Hi, In the script 'spinner_fork_for_question.pl', you're still calling Parallel::ForkManager->wait_all_children, which I stated in my previous e-mail is a blocking call. Are you going to remove that method call at some point?
Also, can you simplify 'spinner_fork_for_question.pl'? You could remove the e-mail sending bits, and replace it instead with a call to qx(sleep), or some other external process that will block for a few seconds, and most likely get the same result. The script "gtk3_spinner_fork2.pl" worked for me. Thanks, Brian On Wed, May 28, 2014 at 10:00 PM, tsuyoshi okita <825...@gmail.com> wrote: > Dear Brian and others, > > Thank you for your quick reply, Brian-san. > Instead of forking two process, I've decided to fork only one process for > sending an email,and using run_on_start and run_on_finish for the spinner. > And the same problem still persists. Strange thing is that I added print > "run_on_start" and print "run_on_finish" to each callback subroutines to > make sure they are processed. They do, but spinners do not respond right. > > In other code using Parallel::ForkManager, Spinner worked fine, it spin and > stopped when I expected. but when forked to send an email, spinner gets > grumpy all the sudden. > I attached two test cases, one is a revised but problem still persists code, > and other works sort of fine. > > There is other problem I need your guide for. when I add hide() for spinner > after stop(), spinner does not show at all. Have you had this problem?? > > Thank you again. > > Tsuyoshi > > > On 2014年05月29日 12:04, Brian Manning wrote: >> >> Your sample code uses Parallel::ForkManager, and uses the >> wait_all_children method to wait for the children to end processing. >> From the Parallel::ForkManager docs: >> >> -begin quote- >> >> wait_all_children >> >> You can call this method to wait for all the processes which have been >> forked. This is a blocking wait. >> >> -end quote- >> >> So you are calling a method that says it will block, hence your demo >> script is blocking. >> >> Instead of using "wait_all_children" (which blocks), I would look at >> using the "run_on_finish" callback method instead, and make the call >> to stop the spinner from that callback method. >> >> Thanks, >> >> Brian >> >> On Wed, May 28, 2014 at 7:27 PM, tsuyoshi okita <825...@gmail.com> wrote: >>> >>> Dear gurus, >>> >>> I've been trying let a spinner spins at the same time sending an email by >>> forking and failing. Although a program forks, and one child process >>> suppose to spin the spinner while other sends email, the spinner does not >>> spin until the email is completely sent. How does you all usually handle >>> this spinner thingy??? Am i on a right path?? >>> >>> I attached my test case. You need to add SMTP host name, your email >>> address, and password. >>> >>> Thank you. >>> >>> Tsuyoshi >>> >>> _______________________________________________ >>> gtk-perl-list mailing list >>> gtk-perl-list@gnome.org >>> https://mail.gnome.org/mailman/listinfo/gtk-perl-list >>> > > > _______________________________________________ > gtk-perl-list mailing list > gtk-perl-list@gnome.org > https://mail.gnome.org/mailman/listinfo/gtk-perl-list > _______________________________________________ gtk-perl-list mailing list gtk-perl-list@gnome.org https://mail.gnome.org/mailman/listinfo/gtk-perl-list
