User talk:Shadow1/Jun2007

Archiving bug
Check out this weird archiving bug. It didn't correctly recognize the end of a comment because it contained a line beginning with = inside of a nowiki section. It needs to be improved to recognize that anything within a nowiki section is still part of the comment.


 * The archive addition
 * The talk page removal

-- Cyde Weys 21:14, 16 May 2007 (UTC)


 * Should be fixed now, thanks for letting me know. Shadow1  (talk) 18:39, 23 May 2007 (UTC)

Shadowbot reverted me! Fix it!
Regarding edits to Backhanded compliment ... I don't mind that it reverted me as much as it is now leaving me threatening messages. Not cool. --evrik (talk) 17:45, 24 May 2007 (UTC)

Bot clash
Hi Shadowbot reverted User:SmackBot, not a problem, and I do need to know (otherwise Smacky will keep making the same change). However it would be great if Shadowbot could notify me on my talk page, perhaps via reading the This user account is a bot operated by Rich Farmbrough. off SB's user page. Also you might consider honouring the nobots on the talk page. Regards, Rich  Farmbrough 13:08 24  May 2007 (UTC).
 * We're currently migrating Shadowbot's link-searching backend to a Perl/MySQL based program, and the user-exemption list hasn't been transferred yet, otherwise there wouldn't have been a problem. Nevertheless, I'll see if I can have Shadowbot honor . Shadow1  (talk) 18:47, 24 May 2007 (UTC)
 * Cheers. Rich Farmbrough, 09:57 26 May 2007 (GMT).

Left you a question
Hi. I left you a question about perlwikipedia at User talk:Shadow1/perlwikipedia. All the best, – Quadell (talk) (random) 16:48, 25 May 2007 (UTC)
 * And again. – Quadell (talk) (random) 19:56, 25 May 2007 (UTC)

perlwikipedia problems
Greetings. I'm having trouble getting perlwikipedia to work correctly. I put a request in at User talk:Shadow1/perlwikipedia. If you could look it over, I'd really appreciate it. Thanks, – Quadell (talk) (random) 14:28, 28 May 2007 (UTC)

Signpost updated for May 28th, 2007.
You are receiving this message because you have signed up for the Signpost spamlist. If you wish to stop receiving these messages, simply remove your name from the list. Ralbot 06:52, 29 May 2007 (UTC)

Re: Touregypt.net
Hi Shadow -- I've been approached by a concerned Wikipedian in good standing that Shadowbot blocked a link to this site. Since linking to this site was the subject of a very vicious argument not too long ago, I'm surprised that it was once again rated as a spam site. How did this get relisted? How do I get its address removed from Shadowbot's config file? -- llywrch 15:44, 1 June 2007 (UTC)
 * My apologies for the reversions, I didn't remove the .net rule the last time I went to remove the site from the blacklist. The link should be fine now. Shadow1  (talk) 19:01, 1 June 2007 (UTC)
 * Okay, thanks! -- llywrch 23:03, 1 June 2007 (UTC)

Shadowbot reverted my change.
I made a change to Dave Mustaine, and I need to show some sources or a fanboy will revert the change... this was the change i made... please help! Spazm 13:56, 4 June 2007 (UTC)

Shadowbot (Special Request)
Hey, I realize that you probably can't do this, but I figure it's worth asking about. Is there any way to have Shadow bot change the heading level (like from two equals to three) of what it archives? Cause that would be amazing. Thanks for the info -- daniel folsom  15:19, 4 June 2007 (UTC)

Spamming with links to photobucket
I wanted to come by and apologize for linking to photobucket I did not relieze I was even spamming I was just trying update the ranger pages in reguards to the team up episode of Power Rangers: Operation Overdrive and I forgotten how to upload pics here on wikipedia could you help me out with that please? That was my ip number I will not do it again.
 * Read the Image tutorial as the bot's FAQ page stated. Shadow1  (talk) 12:08, 5 June 2007 (UTC)

Signpost updated for June 4th, 2007.
You are receiving this message because you have signed up for the Signpost spamlist. If you wish to stop receiving these messages, simply remove your name from the list. Ralbot 08:00, 5 June 2007 (UTC)

Re: Touregypt.net
Hi Shadow -- I've been approached by a concerned Wikipedian in good standing that Shadowbot blocked a link to this site. Since linking to this site was the subject of a very vicious argument not too long ago, I'm surprised that it was once again rated as a spam site. How did this get relisted? How do I get its address removed from Shadowbot's config file? -- llywrch 15:44, 1 June 2007 (UTC)
 * My apologies for the reversions, I didn't remove the .net rule the last time I went to remove the site from the blacklist. The link should be fine now. Shadow1  (talk) 19:01, 1 June 2007 (UTC)
 * Okay, thanks! -- llywrch 23:03, 1 June 2007 (UTC)

