Re: Dangerous update command
- From: Mats <matsben@xxxxxxxxx>
- Date: Thu, 29 Nov 2007 01:02:47 -0800 (PST)
On 28 Nov, 20:23, Donald Arseneau <a...@xxxxxxxxx> wrote:
I don't think I understand why you need any kind of update there.
This is the very end of a fileevent handler. Simply returning will
cause Tcl to update and process the next event, whether a gui or a
file event. Is the problem that the gui events are queued up after
all the file events?
Yes, and it will just block all UI interactions while the file
transfer is progressing. So therefore a mechanism is needed to have
fileevents at "lower priority" than UI events. Using 'after idle' and
just schedule one fileevent at a time seems to be working well, but my
question was: is this the right way?
/Mats
.
- References:
- Dangerous update command
- From: Mats
- Re: Dangerous update command
- From: Donald Arseneau
- Dangerous update command
- Prev by Date: Tcl under conditions of high memory usage.
- Next by Date: tclcurl examples please
- Previous by thread: Re: Dangerous update command
- Next by thread: Can someone explain this performance issue?
- Index(es):
Relevant Pages
|