FortiOS 5.6.0 beta

[AtiT]

Hi,

I received an email from Fortinet where are a lot of information including:

Enroll on the beta program for FortiOS 5.6.0 now

I am just wondering whether it is a good idea from Fortinet to develop a new OS. The 5.4 is quite new and has some bugs. It would be better to focus on bugfixes on 5.4 than creating something new in a short time after 5.4 - I think.

[NotMine]

Second this.

[tanr]

I believe they're now planning a 5.4.2 bug-fix only release.

However I really want to see a more stable 5.4.x line before they start adding new features (whether to 5.4.x or to a new line).

[NeilG]

AtiT wrote:

Hi,

I received an email from Fortinet where are a lot of information including:

Enroll on the beta program for FortiOS 5.6.0 now

I am just wondering whether it is a good idea from Fortinet to develop a new OS. The 5.4 is quite new and has some bugs. It would be better to focus on bugfixes on 5.4 than creating something new in a short time after 5.4 - I think.

Since I don't have any visibility into what goes into their code versioning vs infrastructure I couldn't say which would be more stable. However you may not be aware that sometimes problems and bugs are due to underlying design/architecture choices - which may have been wrong. So a fix in say the 5.4.x line might involve far more changes than you might think, and if the fix has to be shoehorned in, it might be even less stable or break something else in an unexpected location.

I personally have yet to get the "learning mode" to do ever generate a report, which I guess is a bug - but I really don't think I care about the feature anyway.

However the inconsistency of web filter overrides not applying in a predictable way is a bug that I care about. The fact that often the "fix" is to recycle the process seems to imply that the problem is localized and probably can be fixed in a point release (5.4.2 maybe).

On the other hand, getting logging to output in a consistent fashion using Common Event Format standard would impact every system that sends non-debug output, so if the first pass of that occurs in 5.6 then I am all in favor of engineering working on 5.6 beta. 

What are the bugs in 5.4.1 you think they need to focus on fixing before working on 5.6?

-N

[iJake]

I believe 5.6 is mainly to expand the range of products it can manage, as 5.4 is fairly limited (when compared to the vast range of products Fortinet offers).

[dpmcintyre]

Fortinet has already stated that they've changed their versioning scheme such that a release such as 5.4 will have zero new features ever added to it, and there will be just bug fixes released on 5.4 now as it is fixed at where it is. 

Any new feature will have to go into 5.6, so it makes sense that they are starting development on it now, so as to add any new feature what-so-ever, they have to roll it out there, unlike in the past, say with 4.x where each release got a grab bag of patch fixes and new features, it is more structured now. 

[MikePruett]

Yeah, I'm hoping they start pumping out 5.4.x updates like they are going out of style to fix the bugs everyone is experiencing.