You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@httpd.apache.org by "Ralf S. Engelschall" <rs...@engelschall.com> on 1998/04/02 08:06:53 UTC

Global Renaming: Let us start!

In article <19...@devsys.jaguNET.com> you wrote:

> I, for one, am tending to think that NOT releasing 1.3b6 at all until
> this whole issue has been solved is a Good Idea. This means,
> of course, that some sort of "documented" and "name-space friendly"
> API will and must exist for 1.3.0

Yes, let us immediately stop the discussion about Roys veto and look forward
what we now can do to bring 1.3b6 out and not wasting time. I'm already
started to write the "global renaming script", so my plan is this:

1. I finish writing the "rename" script

2. I insert it and a rename.cf file somewhere into the CVS which holds
   "<old-symbol> <new-symbol-with-prefix>" lines and which should be both used
   to sync us with the prefix, what's API and what's not and finally can be
   used both as the input for the "rename" script and to initially create an
   apifunc.html document where we fill in documentation for the then official
   API functions.

3. More then one of us run the script over a fresh CVS checked-out tree and
   look if all works well. C comments and variables and other stuff are a
   little bit problematic. Function renaming is trivial. I plan to make the
   script semi-automatic, i.e. all 100% correctly identifyable syntax is
   automatically renamed, all other stuff needs Y/N keystrokes. To avoid as
   much manual intervention I currently try parsing ctags output or help from
   the mkid tools.

4. If the renaming is stable and correct, I'll run it
   finally and commit the renamed sources.

5. Stop adding stuff and kick out 1.3b6 after a short test period

6. Bugfix, Bugfix, Bugfix... and a longer test period

7. Consider 1.3.0 to be kicked out!

Even if the group decides not to run the renaming for 1.3b6 the script is
needed, so I go on working on it. At least for later time it _is_ needed.

                                       Ralf S. Engelschall
                                       rse@engelschall.com
                                       www.engelschall.com

Re: Global Renaming: Let us start!

Posted by Rodent of Unusual Size <Ke...@Golux.Com>.
Marc Slemko wrote:
> 
> I have said this before and will say it one last time before I give
> up: 1.3 is not the place for this sort of screwing with the API
> nor is it the place for some magical API that suddenly becomes
> documented, etc.  We will waste more time over this, and already have,
> than it is worth and will not get much out of it in the end.

I agree 100% with all of the above.  API documentation for 1.3.*
is a "nice-to-have" IFF someone wants to invest its own personal
effort - not the group's, except for feedback.

> 2.0 is damn close to dead in my mind.  I'm starting to think that
> what is necessary is two or three people, a couple of weeks, a nice
> cabin in the mountains with a few computers, and no new-httpd.

I haven't yet gotten to that point about 2.0, but the life-support
teams *do* seem to be organising for withdrawal.. <g>

#ken	P-)}

Ken Coar                    <http://Web.Golux.Com/coar/>
Apache Group member         <http://www.apache.org/>
"Apache Server for Dummies" <http://Web.Golux.Com/coar/ASFD/>

Re: Global Renaming: Let us start!

Posted by Manoj Kasichainula <ma...@io.com>.
On Thu, Apr 02, 1998 at 06:11:38AM -0500, Jim Jagielski wrote:
> I'd agree, except that (1) this doesn't avoid the whole hide.h/HIDE
> issue. It still needs to be dealt with (2) hide.h/HIDE as it now
> exists is a hack (3) depending on whether HIDE is defined or not,
> the actual definition of the API changes, which isn't a good thing
> IMO and (4) 1.3 will be out for a LONG time... 1.3.1 or whatever ain't
> the time to fix it. This really should be a 1.3.0 issue

Excuse me for the intrusion, but to me, it seems that what y'all need
is a list of the things you must have 1.3. Then *stop* there. No more,
no less. Maybe an idea freeze in a week (or right now), and a feature freeze
a few of weeks from now.

-- 
Manoj Kasichainula - manojk at io dot com - http://www.io.com/~manojk/
"And the moral of today's story is: Lather, Rinse, Repeat." - Yakko Warner

Re: Global Renaming: Let us start!

Posted by Marc Slemko <ma...@worldgate.com>.
On Thu, 2 Apr 1998, Ralf S. Engelschall wrote:

> 
> In article <19...@devsys.jaguNET.com> you wrote:
> 
> > I, for one, am tending to think that NOT releasing 1.3b6 at all until
> > this whole issue has been solved is a Good Idea. This means,
> > of course, that some sort of "documented" and "name-space friendly"
> > API will and must exist for 1.3.0
> 
> Yes, let us immediately stop the discussion about Roys veto and look forward
> what we now can do to bring 1.3b6 out and not wasting time. I'm already
> started to write the "global renaming script", so my plan is this:
> 
> 1. I finish writing the "rename" script
> 
> 2. I insert it and a rename.cf file somewhere into the CVS which holds
>    "<old-symbol> <new-symbol-with-prefix>" lines and which should be both used
>    to sync us with the prefix, what's API and what's not and finally can be
>    used both as the input for the "rename" script and to initially create an
>    apifunc.html document where we fill in documentation for the then official
>    API functions.

I have said this before and will say it one last time before I give
up: 1.3 is not the place for this sort of screwing with the API
nor is it the place for some magical API that suddenly becomes
documented, etc.  We will waste more time over this, and already have,
than it is worth and will not get much out of it in the end.

2.0 is damn close to dead in my mind.  I'm starting to think that
what is necessary is two or three people, a couple of weeks, a nice
cabin in the mountains with a few computers, and no new-httpd.

Oh well.