User talk:Verdy p/archive10

Archives: 2008–2009 ‧ 2010 ‧ 2011 ‧ 2012 ‧ 2013 ‧ 2014 ‧ 2015 ‧ 2016 ‧ 2017 ‧ 2018 ‧ 2019 ‧ 2020 ‧ 2021 ‧ 2022 ‧ 2023 ‧ 2024–

---

My main discussion page is on the French edition of Wikipédia. But you may write here, as I will come regularly on Meta too. See: w:fr:Discussion Utilisateur:Verdy p. Thanks.

Arabic version of Wikimedia Catalyst Programs PDF file requested

Hi, There is an Arabic version of the Wikimedia Catalyst Programs PDF file which I want to add to the Arabic page of this page Global Development/Arabic Language Initiative strategic plan. How to do this? Haytham^Abulela 11:30, 30 October 2017 (UTC)

Access to nonpublic information policy/noiddraft

Hi! This is redundant and may be deleted (with all translations)? --Kaganer (talk) 22:56, 4 January 2018 (UTC)

I am not the author of the original which was created at the same time as the policy in discussion. I just helped to translate it by preparing it because it was requested.

Not redundant but possibly to rename for archiving (because it was a draft and has interesting discussions associated).

The policy was published long after, this is not exactly the same, but this is still useful as it uses less technicalk terms and explains the principles without entering into tricky details.

If this is not archived, may be it needs a review to make it more accessible and in sync with the decided policy.

So ask to the OSMF working group about their advice. I think it's good to keep archives of discussions as all OSMF policies are subject to rediscussions and we'll need to reference past talks to explain some past decisions or if decisions that occured at one time may no longer be pertinent but arguments discussed in the early draft would be useful again even if they were not kept initially: the support for various proposals that were not kept initially may have changed significantly; or counter arguments that were then useful coulds be useful again and we don't need to reinvent the wheel for saying it again: archiving (but keeping it readable with a warning that it was an early draft) is the best option IMHO. People will also be able to find again the arguments they already exposed. Such discussions are useful to assert that this policy was really discussed openly and were not made by just a few people. Our community of users has evolved significantly over years. verdy_p (talk) 00:38, 5 January 2018 (UTC)

Ok. As first step, this was marked as "Historical". --Kaganer (talk) 13:11, 29 January 2018 (UTC)

Ready for translation: Editing News #1—2018

Hi! This is an invitation to join translators working on the upcoming issue of the multilingual newsletter for the visual editor, which will be widely delivered when it's end of Friday in Central Europe. Thanks in advance for your help with this: getting to interact with fellows so skilled like the translators are is among my favorite things about my job :) Best, Elitre (WMF) 10:29, 27 February 2018 (UTC)

PS: Here are some instructions. Please go to the translation page: your language should be available from the drop-down menu on the right. Once you've selected it, you'll see the document in English side by side with any translation work already done in your language. You can add new translations or modify existing ones. Please let us know about difficulties you experience with the translation memory system.

Translation: VisualEditor newsletter October 2018

Thank you for offering to help with translations. m:VisualEditor/Newsletter/2018/October is ready to translate now. /Whatamidoing (WMF)

18:20, 24 October 2018 (UTC)

Please IP user (moderated here): don't pollute my page by replying (with false statements and once again insults) to a message actually posted by an bot I have authorized! Visibly you don't know what this wiki does, and who makes what here, and how things works, how things are discussed or decided collaboratively. It's not an error by you as a newer, it's insistance to repeat it again and again. verdy_p (talk) 03:31, 6 November 2018 (UTC)

Sorry but I'm really not "117.15.55.182"

I would love to share you some brief informations, that my access point password was too short, and I've checked my router log, there are some foreign devices which I don't know these Mac addresses before, I recently changed my AP password, in order to avoid re-happening of this panorama. --Liuxinyu970226 (talk) 04:38, 6 November 2018 (UTC)

I also revoked and blacklisted these foreign-Mac addresses devices from accessing my AP again. --Liuxinyu970226 (talk) 04:39, 6 November 2018 (UTC)

You can say what you want (or don't want), an IPCheckUser-admin can state if that's probably true, or undeniably false. Even if the test turns as "probably true" at one time, at any time it could definitely turn to be really false. The IPCheckUSer-admin has very powerful tools to determine it two users (an IP and a registered user, or two IP users) are undeniably the same (even if its IP changes over time!).

