clif-legacy issueshttps://gitlab.ow2.org/clif/clif-legacy/-/issues2019-01-03T21:54:23Zhttps://gitlab.ow2.org/clif/clif-legacy/-/issues/317630CLIF plug-ins for Eclipse do not load on Windows 102019-01-03T21:54:23ZBruno DillensegerCLIF plug-ins for Eclipse do not load on Windows 10When running the 2.3.7 CLIF Eclipse distribution 64 bits for Windows on Windows 10, CLIF plug-ins do not load. CLIF splash screen is displayed, but then no CLIF feature is available (views, perspectives, menu, wizards, editors). Could no...When running the 2.3.7 CLIF Eclipse distribution 64 bits for Windows on Windows 10, CLIF plug-ins do not load. CLIF splash screen is displayed, but then no CLIF feature is available (views, perspectives, menu, wizards, editors). Could not find any error in the logs.
The same distribution runs fine on Windows 7.https://gitlab.ow2.org/clif/clif-legacy/-/issues/314877late or no completion of ISAC scenario after end of load profile2017-12-13T16:33:14ZBruno Dillensegerlate or no completion of ISAC scenario after end of load profileA scenario may not complete while the execution time specified by the load
profile has elapsed, even with the "force stop" flag enabled. Of course, this
can occur when the number of virtual users has been changed manually during the
test...A scenario may not complete while the execution time specified by the load
profile has elapsed, even with the "force stop" flag enabled. Of course, this
can occur when the number of virtual users has been changed manually during the
test execution (change parameters command), which is completely normal, since
this operation cancels the load profile.
But this may also occur when scenarios contain a never-ending loop (e.g. while
true...) with no timer call (aka think time). In case there is no loop but just
a sequence of actions without timer, the "force stop" flag is still ignored,
and the actual completion will wait for all virtual users to execute the very
last instruction of their behavior (exactly as if the "force stop" flag had not
been set). This situation is actually a bug.
A very simple workaround is to include a timer call, even with a very short
think time, in the behavior section where the virtual users execution might be
dropped. This will make the "force stop" flag to be taken into account.