Again and Re: my uthreads execute serially

[Wolfgang Lipp]

Gletschersorry for the somewhat cluttered appearance of the previous message. i posted the same question to python-list at python.org
("gatewayed bidirectionally into comp.lang.python", jesus), and Mike Fletcher <mfletch at tpresence.com> was so kind to answer in no
time @ all. here's the answer and the posting again. -w

..............................

Same here, works fine with 1.5.4 (1.5.2) version of stackless, but doesn't work for the 2.0 version.  No idea how I missed that, but
it's definitely failing to switch contexts without an explicit switchContext.  I have no time today to track it down this weekend,
maybe Chris can take a look? (Note, adding a switchContext does make the threads run in parallel, but that shouldn't be necessary).

uthread9 does complain on my machine, but I'd though it was the posting an exception to another thread test (which prints out the
traceback).  Should have had an explicit test for such a simple thing I suppose.

Sorry about that,
Mike

-----Original Message-----
From: Wolfgang Lipp
To: python-list at python.org
Sent: 1/12/2001 11:06 AM
Subject: my uthreads execute serially



can anybody help me with this problem? i am using
stackless python 2.0 on win98se, on a laptop. when
i execute the following program:

    import uthread9 as uthread
    import time

    def timer_elapsed( started ):
        return time.time() - started

    def timer_over( started, stoptime ):
        return timer_elapsed( started ) > stoptime

    def timer( letter, stoptime ):
        started = time.time()
        print "%s is started" % letter
        while not timer_over( started, stoptime ):
            pass
        print '%s is done after %3.2fs' % ( letter, timer_elapsed( started ) )

    uthread.new( timer, 'a', 5 )
    uthread.new( timer, 'b', 3 )
    uthread.new( timer, 'c', 2 )

    uthread.run()

my output looks like this:

    a is started
    a is done after 5.05s
    b is started
    b is done after 3.02s
    c is started
    c is done after 2.04s

which clearly indicates that the threads do *not*
run in the parallel. moreover, the outputs from
all the examples on
http://world.std.com/~wware/uthread.html also
point into interpreting them as being executed one
thread after the other. since all the snippets as
well as the test suite u1.py from that page work
without complaining, i guess that there is no
problem with my installation.

furthermore, i tried a few examples using the
threading module, and the results appear to be ok.
because of this and because uthreads are not, to
my understanding and unlike the threading module,
dependent on the os, it can't be my laptop that's
to small to fit in threads :-| ...
can anyone help me parallelize my uthreads?


yours,

-wolf