[perl #41601] [BUG] parrotbug b0rken
Klaas-Jan Stol
# Please include the string: [perl #41601]
# in the subject line of all future correspondence about this issue.
# <URL: http://rt.perl.org/rt3/Ticket/Display.html?id=41601 >
hi,
parrotbug seems to be broken on win32.
regards,
klaas-jan
Paul Cochrane
Actually, I think it's broken on all platforms. I tried using it
several months ago (on linux x86) to submit a bug report and that
didn't work either. When people say "use parrotbug to submit the
report" they usually mean "post to parr...@perl.org".
Since there is confusion between the parrotbug program and the
parrotbug list, and that the parrotbug program isn't used very much
(if at all) is it worthwhile keeping it around? Me might be better
off without it. Just a thought...
Paul
Jerome Quelin
> >parrotbug seems to be broken on win32.
> Actually, I think it's broken on all platforms.
as parrotbug (the program) father, i'd like to add that i don't have
time to work on it. (and it seems to show up)
anyone wanting to work on it or even adopt it would be welcome.
regards,
jérôme
--
jqu...@gmail.com
Paul Cochrane via RT
> parrotbug seems to be broken on win32.
What output are you getting (if at all)?
Paul Cochrane via RT
messages. Since one needs to be online to send messages (if and when
it works) why not let people just use parr...@parrotcode.org? Then
we can reserve the parrotbug program for generating bug reports offline
for later reporting via parrotbug@parrotcode. I.e. is it ok for me to
remove the send-capability of parrotbug and therefore try to fix the
b0rkenness?
Comments most definitely welcome!
Paul
Paul Cochrane
> the output of perl -V, which is often invaluable in categorising or
> identifying the cause of bugs. e-mail to per...@perl.org doesn't always
> have this.
>
> perlbug has the option to save the bug report to a file, which can then
> be mailed later. (Or, immediately copied to a machine that does have
> outgoing mail enabled, or won't reveal your work e-mail address, or whatever)
parrotbug also has the save option, and it saves a lot of
configuration information. I'm asking if the "send" feature can be
removed, which is the bit which seems to be broken. We could always
revisit having a "send" feature from parrotbug, but it seems somehow
redundant when we have parrotbug@parrotcode...
Paul
Nicholas Clark
In Perl 5 land, the advantage of the perlbug program is that it captures
the output of perl -V, which is often invaluable in categorising or
identifying the cause of bugs. e-mail to per...@perl.org doesn't always
have this.
perlbug has the option to save the bug report to a file, which can then
be mailed later. (Or, immediately copied to a machine that does have
outgoing mail enabled, or won't reveal your work e-mail address, or whatever)
Nicholas Clark
Nicholas Clark
> parrotbug also has the save option, and it saves a lot of
> configuration information. I'm asking if the "send" feature can be
> removed, which is the bit which seems to be broken. We could always
> revisit having a "send" feature from parrotbug, but it seems somehow
> redundant when we have parrotbug@parrotcode...
My feeling on perlbug is that the "send" feature is applied laziness - bug
reports may well be more likely if the extra step to send the e-mail is
eliminated, even if the e-mail address is clearly advertised.
But this isn't my call here.
Nicholas Clark
Jerome Quelin
when i wrote parrotbug, i mimicked a lot perlbug.
and iirc, there's an option --send to send the report. and if no option
provided, it will prompt you on what you want to do next...
so, if sending the report fails, maybe parrotbug should default to
dumping the report instead, and warn the user saying the report has not
been send, and is saved in file /foo/bar, charge to her to send it to
the correct email address.
once again, i don't have time to fix it. feel free to adopt it.
jérôme
--
jqu...@gmail.com
Paul Cochrane
> when i wrote parrotbug, i mimicked a lot perlbug.
> and iirc, there's an option --send to send the report. and if no option
> provided, it will prompt you on what you want to do next...
>
> so, if sending the report fails, maybe parrotbug should default to
> dumping the report instead, and warn the user saying the report has not
> been send, and is saved in file /foo/bar, charge to her to send it to
> the correct email address.
has been sent, when in fact it didn't. Perhaps sending should *also*
save, and make a comment that if the send didn't work, attach the
report to an email in a standard email client. Does that do what we
want?
> once again, i don't have time to fix it. feel free to adopt it.
Thanks for your comments Jerome, I'm trying to adopt it ;-)
Paul
Chromatic
Works for me. Opening a pipe to Sendmail isn't exactly Windows friendly, and
I don't believe we've required Mail::Send elsewhere for Parrot.
-- c
James Keenan via RT
ideally).
Is that so? Can we close this RT?
thanks
kid51
Will Coleda
were present. (Right now it fails, though in a somewhat graceful
manner.)
--
Will "Coke" Coleda
Will Coleda via RT
I installed Mail::Send in strawberry perl and tried to use parrotbug to send a bug. It reported
success, but I fail to see either of the two test messages I sent. If they don't show up by
tomorrow, I think w can consider this borked on windows, and perhaps update parrotbug to
always take the failure route on windows and just let the user mail their own report after
saving it to disk for them.
--
Will "Coke" Coleda