You are viewing a plain text version of this content. The canonical link for it is here.
Posted to cvs@httpd.apache.org by sf...@apache.org on 2013/06/10 23:41:07 UTC

svn commit: r1491612 - /httpd/httpd/branches/2.4.x/STATUS

Author: sf
Date: Mon Jun 10 21:41:07 2013
New Revision: 1491612

URL: http://svn.apache.org/r1491612
Log:
comment

Modified:
    httpd/httpd/branches/2.4.x/STATUS

Modified: httpd/httpd/branches/2.4.x/STATUS
URL: http://svn.apache.org/viewvc/httpd/httpd/branches/2.4.x/STATUS?rev=1491612&r1=1491611&r2=1491612&view=diff
==============================================================================
--- httpd/httpd/branches/2.4.x/STATUS (original)
+++ httpd/httpd/branches/2.4.x/STATUS Mon Jun 10 21:41:07 2013
@@ -198,6 +198,8 @@ PATCHES PROPOSED TO BACKPORT FROM TRUNK:
       trunk patch: http://svn.apache.org/r1491155
       2.4.x patch: trunk patch works
       nd: why would you do that in a stable branch?
+      sf: Because it is only annoying and serves no purpose anymore. If you
+          want, we can make it a minor MMN bump for adding a "new" API.
       +1: sf, covener
       -1: nd
 



Re: svn commit: r1491612 - /httpd/httpd/branches/2.4.x/STATUS

Posted by "William A. Rowe Jr." <wr...@rowe-clan.net>.
On Wed, 12 Jun 2013 21:18:05 +0200
Stefan Fritsch <sf...@sfritsch.de> wrote:

> It is a useful function. Disallowing it breaks ISO-C compliant code 
> for no good reason. There has been a complaint by a module author. 
> Allowing it does not use any memory because the function is already
> in libc. Working around the missing function requires additional
> code, which uses memory and may contain bugs. What more reasons do
> you need?

I'm +1... we should only perform such abuses of stdc / clib within
the scope of our own -private- headers.  Modifying these in our headers
meant for public consumption is a very irritating thing to do.  If
this were hidden in some CORE_PRIVATE scope then I would have been fine
with leaving things as is.


AW: svn commit: r1491612 - /httpd/httpd/branches/2.4.x/STATUS

Posted by Plüm, Rüdiger, Vodafone Group <ru...@vodafone.com>.

> -----Ursprüngliche Nachricht-----
> Von: Rainer Jung []
> Gesendet: Freitag, 14. Juni 2013 18:08
> An: dev@httpd.apache.org
> Betreff: Re: svn commit: r1491612 - /httpd/httpd/branches/2.4.x/STATUS
> 
> We do cosmetical changes in stable branches like e.g. reducing code
> drift between trunk and 2.4 or fixing typos in comments. We do it if we
> see only very little risk and anticipate at least some benefit now or
> for the future. In the case discussed here I'd say the risk is very
> small and the benefit of a more complete c89 support for module authors
> is real (but of course not big). I like the addition of the MMN bump
> suggested by Jeff.

+1

Regards

Rüdiger


Re: svn commit: r1491612 - /httpd/httpd/branches/2.4.x/STATUS

Posted by Jim Jagielski <ji...@jaguNET.com>.
On Jun 14, 2013, at 1:20 PM, Eric Covener <co...@gmail.com> wrote:

>> Of course. That's not the question (at least IMO). The question
>> is that when a cosmetic change also results in a functional
>> change (and we wouldn't be suggesting MMN bumps if it wasn't),
>> that it becomes a change that should be proposed as a backport
>> and not willy-nilly added. We have a process, and should
>> discourage attempts to bypass it on a stable trunk.
> 
> This change followed the process. It currently sits in 2.4.x/STATUS.

I know... I voted on it :)

I was speaking in generalities... sorry for the confusion.

Re: svn commit: r1491612 - /httpd/httpd/branches/2.4.x/STATUS

Posted by Eric Covener <co...@gmail.com>.
> Of course. That's not the question (at least IMO). The question
> is that when a cosmetic change also results in a functional
> change (and we wouldn't be suggesting MMN bumps if it wasn't),
> that it becomes a change that should be proposed as a backport
> and not willy-nilly added. We have a process, and should
> discourage attempts to bypass it on a stable trunk.

