Re: data compression program

In article <1145847805.843092.18740@xxxxxxxxxxxxxxxxxxxxxxxxxxxx>,
"Lisa Michelle Smith" <smith_lisa_michelle@xxxxxxxxx> wrote:

[Quoting fixed]

Barb Knox wrote:
In article <1145768858.772528.240...@xxxxxxxxxxxxxxxxxxxxxxxxxxxx>,
"Lisa Michelle Smith" <smith_lisa_miche...@xxxxxxxxx> wrote:


I have designed an amazing data compression program for my computer
company. It compressess data in any format to half its size without
losing any information.

If it losslessly compresses any data to half the size, can you run it
twice to compress to one fourth the size? Three times to one eighth?
What's the limit?

The great thing about it is there is no limit to how many times you can
compress it. That's what makes it better than the competition.

If that were true then you could losslessly compress any file down to
just 1 bit, by repeatedly running your algorithm. That is clearly
impossible. Therefore your claim (as stated) simply cannot be true.


I want to get a patent for this. How do I do
this? Would my company own the patent?

If and only if your written employment contract with them specifies
that
they own anything you develop (which is a common contract term).

Then I should get a lawyer.

As others have suggested, before spending time and money on a lawyer you
should refine your claim regarding what your algorithm accomplishes.
Your initial claim is impossible; but maybe your algorithm does perform
well enough that you can still make significant claims about it.


If so, should I quit my company and not tell them about my invention?

The common contract terms say something like "anything developed during
the term of employment", which means they would own it whether you told
them now or not. But IP law is complex. For example, if you developed
it in your own time at home using your own equipment then you might
have
a stronger claim (again, depending on the terms of your employment
contract). See a lawyer.

I'll talk to a lawyer. Do you or anyone know any good ones in
California?

So far, they don't know. I don't want anyone to steal it from me.
No one knows about how it works except
for my boyfriend and he doesn't really understand it because he never
went to college like me.

Did you study Shannon's theorem in your college?

No. I majored in information science/liberal arts, not math, so we
didn't have to learn any technical theorems.

One big benefit of knowing some theory is that there are various strong
impossibility results. For example, a consequence of Shannon's theorem
is that NO algorithm can losslessly compress truly random data (and BTW,
that's a useful characterisation of what "random" means in the first
place).

Just out of curiosity, did they teach you about the unsolvability of the
"halting problem"? IMO, they should have.
.

Relevant Pages

  • Re: More of an Algorithems question
    ... >How can I proove that no Algorithm can compress every file of length ... However there is an algorithm that will losslessly compress every file ... Prev by Date: ...
    (sci.math)
  • Re: More of an Algorithems question
    ... >>>However there is an algorithm that will losslessly compress every file ... >> n-1 bits. ... Prev by Date: ...
    (sci.math)
  • Re: data compression program
    ... Barb Knox wrote: ... That's what makes it better than the competition. ... If that were true then you could losslessly compress any file down to just 1 bit, by repeatedly running your algorithm. ...
    (comp.theory)
Loading