to not come comment, and that could be as simple as simple as posting a link to some discussion page on Wiktionary, asking for input, and then posting the link to that notification back on your proposal. You'd want to do this once you started the project, anyway, so why not begin now? This seems like a useful project, and I think you'd probably generate some more interest and enthusiasm that way. The committee will begin scoring proposals on 23 Oct, so it would be useful to do this before then, if you're still interested. Cheers, Siko (WMF) (talk) 23:24, 18 October 2013 (UTC)Reply


Hi Rillke,

I appreciate the frugality of this request, but I'd like to make sure it is realistic. If this project takes your time away from other valuable work, I'd wonder if it may be worth a bit more than 100 EUR, and to ensure the project would be completed we should consider at any sort of budget increase? I'm also wondering if, since IEGs are generally 6 month projects, you think you'd prefer to use the full 6 months to complete this, rather than the 4 weeks you have specified? Interested to hear your thoughts. Cheers! Siko (WMF) (talk) 23:32, 18 October 2013 (UTC)Reply

Hi Siko, the 4 weeks were just a model calculation. When involving communities, it is often unpredictable how long it takes getting feedback; feature requests or suggestions are another element of uncertainty. The proposed time investment was for a "basic version" (porting, in-page integration) only, indeed. -- Rillke (talk) 18:14, 22 October 2013 (UTC)Reply


Hey, I’m an administrator at English Wiktionary and I think this project is a great idea. We have a template for audio pronunciation requests; you may want to combine your software with it so it pregathers information from the template’s parameters. Ungoliant MMDCCLXIV (talk) 19:40, 6 March 2014 (UTC)Reply

Hi Ungoliant MMDCCLXIV, thank you for the pointer to the template; this is exactly the kind of support that I'll need during development. I'll have to gather related templates from other Wiktionaries and then decide the best way to implement pre-filling fields and the program flow followed by documentation describing how to add machine-readable data to these templates. Would you like to become one of the Consultants? If not, it would be an honour if you add your name to the volunteers section. -- Rillke (talk) 21:32, 6 March 2014 (UTC)Reply
I’d love to help in any way that I can. Being a consultant would be nice, but I only have beginner knowledge of JavaScript so if that is a problem I’m fine with being a volunteer. Ungoliant MMDCCLXIV (talk) 22:22, 6 March 2014 (UTC)Reply
Ungoliant MMDCCLXIV, that wouldn't be a problem at all. Essentially it's reporting errors from the JavaScript error console, if they appear there, for example. Much more important is that you're involved in Wiktionary work and are able build a bridge over the gap between me and the community's daily work, involving communication of issues and desires by the community. So, if you agree with the Individual Engagement Grants conditions, you can even get a compensation of your work as a grantee. I would really like to see that happening because it will be, indeed, some work (e.g. writing parts of the reports about the community's impression/reception). You may adjust the Budget according to your location's living costs. Furthermore, if the proposal is being selected, it's always a nice reference to point to, when, for example asking for refunding travel and accommodation costs for Wikimania or even other application stuff. I also saw that you're able to speak a lot of languages which is definitely a plus. -- Rillke (talk) 23:02, 6 March 2014 (UTC)Reply
I’ll join up then! Should I add my name to the participants section?
I must say I’m very excited about this project. Back in 2012 I added an audio pronunciation to Commons ([1]) and the process was a nightmare. Ungoliant MMDCCLXIV (talk) 00:05, 7 March 2014 (UTC)Reply
Yes, please add your name to the box. /s/(please apply on talk page)/[[User:Ungoliant MMDCCLXIV]]. I am so glad that you, as an experienced and curious user will help me to tackle this project. Welcome on board. -- Rillke (talk) 00:18, 7 March 2014 (UTC)Reply
@Rillke: where is the best place for us to communicate with each other? Ungoliant MMDCCLXIV (talk) 00:32, 7 March 2014 (UTC)Reply
@Ungoliant MMDCCLXIV: What do you prefer? On-wiki (meta, wiktionary, commons), e-Mail, IRC, Skype? Note that I live in Europe so it can be sometimes hard (assuming you're living in Brazil) when I am too tired in case we choose real-time communication. -- Rillke (talk) 00:37, 7 March 2014 (UTC)Reply
I prefer e-mail. You can also send me a message at enwikt if there’s anything urgent, as I check my watchlist there very often. Ungoliant MMDCCLXIV (talk) 00:44, 7 March 2014 (UTC)Reply
Okay, I am going to send e-Mail as soon as there are updates. -- Rillke (talk) 00:47, 7 March 2014 (UTC)Reply

