+ From:
+ Behalf Of
+ Michael Van Canneyt
+
+ On Tue, 13 Mar 2001, Marc Weustink wrote:
+
+ > Hi,
+ >
+ > Can somebody tell me what the status is on the Linux
+ implementation of
+ > Synchonize, since it looks like the current implementation
+ isn't finised :(
+
+ The status is 'unknown'.
+
+ At the moment, I would advise not to use threads. We are
+ working on a decent threading support at the level of the
+ system unit (i.e. threadvar and so on)
+ and as long as this isn't finished, using threads is here I haven't needed threadvars till so for so... But without sync it is even less usefull.

Do you have any other option to "listen" to a pipestream if new data arrives ?

Marc

Marc Weustink wrote

Hi, Can somebody tell me what the status is on the Linux implementation of Synchonize, since it looks like the current implementation isn't finised :(

The status is 'unknown'.

At the moment, I would advise not to use threads. We are working on a decent threading support at the level of the system unit (i.e. threadvar and so on) and as long as this isn't finished, using threads is .

Hi,

Can somebody tell me what the status is on the Linux implementation of Synchonize, since it looks like the current implementation isn't finised :(

Marc

Sure, use the tasyncio object. See also the socket server demos of the FCL.

Michael.

I see 2 disadvantages: It's only implemented on linux and it has it's own "main" loop, for now I'll try something else.

Marc

Marc Weustink wrote

>Sure, use the tasyncio object. See also the socket server demos of the FCL. I see 2 disadvantages: It's only implemented on linux

for now, at least. A Win32 version will follow soon.

and it has it's own "main" loop,

of course

GTK uses a similar concept internally, but although it isn't difficult to hook into this, you might have some problems to support async I/O in a platform independent way - as the result would duplicate the existing AsyncIO unit(s) on many platforms.

- Sebastian