Shadowbot reverted my change.
I made a change to Dave Mustaine, and I need to show some sources or a fanboy will revert the change... this was the change i made... please help! Spazm 13:56, 4 June 2007 (UTC)

Shadowbot (Special Request)
Hey, I realize that you probably can't do this, but I figure it's worth asking about. Is there any way to have Shadow bot change the heading level (like from two equals to three) of what it archives? Cause that would be amazing. Thanks for the info -- daniel folsom  15:19, 4 June 2007 (UTC)

Spamming with links to photobucket
I wanted to come by and apologize for linking to photobucket I did not relieze I was even spamming I was just trying update the ranger pages in reguards to the team up episode of Power Rangers: Operation Overdrive and I forgotten how to upload pics here on wikipedia could you help me out with that please? That was my ip number I will not do it again.
 * Read the Image tutorial as the bot's FAQ page stated. Shadow1  (talk) 12:08, 5 June 2007 (UTC)

Signpost updated for June 4th, 2007.
You are receiving this message because you have signed up for the Signpost spamlist. If you wish to stop receiving these messages, simply remove your name from the list. Ralbot 08:00, 5 June 2007 (UTC)

Shadowbot3 Consistency of Running
Anything we can do to help get this bot to run daily? -- Auto ( talk / contribs ) 18:56, 6 June 2007 (UTC)
 * I'm working on getting it onto the Toolserv. Other than that, you could buy me a new machine :) Shadow1  (talk) 19:28, 6 June 2007 (UTC)

It hasn't archived my talk page since May 27, and I have it set for Archive Age-4. What's going on? Is the bot not functioning properly? I know I have the correct code, as I have not changed anything. Any ideas as to what may be wrong, or is this universal? (talk page source here, last edit by Shadowbot3 here. (If I had the funds, say, if I had more money than God, I would buy you a new machine...) -- tennis man  23:37, 6 June 2007 (UTC)
 * Running a job right now. Some users' talk pages had Unicode characters, which Perlwikipedia didn't support until... about 30 seconds ago. Shadow1  (talk) 13:24, 14 June 2007 (UTC)

Another problem with Perlwikipedia on Windows
Greetings, Shadow1. As you know, I'm running the latest Perlwikipedia on a Windows system with ActivePerl, and I've run into a problem. When I try to write to a file using all standard ASCII, everything works fine. But if I try to write any extended-ASCII characters (such as Chinese characters), the edit call has no effect.

As an example, I have run the following code: my $wiki = $pw->get_text('User:Quadell/a'); #This page has only standard characters $wiki .= "\nA small change."; $pw->edit('User:Quadell/a', $wiki, "Testing"); This works as expected. But if I use User:Quadell/b instead, which has a Korean character, the command has no effect. The practical upshot of this is that I can modify very few Wikipedia pages, since most of them have at least inter-wiki links with extended characters.

I have tried the same task on a Linux system running an old version of Perlwikipedia (before you changed ->content to ->decoded_content), and it works just fine. I haven't been able to test the latest version of Perlwikipedia on Linux, because for some reason I can't install XML::Simple. I'm curious to know whether you are able to use the latest Perlwikipedia on Linux to modify a page like User:Quadell/b. – Quadell (talk) (random) 20:15, 9 June 2007 (UTC)
 * Nope. I'm guessing that I need to set the encoding properly, but Perl segfaults whenever I tell it to use Unicode. I'll keep working on it. Shadow1  (talk) 15:38, 12 June 2007 (UTC)


 * I don't know much about Perl but when I customized Perlwikipedia for a none-wikipedia wiki with german umlauts I found a workaround that works for me. I noticed that in line 312 in Perlwikipedia.pm (subroutine get_text) the internal variable $1 delivers unicode but for some reason I don't really understand in $wikitext it is converted into ASCII. So I did the following:

use Encode; # added $wikitext = encode("utf8", $1); # changed from $wikitext = $1


 * Everything else seems to work fine. Unicode gave me no trouble with these modifications.
 * André, 88.72.249.10 09:09, 14 June 2007 (UTC)
 * I've updated Perlwikipedia to automatically encode the output of get_text and the page text of edit as UTF8. That should resolve any issues. Shadow1  (talk) 22:13, 14 June 2007 (UTC)