Extension work: MW is in GSoC this year[edit]

Mediawiki is in GSoC this year. Perhaps consider adding your idea to mw:Google Summer of Code 2014. Gryllida 23:54, 7 March 2014 (UTC)Reply

Yeah, why not, thanks for the link. (BTW, this wasn't really my idea ;-) -- Rillke (talk) 08:31, 8 March 2014 (UTC)Reply
Unfortunately, I didn't manage to do that. Last year, I was asked to mentor my proposed projects and this doesn't fit in my schedule. -- Rillke (talk) 21:19, 14 April 2014 (UTC)Reply


This proposal appears to be, at least originally, mainly focusing on Wiktionary. However it may or may not be possible to design means of writing interactive wiki interface on-wiki with nice things such as upload which any sister project can use (including Wiktionary, where simply adding or editing an article is manual work (especially for someone with no linguistic background) and some sort of interactive interface would aid it). I had somewhat clumsily put the idea at mw:Requests for comment/UploadWizard: scale to sister projects. What do you think? Would the extra effort be worth the output? Gryllida 23:59, 7 March 2014 (UTC)Reply

I will make sure to keep it modular, i.e. separate upload, wikitext-interpreter, wikitext-editor libraries that are finally controlled by some application script tying up the loose ends. The proposal clearly outlines the scope as Wiktionary enhancements for uploading pronunciation files.
The issue when allowing arbitrary uploads is with licensing: When it's not own work, there is a lot to consider. I believe that a very simple VisualEditor file uploader plugIn or some upload interface shown if some template is detected is technically feasible but it will be troublesome to get people using it only for uploading their own work. When, I've integrated TemplateData support into Template:TemplateBox (so Commons could continue using TemplateBox), I got some very unpleasant feedback by the product manager: I'm not sure why you think derailing Commons to adopt mis-usage of TemplateData that makes it useless is a good idea, though.
Reading this, you'll probably understand that I am hesitative commenting on this RfC. There are also some UploadWizard-patches pending for review and they depend on the developer's good-will to get them finally merged.
As for the interactive interface, do you imagine something like Help:Watchlist messages/Wizard (data from Template:WatchlistNotice/Wizard/Data) or Template:Dashboard/Widgets/Add blacklist user? When creating stuff like that, you have to be extremely careful regarding security. An attacker must not be able to screw up a wiki page in a way that it can't be deleted utilizing the user interface or, for examle lets users uploading files under the assumption they could license them as all rights reserved, use for Wikipedia only permitted. The easiest way to ensure nothing bad like that happens, was to write tools for very specific tasks.
Is is, however a fact that we have now 2 programming languages floating around on-wiki (JS and Lua) and we're using 4 ways to store data (Lua table, JSON, Wiki table/templates, Wikidata) which is very confusing.
What can be done:
  • Making an IEG proposal for a wizard creating a whole Wiktionary entry.
  • Grants:IEG/Upload from context (perhaps also as without VE-binding; I wouldn't rely on UploadWizard anymore because there are too many stacked changes and browser support for easier upload methods is growing)
-- Rillke (talk) 09:36, 8 March 2014 (UTC)Reply

Some details[edit]

Hi Rilke. First of all, thanks for proposing your idea and all the efforts you're doing about it.

I just want to make sure I completely understand what this extension / gadget will do before giving my mind. If I understand, after uploading a sound file to Wikimedia Commons we will be able to add pronunciation tracks on Wiktionary. But do we have to split the talk word by word before uploading at Commons, or it will be automatically done by your software ?

Regards, -- Quentinv57 (talk) 09:01, 8 March 2014 (UTC)Reply

Hi Quentinv57.
Thanks for asking; a workflow was truly missing. It is available at Grants:IEG/Finish Pronunciation Recording#Workflow (Draft) now.
When you say split the talk word by word, what are you conferring to? Do you usually record multiple pronunciations into one track? Do you think there is a need for that after reading the intended workflow? -- Rillke (talk) 11:29, 8 March 2014 (UTC)Reply
No, I just didn't understand at first what this project really aimed at. The page Grants:IEG/Finish Pronunciation Recording illustrates it clearly, whereas your message on community pages doesn't. I'm just saying this because you're not getting enough community input here yet (I've not seen somebody else from frwikt). And it's not because your idea is not interesting. Hope my comment will help. -- Quentinv57 (talk) 14:03, 8 March 2014 (UTC)Reply
It does. I will draw up a draft and create a video (because there's not slideshow feature available on all wiktionaries) which I am going to add to the messages. Thanks for adding some clarification entre Rendre accessible à tous l'enregistrement des prononciations. -- Rillke (talk) 23:36, 8 March 2014 (UTC)Reply

Add the pronunciation on every Wiktionary project[edit]

Do you think that the following scenario would be possible ?

