Hendrik Boom <hendrik@???> wrote:

> On Sun, Jul 23, 2017 at 09:17:29PM +0100, Simon Hobson wrote:
>>
>> So the analogy is, I wouldn't expect "support" for all this "new"
>> stuff on an old vehicle. Similarly, as other have suggested, if I
>> was running very old hardware, I'd probably not be too worried about
>> being able to run all the latest and greatest software on it.
>
> And why think that the latest software is the greatest.

I don't - it's a turn of phrase "latest and greatest". I forget about the number of people here with different cultural backgrounds.

> Bugfixes make
> old software better, and so in that sense later is greater, but for
> the most part I see a lot of new software be buggy and bloated.
> I wouldn't want a lot of this new-fangled stuff on any machine, new or
> old.

I agree. I frequently look at the newer versions and think something along the lines of "what a pile of ****", especially with things like MS changing the UI whenever there's an R in the month !

>> So there's an argument for dropping support for an old and little used
>> architecture for NEW VERSIONS - leave the older versions in the
>> repos so that people can still install a system, but make it clear
>> that this won't be the latest and greatest version. There then comes
>> the issue of ongoing bugfixes - and my assumption would be that only
>> serious and/or security related bugs would get fixed in it.
>
> In a world where most of the new hardware contains unauditable
> firmware, that's tantamount to giving up any hope for security.

It's all a balancing act. How much effort to put into updates for "old" stuff - do you keep supporting version1, 2, 3, ... when the package is up to whatever number it is ? And what level of support - security fixes, other bug fixes, feature upgrades ?