Re: Disobeying jet engines - why?

John Larkin wrote:
On Mon, 28 Jan 2008 10:41:41 +1300, Terry Given <my_name@xxxxxxxx>
wrote:


John Larkin wrote:

On Sun, 27 Jan 2008 22:44:59 +1300, Terry Given <my_name@xxxxxxxx>
wrote:



John Larkin wrote:


On Fri, 25 Jan 2008 15:38:15 +0000, Martin Brown
<|||newspam|||@nezumi.demon.co.uk> wrote:





I do embedded realtime apps that have no OS. I've written a few
RTOS's, but haven't needed to use one in a long time.

If the code is...

START: do thing1
do thing2
do thing3
goto START

where is the OS?

The OS is the bit that should step in when one of thing1, thing2 or thing3 fails to complete. Most non-trivial embedded applications have some kind of RT kernel underpinning the allocation of resources.


I doubt it. You can do arbitrarily non-trivial apps with just state
machines running in a loop. The only RTOS would then be a hardware
watchdog timer.

The code blocks can themselves use whatever resources they want,
assuming they are neither hostile nor stupid. The rare
allocation-of-resources situation can be handled by code blocks
locking out interrupts, or with simple flags.

If ever you assign a programmer to do a deep-embedded app, and he says
"first, we'll have to write/buy an rtos", fire him instantly.


John

LOL! I had a programmer write code for a user interface module. 4x20 LCD, 3 LEDs, 5 buttons and a MODBUS interface (it was a master). Used a 8951 with 64kb flash and 4kb SRAM. He wrote an RTOS, which wasnt a bad start, although not really necessary. but ran out of RAM, and used most of the FLASH, which was supposed to store drive config data - I expected the code to take 4kb max.

turns out his C compiler was using 32-bit numbers for BOOLEANS! and it got a lot worse. It was a really fancy RTOS, multi-threaded & multi-tasking. And got binned. I learned a lot about writing clear specs

from that debacle. The replacement programmer got all the code in 8kb


including the GNU real-time debugger. and used a fraction of the RAM.

Cheers
Terry


When I was young and naiive I hired a guy to do a fairly simple
parts-list thing. He said "if first I write a database manager, your
probem will be trivial and just take a day more." He fiddled with the
db manager for months and never finished anything. So it dawned on me
that he didn't ever give a damn about my problem, he just wanted to
fiddle with databases. Similarly, there are lots of programmers who
aren't interested in your dinky buttons and LCD's, they really want to
get into context switching and schedulers and dynamic memory
assignment algorithms.

Fire them.

John


we did exactly that. I kept a copy of his RTOS, as it was kinda sexy, but gross overkill. his previous job was coding for expensive laser copiers, where he had Mb of RAM.....

the moral of the story is: never hire computer scientists.

Cheers
Terry



As Phil Hobbs has noted, and field of study that includes the word
"science" in its title isn't one.

John

Indeed. And any computing project with the word "integrated" in its title is doomed to be a costly failure.

Cheers
Terry
.

Quantcast