  • A user is in the french Wiktonary and notices that the pronunciation of the word wikt:fr:prisme is missing
  • He uses the gadget to register the pronunciation
  • The pronunciation is then uploaded at Wikimedia Commons
  • It is added on the page wikt:fr:prisme
  • Every interwiki links are checked. For example, it is missing at wikt:en:prisme so added there too.

Quentinv57 (talk) 14:19, 8 March 2014 (UTC)Reply

  • That's a great idea for phase II. Checking interwikis is a fairly easy game, but inserting it at the right place in the target entries will be a major challenge. If there is a way to automate that -- great. Otherwise, I'll probably offer displaying all these entries where auto-adding failed and let the user drag&drop the file to the correct position (of course without the need to navigate around, all using AJAX).
  • I even don't really dare to ask you, seeing what great tools you built on toolserver (everyone in wiki world knows them), but would you like to join this project? Your hints were very valuable to me and you could fill the field where I am blind to recognize important matters. I would feel somehow relieved if you would. -- Rillke (talk) 21:52, 8 March 2014 (UTC)Reply


License CC BY-SA 3.0 applies to my work on the "Pronunciation Recording Extension"--Rahul21 (talk) 17:15, 17 March 2014 (UTC)Reply

Thank you, Rahul21, that makes everything easier. I'll make sure that you get proper attribution. Do you have time to improve the extension this year? Wikimania, again takes part in GSoC. But you would have to hurry. There are some deadlines... -- Rillke (talk) 17:19, 17 March 2014 (UTC)Reply
Yep, this should smooth things out a bit. I mostly contributed by reviewing Rahul's work and making suggestions (code and otherwise), but the same goes for all of my work on the project. That should cover all the newly developed JavaScript (external libraries are discussed above).
Rillke, I'm glad to see you picking this up, preparing, and engaging the community. Superm401 | Talk 18:06, 17 March 2014 (UTC)Reply
Also, if you can ensure that everything in this project going forward is dual-licensed under GPLv2, I'd appreciate it. That will allow us to copy/move things to extensions (this one or other ones) more easily. Superm401 | Talk 18:10, 17 March 2014 (UTC)Reply
I will. Promise. -- Rillke (talk) 19:22, 17 March 2014 (UTC)Reply

Discussion on budget[edit]

I have a few comments to the presented budget:

  1. place for coding. I dont know how IEG is set, but are there some alternatives, for such place in the area?
  2. Who will be the coder? Rillke?
  3. Why we need payed testers? What about to leave the gadget and fix the bugs and enhacements announced by volunteer users?--Juandev (talk) 16:58, 30 March 2014 (UTC)Reply
Hi, Juandev,
  1. Is this question referring to Food, drink and a nice, calm place for efficient coding for 4 months 5 hrs/week? If so, it means that I do not charge a full salary but the cost of living where I live for the time I am spending coding. Could you clarify this question?
  2. Yes. I may decide to devote parts to third parties if they are interested but most code will be written by me and, last but not least I am managing project creation.
  3. Feedback for creating a viable product is a must-have. My experience at Commons is that getting feedback is possible from volunteers but often not sufficiently and not within a fixed time frame. But for IEG, I had to rack up such a 6-months-time-table and having some users who are highly motivated and whom I can nag and ask to write parts of the midterm and end report summarizing the community's reception of the tool is required. Also, sadly, I am not that familiar with Wiktionary and that's why it's good having people whom I could ask dumb questions at any time and who will respond in a constructive manner. The WMF also has community liaisons, even uses (see section "Over 20,000 happy customers" which lists Wikipedia) and I see no reason why I shouldn't do the same. It will be work for them, differently from what they are currently doing (volunteering), so they should receive compensation for their time. I am not inclined asking totally randomly selected people and funding sites like, that's why I want to recruit from experienced Wiktionary users.
-- Rillke (talk) 18:43, 30 March 2014 (UTC)Reply


