Slashdot Mirror
Amiga OS Ch-Ch-Ch-Ch-Ch-Changes
First we heard that Amiga's new OS was going to be based on the
QNX realtime OS. Then we read
that the new Amiga OS would use a Linux kernel, and
QNX issued
this proclamation. Then Amiga issued
this one.
Now Zenn
sends us a link to
this article
from BeDope
that claims the new Amiga OS will really be based on
BeOS. Okay, the BeDope story is a satire, but this Amiga
thing has gotten out of hand. Has Amiga really
settled on the Linux kernel? I wish they'd make up their
minds. I am one of the millions who would (sigh) love to
have an Amiga again...
QNX probably wouldn't have been a good choice
for an OS foundation for the Amiga. I took a
quick browse to the QNX website, and they make
a lot of neat stuff. The relevant products to
the decision probably were QNX and QNX/Neutrino.
Each has serious problems as a foundation.
QNX is tied to x86. This alone is a serious
problem, as presumably Amiga would want to go with
a high-performance, clean CPU (e.g. Alpha, PPC,
MIPS, ARM). QNX also lists very few choices as to
what hardware you can use. This may or may not be
a concern in a relatively closed system (in the
sense that the system would be sold mostly
complete), but considering the really keen
hardware Amigas had when they were new, their
target market would probably want something like
a Voodoo3 videocard (for instance), something
not supported by QNX, as most of their hardware
supported looks fairly old.
QNX/Neutrino, unlike QNX, is portable, and it
looks like it supports at least the x86, some
PPC, some MIPS, and a few obscure CPUs. As such,
it probably would be a better choice than QNX
for an AmigaOS, but as stated on the webpage,
most of its components are very minimal.
QNX and QNX/N both look like they'd take a lot of
work to make a consumer product. Is Linux a better
choice? Maybe. What does Amiga need to do? Here
are some ideas:
Select a set of hardware peripherals that are
relatively inexpensive, are made by a vendor
friendly to third-party OS's, and ideally are
close to best-of-breed, and offer their vendor
an exclusive contract where all Amiga-branded
machines come with said hardware in exchange for
price cuts and input into design of said products.
Hardware worth considering:
Voodoo3 Graphics
Alpha CPUs
Ensure that Linux/glibc binaries can run
Write an X server that will talk to the
AmigaOS graqphical layer to allow
X programs to run (within a window?)
Write a really good emulator for the 68k-based
Amigas, ideally capable of running dirty programs
In the end, I imagine we'll probably see something
that doesn't look much like a Linux system as
we know it (i.e. non-X GUI, not Unixy), probably
making extensive use of kernel modules to avoid
GPL issues and keep things closed source.
For every problem, there is at least one solution that is simple, neat, and wrong.
BeDope is satire, dude.
---
"'Is not a quine' is not a quine" is a quine.
"'Is not a quine' is not a quine" is a quine.
Quine "quine?
if you were implementing UNIX on top of NT (you can do that too, but it's messy).
Actually, if you look at the product architecture drawing for the Interix UNIX-on-NT product, it doesn't look messy at all. It sits right on top of the NT kernel and not on Win32.
--
Business. Numbers. Money. People. Computer World.
IMHO, the only reason apple has survived (before the Imac revival) is due to all the advertising hype.
Actually Apple survived because they realized that as a nitch platform the only way was to charge a big premium (such as $10,000 for a stripped IIfx) and sink those 60% margins into lots of R+D and marketing. Apple survivied the lean years primarily on things they thought up during the fat years (things like UI Research, QuickTime and the 68K emulator).
Commodore/Amiga had a different plan - sell the machine for less money (lower margins) and then find that you have no money left over for R+D and marketing. Thus the platform never really technically advanced or gained market share. After a few years everyone else has caught up on Amiga's technology, and there was no reason left to use it.
Bottom line - Amiga might have survived if they followed Apple's lead and made people pay through the nose when they had the technological advantage.
--
Business. Numbers. Money. People. Computer World.
This is the text of an email I sent to the president of Amiga. Please pardon the line length formatting; sadly, Slashdot doesn't yet have terribly good tools for fixing it.
Mr. Collas --
I read with great interest your announcement that future versions of the
Amiga operating environment will be based on the Linux kernel. The
importance of the Amiga system as an alternative operating system is well
known among users of Linux-based systems. As a Debian GNU/Linux system
administrator with some knowledge of the benefits of the Amiga, I can only
expect that the two will complement each other nicely.
