What is a bug, what is a bug report?
- From: rjf <fateman@xxxxxxxxxxxxxxx>
- Date: Sun, 27 Jan 2008 13:24:34 -0800
nano bagonghi wrote:
Wrong. It is an almost universal observation from people who receive bug reports professionally or as a courtesy as part of maintaining free software, that the vast majority of so-called bugs are simply "user error" and have almost no informational value whatsoever. The small bit of information they might contain is "you tried to make your program idiot-proof, but you must expand your definition of idiot."
A bug is a bug. I don't care who or how reports it.
Even little information is better than none.
An unchecked "bug report" sent to a few hundred people on sci.math.symbolic is simply a disservice to the community.
This is not VB's mode of operation, which sometimes is useful, but sometimes is merely determining how some meaningless piece of math is treated in different versions of software. A more sophisticated version of the child asking each of the math teachers in his school what happens when you divide by zero.
Getting back to the useless bug report, let me give you an example from the (now out-of-service) TILU, a Table of Integrals LookUp program which ran on-line for 5 or 6 years and collected hundreds of thousands of user-supplied inputs, and maybe a hundred bug reports. With a very few exceptions, the bug reports were user error, uninformative except to tell us that we should change the input parser so that whatever syntax some moron might use, we should try to figure it out.
In terms of real bugs, there may have been one or two actual mathematical errors from typos in the table, or some subtle pattern matching issues that were raised. Sometimes TILU gave an unwieldy answer, but correct. (The idea was that TILU would be an online service accessed by CAS, each of which would do simplification in its own way, not that any of them used it...)
So what kinds of things happened?
cos x was initially treated as cos*x , with integral cos*x^2. Surprisingly, many people accepted this as the answer; no one reported it as a bug!
Most of the valid "bugs" were items like "your server seems to be down now" (when it was -- usually something like a power failure in the building...)
More details of an early survey in an online paper http://citeseer.ist.psu.edu/416153.html
Microsoft deals with bug reports by telling people to subscribe to some newsgroup, and hope that some other person not paid by Microsoft will help "fix" the bug, or more likely point out the error in the complainer's way.
I have tried this method, and I suppose it works for sufficiently moronic questions and non-bug reports. (The typical advice starts with reboot your machine. It proceeds to .. reload your operating system.) It does not allow for the reporting or fixing of serious bugs, except by first convincing a battery of others that you actually DO know what a bug is, and you have isolated it. This is, as I see it, an excellent reason not to use Microsoft software for any technically sophisticated computing.
RJF
.
- Follow-Ups:
- Re: What is a bug, what is a bug report?
- From: nano bagonghi
- Re: What is a bug, what is a bug report?
- Prev by Date: Re: Wolfram Research QA process defect: Bug in Mathematica 6 - Integrate - 64 (Log, Cos, false convergence, multiple regression bug)
- Next by Date: Re: What is a bug, what is a bug report?
- Previous by thread: Re: Wolfram Research QA process defect: Bug in Mathematica 6 - Integrate - 64 (Log, Cos, false convergence, multiple regression bug)
- Next by thread: Re: What is a bug, what is a bug report?
- Index(es):
Relevant Pages
|
