I'm sorry but we don't actually support the Opera Mini browser (https://www.mediawiki.org/wiki/Compatibility#Grade_A) and we do not support the desktop site on mobile devices. If the table of contents is important, I would recommend disabling extreme data saving mode on Wikipedia. I am sorry I cannot be of more help with this matter.

Dear Jdlrobson,

In extreme data saving, many browser features in Opera Mini are disabled. I would assume the loss of table of contents is a feature - as this mode is primarily designed to get the most data savings possible and the table of contents is arguably not an essential part of the article.

Discarding - but is way to old.

Can you please provide more information, such as the data you look at, and what you'd be expecting?

This relates to Data Quality, that's why the priority changed.

This is not an ops week task. I think this should be groommed with wikistats maintainers and prioritized accordingly. Moving back to incoming.

Opera 12.17 fell out of basic support 3 years ago, see https://www.mediawiki.org/w/index.php?title=Compatibility&type=revision&diff=2636643&oldid=2620648

Successfully resolved a long time ago, with https://www.mediawiki.org/w/index.php?title=Compatibility&type=revision&diff=2636643&oldid=2620648 as closest.

Updated browser matrix in https://www.mediawiki.org/w/index.php?title=Template:Compatibility_browser&type=revision&diff=4468318&oldid=4429345&diffmode=source.
Successfully resolved. ✨

No further objections raised here or on wikitech-l thread, so will go ahead with bumping the requirements.

Android support is a bit unclear. Even though caniuse returns Android 2.1-4.4 not supporting TLS 1.2, Browserstack's testing on Samsung/Google Nexus/HTC One M8 device( emulator)s with Android 4.3 or 4.4 access enwiki without limitations.

The principal thing this task was driving at beyond IE11/ES6 was support for archaïc Safari versions, which are not captured in any of those tasks.

The following related tasks are more specific and have been resolved since and/or are currently actively discussed instead. Consider participating there instead.

Email sent out to wikitech-l & design.public for further comments. Countdown until 18 March for objections, if none we'll bump the requirements then.

@Izno The main difference is, that we can't say for sure for Firefox 24-26, that they have the correct ciphers enabled. The Firefox 24-26 stats are showing between 280-400k per week, averaging at about 340k out of 3.6 to 4 billion total for weeks of 2021.
For completion, IE 9 10 have about 2 million views per week in the same timeframe.

Since it's not clear to me, what is the difference between Firefox 24-26 and IE9/10 as in the task description? They both describe being able to use TLS1.2 but the former is supposedly an advanced feature (Firefox users, I would venture, would be more willing to turn such a thing on..., especially if they're still on these versions).

Hey folks, I just wanted to chime in here and say that from a product perspective, paying any cost or effort to continuing to support browsers that we don’t support on our flagship products feels like something we should stop. I understand that there are 3rd party wikis that might choose to make different decisions about TLS support, but the value of increasing development velocity and improving user experience on Wikipedia (visited by 1.5B devices a month) vastly outweighs the negative consequences. Let me know how I can help communicate this if you feel a communication for this is necessary.

@Krinkle Updated stats in task description should be sufficiently clear to move this forward. :)

Volker wrote on 23 Nov 2020:

Adding this task to our Web team kanban board for possibly looking closer at the browser stats.

it's 2020, support for generic web notifications is wide enough now.