An IP is never completely "private" or "anonymous", even those passing through "anomizing proxies" (or VPNs) and changing their ISP, because they all have detectable holes, or most frequently by users do not know how to NOT transmit personally identifiable data, as this is extremely challenging on the Internet and so this requires very complex and costly means; even if the technology employed by the user is perfect, the last imperfect part is the user itself, arguably identifiable by his behavior and actions (there's not an infinite number of people on Earth, and if a statistics proove that a user is identifiable with less than 1 chance to be wrong in seven billions cases, the user is perfectly identified statistically), or its unique personal faults/errors, or the language and expressions he uses (this is also the base to assert copyright violations). Even his promoted ideas as characteristics usable statistically to proove "scientically" that two users are actually the same. But humanes are unable to constantly remain in the undetectable statistics, and IP addresses do not change so often. We can even detect that an IP was stolen (e.g. harvested by a bot, by identifying the bot itself with its unique behavior).

The only way to be anonymous on the Internet... is to be NOT connected to it, i.e. to not use it at all (in which case that user cannot abuse any one else). verdy_p (talk) 22:45, 6 November 2018 (UTC)

So now you pretend that someone using the same kind of behavior and promoting the same ideas as you, has "stolen" your IP access and this was not simply you ?(this is still the most highly probable explication: stolen IP are not used for the same purpose, except if this is done to steal your identity and then you have a severe problem and should call the police to protect you, because you'll turn to have extremely severe problems if you're impersonated, which will ruin your life for many years if you don't act urgently; changing your settings on your devices will not be enough to protect you from what the harvester wanted to do, not necessarily to offense you directly, but to leave the rest of world thinking that you are the same person, and you serve to the harvester to "hide" his crimianl acts: impersonating someone else, on the Internet or elsewhere, is a crime in most countries, and if someone impersonates you, you MUST act legally, acting alone will not protect you at all, you'll have to proove again and again you won legal existence if the criminal harvester succeeds in obtenaining real legal identity proof, and if he does other criminal acts, you'll be prosecuted instead of him and you'll have little or no defence!).verdy_p (talk) 22:55, 6 November 2018 (UTC)

But given that you are immediately reacting here (on this page) about the abuse that was made this time in my page when I moderated it, the synchronism of events, and the fact that I did not cite you on this talk page, just conforts my idea (as a certitude) that YOU are that IP user ! An IPCheck admin will easily confirm it with extremely high level of certitude (an even higher certitude than mine).

(Note: I have the full right to to moderate my own talk page and deleting everything posted here by anyone else, but no one else than me can falsify this page or delete my own content or comments posted on it, unless this is just to fix minor technical problems, such as edit links that were redirected, or contents that cause severe technical trouble to this wiki such as frequent crashes of the server, even if this is caused by a bug in the software).

So if you thought that disconnecting from your Wikimedia account, and rebooting your internet access router before posting in Wikimedia would be enough to "hide" your identity, you're completely wrong. Never, ever try to impersonate anyone on this wiki, on the Internet or anywhere else (if you do that, someone else will do that to impersonate you by harvesting your identity, and it will be actually easier for the harvester that he is the legitimate owner of the identity that you have compromized yourself definitely by using impersonation: the harvester are extremely powerful to assert that the identity stolen is theirs, once they get it, they will not compromize it and will use that stolen identity extremely abusively, without any care for your own security: you'd have become their puppet slave for their service).

