Elfyn McBratney <[EMAIL PROTECTED]> tapota : > [Bringing this up here instead of re-opening the bug.] > > Hello, > > On Sat, Jul 24, 2004 at 11:58:23AM +0200, Mathieu Roy wrote: >> [...] >> Well, in some sense, it is not particularly problematic that users >> reports bugs directly to savane, even if they are due to a specific >> installation. > > I disagree, and I believe it is problematic. If I or another savannah > hacker hadn't had spotted this, we wouldn't have known about it, > regardless of it being fixed or not.
If a bug is specific to an installation, the bug is likely to be closed with a message saying it is related to a specific installation. It should lead the bug submitter to forward the issue to the people that run the installation (in fact, the submitter would be advised to do so if he wants his problem to be fixed). We cannot force people that install Savane to ask people to submit bugs directly on Savane tracker. But I have the strong feeling, and some experience in fact, that duplicates bugs trackers just makes harder to follow issues. More people you have in the middle, biggest chances you have a bug is forgot because one link forgot to do his job. One thing people that run Savane can do is to tell users to use their local administration support trackers in any case, and on a case per case basis forward items that looks like bug to our tracker. But they should be able to spend time forwarding bugs, etc. In this case, https://gna.org/bugs/?func=detailitem&item_id=555 the only strange thing is the fact the item is closed without notice: I did closed it while the exact same bug was closed on savannah with a meaningful message. I just did not duplicated the message. If I had to post a message, that would be "tell admins to upgrade their installation" :) In the end, I would say that there's no big deal having users reportings bugs directly to savane bug tracker. It is just like Debian users that report a bug in software directly to software developers. In some cases, it could be considered as nuisance; it could in other cases be consider as a gain of time. It depends mainly of developers opinion. In the current situation, I consider it to be perfectly fine. But indeed, it is partly a matter of taste. Well well... > If this bug were more specific i.e., to do with custom code, and the > bug had been closed, we wouldn't even know it existed. In fact, if the item had been tagged invalid, you would have known: the user would have be advised to submit the item to you. Because we would have made clear that the bug is not related to Savane. > So, I'd appreciate if someone could let us know if this happends again > and we don't reply to the bug (as we do when people report Savane bugs > to Savannah Administration). Hum, you're not dealing with *sigh*. I dont think we need to establish conventions to be sure we will not forget to reply to a dozen mail sent. There's no pope to beg to get followers taking care of your messages. :) Be sure that we would have warned you if we estimated it to be possibly relevant. It is not like there's was 1000 savane installations on earth and 1000000 bugs reports per day so we cannot forward a mail once, when appropriate. But in this case, I knew you (at least Sylvain; otherwise I dream awake) we're aware of the issue. Best Regards, -- Mathieu Roy +---------------------------------------------------------------------+ | General Homepage: http://yeupou.coleumes.org/ | | Computing Homepage: http://alberich.coleumes.org/ | | Not a native english speaker: | | http://stock.coleumes.org/doc.php?i=/misc-files/flawed-english | +---------------------------------------------------------------------+
