[1407] in testers
Re: vax 7.2G: session_gate, end_session
daemon@ATHENA.MIT.EDU (daemon@ATHENA.MIT.EDU)
Thu Dec 13 00:42:33 1990
Date: Thu, 13 Dec 90 00:42:16 -0500
To: testers@ATHENA.MIT.EDU, brlewis@ATHENA.MIT.EDU
In-Reply-To: testers[1405]
From: Richard Basch <probe@MIT.EDU>
Subject: vax 7.2G: session_gate, end_session
To: testers@ATHENA.MIT.EDU
Date: Wed, 12 Dec 90 14:39:00 EST
From: Bruce R Lewis <brlewis@ATHENA.MIT.EDU>
System name: euphrosyne
Type and version: CVAXSTAR 7.2G (1 update(s) to same version)
Display type: SM
What were you trying to do?
let session_gate and ens_session interact running as daemon
What's wrong:
session_gate write /tmp/session_gate_pid.01
end_session relies on /tmp/session_gate_pid.1
What should have happened:
session_gate should add .1 extension instead of .01
end_session should take advantage of the XSESSION variable
Last time I looked at session_gate.c I overlooked another place where
the 5-digit pid was hard-coded in itoa().
*** session_gate.c Mon Dec 10 16:13:21 1990
--- /tmp/session_gate.c Wed Dec 12 14:31:41 1990
***************
*** 101,107 ****
for (i = 0; powers[i]; i++)
{
digit = (x/powers[i]) % 10;
! if ((pos > 0) || (digit != 0) || (i == 4))
buf[pos++] = '0' + (char) digit;
}
buf[pos] = '\0';
--- 101,107 ----
for (i = 0; powers[i]; i++)
{
digit = (x/powers[i]) % 10;
! if ((pos > 0) || (digit != 0) || powers[1+i])
buf[pos++] = '0' + (char) digit;
}
buf[pos] = '\0';
Fixed... however, I think you made a mistake in your patch... I think
you want to testing for powers[i+1]==0.
-Richard
