Le April 30, 2008 03:32:10 pm Thijs Kinkhorst, vous avez écrit : > On Saturday 5 April 2008 09:32, Filipus Klutiero wrote: > > Package: signing-party > > Version: 1.0-1 > > Severity: important > > > > I never used springgraph before, but it looks like its invocation syntax > > changed. It seems it now ignores file arguments and only uses stdin and > > stdout. The manpage wasn't updated, and -h doesn't document what > > springgraph does. I must have taken some 20 minutes to realize it wasn't > > hanged but waiting for EOF. > > I do not follow a bit of what you mean. > > The man page has the following synopsis: > SYNOPSIS > springgraph [OPTIONS] < input-file > output-file.png > which I think is quite clear about that it uses standard in and out. Ah, OK. I red "springgraph [...] < input-file > [...]" as "springgraph [...] <input-file> [...]". Thanks and sorry about that. It's just that normal manpages don't mention specific ways to invoke the program, so I was confused.
This could be fixed by adding the general syntax springgraph [OPTIONS] and, if you wish, other possible combinations. But I think this is superfluous, just the general syntax should be fine. However, the manpage needs to explain where springgraph reads and writes. Alternatively, the program itself could be changed to handle arguments properly (like, say, grep and iconv). As I understand it, springgraph simply flushes arguments, which doesn't help to reduce confusion. > > The -h text, although not specifying how to supply input to the program, > doesn't claim that springgraph accepts filename arguments. Right, and I didn't write that. Thanks for fixing. > > You conclude that its invocation changed yet you've never used it before > (???) and nothing in the help or manpage claims that you can pass filename > arguments. Not exactly, as I wrote, it *looked like* the syntax had changes. But yes, as I explained above, I was confused by the manpage. > > You file this as severity important. Right, how is you not reading the man > page, or the help text not explicitly stating how to pass input, in any > sense an important bug? I wouldn't have written that the manpage wasn't updated if I hadn't checked it. Given your clarification, I do not think anymore that this is an important bug. I don't think I would be alone to be confused, but I suppose not everyone will be. IMO this is a normal bug. > > Please don't bother us again before at least trying to verify the bug you > want to file and reading up on bug severities. Thanks for your effort > though. Please don't bother bug reporters with such requests again before at least trying to verify that your software is not responsible for their confusion and reading on courtesy. But thanks for your work anyway (honestly). -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
