Andrew, almost. Mail me in private and I'll have a look at sending you the missing code.
Rene
-- Ing.Buero R.Tschaggelar - http://www.ibrtses.com & commercial newsgroups - http://www.talkto.net
"Steve at fivetrees" schreef in bericht news:SMadndau snipped-for-privacy@pipex.net...
property.
yuk! -
control
Nah, a ruined batch will be mixed with the next batches, in a 1:10 or 1:100 whatever is acceptable. Skip that definitely stuff. And that few millions $ was street value, not ingredientens value.
Give me a break, please.
Now, for a change, I want to hear from someone who has actually and proven killed a person, by writing crap software....
Possibly, end of thread, but I am all ears.
-- Thanks, Frank. (remove 'q' and '.invalid' when replying by email)
Do you write software? With the attitude you're showing here, you look like the best candidate to be such a person, so far.
"larwe" schreef in bericht news: snipped-for-privacy@t31g2000cwb.googlegroups.com...
at best I have killed an animal, but not that I know off.
(which would have been bad enough...)
But I would be the first to admit writing crap software where it didn't matter, if that was your point.
And no, I don't add disclaimers to my source files that the stuff can not be used for other purposes than intendend. Do you?
-- Thanks, Frank. (remove 'q' and '.invalid' when replying by email)
I believe the classic case was the Therac:
-- John Devereux
Sure, Therac is the one everybody learns about, but Frank wants someone with actual blood on his hands to step forward and explain how he killed someone with software.
Well if I were such a person, I don't think I would post the details to usenet! :)
-- John Devereux
Such a person would be like Marley's ghost - coming forward to save somebody who's on the path to eternal damnation (i.e. designing Windows into an embedded system).
"larwe" schreef in bericht news: snipped-for-privacy@u72g2000cwu.googlegroups.com...
Right, just to make my point that it doesn't happen all that very often.
You guys are serving the soup hotter than it is.
-- Thanks, Frank. (remove 'q' and '.invalid' when replying by email)
"larwe" schreef in bericht news: snipped-for-privacy@g10g2000cwb.googlegroups.com...
someone
Would it be wrong to use windows for a digital picture frame?
Windows is perfectly all right for a gazillion of applications, if you know what you are dealing with.
BTW, about the Therac accidents:
"Related problems were found in the Therac-20 software. These were not recognized until after the Therac-25 accidents because the Therac-20 included hardware safety interlocks and thus no injuries resulted.".
Didn't I write earlier: "Property damage may be at risk more often, but 'failure to act' should not lead to (large) damage anyway, with any sensible system."
It seems that the Therac-20 machine had undiscovered software related problems, but was a sensible system because it had hardware safety interlocks. The real accidents were all thanks to the Therac-25 system.
Now you can say, "Ok, point taken".
The bottom line remains: Windows is perfectly all right for a gazillion of applications, if you know what you are dealing with.
In particular for embedded stuff, because embedded stuff is often pretty much wrapped up in boxes that are not messed with.
Exercise: take a new windows PC out of the box, start it up and let it run with the clock on the desktop. It will keep running until the hardware fails.
Bottom line: Windows is perfectly all right for a gazillion of applications, if you know what you are dealing with.
-- Thanks, Frank. (remove 'q' and '.invalid' when replying by email)
In article , larwe writes
There are many cases where software has killed people Patriot missiles for one.
-- \/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\ \/\/\/\/\ Chris Hills Staffs England /\/\/\/\/ /\/\/ chris@phaedsys.org www.phaedsys.org \/\/\ \/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/
Oh no! I have designed WinXP Embedded into several systems. Time to market and cost would have killed the products without it.
These are digital media players and the worst thing that could happen to users would be a 1 minute wait without music while it reboots. They have been very reliable in practice with some of them running 24/7.
That said, I wouldn't dream of using Windows for anything critical.
-Mike
PRECISELY the kind of application where it makes sense to design in a consumer OS.
I've been down this path myself with Flash support - there's no sensible way to support Flash unless you use a consumer OS (or unless you're really enormous and big enough for Adobe to take notice).
