Mmm, I’m thinking that not being an OS, exploits in GNome 2 are likely to be less critical than exploits in the actual OS (?) … and the differences between Gnome2 and 3 are so great that I would wonder about how many exploits are likely to be common to both?
[whereas with Windows they tend to build on what they have release by release, and any exploits are likely to common and more serious]
IMHO it’s a fundamental difference between the Windows and Unix ethos’s (plural for ethos?) …
- Unix : lets implement something that reliable and secure, then build user services on top
- Windows : let’s build a user experience for x86 hardware (note the lack of the word ‘security’ here … or ‘robust’ for that matter )
The net result seems to be that Windows provides for a better (subjective?) UI experience on x86 hardware, but tends to be less reliable, less secure, and more difficult to port to other processors / architectures.
In 1992 I had a conversation with my account manager about SCO’s UK pricing and sales strategy and told him in no uncertain terms that unless SCO did something fairly radical fairly quickly, Linux was going to wipe it off the map (bear in mind Linux was still fairly young at this point). Sure enough today, SCO’s market share seems to have become somewhat depleted … despite the best efforts of M$ Baystar capital(!) SCO–Linux disputes - Wikipedia
By the mid 90’s it had become apparent that Windows would eventually go the same way for reasons mentioned above … unless they re-started from the ground up (or unless the ground opened up and swallowed Linus) Linux was always going to be a fundamentally better design than Windows, hence Windows would eventually fade in favour of Unix-based OS’s.
The bit I find really interesting is the timing, M$ seem to have seen the writing on the wall and are in the process of onlining everything, and indeed favouring Office 365 over Office on a PC. Yet they are “still” flogging the dead horse that is ‘Windows’. (and I read recently that after losing their shirts on Windows RT to the tune of over $1Bn, they’re now working on Windows RT version 2) I have to wonder how much if this is ‘developing windows’ and how much is ‘keeping users engaged while they get all their other stuff up to speed’.
When they finally reach a point where their online offerings are good enough to supersede their local offerings, I’m wondering what the maintenance window will be with regards to all existing Windows installations which may effectively all reach ‘end of life’ at the same point in time?? In this particular instance with Windows XP, they face the cost of continuing to pay to maintain a very old OS, versus the potential to make $12Bn on Windows XP → Windows (n) upgrades … When the day comes, the choice will be between paying to maintain (n) versions of Windows on different architectures, or $(n)Bn to have everyone upgrade to a ‘Dumb’ OS running a subscription to M$ online services … I wonder which way Redmond will jump? (at this point I’m looking at all the people who have just paid £200 for a copy of Windows 7 or 8 … )
But here’s the really interesting bit, M$ will always need some sort of OS to launch from to access their online services, even if it’s an embedded OS of some type. It seems fairly obvious (to me, not least from RT sales) that this OS won’t be Windows based, so, what will it be ??
There are actually (IMHO) better versions of Unix out there than Linux, and in particular I’m thinking of QNX (which is what BB use to power their playbook) but which has actually been around for quite some time, longer indeed than Linux has. As QNX is commercial and “purchasable” in one way or another, it almost seems beyond belief that M$ doesn’t already have an OS like this up their sleeve ready to take up the slack … as things stand, that dumb OS launchpad that’s going to be on every device looks like being Android! … or should I say … Linux !!
Given M$'s history with Linux (and I refer to the Wikipedia link above) it seems, well, ‘risky’ for M$ to rely on Linux in such a context for their ‘entire’ business model … perish the thought that one day a bug in that Dumb OS might prevent people from accessing M$.com … (!)