  1. I think your reply is perfect. You got money to pay your appartment, rather than take money for work.

Anyway, thank you for your answers.--Juandev (talk) 00:06, 31 March 2014 (UTC)Reply

Gadget or extension[edit]

Why is it a gadget and not an extension? Gadgets needs extra settings in proper mw settings as far as I know. So it would be better to have an extension.--Juandev (talk) 16:58, 30 March 2014 (UTC)Reply

  1. Extensions work is not funded by IEG.
  2. Extension code must be code-reviewed.
    1. This is always a tedious process.
    2. Code-review depends on good-will by MediaWiki developers.
    3. Code-review is often slow. I wouldn't be able to complete the project within 6 months, nor would I be able to quickly deploy fixes or improvements based on the community's feedback.
    4. Deployment schedules would even further slow-down the development.
The gadget, to be installed as a gadget at a Wiktionary, will presumably require amending four pages. MediaWiki:Gadgets-definition, MediaWiki:Gadget-Pronunciation Recording, MediaWiki:Gadget-Pronunciation Recording.js, MediaWiki:Gadget-Pronunciation Recording.css.
I don't get the point about extra settings in proper mw settings. Gadgets can provide their own preferences interface and can store these preferences in the contributor's user account. If you do not believe that, have a look over to commons:Special:Watchlist hover the area below the heading displaying news and press the config notices button.
The only disadvantage gadgets suffer from is that there is no means provided by MediaWiki for a translation system, thus I have to implement that my own and that they tend to be loaded after other code, so there might be a visible delay until the "record pronunciation" link appears. -- Rillke (talk) 19:01, 30 March 2014 (UTC)Reply

Well, the point is about those 4 pages, which must be edited by an adminstrator. So to adopt a gadget is two level process, whill to adopt an extension is on level. At first community must agree that. In second case developers, would install it to proper project, in first you need a skilled willing administrator. If it is not there, no fruits from a piece of script.--Juandev (talk) 00:01, 31 March 2014 (UTC)Reply

Okay. If I see this becomes an issue, I will provede a button at Commons like that:

which has to be pressed by an admin of that wiktionary and 3 pages will be created on the fly and one will be suggested to be edited. Or, in case there is community consensus but no admin around, I will ask for the editinterface right to implement community consensus at Steward requests/Global permissions. In my own interest, the JS page will be just a bare minimum loading the code from a central repository. CSS may be completely superfluous.
Did you ask skilled willing administrator at and they had concerns? In any case, please report to me, if you have or see that a wiktionary has issues getting it installed. I am quite sure, this can be solved. -- Rillke (talk) 08:56, 31 March 2014 (UTC)Reply

Well, here I was talking more like about politica and motivation than technical problems. I have a personal experience from some Czech projects that installation of such tools is impossible due to the lack of time/interest of local admins. Than if you ask developer for help, theyll redirect you back to the project, that gadgets and such is a local issue. Hope we can use this gadget at least at en.wikt:-)--Juandev (talk) 16:07, 31 March 2014 (UTC)Reply

I would be surprised, if not. An en.wikt admin is listed in the grantees section. -- Rillke (talk) 16:36, 31 March 2014 (UTC)Reply

My help[edit]

I add myself as a volunteer. I can help with testing, advertising and pronunciation rallyes.--Juandev (talk) 17:01, 30 March 2014 (UTC)Reply

Thank you. -- Rillke (talk) 19:02, 30 March 2014 (UTC)Reply

I love it[edit]

is like but with GNU licence, i hope this project is carried out, is very impractical current form --Esceptic0 (talk) 19:53, 31 March 2014 (UTC)Reply

Thank you. The guys at are using Flash by default. -- Rillke (talk) 21:13, 31 March 2014 (UTC)Reply

Notifications to Commons?[edit]

Has the Commons community been notified of this project? Obviously Rillke is a member of the community, but other members might want to share related ideas. whym (talk) 22:58, 31 March 2014 (UTC)Reply

