User talk:Dvandersluis/Archive 2
This is an archive of past discussions about User:Dvandersluis. 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. |
Archive 1 | Archive 2 | Archive 3 |
CbmBOT running
Is CbmBOT running? The stats for category cleanup by month haven't been updated for a couple days. RJFJR 14:00, 8 May 2007 (UTC)
- Sorry about that, I made a change elsewhere in my codebase, and it stopped CbmBOT from running, without my realizing it. I just ran the bot, and it should continue it's nightly update tonight. —Daniel Vandersluis(talk) 14:45, 8 May 2007 (UTC)
- Thank you. The statistics the bot produces are appreciated. RJFJR 16:19, 8 May 2007 (UTC)
- Not a problem. I've actually been thinking about expanding the stats given in the table (something like ±x since yesterday) but haven't gotten to it yet. Also, the bot is going to be replaced by User:StatisticianBot in the near future; the function will be exactly the same as now, but I am moving all my related bots under one umbrella account. —Daniel Vandersluis(talk) 16:22, 8 May 2007 (UTC)
- Thank you. The statistics the bot produces are appreciated. RJFJR 16:19, 8 May 2007 (UTC)
Main GA page
Would it be possible to adapt the statisticbot to also total up the amount of article on the main GA page also? It would also save us alot lot of time from having to update the count manually. Tarret 21:53, 9 May 2007 (UTC)
- This can likely be done, but would have to be approved as an additional task at WP:B/RFA in order to be added. Once the current tasks are approved, we can take a look at implementing this as well. —Daniel Vandersluis(talk) 13:36, 10 May 2007 (UTC)
GA bot - error or someone fooling around?
G'day - For your information I note as a GA assessor that this is coming up on the GA candidates report at
Holds over 7 days old
1. Sunshine (2007 film) (+13642 days)
Not sure if it is the bot or someone fooling around? Cheers!--VS talk 08:48, 16 May 2007 (UTC)
- Sorry about that, I messed something up in the code yesterday without realizing it, and it caused the bot not to parse the page properly. It's been fixed, and I re-ran the report. —Daniel Vandersluis(talk) 12:49, 16 May 2007 (UTC)
Hey there, no apology required - it's a great bot and I was just letting you know about the minor error. Best wishes.--VS talk 13:06, 16 May 2007 (UTC)
Two bot notes
Daniel, a couple of things. Would it be possible to have the backlog list formatted as
first item
, second item
, third item
, fourth item
, fifth item
and similarly for the 6-10 list? That might help avoid the "trailing comma" mistake that people often make when they do a manual update, and it wouldn't change the formatted output.
Second, there was a discussion on the GAC talk page about using the GAC category as a way to track down nominees for the count. I hadn't thought of that; is there any technical problem there or could this go on the list of possible enhancements for StatBot? The idea was that this could be used to identify nominees that weren't listed correctly on the GAC page, and would generate another section of the exception report.
By the way, where should I post notes like this; here or at User talk:StatisticianBot?
-- Mike Christie (talk) 16:58, 17 May 2007 (UTC)
- Here is good, I don't generally log into to StatBot myself so here I get the new message notice. I did notice the discussion about counting the category and finding articles listed there but not on the GAC page (I assume vice versa wouldn't happen). It should be fine, but I need to modify the bot to be able to count categories first. I will also have to apply to add a task to the bot, but that shouldn't be a problem I don't think.
- I could change the format, but note that that would put a space in between article link and the comma. Something like this: X , Y , Z ... are you sure that's what you want? If you do want it like that, maybe a different punctuation (or a bullet) would be a better separator? —Daniel Vandersluis(talk) 17:11, 17 May 2007 (UTC)
- Hmm. Hadn't realized there would be a space there; best to leave it as is, then, I think. The bullets (etc.) wouldn't work well because of the inclusion in the GA task template. Re "vice versa"; hey, who knows what people will do when they're confused. I wouldn't be prepared to bet you won't see any of them; it would indicate they'd done the GAC page part but not added the right template to the talk page, wouldn't it? Quite possible. Thanks -- Mike Christie (talk) 17:17, 17 May 2007 (UTC)
- By bullets, I meant something like X • Y • Z ... that way, the spaces are supposed to be there. Another option would be —'s or –'s. With a separator like this, the separator can be placed as the first character of each line and it will still appear properly.
