"Not very useful" is not a phrase that makes sense in this context - we are talking about computational theory, not practical, efficient devices. There is *no* efficient way to address unlimited memory, since no pointer is big enough on its own. But it is perfectly possible to imagine ways for an AVR to indirectly access unlimited memories. After all, it's not hard to figure it out on a Turing machine, and it doesn't even have the convenience of being able to do binary arithmetic.
I mean a sort of squared version of such a spiral, going through specific points. The idea is to find a way to number all the points on your 2D great in a list that covers all the points. Try drawing something like this to get started:
The reference was to a 2D grid of points, not the continuous 2D plane. The number of points on the continuous plane is the same as the number of points on a continuous line - aleph 1. The number of points on a 2D grid is that same as the number of points on a 1D dotted line - aleph 0. Proving that (aleph 1)^2 is equal to (aleph 1) is a little harder than proving that (aleph 0)^2 is equal to (aleph 0), especially when keeping the post in straight ASCII...
Infinite. Igg. I dimly recall a brief period many years ago when I spent some time wandering through hotels with infinite numbers of rooms and diagonal hallways (or was it elevators?). I quit when I learned that even though the man at the front desk had an infinite number of guests who had to be shuttled in and out of an infinite number of rooms, he still had time to tell an infinite number of stories, and they were all infinitely boring.
At that time, the "size" of the "smallest" infinite set was Aleph-0, the size of the set of integers (also the positive integers, the odd integers, the perfect squares, yadda yadda...). The presumed next larger "size" was 'c', that of the set of real numbers, and The One Big Question in that neighborhood was 'c' was the same as 'Aleph-1', that is, whether there was an infinite set whose size was "larger" than Aleph-0 but smaller than 'c'.
Don't know if that one was ever solved.
Anyway, if you're looking for a mathematical answer, I think I'd suggest starting with Murphy's Law: "Anything that can go wrong will go wrong, and at the worst possible time". Even if Murphy's Law is recursive, as long as you can prove that its _domain_ is countable (that is, if "anything" can be restricted to an Aleph-0 set) then you only have a countable number of lawsuits to worry about (if Murphy's Law is combinatoric, I'm slightly less certain of this).
In order to handle them, you will, of course, need a countably infinite number of lawyers. The good news is that the set of lawyers here in the U.S. is already approaching that size.
Hope this helps...
Frank McKenney
-- ...the fact that some geniuses were laughed at does not imply that all who are laughed at are geniuses. They laughed at Columbus, they laughed at Fulton, they laughed at the Wright Brothers. But they also laughed at Bozo the Clown. -- Carl Sagan
-- Frank McKenney, McKenney Associates Richmond, Virginia / (804) 320-4887 Munged E-mail: frank uscore mckenney ayut mined spring dawt cahm (y'all)
Quoting this webpage (see last chapter before "Conclusion") :
formatting link
| In the 1930ÿs, Kurt Gödel showed that the continuum hypothesis can't be | disproved from the axioms of set theory, and in the 1960ÿs another | mathematician named Paul Cohen showed that it cannot be proved, either.
Wikipedia says the same:
formatting link
--
Frank Buss, fb@frank-buss.de
http://www.frank-buss.de, http://www.it4-systems.de
Nope, they are the same infinity (aleph 0). There may well be infinitely more ways to get the software wrong than there are of getting it right, but the sum turns out the same in the end.
To get bigger than the countably infinite number of correct implementations, you need to move to uncountably infinite (i.e., aleph 1 or above, assuming the continuum hypothesis), which is more than the number of computer programs.
As Steve and Clay mentioned, I think you just need an E&O policy. We have one because some clients ask for it (like yours seems to be), and having one lets you check that box off.
And, yeah, everybody wants to put risk on somebody else's shoulders, and that seems to be how many corporate lawyers justify their existence. Asking for insurance against any possible failure or infringement against current and future patents, alien invasion, meteor impact, etc., seems to be pretty routine these days. We just say, "Sure, happy to include you as a named insured on our policy" and then they get all happy and smile again and their lawyers stop bugging them. If a client asks for that I don't think it's unreasonable to pass the cost of the policy along to them.
Eric Jacobsen Minister of Algorithms Abineau Communications
How do you pass the cost of the policy along to those who demand it, without penalizing those who, say, would be happy with liability limited to the cost of work done?
Best regards, Spehro Pefhany
--
"it's the network..." "The Journey is the reward"
speff@interlog.com Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com
In most cases something resembling "market pricing" applies. Customers who tend to ask for indemnification, or stricter contract terms in general, do not attract bids as favorably low as customers who are more relaxed, so the contractor usually has room to charge them more.
But having liability limited to "cost of work done" is pretty uncommon in most contracts unless they are of the straight, time-and-materials, "best efforts" variety.
