On Thu, 11 May 2000, Olaf Titz wrote:
> > It's certainly intended to be stable, but it's called "pre" if we're not
> > quite convinced yet. This "pre" suffix just seems unnecessarily confusing.
>
> And it makes KERNEL_VERSION ifdefs impossible when the changes happen
> between two pre versions, like the conversion of name to an array in
> struct net_device currently underway. This is the main reason why I
> would like to abolish those suffixes altogether or make them strictly
> numeric.
Yes, exactly. It's easy to understand how to compare versions that are all
numeric. (The trend seems to be towards two or more dot-separated integers
than using real numbesr.) It gets more confusing when adding suffixes like
"pre", alpha/beta tags, lettered subversions, etc.
My original thought was "how could we get rid of this 'pre' numbering?" --
sometimes .9x numbers are used to represent near-releases, as when Linux
jumped from 0.1x to 0.9x range rapidly. Unfortunately, there were so many
releases before the final 1.0 release that versions like "0.99pl26" had to
be used, again too confusing.
Since the ".9" idea seems somewhat intuitive to people, I was thinking just
to extend the versions with another number, using xx.xx.9.xx for release
candidates. Making the version number a vector of integers allows for
unlimited release candidates before the final release.
Working backwards, I tried to come up with a potentially useful scheme for
4-part version numbers on development kernels that would attempt to encode
the development phase into the version number, allowing for more precision
than simply "this is a release candidate". Hence, my (somewhat arbitrary)
proposal. There's a lot of ways to possibly tune the sequence; I posted a
first draft as it came to mind, but I'll probably try posting a revised
version sometime soon.
Deven
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Mon May 15 2000 - 21:00:18 EST