Re: Any limits on Rcs version numbers?

[Aaron S. Hawley]

Hal,

RCS version numbers are obviously stored and i believe internally
represented as strings, so any of the usual machine arithmetic limits are
irrelevant.

did you try this?

touch foo
ci -i365000000 foo

so revision numbers wouldn't appear to be a problem.  who knows if there
aren't any other design imposed bug limits in RCS.

these sorts of problems are what the usual database management systems
(DBMS) are for.

good luck,
/a

On Sun, 7 Dec 2003, Hal Morris wrote:

> Are there any limits on rcs version numbers?
>
> I.e. can I do 100 deltas a day for 10 years and it won't break down?
>
> I'm thinking of using rcs in the implementation of an unorthodox sort of
> database.  The data will take the form a large Perl structure, which
> is output with Data::Dumper and input just by 'do'-ing the output file.
> In fact I'm already using something like this, but wondering how far I can
> push it.
>
> (You must for sure do this
>
>    $Data::Dumper::Sortkeys = 1;
>
> if you want it to work reasonably well -- otherwise you add an element
> to a hash which may cause the hash table to be reallocated, and then
> every element gets rearranged randomly).
>
> Please at least cc me on replies; I don't always watch the list closely.
>
> Thanks, and best wishes,  Hal

-- 
"But, look, the [Israeli security] fence is a sensitive issue, I
understand that. And the Prime Minister [Sharon] made it very clear to me
that it was a sensitive issue.  And my promise to him is we'll continue to
discuss and to dialogue how best to make sure that the fence sends the
right signal."
        -- Ariel Sharon's Play Friend, "George" [W. Bush]