Huge thanks to our Platinum Members Endace and LiveAction,
and our Silver Member Veeam, for supporting the Wireshark Foundation and project.

Wireshark-dev: Re: [Wireshark-dev] What is Wireshark 1.6.0 ?

From: Gerald Combs <gerald@xxxxxxxxxxxxx>
Date: Fri, 12 Aug 2011 08:46:09 -0700
On 8/12/11 1:52 AM, news.gmane.com wrote:
> 
> "Gerald Combs" <gerald@xxxxxxxxxxxxx> wrote in 
> message news:4E43F85A.6070809@xxxxxxxxxxxxx...
>> If you take a fresh SVN checkout (*not* an export) of
>> http://anonsvn.wireshark.org/wireshark/releases/wireshark-1.6.0 and run
>> the following:
>>
>>  perl make-version.pl --package-version
>>  ./autogen.sh
>>  ./configure
>>  make
>>  make dist
>>
>> You should get your own copy of wireshark-1.6.0.tar.bz2 which is ready
>> to build on Windows, OS X, Linux, and UNIX. If your version of the GNU
>> build chain is close enough to the one running on the Buildbot builder
>> it will *probably* even be identical to the one available for download.
> 
> But would this reproduce the "offically released source files" as mentioned 
> in section 3.3.4 of WSDG?

As I said, *probably*. I'm not going to guarantee that the Autotools
chain on your system will produce the exact same results as the chain on
the builder.

> Is the SVN release branch official or the wireshark-1.6.0.tar.bz2?

wireshark-1.6.0.tar.bz2 is official, in the sense that that's what is
published for download and listed (with md5, sha1, and rmd160 hashes) in
the release announcement. It is derived from the SVN release branch
using the steps listed above.


> Why should I use SVN checkout instead of export? The checkout generates a 
> lot of .svn directories that I only need when I modify the files and want to 
> check in probably later.

The .svn directories include a revision number which is used in a number
of places in the release packages. It's fetched with "make-version.pl
--package-version".


> I just want to find the best solution to get the header file and 
> supplementaty files that are necessary to compile my dissectors. Your 
> proposal to "make" the whole thing unfortuanately doesn't work because a 
> problem with the GeoIP library (see other thread).

I'm not proposing that you do anything. I'm trying to fill in gaps that
you seem to have already filled in with invalid assumptions.