This change followed the process. It currently sits in 2.4.x/STATUS.

Re: svn commit: r1491612 - /httpd/httpd/branches/2.4.x/STATUS

Posted by Jim Jagielski <ji...@jaguNET.com>.
On Jun 14, 2013, at 12:07 PM, Rainer Jung <ra...@kippdata.de> wrote:

> 
> We do cosmetical changes in stable branches like e.g. reducing code
> drift between trunk and 2.4 or fixing typos in comments.

Of course. That's not the question (at least IMO). The question
is that when a cosmetic change also results in a functional
change (and we wouldn't be suggesting MMN bumps if it wasn't),
that it becomes a change that should be proposed as a backport
and not willy-nilly added. We have a process, and should
discourage attempts to bypass it on a stable trunk.

FWIW, I'd be +1 with the bump.


Re: svn commit: r1491612 - /httpd/httpd/branches/2.4.x/STATUS

Posted by Rainer Jung <ra...@kippdata.de>.
On 14.06.2013 17:44, André Malo wrote:
> On Friday 14 June 2013 17:34:26 Rainer Jung wrote:
>> On 14.06.2013 16:41, André Malo wrote:
>>> On Wednesday 12 June 2013 21:18:05 Stefan Fritsch wrote:
>>>> On Tuesday 11 June 2013, André Malo wrote:
>>>>>>        trunk patch: http://svn.apache.org/r1491155
>>>>>>        2.4.x patch: trunk patch works
>>>>>>        nd: why would you do that in a stable branch?
>>>>>>
>>>>>> +      sf: Because it is only annoying and serves no purpose
>>>>>> anymore. If you +          want, we can make it a minor MMN bump
>>>>>> for adding a "new" API. +1: sf, covener
>>>>>>
>>>>>>        -1: nd
>>>>>
>>>>> Long discussions in STATUS are kinda tedious ;-)
>>>>>
>>>>> Well, I think, such changes are what trunk is for. Why not simply
>>>>> leave  everything below as-is? Even more if it removes only an
>>>>> annoyance? Or is there a real technical reason I'm just not seeing
>>>>> right now?
>>>
>>> [...]
>>>
>>>> Or, is there a real technical reason to keep it broken in 2.4?
>>>
>>> Annoying rhetoric games aside - we went from "only annoying" to "broken".
>>> What is it now?
>>>
>>> No other opinion on this?
>>
>> As far as I understand the matter, the block removed by the above commit
>> would throw a compiler error if code uses strtoul() and includes httpd.h.
>>
>> The motivation was that at the time the block was introduced some
>> supported platforms, e.g. SunOS 4, did not have strtoul(), so it was
>> helpful to throw that error even when compiled on other platforms to
>> ensure compatibility.
>>
>> The error will also be thrown for any 3rd-party code that uses strtoul()
>> and includes httpd.h when being compiled - so practically all modules
>> are prohibited to use strtoul().
>>
>> The function is part of C89 which we assume as a minimum when compiling
>> Apache 2.4. So I do not see any positive effects of the old block. I do
>> recognize, that it breaks module compilation for modules using
>> strtoul(). So the proposed change removes an obstacle for full CC89
>> support in modules. Furthermore I can not imagine any risk of breaking
>> stuff that worked before removing the block.
>>
>> Based on that I am +1 to remove the block.
> 
> I agree, that the block should simply die. However, I question the value of 
> doing cosmetical changes in our stable branches (which is the justification 
> in STATUS).
> 
> So the question in this specific case is: is it a cosmetical change or not?

We do cosmetical changes in stable branches like e.g. reducing code
drift between trunk and 2.4 or fixing typos in comments. We do it if we
see only very little risk and anticipate at least some benefit now or
for the future. In the case discussed here I'd say the risk is very
small and the benefit of a more complete c89 support for module authors
is real (but of course not big). I like the addition of the MMN bump
suggested by Jeff.

