Ninety minutes ago, I asked this question here on Stack Overflow

I have just recently acquired some new evidence on the question. Not only is the UART giving me bogus data out the port, but it is re-interrupting me, and I believe it's continual; like recursion. This is a new chip for me and I didn't write the code that I'm debugging.

So now I have this very useful new data to show the other guy. It is probably related to the bogus data problem. I have trapped a _general_exception_error and pinpointed it to a bus exception, and I found the exact spot where it's happening, and yes, it's fitting together now.

Still, I don't know exactly what to do to correct the problem. What's more, the other guy's previous answer was so close that I can predict that if I showed him the rest of this evidence, he would instruct me as to the exact cause and most probably a simple fix for it.

So, do I now...

  • Delete my original question and ask it a second time as a new question with this new supporting detective work?
  • Edit the original question to add in the newly discovered debugging evidence?
  • Leave this question as is and ask a second separate question with the newly enhanced extra evidence?
  • Something else?

For that matter, does this kind of stuff extend beyond the intent and design of stackoverflow.com? It's still one question. I've just discovered a second facet which is probably affecting the observed behavior. I believe that the threaded discussion thing is what I'm trying to avoid, while I want to solve this problem.

Just trying to do the right thing here, while hopefully making a buck for the boss.


2013-FEB-24 00:51 (UTC)

I have just learned that the PIC32 has a bit in a register called U1RXIF which appears to be extremely similar to the bit that I'm using in the PIC24 to stop this exact same behavior; i.e., if I leave that bit set, I'll get the exact same thing on my side (I know this, because I debugged this precise behavior when I was writing the code on my side).

In true Microchip Technology fashion, that have it mentioned in the family data sheet, but not explicitly documented it in the PIC32 manual the same way as the PIC24.

Somebody on planet earth (probably here on Stack Overflow) must know how I clear that bit.

I think I know what to do now, but I've never actually done exactly this on this exact chip, so I would really like to get some input from someone who has done exactly this on exactly this chip. I'm probably right, but you know how that goes.

Do I edit the Stack Overflow message again??? Would this be creating a chameleon question?

Nerd psychology: I'm getting excited about tracking down this bug and finding an answer. I do not want my excitement to trump the protocol, and decorum that this place has established; evidently, for good reason.

Amidst all this discovery, the single question remains: How do I stop the recurring ints?

So is this a chameleon question?

  • first thing is to make sure you aren't making is a “chameleon question” – gnat Feb 23 '13 at 6:00
  • @gnat I was unaware of the concept. I have edited the question with the new debug evidence. I put comments after other members' comments. I hope that this is not creating a chameleon question. Hey, anybody reading, please give me feedback, is my question still clear and focused ? It is supposed to be: How do I stop that UART from generating infinitely recurring interrupts on the PIC32 ? There's a bit somewhere that stops that, I am so confident. Now, to figure out which bit that is. – User.1 Feb 24 '13 at 0:18

What I would do is edit the question in a way that it is totally clear that this is new input to the problem. You do not remove old inputs but only add to get more precise answers toward the resolution of your problem and it regroup all the info for future users having the same problem.

Also, as @Andrew Barber pointed out, it would be a good Idea to add a comments to existings answer, explaining that the question has been modified. This way it gives answerers a notifcation to consider modifying their answers. Also, it notifies the future users coming to the question that the question has been modified and that some answer might be outdated, and not directly related to the latest version of the question. This way, they won't downvote the answerer because they think the answerer misunderstood the question.

  • 3
    I would also consider adding comments to the existing answers, explaining that you modified the question somewhat. This gives the answerers a notification that they might want to modify their answers to match, and also notifies other readers that the question changed a little, so they don't down vote because they think the answerer misunderstood the question. – Andrew Barber Feb 22 '13 at 23:46
  • That is totally a good idea! – Hugo Dozois Feb 22 '13 at 23:49
  • Feel free to add it to your answer :) – Andrew Barber Feb 22 '13 at 23:49
  • @AndrewBarber just did exactly that, I think/hope. I welcome feedback from you or anyone here on whither or not I'm doing this right. I want to get this app working properly. p.s., I did not write the code I'm debugging. – User.1 Feb 24 '13 at 0:19

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .