Hi,
There are many many ways to format a version number. The developers looked at
various options and decided upon an even/odd number scheme for releases and
development versions. The agreed upon compromise is discussed here:
http://wiki.wireshark.org/Development/ReleaseNumbers
It's not perfect, but what numbering scheme is? This will do for Wireshark for
now.
Thanx,
Jaap
Andrew Hood wrote:
Jaap Keuter wrote:
Hi,
So isn't it time now to 'correct' the development release number from 1.0.99
to 1.1.0?
I've never liked this sort of version numbers since string comparisons
don't sort them correctly.
Consider GTK's:
#define GTK_CHECK_VERSION(major,minor,micro) \
(GTK_MAJOR_VERSION > (major) || \
(GTK_MAJOR_VERSION == (major) && GTK_MINOR_VERSION > (minor)) || \
(GTK_MAJOR_VERSION == (major) && GTK_MINOR_VERSION == (minor) && \
GTK_MICRO_VERSION >= (micro)))
#if GTK_CHECK_VERSION(2,4,10)
and Motif's:
#define XmVersion (XmVERSION * 1000 + XmREVISION)
#if XmVersion>2003
all caused by pseudo numeric version numbers.
1.0.99 would logically be followed by 1.0.100, but "1.0.99" > "1.0.100"
Can we start from now at "1.01.000"? A hundred point releases each with
a thousand patch levels should be enough for anyone. (With apologies to
Bill Gates.)
