From RationalWiki
Jump to navigation Jump to search
Error icon.svg No longer works!

Look at CaptureBot4 for its successor.

CaptureBot //   

Rare photo of Capturebot2 in its natural habitat

Capturebot2 takes screenshots of webpages and uploads them to RationalWiki. Normally, it's used on WIGO:CP with Conservapedia, however it can be used on any page and with any external link.

tl,dr: click here

Detecting links
To be detected by Capturebot2, external links have to be surrounded by special html code: the <votecp> tag does this for all external links to, and the <capture> tag can be used for any other external link and outside polls. These tags will create the small img links.
Monitoring pages
Capturebot2 monitors only a few pages (listed in the watch option). It does this by getting a list of recent changes every minute. To improve performance, it remembers the last revision, and only rechecks a page if there is a new revision.
A page should only be added to Capturebot2's list if the <capture> or <votecp> tags are used on it often. For other pages, you can use the capture command to scan the page once, or the sandbox to capture one webpage.
Capturebot2 will only upload images for red img links. If the image already exists and you want a reupload, you will have to delete the previous image. You will also have to get Capturebot2 to recheck the page: either use the capture command, or if the page is watched by Capturebot2, make a dummy edit.
Failed captures will still have a page made, without the file, to prevent the bot from trying to upload the image every minute. These are put into Category:Failed screencaps. In case of a failure, try capturing the page and uploading the image manually, or modify the link (change links to wiki pages to difflinks, reduce the number of returned results in log links, etc.).
Capturebot2 will automatically add CP screenshot to screenshots of Conservapedia. In any other case it will add a Nolicense tag, and it is the responsibility of the editor requesting the screenshot to add the proper license tag disregard that I suck cocks.

Please do not block Capturebot2 unless absolutely necessary. Use the sleep command instead.

Capturebot2 is operated by Nx (talk).

Easiest way to capture links, just add <capture>LINK</capture> to this page.
Command console
Currently supports the following commands:
Status: returns Online or Sleeping depending on state. In inactive state, Capturebot will not upload pictures, but will monitor the command console for changes. If Capturebot does not reply within a minute, then it is offline.
Sleep: deactivates Capturebot
Wake: reactivates Capturebot
Capture page: tells Capturebot2 to check page for any links to be captured (use in conjunction with the <capture> tag)
Lines starting with * are comments.
Options must be on a single line
Currently supports the following options:
watch: list of pages to patrol
Format: watch = page1 & page2 & page3
Note: use spaces in page names, not underscores.
match: whether to check only pages specified by watch (true), or all pages that show up on recent changes (false)
Format: match = true or match = false
Lines starting with * are comments.
Source code
Main application
Webkit2png script, original by Roland Tapken, slightly modified version from here. The script has been further modified to support multiple uploads (to prevent starting and stopping the X server for every image), running javascript on the page before capture (to hide the mediawiki sidebar), and limit images to 32766 pixels in either direction, since Qt cannot handle larger images.