On Tue, 4 Aug 2020 at 18:20, Shayan Doust <he...@shayandoust.me> wrote:
> Hello Andreas, > > Thanks for the nitpicking! > > I've had a moment of inactivity with this package, so I forgot why the > patch was there. I now remember the patch I put there simply disables an > erroneous assertion: > > AssertionError: Tuples differ: (3030000, 2) != (3030000, 2, > void(b'\x00\x00\x00\x00')) > > I did this, and dh_missing, to speed up the packaging and should have > really removed those before I emailed. Although, I am not sure how I > can contact upstream and should have just written an email here about > it. I still think this is a very trivial assertion issue, and the tuples > compared are very similar with the exception of the extra void(...) > element. Out of 63 tests, maybe it's still worth keeping the patch > The first two values are still the same - so it is worth comparing atleast these values instead of disabling altogether, right? > (given I add the description) until a future fix, or someone in the team > can rectify this. What do you suggest? I dug in a bit - it looks like that there's an issue with typecasting in the code. Upstream can shed more light on this, it makes sense to open an issue IMO. Also, I've pushed in the above mentioned 'fix' (?) and a couple of other minor changes, please $git pull :) > > As for the manpage, the only other method really is writing manpages for > the two scripts. Apart from the spelling mistakes which can be fixed, > the manpage has quite a featureful and meaningful content of both the > scripts and the library usage - although I agree this is a replication > of the HTML documentation. Do you suggest I simply just write the > manpages for the two scripts? > You can do that, but to save your effort you might consider using ronn[1] or marked-man. [1]: https://packages.debian.org/unstable/ronn [2]: https://tracker.debian.org/pkg/node-marked-man Kind Regards, Nilesh
