|
Authored by: mrisch on Sunday, June 10 2012 @ 08:14 PM EDT |
I don't know that sticking a new roll in creates a new
machine. I tend to doubt it. I write about the loom in my
America's First Patents article and point out the very
issues you raise here.
But, I will note that patents cover methods, so the question
is whether putting a new roll in a pianola is a new way to
solve some practical problem. I say no (and probably not
with the loom either). But if putting software in a machine
is a new way to solve a problem, then it is a new use for
the machine, and that is a process worthy of consideration.
And I knew going in that most people wouldn't agree with me
:)
As for the hardware - of course you can have software
written to run on different platforms, but I assure you that
Windows 3.1 did not load on a Mac SE, and I assure you that
Linux will not run without the proper kernel, and Java will
not run without the proper virtual machine. And given the
many, many patents and journal articles I have read, many
solutions are not implementable broadly due to resource
constraints. Stuff that I can do with my laptop I just can't
do on my phone. Some stuff, at least, and the list is
shrinking, but not in the mid 1990's when this explosion
hit.[ Reply to This | Parent | # ]
|
- Stuff that I can do with my laptop I just can't do on my phone. - Authored by: Wol on Sunday, June 10 2012 @ 08:35 PM EDT
- Hardware too small or slow or programmers unwilling errors... - Authored by: Anonymous on Sunday, June 10 2012 @ 11:09 PM EDT
- software has to be written and compiled for specific hardware - Authored by: Anonymous on Sunday, June 10 2012 @ 11:50 PM EDT
- software has to be written and compiled for specific hardware - Authored by: Anonymous on Monday, June 11 2012 @ 03:16 AM EDT
- Patent Protection is too Broad and too long to work for software - Authored by: RMAC9.5 on Monday, June 11 2012 @ 03:22 AM EDT
- The Jacquard loom - Authored by: Ian Al on Monday, June 11 2012 @ 07:07 AM EDT
- Linux will not run without the proper kernel - Authored by: Wol on Monday, June 11 2012 @ 07:23 AM EDT
- A virtual machine is not hardware - Authored by: Anonymous on Monday, June 11 2012 @ 09:04 AM EDT
- software has to be written and compiled for specific hardware - Authored by: PolR on Monday, June 11 2012 @ 10:36 AM EDT
- software has to be written and compiled for specific hardware - Authored by: drakaan on Monday, June 11 2012 @ 11:22 AM EDT
- If software is a new way to solve a problem, then it is a new use for the machint? - Authored by: Anonymous on Monday, June 11 2012 @ 11:44 AM EDT
- Wrong on EVERY COUNT, Mr. Risch. - Authored by: Anonymous on Monday, June 11 2012 @ 12:00 PM EDT
- Wrong on EVERY COUNT, Mr. Risch. - Authored by: mrisch on Monday, June 11 2012 @ 01:32 PM EDT
- Wrong. - Authored by: Anonymous on Monday, June 11 2012 @ 02:03 PM EDT
- And regarding the process of books/arts/music: - Authored by: Anonymous on Monday, June 11 2012 @ 02:05 PM EDT
- And your point #2 is not germane. - Authored by: Anonymous on Monday, June 11 2012 @ 02:08 PM EDT
- The general purpose computer does not have a known use for stuff that hasn't been invented yet. - Authored by: Wol on Monday, June 11 2012 @ 02:17 PM EDT
- Wrong on EVERY COUNT, Mr. Risch. - Authored by: drakaan on Monday, June 11 2012 @ 02:57 PM EDT
- Wrong - Authored by: jesse on Monday, June 11 2012 @ 03:59 PM EDT
- Wrong on EVERY COUNT, Mr. Risch. - Authored by: PolR on Monday, June 11 2012 @ 05:37 PM EDT
- Wrong on EVERY COUNT, Mr. Risch. - Authored by: ThrPilgrim on Monday, June 11 2012 @ 07:29 PM EDT
- software has to be written and compiled for specific hardware - Authored by: NigelWhitley on Monday, June 11 2012 @ 12:17 PM EDT
- "new use for the machine"? - Authored by: Anonymous on Tuesday, June 12 2012 @ 10:59 AM EDT
|
Authored by: Anonymous on Sunday, June 10 2012 @ 11:42 PM EDT |
Debian supports multiple architectures (platforms) without the authors of
most of the packages having to do anything extra for that cross platform
support. Most of the software authors don't even have access to most of the
hardware platforms their packages run on. The Debian package maintainers (a
completely separate group of people) worry about that. I have written Python
software which runs on x86, x86-64, and ARM without any changes. I didn't even
have access to the ARM hardware for testing. It just worked.
Even
thirty years ago it wasn't true that "software has to be written and compiled
for specific hardware". You could say that it usually was the case, but
it wasn't a case of had to. Today it's usually a case of the economics of
targeting specific markets however, rather than a technological matter. Indeed,
many programming old-timers today are decrying the extent to which software has
become abstracted from hardware.
An interesting twist to this is the
UCSD P-System, which used a "virtual machine" to execute code starting in the
1970s, long before Java was around. You could run your compiled code on multiple
architectures without any changes, and P-System was in fact ported to many
different mainframes, mini-computers, and micro-computers. Western Digital later
designed a chip to implement the p-machine in hardware. In other words, the
abstract representation of the p-machine came first, and the actual p-machine
hardware later. The WD Micro-engine never became very popular however, as the
emulated systems were actually faster than the dedicated hardware.
For
that matter, who says that software has to be "compiled". Plenty of commercial
software isn't. And what exactly does "compiled" mean these days? Static
compiled? JIT compiled? Tracing compiled? Compiled to byte code? The whole
concept of "compiled" is very nebulous these days. The world of software has
advanced a lot since the term "compiler" was created. Your average web developer
today works in an entirely abstract world where he has no clue how the
underlying hardware works or what it is, and indeed believes that state of
knowledge to be the ideal condition!
[ Reply to This | Parent | # ]
|
|
Authored by: David665 on Monday, June 11 2012 @ 11:44 AM EDT |
Your explanation has nothing to do with the case. Sun wanted a Java phone, they
would not accept what Google needed which was a subset of Java, in the middle of
this other Sun patent attorneys came after Google, the agreement fell apart, Sun
CEO said well at least it was a Java phone, Oracle buys Sun, changes its corp.
stance and decides Google did the nasty, the reality is Ellison wants a
percentage of all Google ad income.
The copyright infringement attempt failed, the patents were not infringed
regardless of whether they were good or bad.
All the rambling about Java, JVMs and such have nothing to do with software
patents. Oracle was claiming ownership over that because of copyrights of the
API.
Interesting position but not a valid one for this argument.[ Reply to This | Parent | # ]
|
|
Authored by: Anonymous on Monday, June 11 2012 @ 02:52 PM EDT |
There are (and have always been) limitations to System 360
program compatibility. The first, and biggest, is that only
problem state programs are likely to be compatible. That
limitation means that programs which directly interact with
the hardware may not run (some will, but over time, fewer
and fewer will.)
In essence, the 360 provided user application compatibilty.
It does not mean that an operating system written to run on
a 360 will run on later hardware, or even entirely unchanged
on different 360 models. Changes within a processor family
(e.g. all 360s or all 370s) would generally be invisible to
the user as they would be incorporated into the operating
system like any other fix.
The second limitation is that the program must not rely on
characteristics or features of the hardware such as timing.
So, looping to delay processing by, say, a second will no
longer work properly (besides being bad practice.)
Originally, floating point was optional too, so any program
written to use that would not work on any box (even another
360) without it.[ Reply to This | Parent | # ]
|
|
|
|
|