Would a watchlist notice be appropriate? I think this may be even overkill but posting to all language village pumps also - as Commons will be only affected tangentially by getting some files with machine-created file description pages. So mainly, Commons maintainers are interested? Thus, posting the to the administrator's noticeboard only? Not sure. -- Rillke (talk) 23:15, 31 March 2014 (UTC)Reply
I personally feel short VP posts are ok for notifying this (and perhaps another after it gets started.) Ideally we'd want to notify people who are interested in editing and categorizing sounds, but I haven't seen a noticeboard or similar for that on Commons. whym (talk) 11:52, 2 April 2014 (UTC)Reply
Which kind of audio-editing do you have in mind? Often, it is not really worth editing such short recordings or audio in general; re-recording is often much faster and results in better files if the cause for the quality-issue has been addressed. Additionally, it is my intention offering audio-editing-tools to crop the mouse-click out of the recording - I reviewed those produced with the extension and found it will be neccessary. -- Rillke (talk) 21:43, 14 April 2014 (UTC)Reply
By "editing", I was referring mainly to truncating silent intervals or background noise at the beginning or the end of a recording. According to your explanation, it sounds like Commons volunteers won't have to be bothered very much in this front. Still, as you suggested, it could be worth letting Commons administrators know of this plan of (hopefully a large number of) tool-assisted contributions to Commons, perhaps after the project gets started. A friendly notification in advance would be better than some unexpected bugs or other failures trigger their alarm as the first introduction. I admit this is a minor point compared to the whole objective, but setting the categorization scheme right is another thing that Commons volunteers could help with. whym (talk) 13:12, 21 April 2014 (UTC)Reply
There was a notification, in the meanwhile.[2] --Nemo 14:35, 21 April 2014 (UTC)Reply
I am hoping that the IEG committee is inclined to provide me with funds for audio editing tools (or at least the option for) so we can hopefully avoid having to edit these files. It is, in my opinion, simply impractical editing thousands of recordings; if something valuable is discovered, it could be edited but I see no reason from what this should be different compared to any other pronunciation upload in regard to editing. I agree that Commons administrators and community members dealing with categorization should be notified as soon as I am sure that this is going to become reality. As commons is my home wiki, you can probably imagine that I do not like to get the disgrace of my colleagues by dumping a lot of work onto their plate. -- Rillke (talk) 19:48, 21 April 2014 (UTC)Reply

Community engagement[edit]

Hi Rillke,

I just wanted to take a moment to appreciate you for being so thoughtful about community engagement in this proposal. I love that you're thinking not just about how to build the best gadget, but also about a variety of community organizing strategies to help ensure the tool is useful and used by those you aim to serve with this project.

Cheers! Siko (WMF) (talk) 23:12, 7 April 2014 (UTC)Reply

Learned from CodingChallange, UploadWizard, VisualEditor and a bunch of similar projects. If one is developing for a community these days one cannot hide behind walls anymore. There is no developer-team for PRG, so I have to carefully listen, and try to get as much feedback as possible 'cause I won't get any second and third opinion with ease. Luckily our community is highly diverse and there is always someone who is able and inclined providing one with specific and useful advice. -- Rillke (talk) 09:07, 8 April 2014 (UTC)Reply
Siko, I see that other proposals are under review now; is there something I missed doing here? -- Rillke (talk) 16:43, 8 April 2014 (UTC)Reply
Heh, no, I'm going a bit slower on the tools ones, to give WMF tech staff more time to yell at me if they think I'm marking things eligible that should not be. Yours will be all set shortly :) Siko (WMF) (talk) 23:25, 8 April 2014 (UTC)Reply

Eligibility confirmed, round 1 2014[edit]

This Individual Engagement Grant proposal is under review!

We've confirmed your proposal is eligible for round 1 2014 review. Please feel free to ask questions here on the talk page and make changes to your proposal as discussions continue during this community comments period.

The committee's formal review for round 1 2014 begins on 21 April 2014, and grants will be announced in May. See the schedule for more details.

Questions? Contact us.

--Siko (WMF) (talk) 23:26, 8 April 2014 (UTC)Reply

Russian Wiktionary[edit]

I'd like to offer my help because I'm very excited about the idea. What can I do: to test in old Opera, to check cyrillic compatibility, to incorporate the gadget/extension/whatever in Russian Wiktionary (I'm sysop). May be something else I don't know. Infovarius (talk) 12:02, 12 April 2014 (UTC)Reply