As you may already know, the Linux kernel is released under the GNU
General Public License, a "free software" license intended to ensure that
free software remains free --- that nobody can turn a piece of free
software into proprietary software without the author's permission.
One of the provisions of the GPL states that if you take a piece of GPLed
software, modify it, and release it, you are legally required to release
it under the GPL itself, and to provide your users the source code of your
modifications upon request. That is, "derivative works" of GPLed code
must themselves be GPLed and released with source available.
Because the Linux kernel, and most of the common utilities that form the
rest of Linux-based systems, are GPLed, this means that if Amiga were to
make customizations to them for use in the Amiga OE, those customizations
would have to be released back to the community.
The best way to do this, of course, is for Amiga's programmers and
engineers to participate in the Linux community -- to get those
customizations inserted into the mainstream Linux kernel rather than
"forking" the project and creating a separate Amiga-only development tree.
While the latter is legal under GPL (as long as you release source) it
reduces the benefits that both Amiga and Linux can gain from the free-
software (aka "open source") model. Cooperating and being involved with
the kernel project also would earn the respect and admiration of the
existing Linux user base, whereas forking the kernel would likely elicit
some degree of displeasure, not to say contempt.
Naturally, being involved in mainstream kernel development means accepting
a certain reduction of control over the code your OE uses -- after all,
the kernel development effort is led by Linus and Alan, and they say what
goes in and what doesn't. However, it also means that fixes and
improvements to the mainstream kernel will automatically be compatible
with the one Amiga uses; a forked Amiga-only kernel would require merging
of patches, a procedure that's not painless.
Back on the issue of the GPL -- besides it being *legally* important that
Amiga follow the *letter* of the GPL, it is also *culturally* important
that Amiga respect its *spirit*. The GPL is of great cultural importance
among Linux developers and users, and respect for it is highly valued. It
would be a good move for Amiga to make it clear to the community that you
intend to abide by both the letter and spirit of the GPL in your use of
the Linux kernel in your operating environment.
Conferring with Linux's "leaders" (among whom I'd rank Linus Torvalds and
Eric S. Raymond as the two foremost) would also be a good move in the
sense of showing connection with rather than separation from the Linux
community. Linux works and grows by widespread communication, not by
press releases; opening more lines of communication between Amiga and
Linux can only help.
I wish you luck in your endeavor, and look forward to seeing greater
connection between the Amiga and Linux development and user communities.
Thank you for your attention.
--- [my real name deleted]
First, let me say that arcticle was hilarious. I've been a fan of BeDope for awhile, its much better than segfault.
I just find it funny how many people yesterday thought AmigaOS would be YALD, the only thing they are using is the kernel, it'd be like calling MacOS X, Nextstep just because it uses the Mach Kernel. A kernel does not an OS make.
cheese logs keep my wang warm at night.
One thing most people seem to ignore, though, is that Linux (the OS), is defined by way more than Linux the kernel. An Amiga that uses the Linux kernel, and Amiga system libraries, Amiga shells, an Amiga GUI etc., won't look much like Linux the OS at all.
However, it will likely just be a question of adding the right libraries and binaries to let it run all Linux applications.
And this is the sweet thing about this: They can completely redefine the OS, by building something new on top of the kernel, without loosing Linux compatibility.
Which means that they can provide whatever "revolutionary" features they like, and at the same time offer access to the huge amount of existing Linux software out there.
That is something they wouldn't get with QNX - the amount of available QNX software is a lot smaller.
Contrary to QNX there is also a huge groundswell of support for Linux from third parties, and it will be a LOT easier to convince those third parties to provide Amiga applications, if they can do it in a way that let their apps run on Linux as well (maybe with some Amiga specific functionality taken out).
This can be good for Linux too, because Gateway is throwing lots of money at this, and if they start getting people to port to the Amiga, and the Amiga uses the Linux kernel, and can run Linux applications, a lot of those people are likely to make sure their applications run without the Amiga operating environment too.
In a way, I see what they are trying to do as something similar to OpenStep, which is a set of API's, that run on top of any host OS you care to port it to (ref. www.gnustep.org. It completely redefines the interface to the user, and for the developer, but the host OS is still there, and easily accessible.
