>>> I'm really struggling to see their rationale for this
No doubt they're hoping that all those people who use TechNet licence keys on production servers will now cough up for a "proper" licence.
196 publicly visible posts • joined 27 Mar 2012
Also not trolling, but is the lower frequency of HP/EMC patches because they have less, or because they release them less often? Less frequent patch release can mean a more secure product, or it can mean a product that's insecure for longer (n.b. not trying to have a go at Windows or HP/EMC here, just saying).
Also, bear in mind that if you're using this, you're probably a big Windows server shop anyway, so most of the infrastructure is being rebooted on patch Tuesday anyway.
>>> You certainly wouldn't get better performance from internal cards ...
Thunderbolt 2: 20Gbit/s (bit, not byte).
PCI Express 3: 985 MB/s (byte, not bit) per lane
A modern 16 lane PCI Express 3 card has approximately 6 times the throughput (123 GBit/s). Although not yet formalised, PCIe 4 is twice as fast as PCIe 3.
Mainly due to delivery time and risk minimisation. Even if developing something like SAP internally cost less in the long term (which is a debatable, but entirely separate, point), buying software "ready made" drastically reduces the time before you get to use the software, and have it delivering benefits for the business, and there is no risk that the software will actually never be delivered, which is always possible with in-house stuff and internal management changes.
"When you're at the supermarket you might be quite annoyed if the check out girl started advising you on the nutritional content of your basket and you'd definitely be unhappy if your bill was twice as big as result."
Yes, I would, but that's because your analogy is flawed. The checkout girl is there soley to collect money, not offer advice. I'd be just as annoyed if the person in the recruitment firm's accounts department started commenting on the person I'd recruited and offered staff development advice when they sent the invoice.
I'm not in any way trying to be negative about the intelligence levels of recruitment consultants - my point was that if you have someone who has enough domain knowledge to make a really good recruitment consultant, they have enough domain knowledge to get a better paid (and probably more fulfulling) job elsewhere, within the domain itself, and that's why you don't get "expert" recruitment consultants.
"So to make the same money as you, he needs to place one person like you about every week."
Your maths assumes a 12% placement fee (in my experience, it's more like 25% - 30%) and also assumes zero base salary. Are you really saying that if a recuiter makes no placements in a calendar month, they get zero pay?
I suspect it's becuase if you have enough technical nouse to be a good recruiter, you have enough technical nouse to get a better job.
I've never met a recruiter who is genuinely interested in either the client's business or the candidate's career. Like every other sales person (and that's what they are), they're only interested in their end-of-month commission cheque.
"owned hardware vs non-owned hardware usage regardless of the business model used to lease it"
If it's owned, it's not leased.
"If I call my business's departments 'customers', does that mean I've now got a small public cloud? "
No, because the data is still stored on corporate-owned and -controlled hardware, not hardware owned and managed by a third party. What you choose to call your business departments is irrelevant.
Not the only difference, there's also that opex vx capex thing. With a private cloud, you have the additional burden of the capital expenditure. With Amazon's cloud, you don't have that big, nasty, up-front cost.
Of course, you do have a smaller cost, that goes on, and on and on, month after month after month, and before you know it, you've paid more to Amazon than the original capex "saving".
Every time I look at off-prem cloud pricing, it seems wildy unecomonic in the medium to long term for even relatively low utilisation levels.
Four of the files included in the download are not even digitally signed.
An anti-malware firm wants me to download and run unsigned executables? That's what I call setting a good example!
(Yes, I realise that just 'cos it's signed, doesn't mean it isn't malicious, but it's a good start).
If you're going to be pedantic, at least be correct...
You *can* use a kettle lead to power a computer. What you can't do is use a standard PC power cord to power a kettle, since the PC power cord won't have the cut out required for it to fit in the kettle, but the kettle lead will fit in the computer. Saying that a C15 connector cannot be used for domestic computer applications simply isn't true.
C13/14 - no cutout, rated to 70C, usable in computers.
C15/16 - with cutout, rated to 120C, usable in computers and kettles.
You may be thinking of C15A/C16A, which are a slightly different shape.
