View Single Post
Old 19 October 2016, 15:28   #181
Olaf Barthel
Registered User
Join Date: Aug 2010
Location: Lehrte, Germany
Posts: 157
Originally Posted by Locutus View Post
Ho, ho, i think you misunderstood :-)

I was not saying you should compare AmigaOS in its application area to VxWorks or so. We can all agree that this whole 'AmigaOS for embedded!1' thing people whine about is stupid.

What i meant is development methods.
And here I was, naïvely hoping that just for once, I could finally have nice things for a change

That said, having gone through the hard knocks school of writing 'C' for the Amiga (started in 1988, if I remember correctly) has served me well in all these years.
Yes, i've been paid to port software off of or 'blackbox' even more arcane operating systems then AmigaOS :-D

I fear you will be working with a lot of external harness constructs to facilitate you're tests. Might not be as clean but you have to do with what you have to do.
So it is just as I had expected: the more abstractions the implementation language provides, the more leverage it gives you to perform testing. If you are that close to the CPU with assembly language and 'C', the more primitive your testing methods and tools need to be.

Oh well, it's not as if we haven't been there before.

Well, from a modern perspective the codebase is really bad, not just the various build systems but also a lot of code quality is questionable. The A4091 device driver code scared me.
You should be scared. Now you know why it's in there, and you shouldn't have looked. What has been seen can never be unseen.

This particular SCSI driver is a combination of the unified scsi.device build, which encompasses ST/XT/IDE/WD/LSI hardware, and the rather too complex LSI Logic hardware itself. This SCSI hardware is programmable, and the A4091 device downloads a control script to it during initialization. This is some extraordinarily complex code. The author, Randell Jesup, sat on the ANSI SCSI committee for Commodore, if I remember correctly.

Which is of course one of the core problems, in essence when you look through the RCS data it becomes very obvious how the OS development was a set of very specific personal kingdoms, inter module cooperation is very low. ( a fun paper on this subject).
I do not know how much of the change history you looked through (something more which, once seen, can never be unseen, I suppose).

From my point of view it was possible to see how many hats the different developers used to wear at the time the original Kickstart 1.x was built. You could also see how much interaction there was between the engineers who designed the low level drivers and .resources. This looked like the work of a very small, dedicated team to me. I have worked in, and with teams like that. It looked familiar to me.

What neither of us can see is the history beyond the 1985 starting point of the RCS history. It's possible that such a history exists, maybe in the form of SCCS files. The collaboration between Dale Luck and RJ Mical was said to be very close when graphics.library and intuition.library were "cast". This cannot be seen in the RCS change history, which seems to begin right when Commodore acquired Amiga, Inc.
Olaf Barthel is offline  
AdSense AdSense  
Page generated in 0.04034 seconds with 10 queries