If well done, it also means that if Amiga Inc. in the future should choose to support other host OS's besides Linux, any applications written to the Amiga OE APIs should instantly run in that environment.
Apparently their biggest concern was hardware support. I am posting this from FreeBSD right now so I'm obviously no anti-FreeBSD bigot, but FreeBSD is very much aimed at the server market, not the home market -- sound support is minimal, the range of hardware devices supported is much smaller than Linux, and folks like Creative and Real are supporting Linux with their latest and greatest software and hardware, not FreeBSD. Given that the Amiga has always been about graphics and sound, choosing an OS that doesn't support any sound card made after 1998 doesn't make sense...
The biggest bloat in Linux is userland, not the kernel. This may be where Amiga Inc. can innovate. Who knows, they might even use userland stuff from FreeBSD to avoid the licensing hassles...
-E
Send mail here if you want to reach me.
For those interested, Jim Collas has posted another message about the whole ordeal on the bulletin board at http://www.amiga.com/ubb/Forum9/HTML/000001.html. It has a bit (but not much) more explanation on how they came to the big decision.
Admittedly, it looks rather bad to be flip-flopping on core aspects of a product. But this really isn't as negative as some have played it to be. This company wants to build something innovative -- and successful. Ask Apple... the two don't always go together.
They found that their success factor was unacceptably low, and they were forced to make a change. Heck, I'm glad they're carefully looking at these kind of issues. I'm even more happy they've chosen a Linux kernel.
The interesting thing is, if they both follow through on their announcement, you've got even more innovation and competition than before. QNX is going to continue with developing the operating system. Amiga is going to continue with developing their OS and with a Linux kernel.
If you're one to track win-loss comparisons, it seems like a win for the consumer, a win for Gateway, and a lesser win for QNX.
Of course, it isn't ALL roses. If you're porting or developing software, you've got an OS decision to make. And users will have one to make as well. One side to this to keep aware of is on the hardware end... the Linux version is going to take advantage of far more hardware than the QNX version. Might marginalize QNX if you've got the latest monster video card and QNX won't do it.
I haven't really been following the Amiga, but here's my reconstruction of what's been happening over the last few days:
+ Gateway buys Amiga, because they think tney might be able to use the classic hardware and software in a set-top box or palmtop or something.
+ Gateway figures out that's not really feasible, and that they're going to have to re-write from scratch.
+ However, writing an entire modern OS is a bunch of work, so Amiga decides to create an application environment called "Amiga Operating Environment". Classic Amiga is dropped and it's software will only be run in emulation. I'll assume they're somewhat smart, and they write AOE to run as a POSIX application, so they can port it to different places without too much trouble.
+ Meanwhile, QNX has a robust OS with a jazzy GUI that's been around a while in various vertical markets. They look at the success of Be, and they realize that there might be a market for a consumer version of QNX. They devote "40 engineers" to Consumer QNX.
+ After sniffing around Be (and probably *BSD), Amiga finds QNX, and realizes that there might be some "synergy" between their goals - A new consumer platform running Consumer QNX with AOE on top. They start talking.
+ Amiga realizes that you can buy a $500 Windows 98 computer, and it's going to be difficult to build a consumer computer with razor-thin margins and have both Amiga/Gateway and QNX make any money. QNX can play hardball in the negotiations because they've a working OS and 40 engineers, whereas Amiga has vapor.
+ Amiga looks at all the Linux hype, sees the word FREE, as in Free Beer Software and Free Beer Device Drivers and Free Beer Programmers. They privately tell QNX to screw off. Since they're building a POSIX application, moving it to Linux is possible, it just more delays.
+ QNX gets pissed, and realizes that Amiga won't have jack/shit for another 12 months, where Consumer QNX will be ready sooner. As an appeal to the Amiga fans, they post a bunch of stuff, including sexy screenshots, about being the platform for AOE.
+ Amiga fires back - They're using Linux, it's more popular and free.
+ QNX announces that they're going ahead and shipping ConsumerQNX anyways. They make an appeal to the Amiga fans out there - "We're the *real* Amiga - and we are shipping a product."
+ A big split occurs in what's left of the Amiga community. Some go with QNX as the next best thing to the real Amiga. Others wait for alphas of AOE running on a Linux kernel. Others stick with the classic hardware because both QNX and Linux/AOE are imposters. Some finally give up and go with BeOS or Linux.
--
Business. Numbers. Money. People. Computer World.