Depending on the nature of the backups, a simple restore may not be an option, since that would then wipe out all the stuff that's been changed since the backup. It's more of a "restore somewhere else, and merge the old with the new". Always much more tricky. Much like when there was the big RBS / Natwest cock-up, some (mainly non-IT) people said "just restore from the backups". It's not that simple with systems that (in theory) never go off-line and are always being updated. RBS *had* to put everything back as it was - eBay *should*, but I bet, given their Ts&Cs, they aren't in any way obliged to.
So, if I install Office 2013 on a computer, and then, on progressive weeks, upgrade the motherboard, processor, memory, hard disk and video card, is it still the same computer five weeks later? Have I invalidated my licence? Do I need to reassemble the previous components into a working system to be legally allowed to use the software?
The trouble is, devices like this aren't like cars, where the cost of a single component that fails is orders of magnitude less than the cost of the whole. If something fails on a Surface, it's probably going to be expensive to replace, costing a large percentage of the overall price. Suppose one of the surface mount ICs fails. The labour involved in manually unsoldering a single SMT chip and putting a new one in, plus the cost of the chip itself (assuming they're still being made any more) will probably be more than the device is worth. Since it isn't economic to repair the device anyway, it makes sense to manufacture the device in a way that costs less / makes the device lighter / look prettier etc.
Having said that, for the bits you expect to fail / wear out (batteries, and possibly SSDs), it makes sense to allow those to be replaced easily.
.... before this is cracked?
"A browser plugin is an essential piece of the FIDO plan"
Of course, the coding of the plug in will be to such a high standard that it will be impossible for any rogue web site to read anything from the token (or, heaven forbid, write to it), wont' it?
See also: Java, ActiveX, Flash.
"Clearly the #1 motive for any business is making money. So when a company is keen to move you to a different way of paying, there can only be one who really benefits."
Sure, the company benefits. But that doesn't necessarily mean individual customers suffer. If a price drop of 20% means 50% more sales, cost to the individual consumer goes down, while total revenue for the company goes up. (Yes, I do realise that increased revenue isn't automatically increased profit, if you have to drop unit price to get it).
"RAM is cheap". True, but the margins on something like this are thin - even $50 on another 4GB of RAM could be more than half the profit.
"a recovery partition using a similar amount [20GB]". Why? An ISO for Windows 8 is 3.5GB, and you can, of course, do a full OS recovery from that. Any additional applications won't need a further 16.5GB.
Earth could have had technology-using species several times over by now. We can barely tell what was going on in classical Greek / Roman times, and anything further back that 10,000 years is purely speculation. Any remains of a now-dead advanced civilisation from say, half a billion years ago (a mere 11% of the planet's lifetime) will be long gone, probably swallowed up by the actions of plate tectonics, and buried in molten magma.
"denying logins from a particular ip address or address range is meaningless."
It's there to preveny denial of service attacks against a single account.
"It's better to deny logins globally to that account for x seconds/minutes ".
If you do that, all I have to do is to log in as you with an incorrect password every few seconds, and you can never access your account. If you limit the lock out to the IP addresses that originated the failed login, the legitimate user can still get access.
Or if you get a failed login for a particular e-mail address, simple deny all logins from that source IP for say, five seconds. Hardly a great inconvience to a genuine user making a typo on the password, but makes a remote "dictionary attack" (where the dictionary including all combinations of upper, lower case and digits) of even an eight-character password unfeasible.
Granted, if someone gets hold of the underlying password database, and so can circumvent the connection and time restrictions imposed when connecting remotely, then the shorter passwords are now much weaker than longer ones.
"So the Anti-Theft Tech is useful after you have reported the theft of your laptop"
It's useful *before* the theft too, provided that the thief knows it won't be any use to them once it's been stolen. Much the same as the security code you often have to enter into car radios if their power is interrupted - it doesn't stop the theft as such, it reduces the likelihood of theft in the first place.
Of course, this requires the your average burglar / mugger is up to date on current technology, and can tell the difference between a two-year old laptop and a shiny new Ultrabook, which might be asking a bit much..
I bet the really tricky bit is that Mars is a long way away (I mean, you think it's a long way to the Chemist, etc), and any image that the driver sees on Earth is already several minutes old, and any command sent will take several more minutes to get there. It's not a case of "oops, there's a sand dune coming up, I'll turn left"; it'll be more like "oops, there's a sand dune coming up, damm, I hit it three minutes ago".