Re: Proposal: adoption of mod_policy subproject

[Stefan Fritsch <sf...@sfritsch.de>]

On Wednesday 29 February 2012, Jim Jagielski wrote:
> On Feb 29, 2012, at 10:32 AM, William A. Rowe Jr. wrote:
> > It's the reason why mod_aspdotnet was ejected, and why mod_arm4,
> > mod_ftp and perhaps even mod_fcgid are all on their way out of
> > the project as it is, unless more committers participate in
> > those subprojects.
> 
> The other way to approach it is to simply move them *into* the main
> project, and not as "subprojects"...
> 
> We've seen that subprojects do NOT work. There is more activity
> when they are in the main tree. With the exception of mod_arm4
> (which I've never looked at), I would be MORE than comfy having
> both mod_ftp and mod_fcgid moved into the main tree today.

But it would have to be clear that inclusion into trunk would not by 
default mean that they would be included in the next major release.

Especially for mod_ftp, I think that it requires many more eyeballs 
and a comprehensive test suite before it should be part of a stable 
release. Otherewise, it would just be a source of lots of security 
issues that need to be handled and would tie up resources.

Re: Proposal: adoption of mod_policy subproject

[Stefan Fritsch <sf...@sfritsch.de>]

On Sat, 3 Mar 2012, Rainer Jung wrote:

> On 03.03.2012 15:22, Eric Covener wrote:
>>>> But it would have to be clear that inclusion into trunk would not by
>>>> default mean that they would be included in the next major release.
>>>> 
>>>> Especially for mod_ftp, I think that it requires many more eyeballs
>>>> and a comprehensive test suite before it should be part of a stable
>>>> release. Otherewise, it would just be a source of lots of security
>>>> issues that need to be handled and would tie up resources.
>>>> 
>>> 
>>> Including into trunk, and with 2.4 RTC, then the above is all
>>> assumed...
>> 
>> I think Stefan's legitimate concern is implicit inclusion in 2.6/3.0,
>> not a potential backport into 2.4.
>
> Yes, I think that's what he meant.
>
> IMHO that wasn't a problem lately. When we got closer to 2.4 several pieces 
> were removed, because they were not well tested, not well-documented or 
> didn't work. Determining the stuff to remove from the release wasn't a 
> difficult or controversial discussion. So I'd say no further hard rules on 
> which additions to trunk will make it into a further release are needed.

Yes, I was concerned about 2.6/3.0. But if other people think it won't be 
a problem, then that's ok with me, too.

Re: Proposal: adoption of mod_policy subproject

[Rainer Jung <ra...@kippdata.de>]

On 03.03.2012 15:22, Eric Covener wrote:
>>> But it would have to be clear that inclusion into trunk would not by
>>> default mean that they would be included in the next major release.
>>>
>>> Especially for mod_ftp, I think that it requires many more eyeballs
>>> and a comprehensive test suite before it should be part of a stable
>>> release. Otherewise, it would just be a source of lots of security
>>> issues that need to be handled and would tie up resources.
>>>
>>
>> Including into trunk, and with 2.4 RTC, then the above is all
>> assumed...
>
> I think Stefan's legitimate concern is implicit inclusion in 2.6/3.0,
> not a potential backport into 2.4.

Yes, I think that's what he meant.

IMHO that wasn't a problem lately. When we got closer to 2.4 several 
pieces were removed, because they were not well tested, not 
well-documented or didn't work. Determining the stuff to remove from the 
release wasn't a difficult or controversial discussion. So I'd say no 
further hard rules on which additions to trunk will make it into a 
further release are needed.

Regards,

Rainer

Re: Proposal: adoption of mod_policy subproject

[Eric Covener <co...@gmail.com>]

>> But it would have to be clear that inclusion into trunk would not by
>> default mean that they would be included in the next major release.
>>
>> Especially for mod_ftp, I think that it requires many more eyeballs
>> and a comprehensive test suite before it should be part of a stable
>> release. Otherewise, it would just be a source of lots of security
>> issues that need to be handled and would tie up resources.
>>
>
> Including into trunk, and with 2.4 RTC, then the above is all
> assumed...

I think Stefan's legitimate concern is implicit inclusion in 2.6/3.0,
not a potential backport into 2.4.

-- 
Eric Covener
covener@gmail.com

Re: Proposal: adoption of mod_policy subproject

[Jim Jagielski <ji...@jaguNET.com>]

On Mar 2, 2012, at 2:50 PM, Stefan Fritsch wrote:

> On Wednesday 29 February 2012, Jim Jagielski wrote:
>> On Feb 29, 2012, at 10:32 AM, William A. Rowe Jr. wrote:
>>> It's the reason why mod_aspdotnet was ejected, and why mod_arm4,
>>> mod_ftp and perhaps even mod_fcgid are all on their way out of
>>> the project as it is, unless more committers participate in
>>> those subprojects.
>> 
>> The other way to approach it is to simply move them *into* the main
>> project, and not as "subprojects"...
>> 
>> We've seen that subprojects do NOT work. There is more activity
>> when they are in the main tree. With the exception of mod_arm4
>> (which I've never looked at), I would be MORE than comfy having
>> both mod_ftp and mod_fcgid moved into the main tree today.
> 
> But it would have to be clear that inclusion into trunk would not by 
> default mean that they would be included in the next major release.
> 
> Especially for mod_ftp, I think that it requires many more eyeballs 
> and a comprehensive test suite before it should be part of a stable 
> release. Otherewise, it would just be a source of lots of security 
> issues that need to be handled and would tie up resources.
> 

Including into trunk, and with 2.4 RTC, then the above is all
assumed...