httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Daniel Ruggeri <drugg...@primary.net>
Subject Re: [VOTE] Release Apache httpd 2.4.28 as GA
Date Fri, 29 Sep 2017 13:07:16 GMT
It's fair to reiterate first that most folks here don't care and "get it"... but let me offer
another perspective that folks may have to deal with. Devil's advocate here, and all that.

1.2.4 is released. It contains security fixes
1.2.5 is burned during RM
1.2.6 is burned during RM
1.2.7 is released. It contains bug fixes
1.2.8 is burned during RM
1.2.9 is released. It contains security fixes

Let's say I choose not to build and deploy 1.2.7 since I am not using any of the modules the
bug fixes address.

To the uneducated eye (AKA: the auditor, the security engineer not familiar with httpd's versioning
idiosyncrasies, the manager, etc), it may appear as a serious lapse in patching if five releases
were between deployed versions. "9 is MUCH higher than 4! What have you been DOING all day?"
they might ask.

Another angle is the unspoken angst of upgrading five point releases versus one point release.
It *shouldn't* be scary... but it can be for someone who doesn't fully trust us to avoid releasing
regressions.

Granted, both represent shaky logic that can be fixed by RTFM... but are things I've seen
in the field (and have answered with a 301 to the CHANGES file).
-- 
Daniel Ruggeri


-------- Original Message --------
From: Eric Covener <covener@gmail.com>
Sent: September 29, 2017 7:16:51 AM CDT
To: Apache HTTP Server Development List <dev@httpd.apache.org>
Subject: Re: [VOTE] Release Apache httpd 2.4.28 as GA

On Fri, Sep 29, 2017 at 6:57 AM, Reindl Harald <h.reindl@thelounge.net> wrote:
>
>
> Am 29.09.2017 um 12:35 schrieb Graham Leggett:
>>
>> On 29 Sep 2017, at 12:25 PM, Reindl Harald <h.reindl@thelounge.net> wrote:
>>
>>> it's not about cheap - it's just questionable that after 2.4.12 the next
>>> release is 2.4.16 because it looks not really sane
>>
>>
>> Looks perfectly sensible to me
>
>
> in your world where you know the background, without the context it looks
> like "are they not capable to count?"

What kind of user do you think this confusion affects?  They'd have to
care that N-1 wasn't released, but not be reading the changelog.

-- 
Eric Covener
covener@gmail.com

Mime
View raw message