- You're probably right. It shouldn't be too hard to check either way. I'd make articles on the page that aren't given the right category (I guess I'd have to check Category:Good article nominees currently on hold too?) marked as such under malformed nominations, and the opposite will have their own category, since there is nothing to link to on the GAC page.
- Speaking of malformed nominations, I've been trying to fix them up from the report, but I've been leaving a couple alone -- specifically those with a strange time format (see Canadian Gaelic or Glenn Knight) or with nominator problems. I don't want to subvert people's signatures, but these nominations need to have the data. What should be done in this case? Generic username wikilinks? I've been adding the timestamp where necessary, except for the above mentioned with the strange format. —Daniel Vandersluis(talk) 17:25, 17 May 2007 (UTC)
- OK, those bullets would look good. I suggest we try that and see if anyone dislikes it. For the malformed noms, it seems like there are three options, other than manually fixing them (which isn't a viable longterm strategy, of course): expand your regexps to deal with them; store an internal lookup that assigns the current day/date when you first find them, so you can "translate" them; or read the diffs to find when they were added. I would think the third option is hard and is the most server-intensive--getting a set of additional diff URL's each night to append to an existing set wouldn't be too bad, but parsing each to find what was added would be a huge pain. The regexps seem the easiest way to go. It might also be possible to update the page to append a standard day-date, once it's parsed, as a gentle hint to contributors to please use standard forms. Mike Christie (talk) 10:56, 18 May 2007 (UTC)
- I changed the bot to use bullets instead of commas. I did it manually today, starting tomorrow it will do it automatically. In terms of malformed dates, updating the regexes isn't viable either, because there are infinite formats that a date can be written in. I didn't even know there was a way on Mediawiki to automatically add a date in a non-standard format. The other methods mentioned would only work with an active (patrolling) bot; as StatBot runs only once a day on the GAC page, it can't track all the data that was added that day; it can only work on the whole scope. I was actually thinking about having a second bot to watch for changes, but I do not have the time to do that yet. It's something we can keep in mind for later, in any event. —Daniel Vandersluis(talk) 12:47, 18 May 2007 (UTC)
- Well, yes, I guess there are effectively an infinite number of date formats. Even if you could take care of more of them with regexps, I think it would lead to constant tinkering, and that's undesirable. However, I think the second approach (storing a date when a nom is encountered with an unparseable date) could work OK. After all, the date would be no more than 24 hours out of date, and if that date is only going to be used for displaying aging holds (etc.) in the exception report, and for determining the order in the backlog list, that would give acceptably accurate outcomes, wouldn't it? So a patrolling approach would be unnecessary, I think. Mike Christie (talk) 16:00, 18 May 2007 (UTC)
- I changed the bot to use bullets instead of commas. I did it manually today, starting tomorrow it will do it automatically. In terms of malformed dates, updating the regexes isn't viable either, because there are infinite formats that a date can be written in. I didn't even know there was a way on Mediawiki to automatically add a date in a non-standard format. The other methods mentioned would only work with an active (patrolling) bot; as StatBot runs only once a day on the GAC page, it can't track all the data that was added that day; it can only work on the whole scope. I was actually thinking about having a second bot to watch for changes, but I do not have the time to do that yet. It's something we can keep in mind for later, in any event. —Daniel Vandersluis(talk) 12:47, 18 May 2007 (UTC)
