Wikipédia:Esplanada/geral/Deploying InternetArchiveBot and/or its tools on ptwiki (2jul2017)

or its tools on ptwiki (2jul2017)

Hello everyone on the Portugese Wikipedia. I am Cyberpower678, an administrator on the English Wikipedia and a bot operator. I have come here to ask if the Portuguese community would welcome m:InternetArchiveBot and/or the tools that come with it. Before just asking the question, I will offer some background on the bot and the motifs for deploying this onto other wiki.

What is InternetArchiveBot?

InternetArchiveBot is a very sophisticated bot designed for the sole purpose of combating the ongoing problems of links dying. The bot is able to fully autonomously crawl through Wikipedia's article space and analyze links, that are formatted in various methods, and appropriately modify the wiki markup to preserve the sources used on articles. It uses a very reliable algorithm to determine if sites are dead or not. In addition to the advanced algorithms used to parse reference formatting, and checking websites, the bot is highly configurable to suit the needs of the wiki. I will touch more on this later. However, the bot does not have to be fully autonomous. IABot comes with a bunch of tools that editors can use, which can be found here. These tools let users report bugs found with the bot, help the bot improve it's reliability by letting users report issues with the bot's link checking abilities, finding more suitable archives the bot should use, and most importantly, provide tools where users can run the bot on a single page immediately as well queue the bot on a collection of user specified pages. So if the fully autonomous is not desired, these tools can still be a way to allow users to combat link rot.

Why deploy InternetArchiveBot?

The IABot project was started when the community wish list survey of 2015 had a proposal to introduce a global bot to repair broken links. The proposal made number 1 on the wishlist and thus this Phabricator ticket was created. This ticket is to track if there is consensus and it's deployment progress on this wiki. As of June of 2017, this project hit the 2 year mark in its development.

About InternetArchiveBot

IABot uses intelligent algorithms to parse wiki text as humanly as possible. That means it will handle formatting differently when inside <ref> tags, as compared to when outside. It's able to detect citation templates and handle those correctly. When inside a reference, provided this Wikipedia uses them, IABot adds {{wayback}} or equivalent appropriate archive templates to plan external links. For unbracketed, or untitled external links, it can convert those to citation templates. When outside of references, IABot can directly replace the external link with an archive URL to avoid disrupting the final rendering of the article. It can still handle citation templates normally. IABot avoids editing URLs inside unknown templates, or URLs hidden inside <nowiki> and HTML comment tags. IABot maintains a massive database of URL metadata that helps it to efficiently do its job. The tools mentioned above can also access this database, where users can immediately fix any issues they discover that IABot will then immediately make use of. Essentially the bot will learn as more users help it to become more reliable. The error rate is very low (<0.01%) of the edits are error prone.

Different configurations of IABot

As mentioned above, IABot is very configurable to suit the wiki as needed. IABot is {{nobots}} compliant, and offers a run page in case of malfunction. Because different language Wikipedias have different cultures and work methods, this ideology was considered during IABot's initial development. When IABot is making its run, it can be instructed to make the needed fixes on the main article and report those changes to the talk page. Alternatively, the talk messages can be switched off, or the bot can be instructed to only leave talk messages alerting editors to dead URLs it found on the respective article. IABot can proactively add archive URLs to all URLs, including the non-dead, or simply restrict its changes to the URLs tagged as dead, or even URLs it sees as dead that are untagged. IABot can be instructed to restrict its operations to only links inside <ref> tags or the entire article. A complete description of InternetArchiveBot, including the configuration options can be found at m:InternetArchiveBot.

Supported wikis

InternetArchiveBot is approved to run on the following wikis:

1. English Wikipedia
2. Swedish Wikipedia
3. Norwegian Wikipedia
4. Dutch Wikipedia
5. Chinese Wikipedia
6. Wikispecies

The big question

Would InternetArchiveBot, and/or its tools be a welcomed addition to the German Wikipedia? Keep in mind that if the bot's full autonomy is not desired that these tools exist to give user control over the bot operation. Here are some examples. Showing examples from different wikis to demonstrate the cross wiki compatibility.Chico Venancio (discussão) 18h47min de 2 de julho de 2017 (UTC)

Fully autonomous edits

• Enwiki: [1][2][3]
• Svwiki: [4][5]

User summoned bot edits

Take note that the edit summaries link to the requesting user for attribution.

• Enwiki: [6][7]
• Svwiki: [8][9]

Direct page analysis with the tools

These are edits made using the single page analysis tool mentioned above. The tool offers some options that are not available when summoning the bot. These options override the wiki set configuration the bot uses when running. Logically these options can be set for single page edits since the edit is made from the user account directly and they are responsible for the edit. Such disruption is harder to manage when the bot is being executed on a collection of pages with disruptive options.

• Enwiki: Only alter dead sources and Add archives to all URLs
• Svwiki: [10]

Discussion

Eu acho que seria ótimo. Érico (fale) 18h50min de 2 de julho de 2017 (UTC)

Seems like having that bot here would be great. I stumble upon dead links all the time, and having a bot to do this kind of work would be immensely helpful. HG (discussão) 19h11min de 2 de julho de 2017 (UTC)

Apoio --ArgonSim (discussão) 20h45min de 2 de julho de 2017 (UTC)

Support Longe-needed addition. aq (discussão) 22h13min de 2 de julho de 2017 (UTC)

Apoio Really useful ► Jackgba^Diga! 02h24min de 3 de julho de 2017 (UTC)

Support --Usien6 21h48min de 3 de julho de 2017 (UTC)

Support This is awesome, we can't miss this opportunity. Gato Preto^trovai-me! 22h21min de 3 de julho de 2017 (UTC)

Apoio --Hume42 ✉ 23h07min de 3 de julho de 2017 (UTC)

Support would be great. --DARIO SEVERI (discussão) 23h38min de 5 de julho de 2017 (UTC)

E então Érico, Holy Goo, ArgonSim, JMagalhães, Jackgba, Usien6, Gato Preto, Hume42 e DARIO SEVERI, já temos consenso. Como aplicar?-- Leon Saudanha 15h56min de 27 de julho de 2017 (UTC)

@Leon saudanha: Chamar o autor e o criador do bot: Cyberpower678. Mandei-o uma mensagem, só resta esperar a resposta. ► Jackgba 11h22min de 28 de julho de 2017 (UTC)

O bot foi aprovado unanimemente pelo consenso, e o resultado foi encaminhado para o desenvolvedor. No momento ele está corrigindo bugs para a implantação. JACKGBA ✉ 00h41min de 30 de julho de 2017 (UTC)