On Mon, 2015-01-12 at 14:40 +0000, Rushton Martin wrote:

<snip>

Another Firefox "funny" to be aware of occurs if you have a $HOME shared between multiple machines. Firefox will refuse to start on the second machine whilst the first is running Firefox, it believes that there is already an instance running. Rebooting the second machine will not help. The quick-and-dirty way around this is to log in as a different user (and hence different $HOME) on the second machine.

That sort of stuff gets my back up, wondering who the hell is making these design changes. Not folks that are aware of how stuff is used in the "real world? Not folks who understand the concept of "regression testing"? Not folks who look outside their walls and solicit comment from their victims? I don't know, but I get obnoxiously resistant (often to my own detriment) in such cases and tend to do things like my patch.

Yep, long-term potential negatives, but sometimes it's worth it to me. E.g., a very long time ago, device drivers for network cards did not work well for a particular card I was using. So I grabbed the code and "Holy Crap". What a piss-poor bunch of code. And it didn't handle the hardware properly either. So I got the technical docs for the card, modified the driver, tested every edge condition I could think of and, when satisfied, made a patch and got it ready for contribution.

Well, they changed maintainers while I was doing all that and the new code set was so far off what I was working with that my patch became useless.

So I just used my driver for myself, making my effort worthwhile, but I wouldn't have been so stringent in the testing and commentary in the code if I had known what was happening. So it cost me some extra time and effort.

<snip>

MHO, Bill