June 23, 1969

[JAB]

This day on June 23, 1969 - IBM announces that effective January 1970
it will price its software and services separately from hardware thus
creating the modern software industry.

<https://twitter.com/engineers_feed/status/1407699978017857536>

And more $$$ for their products

[Eli the Bearded]

That was the intention. Microsoft proves that they failed.

It's interesting to see how things have evolved since then. Separating
software from hardware has helped nuture the non-standard OSes, in
particular Linux. It also helped open the path to compatible computers.
If you can buy software separately from hardware (up to and including
from non-hardware makers), then it raises the potential that you can buy
different hardware that runs the same software. For IBM computers the
major obstacle to this was duplicating the firmware: the software that
was built into the ROM of the computer, namely the BIOS. Once that nut
was cracked, the clones proliferated.

Macs have long been much harder to clone. There are two things going on
there. First, the Apple firmware was a lot more sophisticated than BIOS
even in the early models, and it has continued to remain harder to
clone. Secondly, Apple's software license doesn't make it legal for
buying the OS to run on non-Apple produced hardware, immediately
limiting the market to people willing to violate licensing terms. That
cuts out basically all of the big spending customers, schools and
businesses.

Mac emulation has long required ROM files from real Macs. Nothing stops
PC emulation, which means PC operating systems are widely available in
"virtual" providers, like AWS.

Elijah
------
the Phoenix BIOS "clean room" clone process is a famous computer story

[JAB]

On Wed, 23 Jun 2021 18:51:37 -0000 (UTC), Eli the Bearded
<*@eli.users.panix.com> wrote:

>For IBM computers the
>major obstacle to this was duplicating the firmware: the software that
>was built into the ROM of the computer, namely the BIOS. Once that nut
>was cracked, the clones proliferated.

In earlier clone days, there were some glitches when using a cloned
machine....especially if the software program was using IBM's "coding"
for a hardware interrupt request...iirc. I'm not a programmer, but I
recall tidbits on this topic back then.


"The machine was based on open architecture and a substantial market
of third-party peripherals, expansion cards and software grew up
rapidly to support it.
...
...
Because the IBM PC was based on commodity hardware rather than unique
IBM components, and because its operation was extensively documented
by IBM, creating machines that were fully compatible with the PC
offered few challenges other than the creation of a compatible BIOS
ROM.

Simple duplication of the IBM PC BIOS was a direct violation of
copyright law, but soon into the PC's life the BIOS was
reverse-engineered by companies like Compaq, Phoenix Software
Associates, American Megatrends and Award, who either built their own
computers that could run the same software and use the same expansion
hardware as the PC, or sold their BIOS code to other manufacturers who
wished to build their own machines. "

<https://en.wikipedia.org/wiki/IBM_Personal_Computer>

[Eli the Bearded]

In misc.news.internet.discuss, JAB <he...@is.invalid> wrote:

> Eli the Bearded <*@eli.users.panix.com> wrote:
>> For IBM computers the
>> major obstacle to this was duplicating the firmware: the software that
>> was built into the ROM of the computer, namely the BIOS. Once that nut
>> was cracked, the clones proliferated.
> In earlier clone days, there were some glitches when using a cloned
> machine....especially if the software program was using IBM's "coding"
> for a hardware interrupt request...iirc. I'm not a programmer, but I
> recall tidbits on this topic back then.

Yup, that came from subtle incompatibilities in BIOS or selection of
components.

> Simple duplication of the IBM PC BIOS was a direct violation of
> copyright law, but soon into the PC's life the BIOS was
> reverse-engineered by companies like Compaq, Phoenix Software
> Associates, American Megatrends and Award, who either built their own
> computers that could run the same software and use the same expansion
> hardware as the PC, or sold their BIOS code to other manufacturers who
> wished to build their own machines. "
>
> <https://en.wikipedia.org/wiki/IBM_Personal_Computer>

The Phoenix method was I believe the one challenged in court. To do it
without breaking copyright law they had one team of people study the
BIOS in a computer and document the way it works in great detail. They
then took that document to some programmers who had never studied /
looked at BIOS and had them reimplement it from the documentation. The
well-documented process was explained in court and ruled not a copy
under copyright. Just two different things that do the same thing.

That clean room process can introduce differences when one or the other
has a bug in say reseting a value after an operation causing some
subsequent thing to behave differently. The more things that happen and
the more ways those things can interact creates the complexity. Mac
firmware, eg.

The clean room process can be seen as a precursor to how machine
translation works: comparisons of inputs and outputs then applied to
other text.

Elijah
------
and that's why machine translation only works well for common usage

[JAB]

On Thu, 24 Jun 2021 03:49:28 -0000 (UTC), Eli the Bearded
<*@eli.users.panix.com> wrote:

>> In earlier clone days, there were some glitches when using a cloned
>> machine....especially if the software program was using IBM's "coding"
>> for a hardware interrupt request...iirc. I'm not a programmer, but I
>> recall tidbits on this topic back then.
>
>Yup, that came from subtle incompatibilities in BIOS or selection of
>components.

That was in earlier DOS days.... iirc, MS was suggesting to do
these interrupts via their OS...something like that.

Now the OS examines hardware components, including the BIOS; I have
no idea if MS did this for each clone BIOS back in earlier DOS
days....I assume so.

Just from my "headline" scans of a topic I had no interest in, some 40
years ago...which was in a world when new equipment (printers, etc)
had to be configured to work...no plug=n=play.