This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.
Hi there - for some reason MiszaBot II (talk·contribs) isn't working for Wikipedia talk:WikiProject Football although it used to do so absolutely fine. I thought it might be because one user was using unorthodox dates in their signature which the bot couldn't deal with, so I manually archived the posts. However the archive bot hasn't kicked into action despite me making that correction over 24 hours ago. Do I just need to wait a bit more or has something gone wrong? If you could take a look it would be very much appreciated, thanks! Qwghlm (talk) 10:32, 25 October 2008 (UTC)[reply]
Thank you for sorting it out! I take it was the spam link causing problem? I'll bear it in mind in future if archiving doesn't work. Qwghlm (talk) 13:09, 25 October 2008 (UTC)[reply]
The spam link too. But on top of that, a bug related to a recent change in MediaWiki was disabling the bot from creating new archive pages. Миша1318:16, 25 October 2008 (UTC)[reply]
The image was deleted as an unused unfree image, but there is now a link to it, so that the image will be used if you restore it. -- Eastmain (talk) 13:24, 11 November 2008 (UTC)[reply]
Installing when archives exist
I want to start automatic archiving of Halloween once consensus is achieved, but first I want to make sure that there aren't any particular steps I would need to take in reference to the existing archives. Is there a value I need to include to make sure it knows which archive to start with?--otherlleft (talk) 16:46, 12 November 2008 (UTC)[reply]
Set it up as usual, just remember to set the counter variable manually, to 12, if you want it to continue the current archive, or 13 to start in a new one. Миша1319:30, 12 November 2008 (UTC)[reply]
Thanks for MiszaBot III, suggestions
Thanks for running User:MiszaBot III. And special thanks for catching the case where someone is stupid and leaves the destination as the Example user. I was stupid in exactly this way and appreciate MiszaBot III not tossing my archives into a wildly wrong place. I do have two suggestions: 1. If the bot finds an obviously bogus temple (like archiving to the Example page), it could leave a "Hey, I'm ignoring your invalid request" on the talk page in question. 2. The bot makes some minor style changes to headings. example . They have no impact on the rendered page, and they do match the general style, but I was surprised to see a bot on such a specialized mission engaging in unnecessary tidying up. Thanks for running the bot! — Alan De Smet | Talk06:41, 13 November 2008 (UTC)[reply]
I thought about it, but there's a problem of avoiding duplicate notification, etc. Setting up a page for log output is more feasible.
That formatting is standard - when you create a new section, this is how it's always formatted - one blank line before the title and one after, the title itself being separated from title markers (the equal signs) by a single space on either side. And it's not any aditional workload at all. The bot just parses the entire page, breaking it up into abstract objects representing threads, which are then composed back into an output page - it's a side effect of its design. Миша1308:36, 13 November 2008 (UTC)[reply]
While it may be standard, it remains tangential to the bot's core purpose. I'd rather not a bot I invited in to help with one task decide to take up a second task. However, if it's a natural side effect of the design, and would require serious hackery to change, I'm sympathetic. It's not a big deal, just a preference. Thanks again for running it! — Alan De Smet | Talk23:38, 13 November 2008 (UTC)[reply]
Thanks . . .
