- From: Bjoern Voigt <bjoern@cs.tu-berlin.de>
- Date: Fri, 11 Jul 1997 13:49:35 +0200 (MET DST)
- To: www-lib-bugs@w3.org
- cc: Henrik Frystyk Nielsen <frystyk@w3.org>
Library : LibWWW 5.1b
System : SunOS 5.5
Compiler: cc (SC4.0 18 Oct 1995 C 4.0)
Options : -g -D_REENTRANT
I found 6 messages about the problem that ComLine can NOT POST
correctly. The problem also happens in own programs which use the
LibWWW 5.1b
How long we have to wait for an official and correct PATCH? Some
people have to work with LibWWW 5.1b and DO NOT only PLAY with it.
I made the following test:
w3c -form -post 'http://user.cs.tu-berlin.de/~bjoern/test.cgi' VARIABLE1=1
Looking up user.cs.tu-berlin.de
Contacting user.cs.tu-berlin.de
Writing...
Writing...
(... then w3c waits forever !)
I stopped the application in a debugger and saw that w3c waits in the
following function:
( HTEvtLst.c line 804 )
active_sockets = select(maxfds+1, &treadset, &twriteset,&texceptset,wt);
wt is NULL at this point.
I think that all the problems depending on the functions in HTTimer.c.
Especially the function
PUBLIC int HTTimer_next (ms_t * pSoonest);
often causes segmentation faults and bus errors. I also can't
understand why I can not specify an timeout_handler like in LibWWW
5.0a.
The new function HTHost_setEventTimeout() is a bad joke. I can't set
different timeouts for every Host if the Host is not a parameter!
Has anyone tried to use LibWWW 5.1b in a multithreaded program with
multiple requests in each thread? I didn't found anything about it in
the documentation.
Thanks for all help,
Björn
--
Björn Voigt E-MAIL: bjoern@cs.tu-berlin.de Tel: (030) 6 36 66 27
WWW: http://www.cs.tu-berlin.de/~bjoern/
Received on Friday, 11 July 1997 07:59:09 UTC