2013/6/15 Richard Hipp <d...@sqlite.org>:
> On Fri, Jun 14, 2013 at 8:15 PM, Edward Berner <e...@bernerfam.com> wrote:
>>
>> Hello,
>>
>> db_last_insert_rowid() is defined in db.c to return an "i64", but every
>> use of the function stores the result in an "int".
>
>
> I guess it might cause problems - if you created a repository with over 2
> billion artifacts.  A typical busy project (ex: Tcl, SQLite) accumulates
> between 1K and 10K artifacts per year.  So this probably won't be a problem
> any time soon.   I think the assumption of of less than 2 billion artifacts
> probably permeates the code, so this might be a really big patch, though.
> Are you sure it is worth it?

Hm, I would suggest to change the return-type from i64 to size_t. On
32-bit systems that would double the maximum number of artifacts,
with hardly any impact.  See:
       <http://fossil-scm.org/index.html/info/e1cb483a9b>
Of course, some more internal variables have to change type, but
it's a harmless change, and it will mean that on 64-bit systems
there simply will be no limit any more.

The only hassle that needed to be solved first was
64-bit support for JSON on win64, but that's done now.

Richard, Stephan, how do you feel about this?

Regards,
           Jan Nijtmans
_______________________________________________
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Reply via email to