[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#963832: RFS: iotop-c/1.0-1 [ITP] -- iotop-c - simple top-like I/O monitor (implemented in C)



On Thu, 2020-07-09 at 10:34 +0800, Paul Wise wrote:
> On Wed, 2020-07-08 at 20:09 +0300, Boian Bonev wrote:
> > should be fixed in the package - that option should come from dpkg-
> > buildflags.
> 
> As gregor mentioned, this needs enabling in debian/rules.

Already done.

> > I do not get that - the current Makefile uses GNU make extensions
> > and
> > wildcards, but not a single implicit rule. Please clarify which
> > part
> > can be improved.
> 
> I mean you could delete the commands that specify how to build .o
> files
> and build iotop and rely on the built-in rules. It might not be worth
> the effort to figure out how to do this though, especially
> considering
> the portability concerns.

I would rather not do that, for the following reasons:

1. Current explicit rule is 5 lines, removing that will not do any good
2. Quiet build cannot be implemented properly with implicit rules
3. Dependency generation would still require an explicit rule
4. As you have mentioned somewhere above in the mails, implicit rules
tend to have the risk not to work on non-standard systems

> OK, please do check if Breaks works or not if you haven't already. So
> install an old version of iotop py on bullseye and then apt install
> ./iotop-c_*.deb with both Conflicts and Breaks and see what happens.

Indeed Breaks+Replaces works better than Conflicts. There are cases
where instead of removing the old iotop package apt will just upgrade
it.

I have verified both ways and it works as expected.

> > There is no signature upstream, can't fix that too.
> 
> If you have an OpenPGP key you could do signing.

Already did too, took me some time to sort out what is the expected
format of debian/watch...

I think that most of the things are fixed now, please proceed with your
review/upload.

Thanks for all the help!

I was surprised that packaging the program is comparable to the program
creation itself; for some reason I had expectations that would be much
easier :)

With best regards,
b.


Reply to: