Closed GoogleCodeExporter closed 9 years ago
The main issues I see with the legacy code now are
a) there's a lot of calling subroutines simply for the sake of getting the code
out
of the way so the logical flow of the program is more obvious. But, these
subroutines all deal with global variables, so the code is actually MORE
difficult to
truly understand now. It needs to be properly abstracted into a module, more
than
likely, so global variable and side effects are alleviated.
b) the special cases have been factored out in a way that makes the program not
even
start to work until all the databases and tables are discovered. If there's an
error, partial output is not possible. In the real world this matters a lot.
See
0xdeadbeef for the places where this shows up most.
Original comment by baron.schwartz
on 17 Oct 2008 at 7:55
See also the comments on r2410
Original comment by baron.schwartz
on 17 Oct 2008 at 8:35
See also issue 99.
Original comment by m...@pascalhofmann.de
on 27 Oct 2008 at 1:36
This issue is related to issue 444. I'm block it on issue 444 because however
we
re-do MySQLFind and such will influence how we re-do mk-table-checksum
internals.
Original comment by dan...@percona.com
on 3 Jun 2009 at 8:00
Original comment by baron.schwartz
on 24 Jan 2010 at 8:56
Original comment by baron.schwartz
on 24 Jan 2010 at 8:59
Original comment by baron.schwartz
on 1 Feb 2010 at 1:28
This is not PossiblyDone, not even really done at all. :-) One major issue is
that
it gets a list of all tbls before working. It'd be nicer if it went
table-by-table,
streaming-like.
Original comment by dan...@percona.com
on 23 Mar 2010 at 7:03
This issue has surfaced again in issue 1302. I'm closing this very old issue.
Original comment by dan...@percona.com
on 13 May 2011 at 2:57
Original issue reported on code.google.com by
dan...@percona.com
on 17 Oct 2008 at 7:47