23

In Stack Overflow I had a real quandary which I was unable to solve. I'm about at the level on the topic involved where I should be answering questions, not asking them.

This was a two processor system problem, and I had done everything I could. After asking the question on stack overflow, I discovered the next day that the other microprocessor had no code flashed in it (i.e., it was all FFFFFFh instructions; a chip full of NOPs).

So duh, what is the right thing to do now on Stack Overflow?

Should I remove that question? Five people tried to help, and one of them pointed out something good. Still, the question was totally bogus. It was hardware (well, hardware logistics).

Should I, on the other hand, answer my own question? The answer is: Make sure the code in the other microprocessor exists.

Anyway, I feel like the little boy who cried wolf, or chicken little (and the sky didn't fall).

I'm trying to build a reputation there, and I feel like that incident is opposite to my purposes of being here. Still, if others would benefit from the thing, I'm certainly willing to spend 10 minutes and write the answer. I mean, certainly, a design with two microprocessors sharing a wire between them is a common thing in embedded systems these days.

Here is the Question and subsequent answer

2013-FEB-15 UPDATE

It's not stupid or bogus now. The absence of code in the other microprocessor should not affect the TX side of the UART. It should just send bytes out onto a dead wire; something that is okay (useless, but still okay). Now it's a real question, with a real problem. Thanks for the feedback here; both positive and negative.

Current thoughts...

  • Stack Overflow question will stay there. If I get through it, I should answer it myself for others (including me) later to understand it.
  • Hitting every support board (including chip maker's pro support; we're paying them, I think directly, for their chips).

2013-FEB-19 UPDATE

It was very annoying, but I found out exactly what was wrong. There was a lot more going on than just a chip full of NOPs. Can't believe I missed it, but anybody could make the same mistake easily, so, I'm taking the majority opinion that I read here and I will answer it myself. Thank you one and all for the valuable feedback.

Improve this question
4
  • 2
    I'm baffled why there's no link to the question.
    – Dave Newton
    Feb 20, 2013 at 2:56
  • 1
    Okay, do I post it here in the comments or edit my original post again ? I want to do my best to build my reputation and not get the wrong kind of votes, and I don't know what the rules are; something about a secret algorithm or whatever. If I edit my original post too much, will that get me bad points in the algorithm ?
    – User.1
    Feb 20, 2013 at 3:10
  • I don't really care. But no, editing to add useful information doesn't do anything bad, although if it's edited too many times it'll end up as Community Wiki.
    – Dave Newton
    Feb 20, 2013 at 3:46
  • Okay, edited original post. The link is in there now.
    – User.1
    Feb 21, 2013 at 4:49

3 Answers 3

Reset to default
20

Post your own answer explaining what the problem was.

You never know it might help someone else who's experiencing the same symptoms and would never think to check the hardware.

Improve this answer
4
  • 4
    Really? It's more likely to act as a red herring for people experiencing the symptom for a "valid" reason, and it's unlikely that other people whose problem is "haven't flashed the code" will have the exact same symptom. IMHO, it's the canonical "too localized".
    – Oliver Charlesworth
    Feb 12, 2013 at 21:52
  • 2
    @OliCharlesworth - As long as the question is clearly about embedded systems then I don't see it being too localised.
    – ChrisF Mod
    Feb 12, 2013 at 21:58
  • 2
    IMO, it's analogous to the problem being "forgetting to recompile" or "running in the wrong directory" or "haven't turned the power on". The symptom and the solution are essentially unrelated...
    – Oliver Charlesworth
    Feb 12, 2013 at 22:18
  • 1
    @OliCharlesworth, it doesn't have to be 'the' answer, but it definitely is 'an' answer.
    – Benjol
    Feb 20, 2013 at 7:14
5

That is the kind of question that I would vote to close as "too localized", for two reasons:

  1. The question may act as a red herring for people experiencing the same symptom for a "valid" reason.

  2. It's unlikely that people whose problem is "haven't flashed the code" will have a similar symptom.

Improve this answer
8
  • 6
    The questioner acknowledges his post was totally bogus. Encouraging him to delete it does not look like the right course of action, given our current question-ban algorithm. Better let him self-answer his question, and the community decide to close it or not.
    – Frédéric Hamidi
    Feb 12, 2013 at 21:59
  • Oh, clarify: the problem was a UART that wouldn't empty.
    – User.1
    Feb 12, 2013 at 21:59
  • @FrédéricHamidi: I'm not familiar with that; what's the relationship between self-deletion and question-bans?
    – Oliver Charlesworth
    Feb 12, 2013 at 22:15
  • @OliCharlesworth - deleted, down-voted question contribute towards the question ban.
    – ChrisF Mod
    Feb 12, 2013 at 22:18
  • @Oli, I can only refer you to the canonical resource about this topic. The details of the algorithm are kept secret, but deleted posts are explicitly mentioned as counting towards a ban, so encouraging deletion can have unforeseen consequences if the user is "on the edge".
    – Frédéric Hamidi
    Feb 12, 2013 at 22:18
  • @FrédéricHamidi: Thanks. In that case, it seems the solution is that the SO community should be close-voting the question (although there's not a great deal the OP can do to encourage that!)
    – Oliver Charlesworth
    Feb 12, 2013 at 22:21
  • @Oli, if I recall correctly, he can vote to close his own question. But he may as well answer it and be on his merry way.
    – Frédéric Hamidi
    Feb 12, 2013 at 22:24
  • 1
    @FrédéricHamidi: the q-ban criteria only matters if you're at least somewhat close to the edge (the OP is not). Encouraging him to delete a dozen questions would be very bad; one doesn't matter. Not that he can delete it anyway, given it has answers.
    – Shog9
    Feb 20, 2013 at 2:55
1

It's possible that the question should be closed, as unlikely to help others (if you think no-one else would ever make the same mistake - but they might.) Once it has an answer it won't be deleted. And this is a good thing because new users should avoid having ANY deleted posts at all. Some people are finding themselves question banned with as little as two deleted posts if the others have poor scores.

Once you've answered it (with a real Answer, which you can accept, not by editing the question to say "never mind, i'm an idiot) other people may vote to close it. Don't worry if they do, it won't hurt you.

Improve this answer

You must log in to answer this question.

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