On Sat, 24 Sep 2016 at 05:14 Stuart Henderson <s...@spacehopper.org> wrote:
> On 2016/09/24 18:42, Jonathan Gray wrote: > > On Sat, Sep 24, 2016 at 09:37:45AM +0100, Stuart Henderson wrote: > > > On 2016/09/24 16:35, Jonathan Gray wrote: > > > > Update coccinelle to 1.0.5, changes described here: > > > > http://coccinelle.lip6.fr/distrib/changes.html > > > > > > I tried updating this at Cambridge (it seemed fitting ;) - my diff is > > > nearly the same, only packaging changes - > lib/coccinelle/spgen/spgen.opt > > > should be in PFRAG.dynlink instead of PLIST - but I dropped it because > > > the regression tests didn't look good and I didn't know where to start > > > with debugging it. > > > > Ah, indeed running make test gives: > > > > -------------------------------- > > regression testing information > > -------------------------------- > > regression file: tests/SCORE_expected.sexp > > -------------------------------- > > total score > > -------------------------------- > > good = 218/492 > > Current score is lower than expected :( > > (was expecting 451 but got 218) > > For comparison, the old version gave > > : good = 452/480 > : Current score is greater than expected :) > : (was expecting 445 but got 452) > > If it was just a small difference I'd probably just run with it, but > this was a lot bigger than usual. > I have an M in coccinelle/Makefile as part of my attempt to update ocaml-pcre (and its users ocaml-text, ocaml-net) but it's just a revision bump so I'm not concerned about conflicts. No idea if the slightly updated pcre would improve anything but I can try it if that's of interest. Just have to figure out why ocaml-pcre works fine if I build it alone but won't under dpb ... .... Ken
