Re: [ecasound] Compiletime warnings

From: Philipp Überbacher <hollunder@email-addr-hidden>
Date: Sat Sep 04 2010 - 15:10:42 EEST

Excerpts from Kai Vehmanen's message of 2010-09-04 13:26:04 +0200:
> Hi,
> On Sat, 17 Jul 2010, Philipp wrote:
> > jack_client_new() is deprecated since quite a long time. Use
> > jack_client_open() instead:
> >
> this is now fixed in ecasound's git tree and will be in the next release.
> I did not change to jack_client_open() immediately as I didn't want
> to force people to update their JACK version (e.g. Debian provided a very
> old JACK for a long time) for this reason alone. But of course,
> jack_client_new() was deprecated in 2004, so it's about time for really
> stop using it in ecasound as well. Apologies to everyone still using a
> pre-2004 JACK in their systems! ;)

Thanks. Yep, I also think it's about time ;)
There were a few new jack versions and deprecations in the meanwhile :)

> > This is a very common warning since a couple of gcc versions, 4.2 or 4.3.
> >
> > eca-curses.cpp:91:23: warning: deprecated conversion from string constant to ?char*?
> > eca-curses.cpp:95:23: warning: deprecated conversion from string constant to ?char*?
> > textdebug.cpp:160:27: warning: deprecated conversion from string constant to ?char*?
> > textdebug.cpp:179:27: warning: deprecated conversion from string constant to ?char*?
> I believe this is a bug in the ncurses package on your system. On my
> systems (mostly Debian), tigetstr() takes a 'const char*' param (so I
> don't get the warning).

I don't know for certain, I don't know enough C/C++ to understand what's
going on but this kind of warning is really all over the place, there
was hardly a piece of software not affected.
A quick search came up with this post:

I've also seen something somewhere on the gcc page sometime but
naturally can't find it anymore. So if your gcc is old, it might well be
this issue.

> > /usr/include/python2.6/pyconfig.h:1040:0: warning: "_XOPEN_SOURCE" redefined
> > libtool: install: warning: remember to run `libtool --finish /usr/lib'
> > libtool: install: warning: remember to run `libtool --finish /usr/lib'
> > libtool: install: warning: remember to run `libtool --finish /usr/lib'
> This a bug in the python2.6 package (pyconfig.h should not unconditionally
> redefine _XOPEN_SOURCE).
> Despite me dodging the bullet with these, it's still a good idea to report
> any warnings (especially new ones) when building Ecasound.

Ok. It's hard for me to know where a bug comes from, so I keep reporting
Even with my little piece of software I noticed that stuff behaves
different than I expected.

This was especially the case with my little reconnect thingy. I didn't
know jack client/port names with spaces aren't allowed, so it didn't
work with those. Escaping spaces was fun too, '' or "" works with
jack_connect but only \ works with ecasounds jack-connect. I ran into
another unexpected behavior when trying to avoid parsing the connections
when unnecessary. There I found that in some cases jack_lsp returns a
different order of connections for the same graph, consequently a
different string and thus breaking my optimization. It seems like
software is full of unexpected behavior :)


"Wir stehen selbst enttäuscht und sehn betroffen / Den Vorhang zu
und alle Fragen offen." Bertolt Brecht, Der gute Mensch von Sezuan
This Dev2Dev email is sponsored by:
Show off your parallel programming skills.
Enter the Intel(R) Threading Challenge 2010.
Ecasound-list mailing list
Received on Sat Sep 4 16:15:06 2010

This archive was generated by hypermail 2.1.8 : Sat Sep 04 2010 - 16:15:06 EEST