Infovarius, great to read this and thank you for the help you are offering. I will need it! Feel free listing yourself in the volunteers-section if you like (or if you want to spend a lot of time and do professional work, consider also listing yourself as a consultant). Did you test or check whether the old Opera supports the necessary APIs at all? Do you think a flash or silverlight fallback would be helpful? -- Rillke (talk) 17:48, 13 April 2014 (UTC)Reply
For my surroundings Flash is heavily used, so it would be useful. I check the links: second shows that my Opera supports many but not all (probably "dataset & data-* attributes" partially; not supported: "defer attribute for external scripts", "async attribute for external script", "sandbox attribute for iframes", "Drag and Drop", WebGL, "Canvas blend modes", "Ruby annotation" for adding pronunciation, "getUserMedia/Stream API", "Download attribute", "Details & Summary elements", "HTML templates", Shadow DOM, Scoped CSS, "Toolbar/context menu"). And I cannot use the first test :( --Infovarius (talk) 12:54, 14 April 2014 (UTC)Reply
  • Hello, Rillke. I would also like to ask if there any dependents that would prevent the tool from being adapted to other language Wiktionaries? Or will the issue be connected only with localization? rubin16 (talk) 09:21, 13 April 2014 (UTC)Reply
    • Hello rubin16, localization will be the main challenge. I plan to keep localization centralized so the extension that is in development and possibly other projects that would like to offer pronunciation recording could benefit from this gadget. I will take care of RTL issues, complex plural rules etc... -- Rillke (talk) 17:48, 13 April 2014 (UTC)Reply

video introduction[edit]

Thanks for the video, great job explaining what you are doing, and very refreshing to read so much detail in a plan! Jane023 (talk) 20:45, 14 April 2014 (UTC)Reply

Wikimedia UK project[edit]

Wikimedia UK are also working on a similar tool, for the Voice intro project. Please liaise with them. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 10:49, 15 April 2014 (UTC)Reply

Subject: Pronunciation Recording

Dear Wikimedia UK-Team,
Andy Mabbett made me aware [1] that you are working on a pronunciation recording tool.
However, I could not find anything about that published on your site [2] nor on Google [3],
except the blog post "Recording the voices of Wikipedia" [4] which doesn't mention any software development.

Please let me know about your
* progress
* architectural and design considerations
* whether you are willing to share your code or collaboratively create the tool
* whether you consider your reply confidential, allow or oppose publishing that under CC-By-SA 3.0 license.

Note that I am bound to the schedule as set out at [5].

Thanks in advance.

Kind regards
Rainer Rillke

Product management
Pronunciation Recording Gadget


Thank you, Andy. Sent the above e-Mail. -- Rillke (talk) 21:07, 15 April 2014 (UTC)Reply
@Pigsonthewing: No response, yet. Will try it again; friendlier worded, this time. -- Rillke (talk) 14:29, 25 April 2014 (UTC)Reply
Hi WMUK-team,
I am Wikimedia Commons administrator and I plan to create a pronunciation recording gadget[1].

Andy Mabbett told me that you're about to create a similar tool.
Would you be interested in collaborative work? I think it could be quite efficient, we could share ideas, experience or even code, if you like.

I'd be happy to hear from you.

Kind regards
Rainer Rillke

-- Rillke (talk) 12:32, 26 April 2014 (UTC)Reply
Today, the reply: Work done so far by WMUK tech. -- Rillke (talk) 16:06, 7 May 2014 (UTC)Reply


Thanks, I like the idea. How does this scale to other languages? Gryllida 22:37, 19 April 2014 (UTC)Reply

I'm confused about this because the project seems to be already language-independent, and it even got endorsements from members of multiple language editions. Do you have specific concerns? whym (talk) 07:11, 20 April 2014 (UTC)Reply
I first mistook the project name as "Finnish pronunciation recording". Perhaps did you, too, Gryllida? If so, it might worth considering renaming it into "pronunciation recoding tool for Wiktionaries", for example. whym (talk) 07:11, 20 April 2014 (UTC)Reply
  • I did mistake. This resolves the confusion I had (I couldn't understand what's special about that one language). I would like to thank everyone for this wonderful idea: it looks promising in every detail I can think of. Gryllida 07:14, 20 April 2014 (UTC)Reply
I am moving the page, now, for clarity. Gryllida 07:15, 20 April 2014 (UTC)Reply
Thank you, although the goal is bringing a gadget into production and not strictly finishing the incomplete GSoC project (which is an extension), just to clarify. -- Rillke (talk) 12:09, 20 April 2014 (UTC)Reply

Main page and communication hub[edit]