- OK, those bullets would look good. I suggest we try that and see if anyone dislikes it. For the malformed noms, it seems like there are three options, other than manually fixing them (which isn't a viable longterm strategy, of course): expand your regexps to deal with them; store an internal lookup that assigns the current day/date when you first find them, so you can "translate" them; or read the diffs to find when they were added. I would think the third option is hard and is the most server-intensive--getting a set of additional diff URL's each night to append to an existing set wouldn't be too bad, but parsing each to find what was added would be a huge pain. The regexps seem the easiest way to go. It might also be possible to update the page to append a standard day-date, once it's parsed, as a gentle hint to contributors to please use standard forms. Mike Christie (talk) 10:56, 18 May 2007 (UTC)
- Hmm. Hadn't realized there would be a space there; best to leave it as is, then, I think. The bullets (etc.) wouldn't work well because of the inclusion in the GA task template. Re "vice versa"; hey, who knows what people will do when they're confused. I wouldn't be prepared to bet you won't see any of them; it would indicate they'd done the GAC page part but not added the right template to the talk page, wouldn't it? Quite possible. Thanks -- Mike Christie (talk) 17:17, 17 May 2007 (UTC)
StatBot didn't run last night
Just letting you know: looks like there was no GAC report last night -- is there a problem? Mike Christie (talk) 11:30, 28 June 2007 (UTC)
- Thanks for letting me know. I'm not sure what caused it, but I took a look at the logs and the bot couldn't parse the nominations section last night for some reason. I just ran the bot manually though and it worked fine, so I'm not exactly sure what happened. It's possible some intermediate edit removed or altered the special comments that are in the GAC page, but I haven't seen this in the edit history.
- In any event, it appears to be working again (it ran successfully when I ran it manually, as stated). Let me know if you notice any other problems. —Daniel Vandersluis(talk) 13:22, 28 June 2007 (UTC)
Another possible bot function
Daniel, there's another possible function for the bot that I wondered if you would be interested in. Every now and then I take a look at the GAC exceptions and post a note to the talk page of the editor who has an overdue hold or review tag on a nom. Is that something that could be automated? That is, if a nom is in the overdue hold or report section, the bot would parse out the name of the person who put the nom on hold or on review, and would post a note to that editor's talk page saying something like "This is a reminder that you place Some Article on hold for Good Article status seven days ago. Ideally a hold lasts no more than seven days; if there is no ongoing activity aimed at improving the article, please consider passing or failing the article as it stands. Thank you." What do you think? Mike Christie (talk) 14:10, 28 June 2007 (UTC)
- Sounds like an interesting proposal. The bot already parses out the username of the person who is attributed to the GAOnHold or GAReview template, so we have that information. What would have to be done is send the message, ensure that a user doesn't get more than one message per nomination, allow some way for users to opt out (see Template:bots) from receiving automatic messages, and ensure that no more than the allowable per-minute rate of edits was being used. In terms of the first part, it'd be best if there was a template that the bot could use (with a article name parameter, and whatever other parameters that are necessary). The rest would need to be coded to augment the existing function. Of course, this means another request to WP:BRFA that would have to be approved.
