challenge: end of life for 6.2 is premature with buggy 6.3

[Robert Watson]

On Thu, 5 Jun 2008, Chris Marlatt wrote:

> Adrian Chadd wrote:
>> The project is doing what it can with what people are contributing. If
>
> What if it can accomplish the same or more by simply reorganizing what it's 
> already doing? I completely understand the apparent situation - if you look 
> at it from all angles it appears to be no different than that of the people 
> apposed to the recent scheduling changes FreeBSD has made. There's a limited 
> amount of people and time to do everything.

It's an order-of-magnitude question.  The work required to support a release 
for 48 months is more than double the work required to support a release for 
24 months.  The current regular and extended support releases reflect a 
practical balance with respect to how long a release can be support.  We 
provide a much longer timeline of support for *branches*, however, and that's 
generally the support mechanism recommended for people looking for 4-6 years 
of support for a version of FreeBSD.

If you look at what other OS vendors do -- Microsoft is a particularly easy 
example to inspect -- they require occasional large-scale updates while you 
live on a particular branch.  For example, if you're going to keep running 
Win2k for six years, you must install their service packs, not just hot fixes 
for specific vulnerabilities.  Our minor releases on a branch are a *lot* less 
disruptive than service packs for Windows, and are much more conservative.

Robert N M Watson
Computer Laboratory
University of Cambridge