[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: revision number upper boundary?

From: Steve Greenland <steveg_at_lsli.com>
Date: 2005-03-08 23:23:24 CET

On Tue, Mar 08, 2005 at 08:15:43PM -0000, Max Bowsher wrote:
> 2147483647, i.e. slightly more than 2 billion.
>
> That's assuming you are using a 32bit machine, of course.
>
> So, if you make 1 commit every second without pause, that will last you
> until the year 2073, by which time, it's a fairly safe bet that you will
> have moved on to a 64bit machine, and can enjoy the freedom to have up to 9
> quintillion revisions.

ObOffTopic: actually, many C implementations for 64bit OS's keep 'long'
as a 32 bit quantity, using the abomination 'long long' for 64bit
integer types. They do this for ABI consistency issues, attempting not
to break third-party libraries. This was a heated topic during the C99
standardization process, because allowing 'long' to be shorter than
the largest standard type breaks code that actually followed the C89
standard.

So as long (heh!) as the subversion revision is kept typed as
'long int', it will likely stay a 32 bit integer.

(Hmmm, knowing nothing about the subversion wire protocol and/or on
disk formats, that 'long int' would worry me. Seems like it ought to be
int32_t...)

Steve

-- 
"Outlook not so good." That magic 8-ball knows everything! I'll ask
about Exchange Server next.
                           -- (Stolen from the net)
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
For additional commands, e-mail: users-help@subversion.tigris.org
Received on Tue Mar 8 23:26:07 2005

This is an archived mail posted to the Subversion Users mailing list.

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.