2009/4/16 Yann Dirson <ydir...@linagora.com>: > Hello, > > Here is my second version of the layer_change event patch, hopefully > addressing the points raised in reply to the first one. The core code is > unchanged, only changes to perllib and doc/changelog files have been > added. > > We needed to trace and hook some actions to layer changes for e-Sidor[1], > so here is the functionnality to do so. > > This patch also: > * declare "restack" event to FvwmCommand > * declare "reply" event to FvwmEvent > * stop filtering out most extended messages for FvwmCommandS > * minor whitespace cleanup > * regenerate perllib's Constants.pm > * declare MX_LAYER_CHANGE arguments to perllib > > We hope you'll find it useful.
I'm not against this patch, but it would need documentation and be updated to the cvs version of FvwmEvent. I can see a few cases where it would be useful to track layer changes, without keeping full window information in the module, most of them with FvwmEvent. I'm not entirely convinced however that the benefit of a separate event type over tracking layer changes via M_CONFIGURE_WINDOW. So a little motivation on why tracking the layer via configure window is bad in your use case would help to justify the usefulness of the event type. Finally the module interface is documented at http://www.fvwm.org/documentation/dev_modules.php (which is in the cvs module fvwm-web). Any new event type should be documented there as well. /Viktor