On 09/15/2015 11:43 AM, Luca Ceresoli wrote:
Dear Larry,
Larry Finger wrote:
BTW, if a do a change not related to checkpatch warnings, say I remove
unneeded typecasts, and the change has impact on long lines which are
still long after the change, should I also wrap code to fit 80 chars _in
the same commit_? Or defer long lines fixes to a later commit?
I thought I made myself clear. If checkpatch complains about something, fix it now.
Yes, cleaning up this driver would be a major task, which is why I have
not done any of it. In addition, many of the checkpatch warnings were
not present when the driver was first submitted. The script's
requirements are a moving target. My plan is to replace the driver with
one that uses mac80211, and is programmed more cleanly.
Meaning the current driver is going to be removed at some point? Is the
new driver already brewing? Close to be ready?
Yes. Yes. No.
You, however,
are free to hack on this as much as you want. Just do not break it as it
is still needed.
Sure, on my laptop!
At least you are testing. I do not trust people that submit many patches, and do
not have the hardware. I do not always have time to review and test. At least
those kinds of mistakes are self correcting. If a particular patch has to be
reverted because it broke a driver, that person's future patches are suspect. We
had one case a few months ago that was so bad, the person is permanently black
listed. I;m sure he/she blamed it on the "old-boys network", but the truth was
that we could not trust those patches.
Larry
_______________________________________________
devel mailing list
devel@xxxxxxxxxxxxxxxxxxxxxx
http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
