Installing [the latest] FreeBSD .. was HO RAID ..

[doug]

On Fri, 4 Jan 2019, Matthew Seaman wrote:

> On 03/01/2019 19:48, Carmel NY wrote:
>> Matthew, I think you might have some say in how FreeBSD operates. When
>> a user downloads a new version of FreeBSD, lets say version 12.0, they
>> should get that version. Now, if there is a newer version, lets say
>> version 12.1 of 12.2, or whatever, the user should be presented with
>> the option of downloading the latest version. Microsoft does something
>> similar, and I see no reason not to follow that trend. The user should
>> be given the option. If there is a newer version available, why should
>> the end user be handicapped?
>
> I have less say than you seem to think.  Managing the release strategy and 
> support lifetimes of releases is a current topic engaging a deal of core's 
> attention, with a lot of input from the release engineering and security 
> teams (none of which I'm involved with) -- and this is a contentious topic as 
> there are several large blocks of the user base each of which want apparently 
> contradictory things.  Satisfying as many users as possible, without putting 
> the project under an insurmountable support load is going to take some time 
> and effort and a lot of discussion to sort out.

It seemes to me that this thread is covering two disjoint topics, one of 
which I am totally not qualified to speak on. The other just requires an 
opinion so ...

In keeping with the Unix dictum, "I will not stop you from doing something 
stupid because that might also stop you from doing something exceedingly 
clever", I think ...

this is not an issue that should consume large parts of anyone's time until 
experts stop telling me when going from 11.x to 12.x you might want to use a 
clean image. My severs are 40 miles and 3 hours (at the wrong time of day) away. 
What's wrong with going to freeBSD.org and seeing what's what. A bigger problem 
for me is I have some systems that got to be very out of date and I can not get 
a clean update path without being physically present.

Running the freebsd-update process two times for me is a nonevent matched 
against the above problems. I might like a slightly longer support TTL, but the 
problems in that are obvious and, probably insurmountable.