Regards,

Rainer



Re: svn commit: r1491612 - /httpd/httpd/branches/2.4.x/STATUS

Posted by Jim Jagielski <ji...@jaguNET.com>.
On Jun 14, 2013, at 11:44 AM, André Malo <nd...@perlig.de> wrote:
> 
> I agree, that the block should simply die. However, I question the value of 
> doing cosmetical changes in our stable branches (which is the justification 
> in STATUS).
> 

FWIW, I agree. I don't mind such clean-ups and cosmetic changes
on trunk, but functional changes (and this is) require a backport
proposal.


Re: svn commit: r1491612 - /httpd/httpd/branches/2.4.x/STATUS

Posted by André Malo <nd...@perlig.de>.
On Friday 14 June 2013 17:34:26 Rainer Jung wrote:
> On 14.06.2013 16:41, André Malo wrote:
> > On Wednesday 12 June 2013 21:18:05 Stefan Fritsch wrote:
> >> On Tuesday 11 June 2013, André Malo wrote:
> >>>>        trunk patch: http://svn.apache.org/r1491155
> >>>>        2.4.x patch: trunk patch works
> >>>>        nd: why would you do that in a stable branch?
> >>>>
> >>>> +      sf: Because it is only annoying and serves no purpose
> >>>> anymore. If you +          want, we can make it a minor MMN bump
> >>>> for adding a "new" API. +1: sf, covener
> >>>>
> >>>>        -1: nd
> >>>
> >>> Long discussions in STATUS are kinda tedious ;-)
> >>>
> >>> Well, I think, such changes are what trunk is for. Why not simply
> >>> leave  everything below as-is? Even more if it removes only an
> >>> annoyance? Or is there a real technical reason I'm just not seeing
> >>> right now?
> >
> > [...]
> >
> >> Or, is there a real technical reason to keep it broken in 2.4?
> >
> > Annoying rhetoric games aside - we went from "only annoying" to "broken".
> > What is it now?
> >
> > No other opinion on this?
>
> As far as I understand the matter, the block removed by the above commit
> would throw a compiler error if code uses strtoul() and includes httpd.h.
>
> The motivation was that at the time the block was introduced some
> supported platforms, e.g. SunOS 4, did not have strtoul(), so it was
> helpful to throw that error even when compiled on other platforms to
> ensure compatibility.
>
> The error will also be thrown for any 3rd-party code that uses strtoul()
> and includes httpd.h when being compiled - so practically all modules
> are prohibited to use strtoul().
>
> The function is part of C89 which we assume as a minimum when compiling
> Apache 2.4. So I do not see any positive effects of the old block. I do
> recognize, that it breaks module compilation for modules using
> strtoul(). So the proposed change removes an obstacle for full CC89
> support in modules. Furthermore I can not imagine any risk of breaking
> stuff that worked before removing the block.
>
> Based on that I am +1 to remove the block.

I agree, that the block should simply die. However, I question the value of 
doing cosmetical changes in our stable branches (which is the justification 
in STATUS).

So the question in this specific case is: is it a cosmetical change or not?

nd

Re: svn commit: r1491612 - /httpd/httpd/branches/2.4.x/STATUS

Posted by Rainer Jung <ra...@kippdata.de>.
On 14.06.2013 16:41, André Malo wrote:
> On Wednesday 12 June 2013 21:18:05 Stefan Fritsch wrote:
>> On Tuesday 11 June 2013, André Malo wrote:
>>>>        trunk patch: http://svn.apache.org/r1491155
>>>>        2.4.x patch: trunk patch works
>>>>        nd: why would you do that in a stable branch?
>>>>
>>>> +      sf: Because it is only annoying and serves no purpose
>>>> anymore. If you +          want, we can make it a minor MMN bump
>>>> for adding a "new" API. +1: sf, covener
>>>>
>>>>        -1: nd
>>>
>>> Long discussions in STATUS are kinda tedious ;-)
>>>
>>> Well, I think, such changes are what trunk is for. Why not simply
>>> leave  everything below as-is? Even more if it removes only an
>>> annoyance? Or is there a real technical reason I'm just not seeing
>>> right now?
> 
> [...]
>>
>> Or, is there a real technical reason to keep it broken in 2.4?
> 
> Annoying rhetoric games aside - we went from "only annoying" to "broken". What 
> is it now?
> 
> No other opinion on this?

