You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@apr.apache.org by "William A. Rowe, Jr." <wr...@covalent.net> on 2001/06/28 00:56:26 UTC
Re: module magic number
From: "William A. Rowe, Jr." <wr...@covalent.net>
Sent: Wednesday, June 27, 2001 5:17 PM
> rbb observed...
> > [...] whenever an API changes, the module magic number must be bumped.
>
> *should* being the operative word here. Don't see that happening. I've
> bumped it at least when the really, really cryptic stuff has changed, such as
> harrie's scoreboard reorg. I suppose the Apache group should simply bump it
> as a matter of course on each version bump till the API is at least modestly
> stable.
By the way, this didn't bug me until we started calling releases "Beta Quality."
Folks are building modules and distributing them for the beta. So we are remiss
in not keeping track of this. Can we please bump the MMN before each release until
the apr and httpd API's have completely settled out (basically, every time till we
declare a release GA quality?)
It brings up a deeper issue. I'm thinking we actually need a composite magic
number... the apr and apr-util changes we make also impact new-httpd, but there
is no clear way to signal those to the httpd list when apr is mucking things up.
Do you all suppose we ought to have an APR magic number, that assures the httpd
people that the arguments to existing apr functions haven't been flipped about?
This way the apr list can assume responsibility for their changes, while the
httpd RM doesn't have to review that work?
Bill
Re: module magic number
Posted by "Roy T. Fielding" <fi...@ebuilt.com>.
Somebody just needs to commit it -- I haven't been doing it on every
change to the binary API because there were so many. I just bumped it.
....Roy
Re: module magic number
Posted by Cliff Woolley <cl...@yahoo.com>.
On Wed, 27 Jun 2001, William A. Rowe, Jr. wrote:
> It brings up a deeper issue. I'm thinking we actually need a
> composite magic number... the apr and apr-util changes we make also
> impact new-httpd, but there is no clear way to signal those to the
> httpd list when apr is mucking things up.
>
> Do you all suppose we ought to have an APR magic number, that assures
> the httpd people that the arguments to existing apr functions haven't
> been flipped about? This way the apr list can assume responsibility
> for their changes, while the httpd RM doesn't have to review that
> work?
+1. I'd been wondering about that myself as I've been tweaking various
API's.
--Cliff
--------------------------------------------------------------
Cliff Woolley
cliffwoolley@yahoo.com
Charlottesville, VA
Re: module magic number
Posted by Cliff Woolley <cl...@yahoo.com>.
On Wed, 27 Jun 2001, William A. Rowe, Jr. wrote:
> It brings up a deeper issue. I'm thinking we actually need a
> composite magic number... the apr and apr-util changes we make also
> impact new-httpd, but there is no clear way to signal those to the
> httpd list when apr is mucking things up.
>
> Do you all suppose we ought to have an APR magic number, that assures
> the httpd people that the arguments to existing apr functions haven't
> been flipped about? This way the apr list can assume responsibility
> for their changes, while the httpd RM doesn't have to review that
> work?
+1. I'd been wondering about that myself as I've been tweaking various
API's.
--Cliff
--------------------------------------------------------------
Cliff Woolley
cliffwoolley@yahoo.com
Charlottesville, VA