You've been warned, measure the huge risks you take (which are not just risks of being blocked on this wiki, but that could completely ruin your whole life later!) if you ever impersonate someone that have not registered your second faked identity to a legally strong warrant that can assert that you are in full control of the puppet and that you control that puppet in agreement with rules required by the warrant. Wikimedia allows some "sockpuppets" to operate on their wikis, but only in full agreement of Wikimedia missions and public policies and goals (unregistered sockpuppets are forbidden in Wikimedia if you don't have a strong respectable warrant, respected by Wikimedia for its public goals).

If you've lied about your identity (and an IPCheckUser admin prooves it), you'll be blocked in your own registered wikimedia account and all the IPs you've used or any new IP that will be associated later to you, given all what Wikimedia already knows about you and has kept in secured archives (even if what Wikimedia knows is kept as a secret between Wikimedia and you). verdy_p (talk) 23:29, 6 November 2018 (UTC)

@Verdy p: Is this meaning that you will not give me any chances more? Or what should I do to let you trust me again? --Liuxinyu970226 (talk) 01:20, 7 November 2018 (UTC)

It does not depend on me. Admins will look at this, notably because since your message on this page, the abuse has continued on the Narom request page. verdy_p (talk) 05:51, 7 November 2018 (UTC)

Well I'm trying to apologize to you, why you think that that is also an abuse? And FWIW you seems also said some wrong things, for example you think that BCP47 is stable, where as I've poked with langcom members that isn't either, because of the Western Armenian, where it already got ISO 639-3 code hyw in early this year, but the official BCP47 list provided by IANA still contains "arevela" and "arevmda" (though deprecated on 2018-03-24), so which should be the actual standard? Anyway, the Jon Harald Soby isn't having CheckUser right, why you contacted him? Instead of @Ajraddatz, MarcoAurelio, Trijnstel, and Vituzzu: where they are actual CheckUser right holders here? --Liuxinyu970226 (talk) 03:57, 8 November 2018 (UTC)

I contacted Jon Harald because he is directly concerned in the Language Commity, to which this page is addressed (so this page should not be abused, as this abuses them as well)

And I'm right: BCP 47 is stable exactly because it ensures the upward compatibility of codes, that ISO 639 does not, including and notably for hyw'. You don't seem to have understood how BCP 47 works: it keeps ALL its entries (but adds "deprecation" notices in the IANA database, and this was done correctly for "arevela" and "arevmda", but only if there are possible replacements, otherwise removing them would BREAK upward compatiblity). It also imports new ISO 639 codes, but not all automatically. ISO 639 does not care at all about upward compatiblity nad constantly breaks everything. BCP 47 allows smooth transitions to the newer codes without breaking any existant thing even if they use "deprecated codes".

So I suggest you read the main RFC describing BCP 47 precisely, and how it is "synchronized" (partly) with ISO 639 ! In summary we NEVER need to refer to ISO 639 anywhere in the wiki, only BCP 47 is reliable (and will get updated correctly, fixing all the many issues and incompatiblities of ISO 639). verdy_p (talk) 04:06, 8 November 2018 (UTC)

Traduction de la page WikiFranca

Bonjour Verdy p.

Merci d'avoir apporté votre pierre à l'édifice de WikiFranca en ajoutant les balises de traduction à la page modèle WikiFranca/projets. Néanmoins votre système attribuant une balise par ligne plutôt qu'une balise pour tout le modèle complique la tâche des traducteurs et des contributeurs. En effet chaque ajout de projet à la liste doit d'abord être estampillé d'une balise traduction avant d'apparaître sur la page dans laquelle le modèle est inclu. Vous serait-il possible d'effacer vos balises et d'en installer une seule pour tout le modèle?

Merci d'avance --Adelaide Calais WMFr (talk) 16:32, 29 November 2018 (UTC)

C'est au contraire la façon de faire: les traducteurs n'ont pas à retraduire la totalité de la liste si des éléments sont ajoutés, ils sont traduits séparément et sans se soucier de leur mise en forme. Si on ajoute ou retire des éléments, les autres déjà traduits sont conservés, et bien souvent les différents n'ont pas besoin d'être traduits. On profite également de la mémoire des traductions facilitant leur réutilisation d'un seul clic.

