We need a team of wiki-gnomes. Not quite sure how recruitment should
proceed, though. Maybe we should start by asking everyone with edit
permissions? A weekly crawl to identify broken links would be a great help.
On Tue, Aug 8, 2017 at 4:52 PM, Mats Wichmann <m...@wichmann.us> wrote:
> In an ideal world, interested parties fix things on the wiki when they
> find they have problems. And I'm heartened there's been a spate of
> link-fixing contributions recently, system working the way it's supposed
> to :)
> Does it make any sense to think about any kind of more systematic
> maintenance activities? Is wiki (content) maintenance even a possible
> thing? They are generally pretty notorious for this: it's a wiki, you
> can update it, but if you didn't just update it it's probably out of
> date and inaccurate.
> There are really two things...
> - errors. links go stale all the time, is there any point in looking
> for a tool that could crawl the wiki infrequently (on the order of once
> every many months) to test if links outside the wiki still resolve to
> something and report problems?
> - quality. Here are a couple of examples I ran into just this morning
> looking at some things that landed in the inbox.
> The page on beginner errors:
> got an update (broken link fixing only), and I was curious and looked at
> it. It's not terribly useful as is - the notes also indicate it was
> imported from elsewhere. This *could* be a very useful page: as a
> python-tutor participant, I see common themes repeatedly; it would also
> be great to hear from instructors on what trips students up (I haven't
> taught a Python class since 2002 so I don't have any current
> contributions on that front). This page tried to do that, but then goes
> off in strange weeds... "The latter issue has been fixed, but the former
> has not and very likely never will". Python is what it is, how about
> ideas for how to teach people in an anticipatory way to avoid well-known
> gotchas instead of focusing on how the language might or might not be
> "fixed"? Page still has references to "Python 3000", which means it's,
> ummm, fairly dated :)
> One of the resources pages,
> also had link-fix activity. this page suffers from the "list of
> resources" problem, where a bunch of links are just dumped here with no
> ordering (how do you order them?), and once the lists get more than a
> few entries, it's hard to know what to follow. And every so often we can
> expect to go through "moving my link up to a better place" wars like
> we've seen before. I found the Google class listed in two sections,
> both with old (but not yet broken) links to obsolete Google Code
> addresses, cleaned that up on the way past (there's only one copy now).
> Any thoughts on how resources like this could be made more useful?
> Maybe keep a wiki page of "if you feel like contributing to Python, here
> are some pages that could stand improvement"?
> -- mats
> pydotorg-www mailing list
pydotorg-www mailing list