How does the forum feel about being asked to debug program code?

Question

In the last few days, we have had 3 different questions which have the basic structure of "please debug this program for me", followed by a lot of code. In one case over 600 lines, in another one over 460 lines. The third case mercifully said that the code was too long to post.

Are such questions on topic?

We've probably had good questions in the past about the meaning of short code sequences, but this seems to be a new and recent thing. It's not the same as asking "how do I do $WHATEVER", it's being asked to look at actual code that purports to do it but fails, or to explain how the program works.

Answer 1

I assume it's these two question you've referring to:

• Flickering screen when pressing 'S' key in 8060 assembly program
• What does this triangle rasterization code do? What does it use a “depth cueing table” for?

While the first one can be seen with some good will as retro by being a 16 bit COM program, it's still a generic debug my program question; the second in turn completely leaves this territory by being a high level, machine independent algorithm coded in C. It's pure CS theory and as far from retro as it gets.

---

I would agree that the core component of this question is already covered in several answers to the question: Are x86 Programming Questions on Topic. Maybe a bit indirect. So it may help to point it out more directly:

RC.SE is not a debugging service.

The very same way as RC.SE is not a homework service.

There will always be a class of questions where the core is hard tied to a classic hardware, software or combination thereof. And those are not only fine, but very welcome. So we need to filter and protect them.

I do not think that size or language does really make a point for on/off-topicness. What does is how the problem approached and presented: I would expect the problem being analysed and broken down to a level that excludes generic CS and obvious debugging issues and presents a state the one asking can not explain, because the asked for component is not obvious using today's common knowledge (*1).

Using such a common sense classification could easy rule out all four of the questions mentioned. Fair rules/judgement are rarely about form or size but need o look at content and meaning.

---

*1 - This also means showing basic research, like what an address on the CPU programmed is, or that Assembler really executes straight ahead ...

Answer 2

Yes they are on topic, otherwise we have to try to define a maximum number of lines of code we are willing to accept.

Clearly, ten lines of Z80 assembler is going to be no problem, whereas 300 lines of COBOL is no fun at all!

I propose people upvote questions they like, downvote questions they don't like and let the wisdom of the crowd be the judge.