Signpost updated for June 11th, 2007.
You are receiving this message because you have signed up for the Signpost spamlist. If you wish to stop receiving these messages, simply remove your name from the list. Ralbot 03:08, 13 June 2007 (UTC)

Shadowbot3 Consistency of Running
Anything we can do to help get this bot to run daily? -- Auto ( talk / contribs ) 18:56, 6 June 2007 (UTC)
 * I'm working on getting it onto the Toolserv. Other than that, you could buy me a new machine :) Shadow1  (talk) 19:28, 6 June 2007 (UTC)

It hasn't archived my talk page since May 27, and I have it set for Archive Age-4. What's going on? Is the bot not functioning properly? I know I have the correct code, as I have not changed anything. Any ideas as to what may be wrong, or is this universal? (talk page source here, last edit by Shadowbot3 here. (If I had the funds, say, if I had more money than God, I would buy you a new machine...) -- tennis man  23:37, 6 June 2007 (UTC)
 * Running a job right now. Some users' talk pages had Unicode characters, which Perlwikipedia didn't support until... about 30 seconds ago. Shadow1  (talk) 13:24, 14 June 2007 (UTC)

Another problem with Perlwikipedia on Windows
Greetings, Shadow1. As you know, I'm running the latest Perlwikipedia on a Windows system with ActivePerl, and I've run into a problem. When I try to write to a file using all standard ASCII, everything works fine. But if I try to write any extended-ASCII characters (such as Chinese characters), the edit call has no effect.

As an example, I have run the following code: my $wiki = $pw->get_text('User:Quadell/a'); #This page has only standard characters $wiki .= "\nA small change."; $pw->edit('User:Quadell/a', $wiki, "Testing"); This works as expected. But if I use User:Quadell/b instead, which has a Korean character, the command has no effect. The practical upshot of this is that I can modify very few Wikipedia pages, since most of them have at least inter-wiki links with extended characters.

I have tried the same task on a Linux system running an old version of Perlwikipedia (before you changed ->content to ->decoded_content), and it works just fine. I haven't been able to test the latest version of Perlwikipedia on Linux, because for some reason I can't install XML::Simple. I'm curious to know whether you are able to use the latest Perlwikipedia on Linux to modify a page like User:Quadell/b. – Quadell (talk) (random) 20:15, 9 June 2007 (UTC)
 * Nope. I'm guessing that I need to set the encoding properly, but Perl segfaults whenever I tell it to use Unicode. I'll keep working on it. Shadow1  (talk) 15:38, 12 June 2007 (UTC)


 * I don't know much about Perl but when I customized Perlwikipedia for a none-wikipedia wiki with german umlauts I found a workaround that works for me. I noticed that in line 312 in Perlwikipedia.pm (subroutine get_text) the internal variable $1 delivers unicode but for some reason I don't really understand in $wikitext it is converted into ASCII. So I did the following:

use Encode; # added $wikitext = encode("utf8", $1); # changed from $wikitext = $1


 * Everything else seems to work fine. Unicode gave me no trouble with these modifications.
 * André, 88.72.249.10 09:09, 14 June 2007 (UTC)
 * I've updated Perlwikipedia to automatically encode the output of get_text and the page text of edit as UTF8. That should resolve any issues. Shadow1  (talk) 22:13, 14 June 2007 (UTC)

Signpost updated for June 11th, 2007.
You are receiving this message because you have signed up for the Signpost spamlist. If you wish to stop receiving these messages, simply remove your name from the list. Ralbot 03:08, 13 June 2007 (UTC)

Shadowbot3 and unicode
Hi,

It seems that your bot does not handle unicode very well. . Scroll to the bottom and see my signature "球球PK" obstructed. --ChoChoPK (球球PK) (talk | contrib) 15:16, 16 June 2007 (UTC)


 * I just saw this on my talk page as well. I left a note at User talk:Shadowbot3. Mike Dillon 15:20, 16 June 2007 (UTC)


 * Same here: this diff shows the problems. Here are the letters causing trouble on my page: Æ gets converted to Å, ’ to â, ø to Ã¸, and – to â.  I don't mean reverting and manually archiving, if that's easier till you fix it; if you would drop a note here letting me know how you plan to handle this that would be helpful.  Thanks.  Mike Christie (talk) 16:05, 16 June 2007 (UTC)


 * I've blocked Shadowbot3 for 24 hrs as requested on WP:AN. Of course, if it's fixed before 24 hours is up, go ahead and unblock. I probably won't be here to do it. Thanks - Krakatoa  Katie  17:56, 16 June 2007 (UTC)

Unicode patch
Hi, I did some testing myself. On the newest ActiveState Perl on Windows (5.8.8 build 820), using WWW-Mechanize 1.30 and libwww-perl 5.805), the following patch seems to work:

sub edit { my $self    = shift; my $page    = shift; my $text    = Encode::encode("utf8", shift); my $summary = Encode::encode("utf8", shift);


 * 1) [...]