,snip>
I can't find your email in this thread or your web pages, mine is found by sniping the _not_me.
regards Andrew
For applications that have a short product non-critical life it makes sense. Having see the problems of systems that have a particular version of commercial OS and can not be verified on newer versions, partly due to cost of have to source new hardware as new versions do not support the hardware on that platform.
--
Paul Carpenter | paul@pcserviceselectronics.co.uk
PC Services
GNU H8 & mailing list info
For those web sites you hate
Which NOBODY has denied for lots of applications, BUT for critical applications bodies like FDA I still believe actively ban Windows (no source and other issues to do with criticality). Windows is not much use on product life cycles of 10 years or more. I not so long ago had an enquiry on a hospital system where they did not know the age of the system, other than the earliest document was it was moved to a new building in 1968!
You seem to be showing strange ideas of critical use.
--
Paul Carpenter | paul@pcserviceselectronics.co.uk
PC Services
GNU H8 & mailing list info
For those web sites you hate
In article , Frank Bemelman writes
This is not true. For many reasons these problems are kept quiet.
1 Patriot missile system was (is?) inherently unreliable an and has killed the wrong targets (ie people). Probably over 30 2 It is believed that* Chinook FADEC sw has killed *AT LEAST* 30 people that we know of. (Mull of Kintyre incident). 3 It is believed that* the authoritative breaking sw in a high end German car caused some accidents. I do not know if any were fatal they tend to settle out of court and gag. *legal nicety to stop me getting sued.These are just off the top of my head. If I could be bothered to go to the web site that lists them all (I will have to find it as I lost my favourites a week or so back) It usually has one or two cases a week of safety critical systems playing up. Some are fatal and caused by software.
-- \/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\ \/\/\/\/\ Chris Hills Staffs England /\/\/\/\/ /\/\/ chris@phaedsys.org www.phaedsys.org \/\/\ \/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/
"Paul Carpenter" schreef in bericht news: snipped-for-privacy@pcserviceselectronics.co.uk...
building
A critical system does not per definition require hard realtime performance. If your system needs to be hard realtime, you can always add RTX drivers to NT.
Perhaps I should have said more clearly that critical parts should be dealt with externally, such as hardware safety interlocks. But this is common practice anyway, regardless what kind of system you use to control your critical application.
Like mechanical interlocks on elevator doors. If there is no car, the doors won't open, even if the controller tries to because it thinks it is safe to do so. The controller could be a windows box.
-- Thanks, Frank. (remove 'q' and '.invalid' when replying by email)
In article , Frank Bemelman writes
But only on a non critical system. Re SW causing deaths. The USS Yourktown was dead in the water for 3 hours due to a software problem (it was using win NT as you suggested) during that time ALL weapons, defence and radar systems were out of use. Had it happened whilst the ship was in the middle east or the Gulf of Arabia software might have been the cause of several hundred deaths.
So it is OK to use windows because you can depend on other things to save the day?
Safety and mission critical systems and high integrity systems use a belt and braces approach. You don't seem to subscribe to that idea.
What if the mechanical interlock breaks? Mechanical parts do wear out. This is why defence in depth is used. In the scenario you suggest if the mechanical systems break the whole system becomes unsafe.
-- \/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\ \/\/\/\/\ Chris Hills Staffs England /\/\/\/\/ /\/\/ chris@phaedsys.org www.phaedsys.org \/\/\ \/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/
"Chris Hills" schreef in bericht news:u++ snipped-for-privacy@phaedsys.demon.co.uk...
I was just about to ask if the known failures used windows or not ;) I bet the three examples you gave in your other reply didn't use windows. The problems aren't exclusively related to windows. BTW, what does the USS Yourktown use today?
No, just that you should add mechanical/electrical lockouts too, regardless what kind of system you use. The reliability of the OS is only a minor part. A well prepared windows box won't crash all by itself, as long as you don't allow operators to mess around with it, like you can do with a typical desktop PC. And if the application has bugs, even the most brilliant OS won't save the day for you. You'll have to admit that all those windows PC crashing is mostly (if not 100%) caused by people fooling around with it, loading browser plugins or other obscure software, etc.
Mechanical safety devices often default to 'safe' when broken. And then there is regular maintenance that should take care of parts wearing out too much.
-- Thanks, Frank. (remove 'q' and '.invalid' when replying by email)
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.