Mister QL

[bwinkel67]

I have noticed that two articles (TTos.zip and Stella.zip) I posted on the 24th May 2020 have been downloaded over 30 times.

Can be found here: viewtopic.php?f=12&t=2805&p=34121&hilit=Stella#p34121
I kind of missed the post, but eventually found that I had the same documents for ages already. The Stella HTML docs lacked a logo file which I attach here.

The TTos article is pure gold. I'm glad I followed that link to find it.

[Peter]

Does anyone know what happened to Stella? Did it ever materialise? Did anyone ever get it?

There must have been a demo, so not pure vaporware.

I saw the last realistic chance for Stella in a hard realtime OS specialized for FPGA. This would have turned the lack of portability, which Stella as an assembler-written OS had, into a strength.

On FPGA, the 68K platform - to which Tony Tebby was most familiar with - is easily available. And here, the space saving advantages of both the 68K architecture and Stella, would have been of more use than on hardwired CPU/MCUs which come with plenty of memory today. FPGAs usually have only little block RAM, but with an OS like Stella, applications could be squeezed inside. And FPGAs are often used in the context of realtime applications, where the high predictability of Stella could have found it's niche. (Unlike SMSQ/E which is a nightmare when it comes to realtime usage.)

In addition to that, FPGA based peripherals reduce the need to always write new drivers everytime a new chip comes out. Quite important when an OS is developed with very little manpower.

It appeared that Tony Tebby either did not see my points, still thinking in classic CPU contexts, or had already lost any interest in working on operation systems.

[Derek_Stewart]

Hi,

I have read this Stella document over 20 years ago, the Stella operating seemed to in use, but to persons unknown.

Seemed a good read, but I cobncluded that it was a custom operating system job, which maybe the basis of most of SMSQ/E now being used.

But that said, SMSQ/E did not start to work correctly till Marcel got control of the SMSQ/E source code.

Tony Tebby did a great job, but Marcel did it better.

[Derek_Stewart]

I stumbled upon an article from the developer of the original MiST. Includes some interesting thoughts the development board DE10nano used for the MiSTer.
https://www.atari-forum.com/viewtopic.p ... 0ee15ddc00
Also worth reading for those who would like to see future dedicated QL hardware.

Hi,

I do not know if the Mist is better than the Mister, I would of bought a Mist, but the Q68 finally appeared, which suited my needs.

Butbneither Mist or Mister can run SMSQ/E like thw Q68. Buth this does not make them any less, maybe I should buy everything...

[Peter]

Can be found here: viewtopic.php?f=12&t=2805&p=34121&hilit=Stella#p34121
I kind of missed the post, but eventually found that I had the same documents for ages already. The Stella HTML docs lacked a logo file which I attach here.

The TTos article is pure gold. I'm glad I followed that link to find it.

Took me a while to find. Pure gold is always hidden. But reading from Tony Tebby always makes me sad he did not decide to stay. The SMSQ/E work he did for my Qx0 machines was enormous. Comparing todays version with the last one from Tony Tebby, only the commandline history was a relevant improvement for me, while it broke standard ROM size and copyback cache now has issues that I didn't notice back then. Tony Tebby did a good job for native hardware, and most importantly, he did the pioneering.

[bwinkel67]

Took me a while to find. Pure gold is always hidden. But reading from Tony Tebby always makes me sad he did not decide to stay. The SMSQ/E work he did for my Qx0 machines was enormous. Comparing todays version with the last one from Tony Tebby, only the commandline history was a relevant improvement for me, while it broke standard ROM size and copyback cache now has issues that I didn't notice back then. Tony Tebby did a good job for native hardware, and most importantly, he did the pioneering.

I've only started reading it, and skimmed some ahead. I love the relevance of Unix and Sinclair Research's focus on creating a workable version -- I don't know the history of Unix that much though I used it back in the late 80's on Sun Microsystems in college and didn't care so much about efficiency then. It really underscores what we all feel about the QL, the fact that from the start its idea was based in a solid foundation when it comes operating system design, whereas the other micro's of the time kind of winged it (i.e. MacOS, pre-Amiga Commodore, etc...and even the Amiga's OS has its foundational flaws as Linus pointed to). Unix/Linux today is the OS of choice, even if Windows still has the desktop market (side note, I worked for Microsoft from '07 to '09) and I still feel now that I hope some day they do what the MacOS did and focus on the experience and move away form their core. The code base back 10 years ago was the start of Windows 7/8/10 and it was (and likely still is) a mess. I got to add to the nightly builds doing speech recognition work, which was a stressful process as you didn't want to be that person that crashed the build -- there was this rotten gourd that you'd have to keep in your office until the next crash :-/.

[Tinyfpga]

Stella was developed to follow on from the Atari ST cartridge version of SMS2. The Stella core was finished and fully documented. It ran (still does) on an Atari ST. As far as I know it has never been developed to run on any other hardware. One of the problems at the time of its creation, was there was no low cost off-the-shelf hardware it could run on (other than the ST). Within a short space of time low cost computers stopped being built using the 68000 and that was that, until, as Peter states in his post, it became practical to build 68000 systems in FPGA.

The problems associated with developing system software in traditional hardware designs are detailed in an other document Tony Tebby wrote to the QL community sometime after producing SMS2 and before producing Stella.
The document, "A Brief History of SMSQ". is attached.

At the end of the article Tony Tebby writes:-
"How much more is it reasonable to do with SMSQ which, by its need for compatibility with a 10 year old computer, is locked into a 10 year old design? Is there enough interest in the type of operating system concepts pioneered by QDOS to make it worthwhile producing a completely new system? The future depends on your response."

I think Tony was hoping to receive replies. I don't think he received any. About a year after he wrote this article he started to develop Stella.

BRIEF HISTORY of SMSQ.zip

(14.94 KiB) Downloaded 103 times

[Derek_Stewart]

Hi,

I remember the article, but dud feel in a good position to comment or ask valid questions at the time. But I did have and use all the hardware Tony mentions in the Article.

I still have the Atari ST QL Mode 8 Emulator on a STFM and 2 Extended Mode 4 Emulator boards waiting to be installed into Atari computers I have. Which has cartridgd roms.

I was given beta test version of SMS2 to test, mainlly due having an Atari STE connected to a mono VGA monitor that converted to work the Atari. It did wotk nicely

Can Stella be released...