Windows XP and Vista Long Term Support Plan

[Peter Dolanjski]

>
> >    - Monitor user numbers as ESR 52 reaches end of life (2018), continue
> >    security patches for XP/Vista users on ESR 52 branch if user numbers
> >    justify it (exact threshold TBD)
>
> This sounds like the potential to be a commitment to support the
> ESR52 branch for an indefinite amount of time.
>
> As that commitment lengthens, it has the potential to be more of a
> burden than continuing support for XP/Vista would have been (for
> example, if we have security bugs that require major architectural
> changes to fix).  The cost of backporting security fixes increases
> with the age of the branch (time), probably as a worse-than-linear
> function of time.
>
> Is there a way we can avoid this potential indefinite commitment?


What was discussed with release engineering was to take things on a case by
case basis.  In reality this shouldn't be an indefinite time commitment.
By early 2018 user numbers will likely fall below the threshold where it
becomes harder to justify continued investment. If a particular backporting
issue is significant in nature, that may force a line in the sand to be
drawn.  That's sort of a risk that we'll have to accept with this approach.

Peter

On Wed, Oct 12, 2016 at 7:17 PM, L. David Baron <dbaron at dbaron.org> wrote:

> On Wednesday 2016-10-12 15:16 -0400, Peter Dolanjski wrote:
> >    - Monitor user numbers as ESR 52 reaches end of life (2018), continue
> >    security patches for XP/Vista users on ESR 52 branch if user numbers
> >    justify it (exact threshold TBD)
>
> This sounds like the potential to be a commitment to support the
> ESR52 branch for an indefinite amount of time.
>
> As that commitment lengthens, it has the potential to be more of a
> burden than continuing support for XP/Vista would have been (for
> example, if we have security bugs that require major architectural
> changes to fix).  The cost of backporting security fixes increases
> with the age of the branch (time), probably as a worse-than-linear
> function of time.
>
> Is there a way we can avoid this potential indefinite commitment?
>
> -David
>
> --
> 𝄞   L. David Baron                         http://dbaron.org/   𝄂
> 𝄢   Mozilla                          https://www.mozilla.org/   𝄂
>              Before I built a wall I'd ask to know
>              What I was walling in or walling out,
>              And to whom I was like to give offense.
>                - Robert Frost, Mending Wall (1914)
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.mozilla.org/pipermail/firefox-dev/attachments/20161012/7b4a23b9/attachment.html>