I figured it's an issue not with getting the unicode characters from the server, but with posting them back there again. The submit routine of the form object is apparently too dumb to understand Unicode; you need to pass Unicode data into it but declare it as if it were simple byte data.

Up to the point where you pass the data into the "edit" routine, everything seems to be working fine in Unicode anyway.

Since there might be an issue about dependencies on certain versions of the libraries, it might be a good idea to make those explicit in the "use" statements, e.g.:

use HTML::Form v1.54 use WWW::Mechanize v1.30.0; use Encode;

Fut.Perf. ☼ 09:38, 17 June 2007 (UTC)


 * That's more or less the same code that caused the problem. I've been messing around with using the "utf8" pragma and the upgrade call, but nothing seems to be working at present. Shadow1  (talk) 16:27, 17 June 2007 (UTC)
 * Ah, I see. Anyway, what I've found out so far is that the strings all arrive in Perlwikipedia safely and correctly as UTF-8 as they should, in my environment. I'd recommend leaving the strings as they are as long as you have them in Perwikipedia so they can be manipulated there safely, and only do whatever conversions are needed when you have to pass them back. What is happening on your system when you pass them back in "encoded" like this, again the double unicode conversion like yesterday? You could perhaps try the following. Get the HTTP::Form object into which Mechanizer is going to paste the edit string; get its previous content before you paste the new edit string into it; check whether the Form object thinks that string is unicode or not (doing "utf8::is_utf8" on it); then pass the new string in the same format the old one is in. In my case, is_utf8 returned false on that string, so I decided I'd use "encode("utf8", $...)" to pass the new string bytewise. Fut.Perf. ☼ 17:25, 17 June 2007 (UTC)

Warning
"Additionally, if you do not refrain from adding unverifiable information into Wikipedia articles, you will be blocked temporarily for disruption and vandalism."


 * that's a bit mean isn't it?--Emperor Tony X. Liu 19:50, 17 June 2007 (UTC)
 * No, it's the truth. Your edits are classified as vandalism, and you will be blocked per Wikipedia's blocking policy if you continue making unconstructive edits. Shadow1  (talk) 19:51, 17 June 2007 (UTC)

example.org
Hey! Example.org is, AFIK, exactly the correct URL to add per RFC 2606 (http://www.rfc-editor.org/rfc/rfc2606.txt). Shadowbot should not be forcing me to use a real URL when I mean to give an example. --BryceN 21:15, 17 June 2007 (UTC)

Ninja Warrior
Tried to add a fansite to the external links but shadowbot undid it. see here

Any way to stop shadowbot doing this?
 * Yes, please do not add non-notable fansites to the article. See Wikipedia's external links guidelines for more information. Shadow1  (talk) 18:56, 18 June 2007 (UTC)

Signpost updated for June 18th, 2007.
You are receiving this message because you have signed up for the Signpost spamlist. If you wish to stop receiving these messages, simply remove your name from the list. Ralbot 09:14, 19 June 2007 (UTC)

I submitted a purely informational site (no advertising or promo) on the NH Lakes Region - why so quick to claim as spam?
Your bot's immediate dismissal of content is a disservice rather than an assistance if it doesn't take into context the content link submitted. What is your protocol for what is considered "spam" with regards to your bot? I'd like to link this free encyclopedic website of the New Hampshire Lake's Region to those entries concerning the website. Please peruse for yourself -- http://www.lakesregionrelocation.com -- it makes no personal or commercial references whatsoever anywhere on the site.

Thanks very much. — Preceding unsigned comment added by LakesRegionExpert (talk • contribs)
 * Hello, thanks for your remark. I indeed saw you adding the link to multiple sites, and I left you the first warning on your talkpage.  The warnings point you to our spam guideline, our guideline on external links and the policy on what wikipedia is not (in this case the section WP:NOT is of interest, wikipedia is not a linkfarm).  I am afraid your link does not comply with these policies and guidelines.  Hope this explains, but feel free to ask again when you have more questions.  --Dirk Beetstra T  C 15:47, 19 June 2007 (UTC)

