Re: bug reporting workflow is outdated
schrieb Pedro Larroy am 2011-05-22 22:44:
> I think expecting having a working smtp on laptops, workstations, etc,
> is unreasonable these days.
> I suggest that we can make an HTTP based bug reporting method.
>From my experience as a occasional bug reporter, some thoughts that came
to my mind:
It's easier to get the mail setup wrong than to get it right. The
current procedure assumes you have a working mail configuration at the
reporting machine. That assumption is too strong. I need to authenticate
to my mail server, and from time to time I'm in a different network that
needs a changed configuration that I have to get right before using
reportbug should default to the worst case, that is: "I do not have a
clue how the message will get sent to BTS, just Do The Right Thing(tm)".
If someone knows that his/her mail config is working it should be an
opt-in to use that. Sysamins or DD/DMs will likely know what to do,
others should not need to care.
> Will reportbug often have direct Internet access? (You should answer yes to
> this question unless you know what you are doing and plan to check whether
> duplicate reports have been filed via some other channel.) [Y|n|q|?]?
> Do you have a "mail transport agent" (MTA) like Exim, Postfix or SSMTP
> configured on this computer to send mail to the Internet? [Y|n|q|?]?
This should default to No.
bug filed ;-)
If sending a report fails, the message is saved to /tmp, and no advice
is given how to send the message by other means. I'm not afraid of
sendmail-ish programs, but I guess most of the non-developer users even
don't know what it is, or what to do with the message draft.
I understand Pedro's suggestion to implement an additional transport
from bugreport to BTS. I do not see a big difference to the mail
What is the advantage of having a mail-only BTS reporting mechanism? One
thing would be to have a quite reliably working mail adress of the
reporter. There is no difference to a HTTP base reporting system. One
can easily get the mail config wrong, intentionally or by accident. In
fact, reportbug asks you for your name and mail address. If you really
want to be shure you need a three way handshake to verify the mail
adress, either way.
One thing I can think of is spam. How is spam handled with mail
currently? What would be the difference to HTTP?
Finally, BTS has a web frontend to read bug report, which I assume is
the preferred way of access. Why prohibit this way of transport for the
other direction? It does not need to be a web frontend for reporting or
replacing reportbug as preferred way to report bugs, just HTTP transport.
No one would think if reportbug to rely on mail transport for getting
the bug list...
Engineers motto: cheap, good, fast: choose any two
Patrick Strasser <patrick dot strasser at student dot tugraz dot at>
Student of Telemati_cs_, Techn. University Graz, Austria