Boost Testing :
From: Victor A. Wagner Jr. (vawjr_at_[hidden])
Date: 2005-06-02 09:05:10
At 06:03 2005-06-02, Beman Dawes wrote:
>"Victor A. Wagner Jr." <vawjr_at_[hidden]> wrote in message
> > At 13:10 2005-05-31, you wrote:
> >>"Aleksey Gurtovoy" <agurtovoy_at_[hidden]> wrote in message
> >> > Beman Dawes writes:
> >> >> On a Win XP SP2 machine with a hyper-threaded Intel P4, I run the
> >> >> regression
> >> >> tests with this command:
> >> >>
> >> >> python
> >> >> regression.py --runner=Beman --incremental --monitored
> >> --toolsets=cw-9_4,gcc,intel-win32-8_1
> >> >>
> >> >> It looks like the monitor starts running (in a separate command
> >> >> window), but in the morning (after an overnight run) the monitor is
> >> >> gone and I have to kill each dialog box by hand.
> >> >
> >> > Has your run completed by then, though? IOW, were the dialogs holding
> >> > the regression process, or they were just "leftovers"?
> >>They were holding the regression process. The process had not completed,
> >>there was no system activity showing on the task manager performance
> >>display. As I cleared each dialog, there was activity until the next
> >>then no activity until it was cleared.
> > my system (details at the RudbekAssociates link in the results page)
> > behaves differently. I wake up in the morning with several "dialog boxes"
> > waiting for me to click on, but the tests have completed.
>This is definitely not what is happening for me.
>One additional data point. This morning, I happened to be staring vacantly
>at the Command Prompt window when bjam finished. A message displayed to the
>effect "could not kill build_monitor.exe, process not found"
>(I don't have the exact warning because later messages scrolled it off the
>Since I know (from seeing the Command window it opened) that build_monitor
>started to run, it looks like something killed it. Could it have killed
>Could me minimizing the build_monitor Command window be a problem? If I'm
>using the machine for other work, I minimize the build_monitor window to
>give me more screen real estate.
I had _so_ many problems with running the regressions while doing other
work, I finally created a completely separate user(BoostRegression) for
running them. The _only_ thing that happens in that account is running the
tests (timed starts with "scheduled tasks") and a very occasional foray by
me to make sure that some new environment variables have been correctly set.
>Boost-Testing mailing list
Victor A. Wagner Jr. http://rudbek.com
The five most dangerous words in the English language:
"There oughta be a law"