[mapserver-dev] Vote for MapServer 5.4.2 release (and also 5.2.3 and 4.10.5?)

[Stephen Woodbridge]

This is a common problem with all linux distros.
They release some version of your software and it lives with that level 
will that version of the distro reaches end-of-life. Often you can find 
new version that have been backported to your version of the distro.

Typically the package maintainers get patches from upstream developers 
and port them back to the distro versions.

What I would recommend it helping the package maintainer by providing a 
patch, but I would be reluctant to do a full release. We might want to 
make a patch directory in svn where we can check in these patches and 
make them available to people that want to put the effort in to patch 
their version of the source and do the testing.

I suppose one could argue, that with we are going to all that work, why 
not just release it. I think at this point, there is a LOT of benefit to 
moving to 5.4 so people will when they can, but if you have a huge 
production environment with lots of tangled dependencies, it is not 
always possible to just up your source tree, especially if they have 
things like mapscript that might break or at a minimum would require 
extensive retesting.

Anyway, thats my 2 cents.
   -Steve W


Steve Lime wrote:
> Both of the issues raised by the Ubuntu folks are known (I can provide the
> report if anyone is interested). The report was against 5.0.3. See tickets:
> 
>   - http://trac.osgeo.org/mapserver/ticket/2941 
>   - http://trac.osgeo.org/mapserver/ticket/1836 
> 
> Both have been fixed in the 5.4 release, the first (#2941) was fixed 4.10.3
> and 5.2.2 as well. Question I guess is how far back to patch #1836. I'd like 
> to see users move forward if at all possible. I don't know how other projects
> handle patching security issues with past versions.
> 
> Another question is the MapServer Ubuntu package and how to update that. 
> Not something I have any experience with... Ideally only 5.4.2 when released
> would be available.
> 
> Steve
> 
>>>> Daniel Morissette <dmorissette at mapgears.com> 07/21/09 4:44 PM >>>
> FYI, I kept track of the votes and assuming the no votes for #2 are +/-0 
> (since they were not explicit -1), both votes would have passed:
> 
> #1 release of 5.4.2:
> 
> Passed with +1 from DanielM, PericlesN, SteveW, JeffM, ThomasB, AssefaY, 
> TamasS and SteveL
> 
> #2 release of 5.2.3 and 4.10.5:
> 
> Passed with +1 from DanielM, PericlesN, SteveW, ThomasB, TamasS and SteveL
> 
> However I thought we should wait to see what happens with that security 
> issue reported by Ubuntu before going ahead with a release, just in case.
> 
> Daniel
> 
> 
> 
> Steve Lime wrote:
>> So where'd we end up? I saw some folks didn't vote on #2, does that a +0, -1 or something else?
>>
>>>>> On 7/17/2009 at 2:45 PM, in message <4A60D4BD.4070302 at dmsolutions.ca>,
>> Yewondwossen Assefa <yassefa at dmsolutions.ca> wrote:
>>> +1 on vote #1.
>>>
>>> Jeff McKenna wrote:
>>>>> *** Vote #1 ***
>>>>> I propose that we release MapServer 5.4.2 now with the fixes currently 
>>>>> in SVN branch-5-4.
>>>>>
>>>>>
>>>> +1
>>>>
>>>> -jeff
>>>>
>>>>
>> _______________________________________________
>> mapserver-dev mailing list
>> mapserver-dev at lists.osgeo.org
>> http://lists.osgeo.org/mailman/listinfo/mapserver-dev
> 
>