Block
Why the hell did you block me?!?--Emperor Tony X. Liu 16:41, 19 June 2007 (UTC)
 * For ignoring my warnings and edit warring over the page. I hate to break it to you, mate, but you're not the emperor of China, nor did you order the construction of the Great Wall. If you continue to say so, you will be facing an indefinite block to order to prevent further vandalism of the article. Shadow1  (talk) 17:49, 19 June 2007 (UTC)
 * Oh, hmm. Never mind then. Shadow1  (talk) 17:52, 19 June 2007 (UTC)

Shadowbot3 / Perwikipedia
Hi, glad to see your bot is up and running again. What was the solution? I looked again at the version currently on the download site, and noticed you still/again(?) have the "encode("utf8"...)" statement in the get_text sub. If that's the current version, I'd like to repeat my suggestion that it might be better to have that conversion only in the edit sub, when you commit the text back. It doesn't make much of a difference in the case of Shadowbot3, which is basically just about writing text back unchanged, but it will break if somebody uses it in a way where they actually change strings after getting them from get_text and before writing them back. The encode("utf8"...) function results in a text representation that is actually corrupted and unsuitable for any further text manipulation, such as concatenation with other (unicode) strings. (See test script below to see what happens if you combine an "encoded" string with a normal string). Its only use is for pasting it into the edit form and sending it off, so it should be converted as late as possible. --Fut.Perf. ☼ 22:34, 19 June 2007 (UTC)

use utf8; use Encode;

sub teststring { my $s = shift; my $t = $s; print "--\n"; print "testing string: $s\n"; print "Unicode flag set? ", (utf8::is_utf8($s) ? "yes" : "no"), "\n"; print "Character length: ", length($s), "\n"; print "Byte length in UTF-8: ", utf8::upgrade($t), "\n"; };

my $test1 = "“Πάντα ῥέει”"; # 12 characters in UTF-8, 28 bytes my $test2 = "— “O RLY?”"; # 10 characters in UTF-8, 16 bytes

teststring($test1); teststring($test2); $test1 = encode("utf8", $test1); # now 28 single-byte characters, no longer 12 characters. teststring($test1); $test1 = $test1. $test2; # now 28 + 10 characters, all in UTF-8, 71 bytes teststring($test1);
 * Interestingly enough, I was able to trace the error back to the Toolserv. I think some of their modules are outdated, causing them to inappropriately encode the data. I've moved Shadowbot3 back to my server, so it should be fine now.
 * I'll remove the utf8 encoding from get_text pronto, I hadn't even thought of that problem until you suggested it. Thanks for all of your help! Shadow1  (talk) 22:41, 19 June 2007 (UTC)

Shadowbot questions

 * Hello, and thanks for setting up the Shadowbot. If you see this Diff, looks like the bot said it archived pages, and put them on my index page, but neither removed them from the talk page or added them to the archive page.  Any ideas?
 * Also, in the future, how do I stop the bot from "indexing" to the index page, which I now just use for a monthly note, but still want it to archive to the monthly archive page? Thank you for your time.  Yours, Smee 16:09, 14 June 2007 (UTC).
 * Oh well, if you have time to reply, just message me on my talk page. Thanks for all the hard work you put into this project!  Yours, Smee 04:45, 20 June 2007 (UTC).

Rename bots?
I just realized there are more than one Shadowbot with different functions. Would you consider giving them more descriptive names? — Omegatron 00:42, 21 June 2007 (UTC)
 * That would require renaming the account, wouldn't it? I've considered renaming Shadowbot to AntiSpamBot before, but I don't know if any 'crats would accept the rename. Shadow1  (talk) 10:54, 22 June 2007 (UTC)

Question
What did your bot do to the message I posted on User:Mackan's talk page? And why? I wrote a dictionary's name in Japanese as a RS. Would you please undo the change? Oda Mari 18:14, 21 June 2007 (UTC)
 * It was a problem resulting from some odd character encoding issues. I've fixed the characters. Shadow1  (talk) 18:56, 21 June 2007 (UTC)
 * Thank you! Oda Mari 19:07, 21 June 2007 (UTC)

NON "spam" website links
'''Hello! I tried to add a legitimate website link to my company's page and was told it's not allowed because it's spam. This is not spam, it's a website within my division of the company. Please add it? http://www.caaspeakers.com/ (My company is Creative Artists Agency "CAA" -- my website is the "Speakers" department within CAA)'''
 * This is a representative of Creative Artists Agency. I have already added the link to the page, because the IP added it incorrectly. Probably mistaking it for spam. Miranda 01:52, 22 June 2007 (UTC)

Re:Perlwikipedia
I replied a few sections above (you can reply there too, I just thought I'd write here so that you'd notice :) Oleg Alexandrov (talk) 01:51, 22 June 2007 (UTC)