Microsoft sounds support warning bell for customers running Windows 7, Windows Server 2008, Windows Server 2003, and more
-
You give to much credit, especially to the SMB market.
I can honestly say that I never considered mentioning the EOL of an OS to a client when I used to consult (unless it was looming). Back then I was ignorantly expecting most companies would be replacing their computers long before the EOL deadline, moving onto the next platform.
It's quite clear that places like print shows and manufacturing floors, etc, the people making the decisions are rarely told that the computer that runs the 1 million dollar hardware that is expected to last for 15+ years will need to be replaced and the software that connects that PC to the hardware updated - but oh.. the manufacture won't be bothering to make an update of this hardware (assuming they are still around) etc, etc, etc.
So yes, in an ideal world, everyone is aware and makes appropriate plans around the EOL for the software/hardware, but in reality this is rarely the case.
-
@Dashrender Yes but you as the IT person knows to install latest versions, to be aware of support dates, etc. The business only needs to follow IT's advice, they don't need to know all if the details behind decision making. Only need to bring out details when they balk at decision processes.
-
Using the example of a piece of equipment that will be usable for 15+ years - who's responsibility is it to be aware that the OS that runs the computer that runs that piece of gear will be EOL'ed in, let's say 5 years?
I actually find myself in this very spot. We purchased a CT machine 7 years ago. The control workstations are based on XP, the vendor has no plans to update them. I wasn't involved in the original decision to go with this vendor, or really much of that project all. Now here are 7 years later, about half way through the useful life of this equipment and we find ourselves using equipment that could be prone to compromise. I have to attempt to provide a way to secure these machines while at the same time not impacting their ability to function.
This type of thing happens constantly - just look at the ATM machines.
-
In some ways, you are way past the halflife of the equipment. If the equipment is tied to XP, it's very arguable that while the machine might physically work, whoever purchased it decided at the time that the truly useful lifespan would end now. The lifespan of gear is more than just how long the mechanical portions will physically last. You have to look at it holistically. If it is tied to DOS or XP or anything, that's part of that machine's lifespan. Choosing vendors based on the support that they give, the components that they chose.
It's not different than if the machine run on Windows 8.1 but the mechanical portion wore out. You wouldn't point to Windows 8.1 and say "it's nowhere near the end of life for this machine because the OS is still supported." It's a holistic thing. All of the parts have to keep working.
The problem is this is a bigger issue than an IT issue. It means equipment is being purchased either without good planning, vendors are being selected poorly, people aren't taking responsibility for their decisions, management hasn't figured out that they are leaving the right people out of the decisions, etc. Something bigger is wrong and should be addressed.
-
@scottalanmiller you realize that there is almost zero choice in the OS of the device operating modern equipment don't you? It is not like there are even 10's of vendors out there making much of this equipment.
-
@JaredBusch said:
@scottalanmiller you realize that there is almost zero choice in the OS of the device operating modern equipment don't you? It is not like there are even 10's of vendors out there making much of this equipment.
That means that there is a huge market opportunity. It only takes a single vendor doing a good job and companies voting with their wallets for quality to force the market to correct itself.
-
@scottalanmiller said:
@JaredBusch said:
@scottalanmiller you realize that there is almost zero choice in the OS of the device operating modern equipment don't you? It is not like there are even 10's of vendors out there making much of this equipment.
That means that there is a huge market opportunity. It only takes a single vendor doing a good job and companies voting with their wallets for quality to force the market to correct itself.
This is definitely true, but it's been seen time and time again that the cheapest price often wins out over quality. So if you can't do a better job AND have a better price, you're definitely in a much tougher spot. SMBs going back to another thread, think with the hearts not their logic - If I can spend $20K on this machine today instead of $25K (hell even $22K) I'll worry about that OS problem later ( which means never - or really, when it fails me) and then I'll be upset that I have to spend $20K again, etc.
-
@Dashrender said:
This is definitely true, but it's been seen time and time again that the cheapest price often wins out over quality. So if you can't do a better job AND have a better price, you're definitely in a much tougher spot. SMBs going back to another thread, think with the hearts not their logic - If I can spend $20K on this machine today instead of $25K (hell even $22K) I'll worry about that OS problem later ( which means never - or really, when it fails me) and then I'll be upset that I have to spend $20K again, etc.
And it is not $20k for these machines, we are talking about million dollar plus costs for these machines.
-
@JaredBusch said:
@Dashrender said:
This is definitely true, but it's been seen time and time again that the cheapest price often wins out over quality. So if you can't do a better job AND have a better price, you're definitely in a much tougher spot. SMBs going back to another thread, think with the hearts not their logic - If I can spend $20K on this machine today instead of $25K (hell even $22K) I'll worry about that OS problem later ( which means never - or really, when it fails me) and then I'll be upset that I have to spend $20K again, etc.
And it is not $20k for these machines, we are talking about million dollar plus costs for these machines.
