Template talk:Setup auto archiving
| This is the talk page for discussing Setup auto archiving and anything related to its purposes and tasks. |
|
Archives: 1Auto-archiving period: 2 months  |
Default archive size
edit100K might have made sense when this template was created back in 2012, but is painfully little now. Making a moderate increase to 150K, a number that seems fairly commonplace. CapnZapp (talk) 14:44, 11 September 2021 (UTC)
- @CapnZapp I'd err on the side of safety and keep it at 100K: archive pages can take a while to load for some users and it doesn't cause much (any?) harm to have a few more of them. ― Qwerfjkltalk 20:40, 7 November 2021 (UTC)
minthreadsleft
editRaladic I'm not making sense of your edit summaries you use to justify your changes:
The 4 for ToC is old news. This isn't true since WP:VECTOR2022 anymore and even just 1 section will show a TOC.
What is "old news" and what isn't true?
minthreadsleft to 2 in line with more common usage nowadays
Says who? Who's common usage?
Generally, you fail to argue what makes |minthreadsleft=2 better. In fact, what makes it so much better for users of the new Vector skin that this justifies the inconvenience for editors still using legacy Vector?
A setting of |minthreadsleft=4 not only ensures a TOC for users of every Vector, it also leaves a healthy sampling of past discussions. This is considered friendly by many users including myself.
Of course I can see instances where old discussions are undesirable, but here we're talking what recommendations we want to give our users, especially the presumably non-power users that use templates instead of setting up bot config themselves.
I suggest we settle on |minthreadsleft=4 as a good-practice number for the general case. Individual editors can easily override the defaults if they wish or need to. CapnZapp (talk) 11:36, 28 August 2025 (UTC)
- I don’t know what doesn’t make sense to you here.
- As of 2023 there is basically no vector legacy users - Wikipedia:Vector 2022#The Vector 2022 skin on English Wikipedia, so it should not optimize for a case that does not exist. There are some people that use other skin such as mono book, but not the vector legacy skin, which basically is just gone. So you cannot argue that it’s better for uses of vector legacy if there is no statistically traceable users of such remaining.
- As for why four is too much, if you regularly jump to articles that don’t have any archiving set up, which is the point of this template, you can find pages that have four or sometimes even more threads and they’re all from a past decade, which just does not serve users that land on such a page because they will just be like” OK. Why is there all these old discussion?”
- Uses are used to searching stuff nowadays, not reading through endless amounts of old threads manually. So pruning discussions to leaving to behind, but any completed discussion that has concluded and has been untouched for 90 days, aka not a single comment has been made in three months, is likely not going to serve future users who just landed on this talk page, as again, they’re not likely to go manually read over countless threads. And having a large number of really old threads is usually intimidating, because it could mean that people don’t really care about this article. They are likely to use the search archives function of a talk page header however, given that we use uses are taught to use search nowadays, thanks to Google, Bing and co, and now more recently AI prompts which are entirely search prompt based. Raladic (talk) 15:54, 28 August 2025 (UTC)
I don’t know what doesn’t make sense to you here.
Several things.- While I'm sure the adoption rate is greater than the actual data behind your linked graph, are you aware you are linking to a 2023 data set where the English adoption rate for Vector 2022 is 2.32%? The overwhelmingly large pie slice (94.76%) is for the Vector skin. (I won't argue it is solely the legacy Vector skin, but I suspect it is).
- Even assuming Vector 2022 is 44.2%, as a legacy Vector user I find your stance unacceptable. You need to be much more careful - several orders of careful in fact - before you spout nonsense such as
there is basically no vector legacy users
,a case that does not exist
,the vector legacy skin, which basically is just gone
and, especially,there is no statistically traceable users of such remaining
(which you would have found hilarious if you sat in front of my keyboard, looking at Wikipedia using the legacy Vector skin). - But never mind. I'll simply revert your uninformed changes, not even bothering to argue the bigger issue: even if Vector 2022 were all-encompassing, I still don't find your arguments persuasive; that is "4 = bad but 2 = good", so good in fact that we simply must change our long-standing recommendation. CapnZapp (talk) 17:19, 28 August 2025 (UTC)
are you aware you are linking to a 2023 data set where the English adoption rate for Vector 2022 is 2.32%?
No, I wasn't, mea culpa, I assumed that the legacy vector is the small unlabeled red sliver at the top of the graph. So if that's not the case, we should probably not have this graph there as it's really rather confusing since the number you mentioned doesn't appear anywhere on the page, or the graph, so should probably clean that up.which you would have found hilarious if you sat in front of my keyboard, looking at Wikipedia using the legacy Vector skin).
- I didn't say there are none, I said "basically no", just that the graph didn't show more than the red sliver, which I thought was where those were in. So again, as above - based on confusion caused by graph as labeled (so it wasn't spouting nonsense, as much as getting mislead by the page/graph.- Anyway, in light of this, fair enough, if the legacy skin is in fact still used more widely, then the 4 makes sense and the other point is moot for now, so agree that there's no need to discuss it further at the moment.
- Thanks for pointing out the misunderstanding and reverting the change to 4 for default. Raladic (talk) 18:10, 28 August 2025 (UTC)