Re: Mail2 Fwd [Re: sync 30a4 via pilot-link]

[Eskil Heyn Olsen <deity dbc dk>]

On Tue, 1 Feb 2000, Vadim Strizhevsky wrote:

> krollin@palm.com writes:
>  > I haven't tried pilot-link, so I don't know what problems there might be (I'm
>  > not even sure what pilot-link is).  One very helpful developer is looking into
>  > problems with using psuedo-devices and pilot-xfer.  Seems that the HotSync
>  > process on the device side involves starting and stopping the UART more than
>  > once.  This causes Poser to open and close the device it's been told to use.
>  > However, psuedo-devices don't like to be opened more than once, and Poser fails
>  > when it tries to re-open the device.

The problem is excactly that. POSEr opens the device when syncing and
closes afterwards, This makes it useless for many sync manager packages
built on pilotlink.

I'm quite sure that I mailed you a patch for this problem some months ago
( can't check that far back though ), but if, then it was not accepted.
The patch also cut down on the cpu time used byt the sync threads (from
200% on my dual pentium to about 5%).

I've been told that the new source is so radical different, that the patch
no longer works, and I have lately returned to XCopilot, as it's syncing
simply was more pilot-link and resource friendly (and I use it a lot while
developing gnome-pilot). I'm not motivated to reproduce the patch it it'll
never make it's way into the main codebase.

/dev/eskil
---