[reSIProcate] Packaging for Debian

[Robert Sparks]

Thank you Neil!

One quick comment on 2) below.

We have effectively forked ares - resiprocate won't (I think) build  
against ares straight from MIT.

RjS


On Feb 26, 2007, at 10:12 AM, Neil McGovern wrote:

> Hi all,
>
> After spending quite a bit of time with Daniel Pocock this past  
> weekend
> at FOSDEM, I've been convinced to package the reSIProcate suite for
> Debian.
>
> Now, there's a couple of issues that remain before I can upload it,  
> and
> I was hoping you'd be able to help :)
>
>
> 1) Licences and copyright.
>
> It seems that the source for resiprocate contains multiple projects,
> under potentially different licences, and by different copyright
> holders.  Quite a bit of this is ok, but some have issues.
>
> A quick summary:
>
> Name			Copyright				Licence			Problematic
> ----			---------				-------			-----------
> sipdial			Daniel Pocock			none			Yes
> repro			Vovida Networks, Inc	VSL, 1.0		No
> stack			Vovida Networks, Inc	VSL, 1.0		No
> rutil			Vovida Networks, Inc	VSL, 1.0		No
> tfm				PurpleComm, Inc			BSD 3-clause	No
> build-system	Unknown					none			Yes
>
> In summary: sipdial and the build system need a licence, or they're  
> not
> distributable. If the copyright of the others could also be checked,
> that'd be great :)
>
>
> 2) Embedded code copies
>
> tfm and rutil seem to both rely on copies of other libraries shipped
> with resiprocate, cppunit and ares respectively.
>
> This causes us (Debian) some major headaches when it comes to  
> providing
> security support should a hole be found in one of these packages. I  
> had
> a minor heart attack when I saw pcre included, but it doesn't seem to
> need it to build.
>
> Removing the -I line from tfm builds it against the Debian version
> properly, so that isn't an issue.
>
> With rutil/ares, it seems it doesn't properly install the header files
> making rutil fail to build from source without the included files.  
> A bug
> (http://bugs.debian.org/412515) has been filed, and so hopefully this
> should be fixed shortly.
>
>
> 3) Other (minor) issues which could be fixed, but aren't show  
> stoppers are:
>
> tfm seems to have a compile option of -I/usr/local/include/boost-1_33,
> which isn't present. This, fortunately doesn't cause a problem, but
> possibly shoudn't be there.
>
> contrib/db, contrib/getopt, contrib/pcre and contrib/popt don't  
> seem to
> be used at all, and can probably be removed.
>
>
> That's about it for now, I'll continue to try and get these built, but
> it's non-trivial as the source package is building multiple binaries,
> headers and libraries, so it may take some time :)
>
> Many thanks,
> Neil
> -- 
> A. Because it breaks the logical sequence of discussion
> Q. Why is top posting bad?
> gpg key - http://www.halon.org.uk/pubkey.txt ; the.earth.li B345BDD3
> _______________________________________________
> resiprocate-devel mailing list
> resiprocate-devel at list.resiprocate.org
> https://list.resiprocate.org/mailman/listinfo/resiprocate-devel