Compilation errors
grarpamp
We have these errors. Dropping -Werror allows compilation to proceed
and produces working binaries. Yet they should be looked at and
fixed I guess. Thx.
server.c: In function exit_worker:
server.c:427: warning: passing argument 3 of getpeername from
incompatible pointer type
/usr/include/sys/socket.h:133: note: expected socklen_t * __restrict__
but argument is of type size_t *
server.c: In function tunnel_worker:
server.c:817: warning: passing argument 3 of getpeername from
incompatible pointer type
/usr/include/sys/socket.h:133: note: expected socklen_t * __restrict__
but argument is of type size_t *
server.c: In function mainloop:
server.c:1265: warning: passing argument 3 of accept from incompatible
pointer type
/usr/include/sys/socket.h:214: note: expected socklen_t * __restrict__
but argument is of type size_t *
kademlia_rpc.c: In function handle_rpc_find_value:
kademlia_rpc.c:576: warning: passing argument 3 of local_find from
incompatible pointer type
kademlia.h:93: note: expected uint32_t * but argument is of type size_t *
phantomd.c: In function daemonize:
phantomd.c:27: warning: ignoring return value of freopen, declared
with attribute warn_unused_result
phantomd.c:28: warning: ignoring return value of freopen, declared
with attribute warn_unused_result
phantomd.c:29: warning: ignoring return value of freopen, declared
with attribute warn_unused_result
phantomd.c: In function nak_request:
phantomd.c:158: warning: ignoring return value of write, declared with
attribute warn_unused_result
phantomd.c: In function ack_request:
phantomd.c:165: warning: ignoring return value of write, declared with
attribute warn_unused_result
spjsschl
machine running this ubuntu flavor.
Johannes
Johannes Schlumberger
The compile errors should be fixed in this new package. I have been working on a
Debian squeeze which has an older gcc by default with less strict typechecking.
Johannes
Magnus: Can you please exchange our current release with this package?
Magnus Bråding
Done.
Regards,
Magnus
grarpamp
Yes, standard out and err is now clean. Thanks.
How about versioning the tarball. Perhaps call them:
phantom_ref-0.1.1.tar.gz b2ddd276382b4b95b90452eea78f3857
phantom_ref-0.1.2.tar.gz 59daa0166e6ce56fd681479e9aa4b22c
'ref' meaning reference implementation.
Others writing to spec might call theirs:
phantom_other-M.m.p.tar.gz
> I have been working on a Debian squeeze which has an older
> gcc by default with less strict typechecking.
Being the larger parent, I'd likely prefer Debian. At the moment,
Ubuntu is a "see about the hype" test. I should have mentioned
that I also applied all their auto updates with their 'update
manager' thingy in the menus... which may have led to any
particular differences in our environments, I don't spend a ton
of time in Linux to know.
Johannes Schlumberger
> How about versioning the tarball. Perhaps call them:
> phantom_ref-0.1.1.tar.gz b2ddd276382b4b95b90452eea78f3857
> phantom_ref-0.1.2.tar.gz 59daa0166e6ce56fd681479e9aa4b22c
> 'ref' meaning reference implementation.
> Others writing to spec might call theirs:
> phantom_other-M.m.p.tar.gz
For offixial releases this will be done. As soon as I or anyone else takes up
serious development on the code again there should/eill be a repository for it.
I have not had time to set something up so far and currently do not need it.
That is why we are currently going with the tarballs, you are right about the
versioning though, simply replacing the old one was no good procedure. My bad.
> > I have been working on a Debian squeeze which has an older
> > gcc by default with less strict typechecking.
>
> Being the larger parent, I'd likely prefer Debian. At the moment,
> Ubuntu is a "see about the hype" test. I should have mentioned
> that I also applied all their auto updates with their 'update
> manager' thingy in the menus...
These "problems" would have surfaced in debian at some point as well, so fixing
them now is just as well. I will keep testing changes in ubuntu from now on as
well as on debian.
Johannes