Cette façon est documentée dans l'aide sur le module de traduction destinée aux concepteurs de pages à traduire: il est hautement recommandé de faire des unités de traduction les plus petites possibles (mais en évitant de découper les phrases car les langues n'ont pas la même syntaxe et le même ordre pour former des phrases correctes ou gérer les règles d'accords entre éléments séparés: cela nécessiterait un balisage complexe directement exposé aux traducteurs).

Cela facilite donc le travail des traducteurs (contrairement à ce que tu dis), le travail étant confié seulement à celui qui édite la version de base pour baliser les paragraphes un par un et permettre de les traduire chacun isolément et facilement (moins d'erreurs, et pour les traducteurs aucun embêtement pour préserver la syntaxe).

Mettre une liste entière dans une une seule unité de traduction est FORTEMENT découragé (les traducteurs se sont plaints trop souvent du fait qu'une petite modifs de la liste annulait la traduction de la liste en totalité et les obligeait à refaire tout le travail et chercher ce qui pouvait être avant, et détecter ce qui avait vraiment changé.

On voit assez souvent des corrections mineures dans les listes comme la correction de l'orthographe ou la ponctuation d'un seul élément. Les traducteurs avaient déjà traduit la liste en totalité mais quelqu'un corrige la version de base et toutes les traductions sont annulées en même temps et doivent être revues dans toutes les langues sur une unité trop grosse et où il est peut facile de voir les différences: faire une seule balise a l'effet pervers de décourager les traducteurs de contribuer à la conversion de ces pages qui ensuite ne seront plus traduites du tout et n'afficheront plus que la langue d'origine.

Ce n'est pas aux traducteurs à qui on doit demander de faire le gros du travail, mais bien à ceux qui préparent les versions de base à traduire et qui doivent faire, eux, cet effort une seule fois pour que ce ne soit pas un effort multiplié par le nombre de langues et sans possibilité pour eux d'utiliser la mémoire de traduction: on les condamne à devoir explorer les "diffs" très peu lisibles pour beaucoup de traducteurs à qui on ne doit pas demander de comprendre la syntaxe utilisée.

Bref cette façon de faire respecte mieux le travail des traducteurs: chacun fait de son coté ce qu'il fait le mieux dans sa propre langue sans casser le travail des autres. verdy_p (talk) 18:02, 29 November 2018 (UTC)

Note j'ai revu le balisage afin de le simplifier au maximum pour les traducteurs. Certains éléments communs comme "(sur Wikipédia en français)" sont balisés de façon à être pouvoir les réutiliser chaque fois que possible. Les cibles des liens ne sont plus dans les unités à traduire, où ne figure que le texte affiché.

Les langues actuellement traduites sont mise à jour. Si tu regardes tu verras que c'est réduit au strict minimum et c'est la bonne façon de préparer une page à traduire pour faciliter le travail de chacun dans sa langue.

Bref un contributeur souhaite ajouter un élément il le fait sans toucher le reste déjà traduit et le marque proprement. Le travail pour les traducteurs est réduit au strict minimum (un clic puis taper une traduction simple) et ils n'ont pas besoin de comprendre la syntaxe wiki ou la structure de la page, tout est déjà résolu pour eux, ils n'ont qu'à entrer le texte et même pas à s'occuper du reste qui est inséré automatiquement pour eux sur la page produite. Chacun fait le minimum, la maintenance est bien plus facile, et on profite au maximum de la mémoire de traduction pour s'accorder au mieux sur une terminologie existante commune. Et on ne surcharge pas inutilement la mémoire de traductions de tas de choses non facilement réutilisables, on l'alimente au contraire d'éléments utiles à la traduction d'autres pages. verdy_p (talk) 19:20, 29 November 2018 (UTC)