[tin-dev] -C and timeout with low nntp_read_timeout_secs and huge group

[Urs Janßen]

There seems to be still something wrong with alarm() and -C. This time
it kicks in with low nntp_read_timeout_secs (e.g. 30) and entering a
huge group.
e.g. gmane.linux.kernel on news.gmane.io; ~4.8 million articles;
requires ~3GB memory and takes 'some' time (you may want to skip
filtering/threading during tests). Without -C everything works as
expected, with -C alarm(nntp_read_timeout_secs) kicks in:

[ tin -D 1 -vvCg news.gmane.io gmane.linux.kernel ]
>>> [20:20:24.269317] OVER 7380-4846784
<<< [20:20:24.302542] 224 Overview information for 7380-4846784 follows
nntp_command(OVER 7380-4846784) OK
get_server() 30 sec elapsed without response
nntp_close(TRUE) END

ideas?

urs
-- 
Jef Poskanzer:
"When people aren't stupid Usenet is even more useful. Too bad this happens
 so rarely."