On Mon, 16 Nov 2020, Martin Liška wrote: > I've just pushed fix that supports unidiff 0.5.4.
Great, thanks! > > Again, what's the point discussing it? Let's keep to the facts and those > > are that the script does not work with my system which is not unreasonably > > old (released last Sep for unidiff 0.5.4; Jan 2019 for unidiff 0.5.2) or > > odd (a standard Debian distribution). > > No, 0.5.4 was released in May 2017 and 0.5.2 in February 2016 ([1]). > That said, it's reasonable old. Well, still used by a current distribution, and no 4.5 let alone 3 years is not a very long period in a software release lifecycle for a secondary component used with a stable development environment and not the point of said development. You need to combine the lifespan of both the distribution *and* all the components, every new version of which will have had to go through suitable verification and considered stable enough to be accepted into the distribution. [And yet zillions of bugs slip through verification and cause a major headache with any deployment of a new distribution version, to say nothing of random default configuration changes package authors and/or builders seem to make on a whim. I still haven't shaken out all the issues found in the installation I made last month for the effort I have recently been into, and the less time I spend on fixing the environment the more time I have available to actually do something productive, so obviously I am not going to upgrade more often than absolutely necessary.] Anyway, as I say, thank you for taking my fix. Maciej
