Ville Skyttä writes:
Here's why I think it doesn't really matter in this case,
What you've explained is why it's hard to characterize. It still
matters, and bug reports should at least state "The auto-generated
report lies about the build environment because I don't have access to
it, but the following information is correct: ...".
has. Even people who for some reason run Fedora development (I
don't, and unless Jerry does, I bet nobody else here does) would
only try with their environment du jour.
That's OK. If their environment is different and they can/can't
reproduce, that tells us something important. If an XEmacs dev with
any environment *can* reproduce, then we've won. The closer the
environment, the more likely we are to reproduce.
And FWIW, this isn't something that occurs only on one
system. There's at least two of them, and given what I guess is
the popularity of XEmacs in Fedora at the moment, I think with two
active bug reporters who run into the exact same issue at the same
time, it can be fairly safely assumed that it crashes for everyone
running Fedora development currently.
Then there are two people who can provide the information I requested!
I thought [something] was obvious from the original bug report.
Maybe it is, but given the demands on my time (and the general
scarcity of XEmacs developer time) I'm going to offload as much effort
on reporters as possible. The alternative is "file and forget," which
I'd rather avoid. (NB. I do most of my XEmacs work from home over SSH
in a TTY gnuclient. I can't click-through-to-bug-report, it requires
mousing around and an application switch.)
> If the user wants to ship me his environment, machine and all,
> something can be done,
Come on, this is just silly.
Why do you think so? That is the requirement for effective debugging
until somebody who is an XEmacs contributor willing to debug can
reproduce. Nor is it technically impractical: he can ship it to me
virtually via SSH. I have never done that, but I believe Martin has
debugged a number of problems that way.
> but without a way to reproduce locally, I can't do anything
Nobody is demanding you (or anyone for that matter) to do anything.
But I *want* to do something. This is a crash, and you say it's
common on Fedora, at least in the development version(s).
Starting ranting about gcc and/or X instability or (even jokingly)
asking physical machines to be shipped around just diverts people's
attention elsewhere. At least it does that for me
Would it be preferable simply to put it in terms that "without this
information I will leave it to someone else to debug"? That doesn't
seem very user-friendly to me.
XEmacs-Beta mailing list