I would like to ask for a communication hub. I understand that core team will use its own private ways of communication, but I would be happy for any sort of news updates or something like that. Basically it would be news update sold to user talk page or an e-mail.--Juandev (talk) 08:17, 25 April 2014 (UTC)Reply

That's an excellent idea. I think, I will regularly post to multimedia-l maintain a progress page and offer bulletins through MassMessage at Commons. For major releases, there will be posts at commons-l, wikitech-l, wiktionary-l and at the community forum sites at Wiktionary. Let me think about the details this weekend. -- Rillke (talk) 14:21, 25 April 2014 (UTC)Reply
There we go. I used Commons because I can send mass messages and mass e-Mail there without having to ask for permission every time. Please feel invited to add more content to the skeleton! -- Rillke (talk) 09:59, 26 April 2014 (UTC)Reply

Global usage amongst other WMF project[edit]

We were coing with the idea on #wikiversity list, that it would be nice to be able to use this feature also on other projects and possible for other types of media. I would e.g. appreciate to add pictures directly to the page (this is already demanded as an enhancement to Wikimedia developers, unfortunately for VisualEditor only).

But lets talk about recordings as I understand this extension is limited to recordings only. Would it be possible to deploy the botton on Wikiversity? We have there project, where we tend to learn users the language and we link to the pronunciation. If it is missing, it exists as a red link. So also at Wikiversity would make a sence to have such feature. Not only at Wiktionary.--Juandev (talk) 08:17, 25 April 2014 (UTC)Reply

Pronunciation recording is the topic, that's true and a button for Wikiversity would be in scope. However, there will be a documented framework/configuration allowing admins of any Wiki - uploading to Commons will only be possible through WMF wikis - to add "such a button" in the way they like as I will not be able to maintain >300 different usage instances. I can and will, of course, answer questions when Wikiversity admins get stuck installing the gadget or something is not working as expected. -- Rillke (talk) 14:13, 25 April 2014 (UTC)Reply

Something to use now[edit]

Is there any tool, we can already use? I am refering to the title of this page "Finish incomplete GSoC project", which indicates some work has been already done in the area so something mighg been existing.--Juandev (talk) 08:21, 25 April 2014 (UTC)Reply

That one. But I don't know whether it works in Firefox; perhaps some changes to the source code are required to make it working in the latest FF release. -- Rillke (talk) 14:03, 25 April 2014 (UTC)Reply

Aggregated feedback from the committee for Pronunciation Recording[edit]

Scoring criteria (see the rubric for background) Score
1=weak alignment 10=strong alignment
(A) Impact potential
  • Does it fit with Wikimedia's strategic priorities?
  • Does it have potential for online impact?
  • Can it be sustained, scaled, or adapted elsewhere after the grant ends?
(B) Innovation and learning
  • Does it take an Innovative approach to solving a key problem?
  • Is the potential impact greater than the risks?
  • Can we measure success?
(C) Ability to execute
  • Can the scope be accomplished in 6 months?
  • How realistic/efficient is the budget?
  • Do the participants have the necessary skills/experience?
(D) Community engagement
  • Does it have a specific target community and plan to engage it often?
  • Does it have community support?
  • Does it support diversity?
Comments from the committee:
  • Well-targeted and beautifully crowdsource-oriented tool proposal. Very clear use case, great alignment with our ethos of collaboration.
  • Very reasonable and well-thought out budget.
  • Good deadlines and schedules, the project can finish in 6 months and the proposer has the key skills to develop the tool - we feel confident in their abilities.
  • If successful, this grant could be the key that makes the "Wikipedia of sounds" project possible. Could have large impact and fill a long-standing gap in Wiktionary.
  • The software that this proposal would create would be easy to use, it should be relatively easy to sustain, and need little adaptation to be used Wiki-wide. Could attract new contributors.
  • The measurable indicators are somewhat general, but the tasks are clearly defined into the plan section. Success of this project can easily be measured by the number of editors using it to add sound files to articles.
  • There seem to be a number of editors ready to make use of this right away. Others are thinking of new ways to use it. It could be used in any language, so it supports diversity. The project has great demand from the community.

