httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
Subject Re: cvs commit: apache-2.0/src/lib/apr/include apr_errno.h
Date Wed, 05 Apr 2000 16:18:01 GMT
On Wed, 5 Apr 2000, Jeff Trawick wrote:

> > On Wed, 5 Apr 2000, Jeff Trawick wrote:
> > 
> > > > -1 because of this patch and the description you posted earlier.
> > > 
> > > What is the relationship between this change and the part of the
> > > APRDesign patch which you objected to?  I don't see any connection.
> > 
> > You posted an doc change for how to report errors in APR, and a code
> > change for how errors are done in APR, and don't see a connection between
> > them?  
> If I interpret your disagreement with my APRDesign properly, you
> disagreed with the idea that APR should translate OS-specific error
> codes into portable APR error codes.  
> Nothing in my changes made APR translate OS-specific error codes into
> portable APR error codes.  Instead, it made it possible to return
> OS-specific error codes without having them confused with something else.

I saw a doc change that is against the current design of APR, and a code
change that changed error codes.  Silly me, I assumed you were starting a
code change that would in time implement what the docs said.  I strongly
disagreed with the doc change, and want to stop that implementation before
it starts.  Now, if I am wrong about the code change beginning to
implement the docs, then I'm sorry.  But to me that is a reasonable peice
of logic.

> > But what you are doing, AFAICT, is requiring that each platform returns
> > the same value when there is an error.  
> Show me the line of code in my change that does that.  There is no
> such line of code in the change.

And if you read my objection, it was a combination of a doc change along
with a code change that caused my objection.  A doc change and a code
change within an hour of each other that deal with the same topic, are in
my opinion related.  If I am wrong about this, then I am sorry but again,
this seems like a straight forward conclusion.

> > APR_OS_START_ERRNO2 (for platforms that have error codes that overlap with
> > errno.  ALL platforms would use this, including UNIX)
> Where in the heck did that come from?  If this was the consensus,
> surely it should have been mentioned before at least once?  I'm having
> trouble following your logic.  Maybe you are using that name instead
> of APR_OS_START_OSERR.  I don't care what the name is so much, but if
> you want to change the name, say so.  

The name is new, but the concept is the one that Brian Havard brought up
last week or the week before.  Nobody ever put a name on it before.  :-)
This one seemed to make sense to me, because it is basically an alternate
errno value.

> > APR_OS_START_ERROR  (begin APR error codes)
> > APR_OS_START_STATUS (befin APR status codes)
> > APR_OS_START_SYSERR (to map errno codes for platforms that don't have
> > them)
> > APR_OS_START_USEERR (for APR apps that want to layer their own codes along
> > with APR.  [Note, I'm not sure how useful this is, personally])
> This is all old crap that was there before and which I didn't change.
> Why is this an issue now?

It isn't an issue.  It is a part of the overall scheme, which is why it is
included here for completeness.


Ryan Bloom               
406 29th St.
San Francisco, CA 94131

View raw message