- I could probably do this (and would be glad to), however, I am quite busy at the moment, so I cannot definitively say when I will have to time. I will put it on my TODO list, however. —Daniel Vandersluis(talk) 14:45, 28 June 2007 (UTC)
Hi. I'm sorry to bother you, but as a LoCE member, I just wondered if you would be willing to have a look through the Kent article. It is currently a Featured Article Candidate and needs a copy-edit for grammar by someone who hasn't yet seen it. Any other ways to improve the article would also be welcome. Thank you very much, if you can. Epbr123 21:02, 12 July 2007 (UTC)
Surprising exception situation for StatBot
Hi -- the GAC report has an unusual exception; take a look! I'm not sure this will happen often enough to be important to fix, but I thought you'd like to know. Mike Christie (talk) 10:02, 25 July 2007 (UTC)
- I'm not actually sure what I'm looking at; could you elaborate as to what the problem is? —Daniel Vandersluis(talk) 16:48, 25 July 2007 (UTC)
- Oops; wrong link -- sorry. This is what I meant to post; we're down below ten outstanding noms, and the bot is printing templates for the unused places, looks like. Mike Christie (talk) 17:07, 25 July 2007 (UTC)
- Oh, heh. That's a good point. Good job getting below ten nominations! That's a bug that I'll fix, as soon as I get a chance. —Daniel Vandersluis(talk) 22:06, 25 July 2007 (UTC)
- Oops; wrong link -- sorry. This is what I meant to post; we're down below ten outstanding noms, and the bot is printing templates for the unused places, looks like. Mike Christie (talk) 17:07, 25 July 2007 (UTC)
Re: Toronto Meetup
Hello Dvandersluis/Archive 2, I saw you name on the Wikiproject Toronto page and I would like to inform you about a Wiki Meetup that is being organized. If you are interested, feel free to add your input on the Toronto Meetup talk page. Regards, Nat Tang ta | co | em |
Please note that the categories on the [[WP:GAC|Good article candidates' page has been updated today to make the categories match the categories on the main GA page. I predict that this will most likely cause issues with how the bot generates its GA status reports at Wikipedia:Good article candidates/Report, so it will have to be updated. Sorry about the inconvenience. Thanks! Dr. Cash 05:49, 10 August 2007 (UTC)
- Thanks for the heads up. It should actually not break the bot; the categories weren't hard coded, but taken out of the wikitext when the bot analyzes the page. Everything appears to be still working, but if you run into a problem, feel free to let me know. —Daniel Vandersluis(talk) 12:51, 10 August 2007 (UTC)
Please re-register
CbmBOT
Is the CbmBOT updating the counts at Category:Cleanup by month? RJFJR 18:08, 29 August 2007 (UTC)
- It's supposed to be doing so. It stopped updating on August 23rd though, and according to the logs, it's having trouble parsing the October 2005 category. Of course, that category is now empty (but I thought I had covered this possibility before). I will take a look at the code, but the bot will most likely start working again if the category is deleted. —Daniel Vandersluis(talk) 19:00, 29 August 2007 (UTC)
- Okay it's fixed now. The problem seems to be that the text in an empty category is different now than before, and so the bot couldn't parse it. It should be working fine now, and I've run the bot manually for today -- it will resume automatic updating as normal at 4:30 am PST tomorrow. In the future, if you see that the bot has missed an update, let me know -- it's 99.9% a parse error. The bot won't cancel its future runs, so the only way the bot won't update is if it encounters a problem when trying to do so. —Daniel Vandersluis(talk) 19:10, 29 August 2007 (UTC)
- Thank you. I'd only just noticed that the stats weren't updating. RJFJR 19:28, 29 August 2007 (UTC)
- No worries. Just thought I would point it out incase you thought that you needed to wait a couple days on error. Thanks for the report :) —Daniel Vandersluis(talk) 19:33, 29 August 2007 (UTC)
- Thank you. I'd only just noticed that the stats weren't updating. RJFJR 19:28, 29 August 2007 (UTC)
Rc Template
Hey, thanks for changing the red link on {{Roller coaster}}, but I intended that link to have that go there. I was going to create a page much like Roller coaster inversions, with a general history of how all the elements came together, or perhaps cover all the individual element pages that don't have enough material to be alone. Is it not a good idea? ALTON .ıl 00:48, 30 August 2007 (UTC)
- Oh, no problem. Feel free to change it back. I just thought that it'd be better to point to the category because I knew it existed, but if you're going to create an article, that's cool. —Daniel Vandersluis(talk) 01:53, 30 August 2007 (UTC)
You're awesome
Just wanted to let you know that you're totally awesome, even when you're responsible for my slacking off at work.
p.s. Let's go to Wonderland next summer and take pictures y/y? --Forkyfork 13:59, 30 August 2007 (UTC)