|
Authored by: Anonymous on Monday, June 11 2012 @ 03:13 PM EDT |
If you think of the way you would type a word using a limited keyboard (e.g.
numeric only) then the SEQUENCE 223 could be the word "bad". This is 2
then 2 then 3.
When you use compression on a text file (arithmetic compression or any other
type etc.) the numeric sequence in the data stream of 2 + 2 + 3 could also
generate the word "bad" - i.e. the operation, to the computer is
identical.
For this reason T9 should not have really been allowed to be patented as it is
just an interpretation of a previously used, common computing technique. Have a
look at any of the Infocom adventure games as they use extensive text
compression.
Taking this further there is no difference in computer circles between receiving
the values 2 + 2 + 3 in quick order or 2,2,3 as one large blob. In essence
drawing a pattern is just like a discrete sequence of single keys but arriving
quickly (2,2,3 instead of 2, wait a bit, 2, wait a bit,3). It doesn't care where
the data comes from or how fast it arrives, the software is what will perform
the operations on it.
Computer software invariably comes down to the following algorithm:
1. Input 1 or 0
2. Do stuff
3. Output 1 or 0
The software is the 2nd stage, it doesn't care what the input is, nor what the
output has to look like, it just works on 1's and 0's. [ Reply to This | Parent | # ]
|
|
Authored by: Ian Al on Wednesday, June 13 2012 @ 04:16 AM EDT |
To highlight the issue, the memory and the processor can only store and
manipulate binary symbolic values. They are excellent for X-Y co-ordinate stuff,
but useless for lines with angles. How do you save a square shape or a triangle
shape with binary values? It can be done by specifying the X-Y co-ordinates at
the vertices and deciding what characters were displayed at the vertices. The
trouble is, that is how virtual keyboards already work.
If you cannot enter vector values into memory (except using ASCII characters and
the English language: 'a line of length 2cm at an angle of 30° from the
horizontal') then it is not possible to define arbitrary shapes and paths in a
lookup database.
With a high-end desktop, it would be possible to do the word-shapes thing by
converting X-Y co-ordinates on the fly and then doing the lookup/prediction.
However, just the shapes would not do: the absolute position on the screen is
also required to correspond to the absolute position of the keyboard on the
screen. Defining the absolute position of the starting point of the word shape
vectors would also be a software nightmare.
I'll cut this short by saying that the computing task required by the abstract
idea invention in the patent will be beyond the capability of a touch-pad
working exclusively on implementing the patent for many years to come.
Only by ditching the airy-fairy, non-implementable idea of 'word-shapes' can the
inventive concept be implemented.
The patent holder will then sue the virtual key-pad improver and win because the
jury can 'see clearly' that the invented magic is being used in the infringing
touch-pad.
---
Regards
Ian Al
Software Patents: It's the disclosed functions in the patent, stupid![ Reply to This | Parent | # ]
|
|
|
|
|