Thank you for submitting this proposal. The committee is now deliberating based on these scoring results, and WMF is proceeding with it's due-diligence. You are welcome to continue making updates to your proposal pages during this period. Funding decisions will be announced by the end of May. — ΛΧΣ21 23:55, 12 May 2014 (UTC)Reply

Video capabilities[edit]

If it is not very difficult, I would like to ask to consider also to add video recording capabilities to this project (or leave it prepared for a future development). The reason for this are Sign Languages, which are more visual. It would have a big impact if users could upload how to express in video different words.--Micru (talk) 17:29, 20 May 2014 (UTC)Reply

The API is a very similar one and the workflow could be also similar. So, although I am not going to implement it as part of this project, it code could be useful for that as well. -- Rillke (talk) 21:18, 20 May 2014 (UTC)Reply

Round 1 2014 Decision[edit]

Congratulations! Your proposal has been selected for an Individual Engagement Grant.

The committee has recommended this proposal and WMF has approved funding for the full amount of your request, EUR 1450

Comments regarding this decision:
Thanks for a well-developed proposal with such thoughtful community engagement. We look forward to seeing the resulting impact on Wiktionary.

Next steps:

  1. You will be contacted to sign a grant agreement and setup a monthly check-in schedule.
  2. Review the information for grantees.
  3. Use the new buttons on your original proposal to create your project pages.
  4. Start work on your project!
Questions? Contact us.

Interesting stuff regarding TTS[edit]

History of uploading[edit]

Plot of number of uploaded audio-files with pronunciation examples versus time of uploading
. Infovarius (talk) 18:49, 24 August 2014 (UTC)Reply
Thank you, Infovarius! -- Rillke (talk) 20:32, 24 August 2014 (UTC)Reply

Actual status[edit]

So, how we are doing? Do you have any news feed?--Juandev (talk) 18:22, 27 August 2014 (UTC)Reply

I am going to publish a status update this weekend. -- Rillke (talk) 02:35, 28 August 2014 (UTC)Reply

Current usage of pronunciation files (as of 31.08.2014)[edit]

265333 files were analyzed. For each I've obtained a number of wikipedia and wiktionary pages which use it. Here I give a table showing distribution of such numbers.

Number of projects In wikipedias In wiktionaries
0 252482 52955
1 6872 116572
2 1934 47424
3 1019 15311
4 696 10765
5 489 7839
6 371 5695
7 260 3941
8 194 2241
9 146 1215
10 124 583
>10 746 792

There were 4 files that are used in 55 wiktionary pages, and 1 file that is used in 497 wikipedia pages. --Infovarius (talk) 19:32, 7 September 2014 (UTC)Reply

This is a rather amazing usage share, it's much better than the average of Commons (where about half of the files are in use). --Nemo 20:54, 7 September 2014 (UTC)Reply

What's new?[edit]


What's new about the project? I do not see recent information. Is it for soon? Do you need help? Thank you by advance! — Automatik (talk) 19:26, 12 October 2015 (UTC)Reply

@Automatik: Hey thanks for asking! We always need help. What are you interested in particularly? E.g. are you a good graphics designer? Or do you like to design Workflows, ... or do you like testing software? -- Rillke (talk) 01:56, 7 March 2016 (UTC)Reply
@Rillke: I could help for technical implementation and integration if it could help (I already maintain javascript gadgets on fr.wiktionary). I suppose this is the project page on Phabricator for the tool? Is there a TODO list? I may lack time, but I'd try to do what I can. — Automatik (talk) 22:45, 7 March 2016 (UTC)Reply
@Automatik: That's great and exactly what I need. The Phabricator project is for the MediaWiki extension, not for the gadget. The project is hosted at -- issue list is to come (at GitHub), yet I have to fill-in the status of the development of the components. If you tell me your GitHub user name, I can give you some push permissions etc to the repo. -- Rillke (talk) 22:25, 10 March 2016 (UTC)Reply
@Rillke: Thanks for the precision. My username on Github is Botomatik. — Automatik (talk) 02:07, 11 March 2016 (UTC)Reply

Wikimedia France[edit]

"a web interface for recording words and expressions in regional languages is currently being developed in partnership with the APLLOD association (Association for the Promotion of Languages via Lexicography and Open Data)" Grants:APG/Proposals/2014-2015_round2/Wikimédia_France/Progress_report_form --Nemo 12:12, 3 February 2016 (UTC)Reply

Thanks. Currently I see -- Rillke (talk) 01:54, 7 March 2016 (UTC)Reply
Might be related to Grants:Project/0x010C/LinguaLibre. --Nemo 12:19, 31 October 2017 (UTC)Reply