We are running a survey to help us improve the experience for all of our members. If you see the survey appear, please take the time to tell us about your experience if you can.
Hi,
I have this design for a controller but i dont know the program for it?
Where can I find the code for it? I tried google coding but cannot find it. I normally write Pascal(Delphi) but this needs to be in Keil C for a 8051.
I need to get it working quickly. Who will help?
Unfortunately, the ones most likely to leave it on the ground are usually the ones who need it most...
:-(
I'm glad I helped the community by starting such a productive thread ;)
Andy,
Do sardines have hands? (I've noticed very little fish activity: had he been terminally 'smoked' or not?)
--Cpt. Vince Foster 2nd Cannon Place Fort Marcy Park, VA
The one place where you might truly get perpetual "motion"...
(or, at least, going around in circles forever)
How nice of you to think of me, Vince. I'm afraid I don't tend to bother reading your longer posts as they have a rather poor ramble-to-content ratio, but I do notice that your fantasy world is coming along nicely.
Keep up the good work!
Ouch, that hurt.
I'll agree the ratio isn't stellar, but I'm nearly complete with my world. I just need to add in some final touches, and I'll be set.
And of course I think of you: I'm sure we all do.
good work? Its "bullet-proof" (at least in my world it is)
How nice of you to think of me, Vince. I'm afraid I don't tend to bother reading your longer posts as they have a rather poor ramble-to-content ratio,
Ho, yeah. Maybe you should - you may learn something (I do).
One big problem with quality and reliability is that the people requesting the product so often don't know what they want, and often don't want to make decisions or spend time finding someone else to take the decisions.
- What should I do if x happens? - Do what you feel is best...
It is so common that the specifications contains the main requirements, but never ever once mentions any error handling. What to do if there is no phone number? No phone line? Phone line but no dial tone? The line is busy? No answer? Call is forwarded multiple times? Number of retry attempts? Delay between attempts? The supply voltage is dropping during the call? There is an error updating an EEPROM with current state? What restart state to use in case of unexpected watchdog reset? What abilities for a service technician to deduce an installation error? Regulatory requirements for permanent logs? ... In the end, even trivial things like having an equipment make a call quickly results in huge lists of "what if".
As a developer, you would like the bulk of the requirements to specify all these expected or unexpected cases. How can it be possible to do tests against a test specification if 95% of the conditionals in the code isn't anchored to any requirement? Getting the customer or project manager to back-patch the specification to ten times the original volume?
One product I worked with was a complete "regeneration" of an older product. The hardware was obsolete, the software written without a real specification, the developer just out-of-school and with hardly any management or other developers to request feedback from. In short, a good case for restarting from zero. I did track all conditional decisions in the sw against a requirements specification - posting additions where specification leaks was found. A couple of months later, I found that I wasn't in the send list for specification updates - it was management only. I got one update. Management got five. And the specifications (that I didn't get) had several times been totally renumbered and without any change history. I should have seen this comming since the time plan (which I got to see about 50% into the project) specified that the test specification should be written in a single week, starting after the development phase was done. By me. I can't see DO-178B accept that the developer ends the delivery cycle by writing the test specification for qualifying his own code. After first having realized that the specification wasn't even valid.
The above was basically caused by a single individual without any experience with software development at a company I worked for. But I expect that most people have seen management problems now and then in both big and small companies. Don't accept them. You will neither be productive nor feel satisfied with your work.
I prefer to have the design rights to figure out the best way to solve a problem, but there is no joy if I don't from the start know what is required to get acceptance on the final product. And throwing away tested and documented code because the requirements specification was written on a whim is no fun.
I think its time to bring an end to this thread.
Most of you have provided useful information.
Well done!
Only Keil staff or time (that perpetual enemy to man) can put a stop to a thread living its own life.
As a spin-off from previous discussions, these photos has got me inspired to take a closer look at some microsecond moments:
www.telegraph.co.uk/.../High-speed-photographs-by-Alan-Sailer-capture-the-moment-a-pellet-fired-from-an-air-rifle-hits-an-object.html
Per,
What incredible pictures. Thanks for sharing the link.
David.
Great pictures.
Care to give an example of something you've learned from Vince's posts in this thread?
Well,
I learnt that Vince's wife is not a good flyer.
Regarding his posts ...
Usefullness score: 0% Readability score: 5% Ramble score: 95% Confusion score: 75% Soporific score: 100%
That said, this forum would be missing something without his input.
It is wonderful indeed to see that you so vociferously regard Vince's wife a bad flier. Well, maybe you should look better between the lines, Sausage man.