But the most important rule is: all contracts are different. I have not had to buy E&O insurance myself, but one customer demanded general liability insurance. I made sure the first milestone payment in that contract exceeded the cost of an annual GL policy by a comfortable amount...
More, how many ways are there for the client to use my work product, because I don't develop turn-key systems, I do analysis and algorithms that others find hard.
Tell them they have to agree to idemnify YOU against any 3rd party lawsuits. You could tell your customer that since they have the deeper pockets they may purchase insurance for you in any amount they like, with them ageeing to idemnify you above that. As for the customer, I have had time sheets where signing indicates the customer agreeing to the billing hours and stating that all work has been done to their satisfaction.
FIRST: I appologize if this has already been stated. I am not going to read all 31 posts.
Tim,
Tell them they have to agree to idemnify YOU against any 3rd party lawsuits. You could tell your customer that since they have the deeper pockets they may purchase insurance for you in any amount they like, with them ageeing to idemnify you above that. As for the customer, I have had time sheets where signing indicates the customer agreeing to the billing hours and stating that all work has been done to their satisfaction.
I maintain you have absolutely no reason to apologize. The previous posts are not necessarily even available to other readers, and the purpose of quotes is to present an adequate summary for the points discussed. Just make sure you quote material to which you respond, and snip all other.
--
[mail]: Chuck F (cbfalconer at maineline dot net)
[page]:
Try the download section.
Tim, Just how much are you getting quoted for E&O insurance? I have a similar dilemma when people want assurance.
I normally just use a limited liability statement as follows. An additional embellishment might be that any disputes are resolved by binding arbitration in the defendant's state. Only real butthead companies would balk at those ideas... but in this economy, you may not be able to choose customers as well. It may be cheaper to spend a few hundred $ on a lawyer that is competent at drafting good consulting agreements.
Limitations of Liability
The Consultant and Client agree that the Consultant functions as a limited liability consultant: the Consultant's legal liability is limited to the amount of money already paid to the consultant for that task. There are no additional warranties expressed or implied. Any problems relating to a task should be reported to the Consultant promptly and Consultant and Client agree to make reasonable efforts to resolve these problems. Any disputes will be resolved through binding arbitration in the State of Colorado.
------------------------
The preceeding statement is for informational purposes only. If you want one that really works with the law, get a lawyer.
I would worry the language above creates a perpetual obligation to resolve problems. More reasonable (and extremely common) might be to define a post-delivery acceptance interval, and the contractor is obligated to resolve problems only during that interval.
Been there, done that.... never lost nor paid out a single penny; except to our legal team.....
(pists, put a clause in your contract about "ARBITRATION".)
This avoids court and gets a technical panel of experts involved.....
Not possible to prove or disprove, methinks.
There is the school of thought that infinity is just infinity, regardless of how fast or vast infinite is....
I think on such projects you should build a team to make your customers happy. As coders, the wife (PhD in Computer Engineering) and I search for project where me might fail one day...
(Never going to happen; although we search the earth for a chance to fail....)
What we do is finish projects (particularly with embedded prototypes ready for manufacture). We count government agencies, fortune 20, and small startup among our happy customers. What we do is offer an open ended warranty. If the product is not 100% satisfactory, then the customer keeps on paying us (time and materials) until they are happy. Warranties explicit or otherwise terminate 30 days after we receive the last check.
Battle hardened approach, that keeps companies paying time and materials to us, whilst we endeavor to make them happy.
caveat emptor **AVOID ALL ONE PRICE BIDS**
i.e. their 'back-door insurance' aka "indemnify clauses" are only valid as long as they keep paying time and material costs. At some point they accept the product, get on to production and plan the next release (rev) of the product.
We've built hundreds of microprocessor/dsp/fpga/asic based products and saved dozens after others have failed. We provide 100% knowledge transfer, when their team is ready to take over. Both Honeywell and GE have made a fortune purchasing small companies where we pioneered the product for folks. The one thing we do not usually do, is exactly what you do. We do not dream up the product or the theoretical ideas. We just write code for ideas and hardware, where some EE and said "this should work." EE are very poor coders, as most of our clients are EEs. I have both EE and Computer Science degrees beside one in petro-chemical eng.
State machines to full blown embedded linux solutions. My wife is basically a savant with any type of assembler or weird as SOC. She's designing a profiler that is far beyond what anyone has ever dreamed of, to migrate and test video (h.264 code and gait analysis algos) to quantify on various types of processors. The wife gets most of her client via the semi-conductor referrals directly of from the various FAE (engineers) at the Vendor companies like Arrow, AVnet, Future, All American etc etc.
WE'll take that liability every time, THANK YOU VERY MUCH ! Seriously offload your clients and let us do the coding.
You dream and design high level, we code and debug and help the client refine (or in many cases develop) their specifications for a new product, in an dynamic fashion.
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.