[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference bulova::decw_jan-89_to_nov-90

Title:	DECWINDOWS 26-JAN-89 to 29-NOV-90
Notice:	See 1639.0 for VMS V5.3 kit; 2043.0 for 5.4 IFT kit
Moderator:	STAR::VATNE

Created:	Mon Oct 30 1989
Last Modified:	Mon Dec 31 1990
Last Successful Update:	Fri Jun 06 1997
Number of topics:	3726
Total number of notes:	19516

3543.0. "Daylight savings time and XtAddTimeout" by VINO::MCARLETON (Reality; what a concept!) Mon Oct 29 1990 09:55

    I just reset the system time on my VMS VAXstation back by an hour for
    daylight savings time.  I noticed that several time interval related
    operations stopped.  The DECterm cursor stopped blinking and the
    Banner program stopped updating.  I assume that this is because
    XtAddTimeOut must using some kind of absolute time format internally
    which caused the timeout interval to bumped up by an hour.  In my case,
    the change of time was on both the client and the server.  Which clock
    is used for the interval, the client or the server?

    I rebooted my VAXstation just in case there were other timers that
    needed to fire to keep my system sane.

    						MJC

T.R	Title	User	Personal Name	Date	Lines
3543.1		YRDARM::finnegan	Neal, DECdecision - mail to: via::finnegan	`Mon Oct 29 1990 10:22`	4
	This point is brought up EVERY year at this time and it is release noted. The XtAddTimeOut call does not correctly handle system time changes. Neal
3543.2	Gee and we wonder why "software quality" is a joke.	IO::MCCARTNEY	James T. McCartney III - DTN 381-2244 ZK02-2/N24	`Tue Oct 30 1990 14:32`	6
	Why doesn't someone just fix it? I know it got QARed a couple of years back, did it just get closed with "Developer doesn't care." as the answer? James
3543.3	DECwindows not to blame	PSW::WINALSKI	Careful with that VAX, Eugene	`Wed Oct 31 1990 16:14`	14
	It's a VAX C RTL problem. The C RTL incorrectly implements sleep() and related functions by computing the absolute time when it should wake up and then doing a $SETIMR to establish a timer to go off at that time. What is should do is compute a delta time and $SETIMR using that. The use of absolute time causes spurious timeouts to occur in the spring and extra long timeouts to occur in the fall. I think this may even be fixed in the C RTL sources, but during all the release juggling that VMS went through with HICKORY vs. THUNDERBOLT vs. PHOENIX, the C RTL changes that were supposed to go into V5.4 got dropped on the floor and won't appear until PHOENIX. So that's probably the first opportunity that exists to fix this problem. --PSW
3543.4		PSW::WINALSKI	Careful with that VAX, Eugene	`Wed Oct 31 1990 17:43`	4
	According to the VAX C RTL project leader, this bug is fixed in alarm() and sleep() for the PHOENIX release of VMS. --PSW
3543.5	Since Phoenix is now just bits in the wind....	IO::MCCARTNEY	James T. McCartney III - DTN 381-2244 ZK02-2/N24	`Mon Nov 26 1990 15:32`	5
	Any chance of getting these bits into the Iowa release? James
3543.6		PSW::WINALSKI	Careful with that VAX, Eugene	`Tue Nov 27 1990 13:31`	5
	RE: .5 The C RTL project leader is looking into that. --PSW