So going for the lowest cost without doing due diligence is extra bad.
-
@scottalanmiller you are arguing for people to buy a product that does not exist. I have worked for one company and been a consultant for a couple more that do manufacturing. The equipment manufacturers do not update the control systems unless they update the machine also.
In some perfect world, they may do better, but this is reality. Is there a niche there? Yes, but it is not one that could ever be filled by a single company because the control software for every one of these machines is coded specifically for the machine.
-
I worked in manufacturing and we actually wrote our own software to control the machines to get around this. But the better machines were already running off of Solaris and other more serious platforms. Or embedded. Running off of Windows as a starting point is pretty silly.
Do you have an example of equipment that comes from vendors where there is no market option for something kept to a minimally acceptable level.
-
We use these. The software on our current machines is XP only and heavily customised. We could upgrade, but we've been quoted 200 hours of programming at 60 euros/hour. For a small manufacturer, spending 12 grand on something that has no impact on the bottom line and merely mitigates the risk of running XP is a tough call. Market conditions are extremely tough from European manufacturers at the moment. Our competitors are in the same boat since they use the same machines, although many of them are much bigger organisations with much bigger budgets. If you think programming your own software under Solaris is the answer I'm not sure you have much experience of the realities of small manufacturing.
-
@Carnival-Boy said:
If you think programming your own software under Solaris is the answer I'm not sure you have much experience of the realities of small manufacturing.
Maybe small manufacturing just isn't looking at the best tools for itself. That's a general factor in all SMBs. They often overlook what is cheap and good for what is well marketed.
Has anyone thought of assembling a market consortium to deal with this? 200 hours isn't that much programming. It's a lot for a single piece of equipment at a single shop, of course. But just five weeks of work. How much is that machine worth to protect? Would it lengthen the life of the machine? Could it protect against down time? Could you reduce support costs, licensing costs, etc? Could the consortium sell the result to recoup the investment? Get two shops together and it is just 6.000 Euro. Get one hundred shops together and it is just 120 Euro each. If there are any number of these out there and everyone is facing the same issues, there is likely a market for fixing things.
-
Do you know other companies that have these?
-
I've helped support drill shops in Spiceworks who have Solaris control systems. Single drill shops, very small. Solaris is, or at least was, the standard for drill systems.
-
Great out of the box thinking there Scott.
Here's my situation:
We have a Toshiba CT scanner that's running on XP. Toshiba does not have an option for replacing the control workstations for this equipment with another OS. Simply does not exist. Unless we wanted to hire a company to completely rewrite a whole new interface on a new OS we have no options. Additionally once we do that Toshiba will no longer provide support/maintenance on the machine, which means it will no longer qualify as a certified piece of medical equipment, and the insurance companies won't pay for its use.I'm sure there were other CT machines that would could have looked at, but running them on Windows seems to be standard practice. I'm guessing that anyone who runs their machine on something else would be considerably more expensive and probably outside our price range (programming for windows is for some reason considered much cheaper than for say Solaris - at least that's my guess why someone like Toshiba would use it as a base instead of something else).
-
@scottalanmiller said:
200 hours isn't that much programming. It's a lot for a single piece of equipment at a single shop, of course. But just five weeks of work.
Seven weeks. This is Europe not America
I'm sure there's ways and means. I just took umbrage at you suggesting that buying hardware controlled by Windows OS was silly, really bad or lacking due diligence.
-
@Carnival-Boy said:
@scottalanmiller said:
200 hours isn't that much programming. It's a lot for a single piece of equipment at a single shop, of course. But just five weeks of work.
Seven weeks. This is Europe not America
I'm sure there's ways and means. I just took umbrage at you suggesting that buying hardware controlled by Windows OS was silly, really bad or lacking due diligence.
Ha ha. Actually in the US that is only four weeks. A professional day is 10+ in the US (that's a real term.) So normal professional weeks are 50 hours minimum and 60-72 hours not that uncommonly. Probably 55 hours is most common. So in the US it could be as little as three weeks. When I was on Wall St. normal weeks were 65-75 hours not including being on call 24x7.
-
@Carnival-Boy said:
I'm sure there's ways and means. I just took umbrage at you suggesting that buying hardware controlled by Windows OS was silly, really bad or lacking due diligence.
Is there no way to require vendors to support the products in an updated way? No market pressure, contractual obligations, etc? Not that they have been taken, but is there no means to take them? All they would have to do is follow extremely basic standards and even being on Windows they would not have these compatibility problems. As a vendor that makes this kind of software it's extremely easy not to get into these situations (we make not entirely different software for hospitals.)
-
@scottalanmiller said:
A professional day is 10+ in the US (that's a real term.)
Wow. Really? As far as I can tell no-one at NTG does any work at all, you all just hang out on ML all day long? I don't think over 50 hours of work a week is productive over the long term - it just leads to burn-out and low productivity. I have no evidence to back this claim up though.