As far as I understand the matter, the block removed by the above commit
would throw a compiler error if code uses strtoul() and includes httpd.h.

The motivation was that at the time the block was introduced some
supported platforms, e.g. SunOS 4, did not have strtoul(), so it was
helpful to throw that error even when compiled on other platforms to
ensure compatibility.

The error will also be thrown for any 3rd-party code that uses strtoul()
and includes httpd.h when being compiled - so practically all modules
are prohibited to use strtoul().

The function is part of C89 which we assume as a minimum when compiling
Apache 2.4. So I do not see any positive effects of the old block. I do
recognize, that it breaks module compilation for modules using
strtoul(). So the proposed change removes an obstacle for full CC89
support in modules. Furthermore I can not imagine any risk of breaking
stuff that worked before removing the block.

Based on that I am +1 to remove the block.

Regards,

Rainer

Re: svn commit: r1491612 - /httpd/httpd/branches/2.4.x/STATUS

Posted by Jeff Trawick <tr...@gmail.com>.
On Jun 14, 2013 11:36 AM, "Eric Covener" <co...@gmail.com> wrote:
>
> On Fri, Jun 14, 2013 at 10:41 AM, André Malo <nd...@perlig.de> wrote:
> > On Wednesday 12 June 2013 21:18:05 Stefan Fritsch wrote:
> >> On Tuesday 11 June 2013, André Malo wrote:
> >> > >        trunk patch: http://svn.apache.org/r1491155
> >> > >        2.4.x patch: trunk patch works
> >> > >        nd: why would you do that in a stable branch?
> >> > >
> >> > > +      sf: Because it is only annoying and serves no purpose
> >> > > anymore. If you +          want, we can make it a minor MMN bump
> >> > > for adding a "new" API. +1: sf, covener
> >> > >
> >> > >        -1: nd
> >> >
> >> > Long discussions in STATUS are kinda tedious ;-)
> >> >
> >> > Well, I think, such changes are what trunk is for. Why not simply
> >> > leave  everything below as-is? Even more if it removes only an
> >> > annoyance? Or is there a real technical reason I'm just not seeing
> >> > right now?
> >
> > [...]
> >>
> >> Or, is there a real technical reason to keep it broken in 2.4?
> >
> > Annoying rhetoric games aside - we went from "only annoying" to
"broken". What
> > is it now?
> >
> > No other opinion on this?
>
> I think Stefan did it justice in the STATUS remark.  I don't really
> understand the objection. Are you afraid a strtoul will sneak into one
> of our modules, or that a module developer needs this macro to protect
> them from it?
>
> --
> Eric Covener
> covener@gmail.com

If I can interject randomly...

Give it a minor bump so module authors can hide their workarounds for older
httpd in a self-documenting way, and so there's a CYA if some very odd,
existing module workaround breaks unexpectedly...

Re: svn commit: r1491612 - /httpd/httpd/branches/2.4.x/STATUS

Posted by Eric Covener <co...@gmail.com>.
On Fri, Jun 14, 2013 at 10:41 AM, André Malo <nd...@perlig.de> wrote:
> On Wednesday 12 June 2013 21:18:05 Stefan Fritsch wrote:
>> On Tuesday 11 June 2013, André Malo wrote:
>> > >        trunk patch: http://svn.apache.org/r1491155
>> > >        2.4.x patch: trunk patch works
>> > >        nd: why would you do that in a stable branch?
>> > >
>> > > +      sf: Because it is only annoying and serves no purpose
>> > > anymore. If you +          want, we can make it a minor MMN bump
>> > > for adding a "new" API. +1: sf, covener
>> > >
>> > >        -1: nd
>> >
>> > Long discussions in STATUS are kinda tedious ;-)
>> >
>> > Well, I think, such changes are what trunk is for. Why not simply
>> > leave  everything below as-is? Even more if it removes only an
>> > annoyance? Or is there a real technical reason I'm just not seeing
>> > right now?
>
> [...]
>>
>> Or, is there a real technical reason to keep it broken in 2.4?
>
> Annoying rhetoric games aside - we went from "only annoying" to "broken". What
> is it now?
>
> No other opinion on this?

