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

Re: Sanity check: removing src_err

From: Glenn A. Thompson <gthompson_at_cdr.net>
Date: 2002-12-10 08:23:10 CET

Greg Hudson wrote:

>On Tue, 2002-12-10 at 01:32, Glenn A. Thompson wrote:
>
>
>>Hey,
>>How do you plan on detecting DEADLOCK Errors?
>>
>>
>
>Probably the easiest way is a dedicated svn_fs error code for deadlocks.
>
>As to why I dislike src_err: it is an extra argument to every
>svn_error_create call for the sake of a vanishingly small number of
>cases, its value lives in an indeterminate number space, and it renders
>the use of the error system prone to confusing inconsistencies
>
There could conceivably be other errors other than DEADLOCK that would
be retryable.
I guess I could map them onto DEADLOCK or create other ERRORs. I'm not
to thrilled about mapping every error condition into APR space. How
about an assessor to set the field? Passing a DB error code through for
printing, does have some usefulness. I mean Oracle has tons of them.
 They even have a program to given a verbal explanation of the error. I
would actually prefer a char field. DBAs may want these codes. They
couldn't care less about it being mapped onto APR_ERRORs. They may
need to resize a sort area or something similar. Errors can be
meaningful to the right person even if it is in indeterminate number space.

While we are talking about Error stuff:

svn_fs_set_berkeley_errcall never gets called.
I hate it as much as you hate src_err:-)

It seems to make sense to implement some sort of logging routine. Why not do just that.
Call it with a file handle to log to. This could be abstracted for other FS implementations as well.

 gat

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Tue Dec 10 08:22:02 2002

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