. . . and sorry for creating the million-and-first screwup you needed to fix. My understanding of such coding is not terribly advanced, but I really do know that much, and I have no clue how I left the "example" code in the Talk:Halloween page.--otherlleft (talk) 11:28, 13 November 2008 (UTC)[reply]
It (obviously) creates them by itself. The default content for an archive's header is {{talkarchive}}, so it's what yuo're looking for already (though you can adjust it using the archiveheader = variable). Миша1321:34, 13 November 2008 (UTC)[reply]
Everything is in place and the destination page is up-to-date, so there's nothing for the bot to do. When the source page on Wikinews changes, it will update. Миша1300:33, 16 November 2008 (UTC)[reply]
Bot code
Hi,
I'd like to set up an archiving bot on my native Wikipedia (Romanian). What's your policy/license regarding your bot's code? Is there any dedicated page/site for the little guy? If it's closed source, can you give me a few pointers on request? Please reply here and notify me with a {{Talkback}} ("T+" in my sig). Thanks, GutzaTT+00:28, 16 November 2008 (UTC)[reply]
Thank you for the info, I happened to notice your reply (you should appreciate that watching user talk pages for the sole purpose of receiving a reply puts unnecessary strain on the servers, the Talkback template would avoid that). Unfortunately I don't speak Python, but I'll try to make it work as such. Just in case I decide for a rewrite, would you be willing to provide a couple of pointers regarding strategy? --GutzaTT+01:02, 16 November 2008 (UTC)[reply]
Sorry, but I disagree. It is making an extraneous edit to put the template on your talk page is what puts a much greater strain than putting a page on your watchlist (not to mention that you have to remove the talkback later). It has always been considered (logically) that the person initiating the thread is interested in receiving a reply and so should be the part active in seeking a reply (in other words, extra effort is expected from the poster and rightfully so). Not to mention that I'm too busy to figure out how to use yet another social template.
On that note, I'm afraid I won't be able to provide any more details than this simple algorithm (reading the HowTo might explain some terms I use):
Fetch the page to be archived and break it up into the header and a list of threads proper.
For each thread, find the newest timestamp placed in it and set that as the "thread's date".
For each thread that qualifies to be archived (algo function decides, with optional terminators such as minthreadstoarchive and minthreadsleft) evaluate the archive expression to get the archive's name and put the thread there (abort if the archive is not a subpage of the main page, unless a "magic key" is also provided and valid).
Save changed archives (it is important to do this first, because should we fail in the next step, at least we won't end up losing threads to /dev/null).
Save changed main page (updating counter in the template as necessary).
Noted, thank you for the clarification and the info, that's what I was after -- I was wondering whether you were caching the sections locally for comparison, or you were parsing the timestamps. --GutzaTT+15:27, 16 November 2008 (UTC)[reply]
medcabbot
We're making some significant changes to the way medcab requests are filed... not finalized yet, but want to give you a heads-up that some of the code'll probably need to be changed. Dunno where to start... Xavexgoem (talk) 20:04, 18 November 2008 (UTC)[reply]
I just did a manual archive on my talk page because it seems misza hasn't run on it in awhile. I had one section where the only poster had no timestamp in his sig, but I had another section stamped Nov9 that should have been archived as it was well past its time. Any idea what I borked there or is there something going on with the bot?--Crossmr (talk) 08:19, 22 November 2008 (UTC)[reply]
I have minthreadsleft=0 in my template and its been archiving all along until now. So sorry that faq doesn't answer the question. Here is what my talk page looked like just before I did the manual archive, 11 threads, the template is first (and I know it was working before) and minthreads=0. I can't see the problem[1].--Crossmr (talk) 04:13, 23 November 2008 (UTC)[reply]
In the revision you're referring to ([3]), I can see only one thread older than the threshold. In such a case, the default value of minthreadstoarchive prevents it from operating (which is not necessarily bad - the default value of 2 should not be overridden except in special cases, to reduce page history pollution). Миша1310:42, 23 November 2008 (UTC)[reply]
Signpost updated for November 17, 2008 and before.
Because the Signpost hasn't been sent in a while, to save space, I've condensed all seven issues that were not sent into this archive. Only the three issues from November are below.
Any idea why the bot has never archived the page even though the settings have been on the page for months? I can't believe I just realized that the page has never been archived automatically before. Gary King (talk)21:28, 25 November 2008 (UTC)[reply]
Archiving to /dev/null ?!
I recently added the auto-archiving to Talk:Human. However, this recent archive seems to be a problem:
Obviously, we want the archive somewhere other than /dev/null. Looking through the actual archive pages, the latest seem not to contain recent threads. Actually, looking a bit earlier:
This seems to be same problem. How do we fix this? If we can automatically move the erased material to actual archive pages, so much the better. But if manual action to get things going right in the future is needed, let's do that. LotLE×talk07:26, 29 November 2008 (UTC)[reply]
