Re: proposal: inetd improvements.

[Marc Balmer <marc%msys.ch@localhost>]

Am 06.06.10 21:20, schrieb Manuel Bouyer:

> On Fri, Jun 04, 2010 at 03:14:34PM +0100, elric%imrryr.org@localhost wrote:
>> On 1275659377 seconds since the Beginning of the UNIX epoch
>> Manuel Bouyer wrote:
>>>
>>
>>> On Fri, Jun 04, 2010 at 02:46:41PM +0100, elric%imrryr.org@localhost wrote:
>>>> Right, and I'm proposing extending inetd to be able to subsume more
>>>> use cases which makes the requirements a bit more stringent.  I do
>>>> not think that it's a great idea to unnecessarily limit the
>>>> applicability of inetd to light-load services.
>>>
>>> No, what I'm saying is that it's wrong to discourage use of one inetd
>>> feature in documentation when it's really dependant on use case.
>>
>> Okay, so now we have two examples of cases where this feature appears
>> to be superior to max outstanding kids:
>>
>>      1.  you have a wait service and have a typo in the command, and
>>
>>      2.  you have light load services where availability is unimportant.
>>
>> I think that in order to encourage users to use this feature, we
>> should come up a better use case.  These two are insufficient
>> justification for such encouragement.  Especially given that max
>> number of outstanding kids perfectly well takes care of (2).
> 
> I don't want our docs to encourage or discourage use of this feature.
> Just document it. Period.

fwiw, I fully agree.