Developers’ Weblog

Sponsored by
HostEurope Logo

Developers’ Weblog

All 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40

The server became unresponsible because the load went up due to idiotic web crawlers, not respecting robots.txt or ignoring CGIs, hammering httpd(8). After reboot and fsck(8) I’ve configured CGIs to use a rather harsh setrlimit(2) RLIMIT_TIME, a MirBSD speciality. This should prevent repeating this issue.

As a consequence, some requests, for example annotating in CVSweb on large files (acronyms DB for wtf(1)) will now fail or (diffing between revisions on that file) return incomplete results. SOL.

This is why we can’t have nice things.

ObOTRant: The W3C specification requires the XMLSerializer[sic!] shipped in ECMAscript-capable web browsers to produce invalid XML. (The underlying algorithm is permitted elsewhere to instead raise an exception but even though a (pretty bad, e.g. it suggests dropping instead of entity-escaping -- inside a comment node) specification for fixing up such DOM trees preparing for XML serialisation exists, the spec prevents using it. The whole “web world” is just a steaming PoS really…

MirBSD Logo