Busy harddisk while typing or moving pad

[Mumia W.]

Hello. Xpad is a good program, but I have a problem with it.
Every time a pad is moved, my disk drive thrashes. It seems
that xpad is writing to disk a lot during window moves, so
I've created a patch that seems to solve the problem.
Thanks for writing and maintaining such a good program
for us.

[Mumia W. (mumia-w-18)]

I created a new patch with less cruft in it.

[Mumia W. (mumia-w-18)]

My patch might have a bug. If two pads are modified at about
the same time, the first pad probably won't be saved.
Setting TIMEOUT_SECONDS to a relatively low value probably
makes this a non-issue. However, this is a concern.

[Mumia W. (mumia-w-18)]

This patch fixes the original problem and does not cause some pads
to fail to be saved.
The original problem was that disk thrashing occurred whenever a pad
was moved.

[Mumia W. (mumia-w-18)]

This patch works with trunk and fixes the disk-trashing problem.
Also included is a change from
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=576770

fixing a bug that made all pads blank and non-editable.

[niknah (hankin0)]

Here's an updated patch to work with the current version in ubuntu 13.04, xpad 4.1

The patches here introduces a small bug. The contents are not saved if you exit xpad within 4 secs after you edit it. Say you were making a last edit before logging out.

I have made a fix for this in the patch.

[Arthur Borsboom (arthurborsboom)]

Hi Mumia and Niknah,

Thanks for helping out on developing fixes for Xpad.
@Niknah, before I start testing, should this patch also resolve the 'write on keypress' issue?

[Arthur Borsboom (arthurborsboom)]

I confirm I see a huge amount of disk I/O when typing, moving a pad, or moving the mouse cursor. There is also a relatively high CPU load for such simple actions.

Implementing patch on private branch to test the effect of the patch.

[niknah (hankin0)]

Yes, the patch fixes my other bug report. 'write on keypress'

Thanks Arthur!

[Arthur Borsboom (arthurborsboom)]

I have tested the patch.

Since I have an SSD and no HD led, I don't notice this behaviour, but according to iotop the patch reduces the disk I/O a lot.
So therefore, the patch has been implemented in the main branch.
If the regression tests before the release do not show any negative side effects, the patch will be included in version 4.2.

Thanks a mile for writing this patch!
Keep up the good work, and don't hesitate to suggest more improvements!