On 22/09/17 09:58 AM, Wols Lists wrote:
Is that really seriously economic???
I see a lot of recon gear at Morgan Computers (their IJT Direct subsidiary gets pretty much all my printer business), and recon appears to mean "we buy lots of redundant, surplus and ex-lease gear, QA it, and sell it on". If a computer fails QA they may well strip it for parts, QA those and sell them on.
But seriously, if a part fails QA, is it really worth repairing?
Collecting dead parts in the printer industry (ink cartridges etc) makes sense because of the sheer volume. But does it really make sense anywhere else?
it depends. The Deeming Principles for QA of mass production have changes a lot. Process improvements in production, quality improvements in production, have made QA either very statistical or completely pointless. If, when there are statistics, the failure rate is down in the 10^-6, why bother? Along another axis, there is the issue of the whole delivery process line. If the cost of the component or assembly is low enough compared to the cost of repair, then don't bother with repair, just replace. Often without fault analysis. There is a short film strip I saw once that was meant to illustrate something else but does demonstrate this principle. A fighter jet lands on an aircraft carrier and the ground crew flock around it. One team carried refuelling hoses. Another team comes along in phases. The first phase is a guy with a "key" who runs past opening the bay doors along the side of the aircraft that contain the various electronics and other stuff. Immediately following him is a guy with a trolley tethered to his waist who pulls the units out of the pods and chicks them, regardless of their operating status, onto the trolley. Behind him is another guy, again with a trolley, but this is stacked with racks, all in just the right order, and he stuffs them into the slots just vacated. Behind him is a guy that makes sure the racks are all locked in, closes the pod doors and has a key to lock the doors. Meanwhile other teams have refuelled, reloaded the munitions, and the plane is ready to go. The movie was _meant_ to illustrate a point about 'line replaceable units'. The electronics pods didn't have to be single sourced if they had the same interface and function. What it _did_ show is that it isn't worth stopping to test them. Chuck them on the trolley and later take them down to the shop and test them. How is this relevant here? it is not worth disrupting the production like to test every item. At some point its not worth even doing statistical tests. If the competent fails for the customer replace it without question. It's cheaper that way. So: two things emerged from this for me. One was what I term "The Closet of Anxieties". Many firms using Microsoft found that they were upgrading perfectly capable hardware simply because the Windows upgrade demanded new hardware. They didn't have the heart to add the old machines to landfill. One example: A friend of a friend told me about a law office that went though this 'upgrade'. I visited to see what they had. The office manager showed me a room with the "old" (but still fully functional apart from their wiped disks) machines. It was like walking into a graveyard. They were lined up, perhaps two dozen, with the display, looking like headstones. "How much?" I asked. "Nothing. They're free. They are CCA depreciated and off the books now. They are just taking up space. Take one, take two take them all", she said. I could only fit about 10 in the back of the Volvo. It was the monitors that were the problem. I put Linux on them, threw a garden party "with internet service" (it was still a novelty then) and gave them away. Another example: I upgraded to a tower. It had once been a server in the CoA at a place I worked, but the disk was damaged so I bought a 1T drive for a discount store on College. It failed spectacularly the first day. I took it back and they replaced it without even listening to my report on the number of bad sectors I found. I asked what it this one failed. "Bring it back, we;'ll replace it, no charge". That's the economics of it all. I ran badblocks mercilessly on this one! 'Reconditioned' may fail, but so might 'new'. It's the economic model behind production that makes the replacement policy what it is. Yes they could do better post-production testing. But that would be expensive. And we'd pay the cost, since it would apply to EVERY disk drive. It is cheaper all round to do it this way. And, as a side effect, 'reconditioned' drives are cheap. In once sense they've already been paid for. -- A: Yes. > Q: Are you sure? >> A: Because it reverses the logical flow of conversation. >>> Q: Why is top posting frowned upon? -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org