On Wed, Jul 17, 2002 at 04:39:18AM +0100, Matthew Astley wrote: > > No, I meant "pull down" in the sense of stealing the message-id and > subject line, really. I think I expected some different flavour of > confusion. 8-}
If you want to replace wpoison's WNPP entry with whatever you are working on, feel free to. I don't mind about wpoison anymore. > For a change, it's not NTL. http://bugs.debian.org/133790 8-( > > Oh whoops, I should've hit that with the browser first, rather than > taking reportbug's perspective (ie. first message only). D'ya think I > should file against reportbug? Seems a bit trigger-happy, but a "don't > forget to read the whole thing" warning might save some embarrassment > for someone. I don't know, speak with the reportbug maintainer. > > Do you have a sponsor to get it in Debian? > > No. I know a couple of maintainers, but I've not asked for > sponsorship. I don't know exactly what this requires them to do, or > what sort of standing they require. Will read stuff "later". You need a developer that can help you. Not all maintainers are developers. You can identify a developer by his/her address @debian.org or with the database at http://db.debian.org/. I'd ask in debian-devel, debian-mentors and mostly to people who manifested interest in wpoison (see the RFP entry) This URL contains some useful information for sponsors: http://www.internatif.org/bortzmeyer/debian/sponsor/ > Also, I think "tramspap" should have a few more tricks up its sleeve > before being turned loose on the world. For example, I've just > discovered that any sort of multi-threaded crawler will very > effectively DoS the Apache by using up all eight threads. Oops. > > I can't see a sane answer to this one. You _have_ to tie up the Apache > process while you're returning HTML at 1cps. > > Perhaps the tarpitting aspect could be handled by a special Apache > module which accepts a bufferful of data and returns it slowly, while > still accepting the next connection? > > It would appear to be valid to fork off some trivial process to handle > the squirt of data. All that's required is "feed this data out slowly, > until they go away or you run out of data". There's no reason to > accept input, provided the keep-alive has been disabled. Maybe on the lists about apache and security someone can help on this: http://lists.debian.org/ cheers, -- Robert Millan "5 years from now everyone will be running free GNU on their 200 MIPS, 64M SPARCstation-5" Andrew S. Tanenbaum, 30 Jan 1992 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