I think Stefan did it justice in the STATUS remark.  I don't really
understand the objection. Are you afraid a strtoul will sneak into one
of our modules, or that a module developer needs this macro to protect
them from it?

--
Eric Covener
covener@gmail.com

Re: svn commit: r1491612 - /httpd/httpd/branches/2.4.x/STATUS

Posted by André Malo <nd...@perlig.de>.
On Wednesday 12 June 2013 21:18:05 Stefan Fritsch wrote:
> On Tuesday 11 June 2013, André Malo wrote:
> > >        trunk patch: http://svn.apache.org/r1491155
> > >        2.4.x patch: trunk patch works
> > >        nd: why would you do that in a stable branch?
> > >
> > > +      sf: Because it is only annoying and serves no purpose
> > > anymore. If you +          want, we can make it a minor MMN bump
> > > for adding a "new" API. +1: sf, covener
> > >
> > >        -1: nd
> >
> > Long discussions in STATUS are kinda tedious ;-)
> >
> > Well, I think, such changes are what trunk is for. Why not simply
> > leave  everything below as-is? Even more if it removes only an
> > annoyance? Or is there a real technical reason I'm just not seeing
> > right now?

[...]
>
> Or, is there a real technical reason to keep it broken in 2.4?

Annoying rhetoric games aside - we went from "only annoying" to "broken". What 
is it now?

No other opinion on this?

nd

Re: svn commit: r1491612 - /httpd/httpd/branches/2.4.x/STATUS

Posted by Stefan Fritsch <sf...@sfritsch.de>.
On Tuesday 11 June 2013, André Malo wrote:
> >        trunk patch: http://svn.apache.org/r1491155
> >        2.4.x patch: trunk patch works
> >        nd: why would you do that in a stable branch?
> >
> > +      sf: Because it is only annoying and serves no purpose
> > anymore. If you +          want, we can make it a minor MMN bump
> > for adding a "new" API. +1: sf, covener
> >
> >        -1: nd
> 
> Long discussions in STATUS are kinda tedious ;-)
> 
> Well, I think, such changes are what trunk is for. Why not simply
> leave  everything below as-is? Even more if it removes only an
> annoyance? Or is there a real technical reason I'm just not seeing
> right now?

It is a useful function. Disallowing it breaks ISO-C compliant code 
for no good reason. There has been a complaint by a module author. 
Allowing it does not use any memory because the function is already in 
libc. Working around the missing function requires additional code, 
which uses memory and may contain bugs. What more reasons do you need?

Or, is there a real technical reason to keep it broken in 2.4?

Re: svn commit: r1491612 - /httpd/httpd/branches/2.4.x/STATUS

Posted by André Malo <nd...@perlig.de>.
* sf@apache.org wrote:

> Author: sf
> Date: Mon Jun 10 21:41:07 2013
> New Revision: 1491612
>
> URL: http://svn.apache.org/r1491612
> Log:
> comment
>
> Modified:
>     httpd/httpd/branches/2.4.x/STATUS
>
> Modified: httpd/httpd/branches/2.4.x/STATUS
> URL:
> http://svn.apache.org/viewvc/httpd/httpd/branches/2.4.x/STATUS?rev=149161
>2&r1=1491611&r2=1491612&view=diff
> =========================================================================
>===== --- httpd/httpd/branches/2.4.x/STATUS (original)
> +++ httpd/httpd/branches/2.4.x/STATUS Mon Jun 10 21:41:07 2013
> @@ -198,6 +198,8 @@ PATCHES PROPOSED TO BACKPORT FROM TRUNK:
>        trunk patch: http://svn.apache.org/r1491155
>        2.4.x patch: trunk patch works
>        nd: why would you do that in a stable branch?
> +      sf: Because it is only annoying and serves no purpose anymore. If
> you +          want, we can make it a minor MMN bump for adding a "new"
> API. +1: sf, covener
>        -1: nd

Long discussions in STATUS are kinda tedious ;-)

Well, I think, such changes are what trunk is for. Why not simply leave 
everything below as-is? Even more if it removes only an annoyance? Or is 
there a real technical reason I'm just not seeing right now?

nd
-- 
"Solides und umfangreiches Buch"
                                          -- aus einer Rezension

<http://pub.perlig.de/books.html#apache2>