This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Application control - poor decisions

Well I've not been back here for a while but it's time for another post and this time I'd like feedback from you guys about a decision Sophos has made regarding app control.

We use app control extensively and I block most things not related to our business by default (games, peer to peer etc). One noteable change recently was for Sophos to now add a Firefox V7 and higher policy option. Now while I fully understand the reason for this because of the mindless version updates vomiting out of the Mozilla labs at the moment, I don't understand why they don't pick off the major version numbers and especially, the ability to block beta versions. I've about 20% of users at  my organization that use Firefox and of those, about a third watch the firefox website like hawks always wanting the latest, greatest version even if it's not tested and released. Previously, by allowing specific versions and blocking everything else, I had the ability to lock down to only released versions and I also had the ability to lockout old defunct versions that were either to vulnerable or really not fit for purpose (v7 immediately springs to mind!). Now, my users are freely downloading v9 beta, installing it, using it and I have absolutely no control over that with Sophos because they've adopted a v7+ identity only. How bad is that!

I'd like to get some feedback on whether you feel this is the right approach or not. As administrators, we know that the FF version change every 30 days is a big problem and I know there will be a few people out there that don't really care that users can get to higher versions even untested betas and alphas but I and many others do. How does the community feel about this approach?

 Should we at the very least still continue to get individual version control? Should we have the v7+ AND the individual version control?

Matt

:19333


This thread was automatically locked due to age.
Parents
  • Hello Matt and Dan (and all other potential listeners),

    if I understand correctly the arduous part is creating an entry in the first place. In terms of detection it is a stub where detection identities can be plugged in (and I assume these can be distributed with the usual threat detection data updates), the required management data for adding an entry can only get distributed with a "monthly" update though.
    @Matt: I'm repeating myself - the former entries did not enable you to enforce something like "at least 3.16" or "at most 4.3" and major version changes ran "undetected". You were also not able to block any FF5 builds including the first general releases because AppCtrl did not know FF5 and there is no "catch all Firefox". Thus the V7+ entry is not a step backwards, it just makes the deficits which "always" existed more obvious.
    @Dan: I assume it is possible to create - at a later time - a, say, V7-9 and a V10+ entry. If in fact it is intended to "close" V7+ then the + is not the ideal name. And any line drawn (or versions bundled) will probably lead to discussions ("it should be V7-11 not V7-12").

    Christian
    :19613
Reply
  • Hello Matt and Dan (and all other potential listeners),

    if I understand correctly the arduous part is creating an entry in the first place. In terms of detection it is a stub where detection identities can be plugged in (and I assume these can be distributed with the usual threat detection data updates), the required management data for adding an entry can only get distributed with a "monthly" update though.
    @Matt: I'm repeating myself - the former entries did not enable you to enforce something like "at least 3.16" or "at most 4.3" and major version changes ran "undetected". You were also not able to block any FF5 builds including the first general releases because AppCtrl did not know FF5 and there is no "catch all Firefox". Thus the V7+ entry is not a step backwards, it just makes the deficits which "always" existed more obvious.
    @Dan: I assume it is possible to create - at a later time - a, say, V7-9 and a V10+ entry. If in fact it is intended to "close" V7+ then the + is not the ideal name. And any line drawn (or versions bundled) will probably lead to discussions ("it should be V7-11 not V7-12").

    Christian
    :19613
Children
No Data