Not to mention that if there were a bug the maintainer's going to need a bit more detail than "placeholders don't work" in order to reproduce it at the least (and preferably a short snippet, along with any supporting structure, that reliably reproduces the problem).
| [reply] |
| [reply] |
The problem is a large number of times someone inexperienced (or clueless as in this case) finds a "bug", it's not (see dominus' File of Good Advice, specifically #11907 and #11943).
- They've missed part of the documentation
- They're making a naïve presumption on how something works (witness the reports that floating point math "doesn't work" that not infrequently pop up here and on p5p)
- They're using the wrong part of an API, either through ignorance or crufty design ("Oh no, you want keys %h in scalar context to get the number of elements in %h.")
- They're just doing something plain wrong
Even though the odds that they've found a real bug are stacked against them, they're going to need to come up with a decent bug report at some point. It can't hurt to put all the information out in front of a large number of eyeballs. The benefits of doing that here (or on a mailing list, or some other public forum) include:
- If it is one of those things I mentioned above, someone's very likely to point out what the problem is fairly quickly ("No, IEEE floats don't work like that."; "It says in the README that it won't work for nonaffine vreemflitzer transforms.")
- that public explanation may stop n other people from submitting the same "bug" m months down the road when they encounter the same thing
- someone may have a better explanation and come up with a documentation patch that clarifies whatever ambiguity misled them (granted that's useless with people who don't seem to bother reading documentation even when it's spoon fed to them, but . . .)
No this isn't the place to send bug reports, but it's not a bad place to get them in front of some other knowledgeable eyeballs before you really do send one.
| [reply]
[d/l]
[select] |
