Hacker News Reader

Comment by XERQ | Hacker News Reader

I've used FreeBSD since the 6.0 days, and agree that it's absolutely rock solid. The main problem is that the support cycles are way too short, for instance 10.0 is EOL in 4 months (1 year after it was released), whereas RHEL is EOL in 10 years.

View on HN

As others have commented, the effective EOL is not really that short. FBSD 10.0 was released Jan 2014, so 10.0 will be supported for a year or so. Updating to 10.1 (when it is finalized in Q4) will very likely be painless, no incompatibilities, so as a practical matter, EOL is extended at least another year or so.

This has been the history of FBSD since my exposure to 4.x back in 2000. A more recent example, FBSD 8.0 was announced Nov 2009. And now, nearly 5 years later, FBSD 8.4 is still a supported version.

Sure, technically each minor version of 8.x reached EOL after a year. But since keeping an 8.x installation up-to-date has been so seamless, it achieves an EOL >=5 years from introduction.

Anyway, how I see it based on real-world experience.

This is being worked on, in that 10.x is going to be supported for at least 2 years, and in the future there has been talks about 5 years of support per major release. The minor releases will get dropped sooner, upgrading is mostly painless anyway (from 9.1 to 9.2 was very simple) so it shouldn't make much of a difference.

I think some of the reason for that is that enterprise usage of FreeBSD is frequently assumed to be via an intermediary, not directly as an actual FreeBSD install. RHEL directly supports enterprise installs, but FreeBSD is doing something more like producing a reference implementation. The enterprise installs that need N years of support lifecycle are supposed to run on a derived system (one of NetApp's, say), with its own EOL policies.

There is now going to be commercial support for 15 years. There was a discussion with the company doing it on bsdnow.tv recently.

Reply on news.ycombinator.com