ANNOUNCEMENT: Live Wireshark University & Allegro Packets online APAC Wireshark Training Session
April 17th, 2024 | 14:30-16:00 SGT (UTC+8) | Online

Wireshark-dev: Re: [Wireshark-dev] [Wireshark-commits] rev 51118: /trunk/epan/ /trunk/epan/: ad

From: Jeff Morriss <jeff.morriss.ws@xxxxxxxxx>
Date: Fri, 02 Aug 2013 21:43:53 -0400
On 08/02/2013 08:50 PM, Jeff Morriss wrote:
On 08/02/2013 08:48 PM, Joerg Mayer wrote:
On Fri, Aug 02, 2013 at 05:17:33PM -0700, Gerald Combs wrote:
On 8/2/13 5:08 PM, Joerg Mayer wrote:
On Fri, Aug 02, 2013 at 02:53:13PM +0000, etxrab@xxxxxxxxxxxxx wrote:
http://anonsvn.wireshark.org/viewvc/viewvc.cgi?view=rev&revision=51118

User: etxrab
Date: 2013/08/02 07:53 AM

Log:
  g_int64_hash() and g_int64_equal() first appear in GLib 2.22,
make a local copy.

Directory: /trunk/epan/
   Changes    Path             Action
   +45 -0     addr_resolv.c    Modified

How about requiring a newer version of glib for HEAD instead?

GLib 2.22 was released in September 2009, so 2.22 or later might be a
reasonable expectation by the time we make the next major release. If we
do make 2.22 the minimum version we can remove the same check + code
from packet-infiniband.c.

Or even 2.24, which is what Debian old stable provides.

If we do 2.24 we lose support RHEL/CentOS 6:

https://www.wireshark.org/lists/wireshark-dev/201305/msg00028.html

For that reason I'd vote for 2.22 instead.

FWIW I created a page on the wiki to track Glib and Gtk versions and the OS distributions they came in:

http://wiki.wireshark.org/Development/Glib_Gtk_version_tracking

I only put Redhat-based systems in there because that's all I really know.