Quoting Axel Beckert (a...@debian.org): > Hi Christian, > > Christian PERRIER wrote: > > On Sunday, February 05, 2012 I sent you a notice announcing my intent to > > upload a > > NMU of your package to fix its pending l10n issues, > > That's wrong. On Sunday, February 05, 2012 you sent a notice > announcing that according to your plan the "Maintainer uploads the > package when possible".
Please notice that the mail you receive is sent from a standard template. When I'm working on l10n "NMUs", I start up to 3 packages every day. If you consider the various states where packages are (original intent, waiting for a few days to give maintainer some time to react, send NMU intent.....or call for translations in case the maintainer agreed for an NMU...or a call for translations in case the maintainer proposed to upload him|herself), I think I send mails to 5-6 different packages *every day*. As an example, I currently have 31 packages being worked on at the same time....each in a different state, each with a different maintainer, each with different maintenance habits and tools. For each of them, the maintainer's reaction varies quite a lot and I try to adapt communication as much as possible....but I can't rewrite each template manually for each package and exactly adapt to the circumstances of my communication with each maintainer. I try to do this as much as possible but sometimes I fail to do so because: - at the time I send my automated mail, I'm a bit in a hurry and don't check the nature of exchanges I had with the maintainer (that was the case for pconsole) - I never had any exchange with the maintainer - any other reason So, in some cases, the mail that's sent is not exactly adapted to what was discussed with the maintainer, particularly in cases like your package, where the discussion with the maintainer might have been slightly different from package to package. > > > after an initial notice sent on Saturday, February 04, 2012. > > In that notice there was an NMU intent, but no more in the second > notice. > > > We finally agreed that you would do the update yourself at the end of > > the l10n update round. > > No, we hadn't agreed on any date. You just imposed a date. Of course not, I can't impose anything on a maintainer, voyons! The proposal is "OK, let's do a translation update round and at the end of it, I give you a kind of GO sign". This mail is this GO sign. Of course (but I really though it was obvious...and it has been indeed obvious for the vast majority of maintainers up to now), you're free to upload *when you want* and when you can. I'm not in position to impose anything on you. > Additionally I wonder why this (annoyingly formal) reminder is > necessary at all even though everything is already prepared and all > l10n bug reports against pconsole are marked as pending. How can I know? Again, this formal reminder is more a way to tell you "OK, I monitored what translators were sending up to now, everything is OK, I even tested PO files, you can go ahead....and from now on, I will reduce my attention on your package". (indeed, I put the package in "wait" state on my radar which basically means that I wait for N days before looking at the package situation again....after these N days, I manually check the situation, I manually check how sensitive is the maintainer and sometimes I send a gentle reminder just to check that (s)he hasn't forgotten about my suggesiton to upload) As you see, I take a *lot* of care to try adapting to each and every maintainer's habits and, believe me, there are many. You think the template is too formal. Guess what? Some maintainers, in the past, though it wasn't formal enough (or too harsh, or whatever). For sure, being formal with someone I already know, with someone that prepared the one and only fondue we had at DebConf and with someone I had very good exchanges with, may sound a little bit too formal. Indeed, when I sent the last mail, I didn't even remember that "OK, pconsole is maintained by Axel, so I should maybe add a personal note to my formal template...." Apparently, this didn't fit exactly your own expectations but I think we reached the goal I had : deal with pending stuff. > Not sending that reminder would have spared some of your and my time > and we both would be in a better mood now... > > And if you hadn't predicted a Vietnamese patch in [0] (but none came), > I likely would have uploaded it already last night. > > [0] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=628418#25 Guess what? I'm not online now. Did you ever imagine that some people do not work online and can't use such information? Probably not. Exactly the same situation: you can't adapt your communication to the exact situation the person you're talking to is. Yes, I predicted a translation that didn't came in (I think from memory that it was Indonesian more than Vietnamese, because the Vietnamese main translator is currently very ill). Guess what: I'm not in translators' shoes, I do not control them. I though it would come because this translator is currently quite active and, apparently he could make it. That will be for another time. Maybe he'll even send his translation one day after you upload, that happens. How can he know? > > > To help you out, here's the patch which I would have used for an NMU. > > Why should I need help in form of a patch when everything is already > committed in git since yesterday[1]? How the fsck can I know? (now I'm angry, really) Do you really expect me, Axel, to know how each maintainer works? OK, *you* work as we expect many maintainers to work, you use a VCS, you commit things as they come and you don't need a patch. So, for $deity's sake, just ignore my patch and go ahead. Or do I need to reword this to something like "To help you out in case you need it, here is the patch I would have use myself. Please don't be angry if you committed everything in ${VCS} and don't need my little patch, it's just meant to help". > P.S.: In your notices there was never a timezone given, so giving > times like "midnight" is useless. Additionally "midnight of $day" > itself is ambiguous because it is unclear if you mean the night > between $day-1 and $day (i.e. 00:00) or the night between $day and > $day+1 (i.e. like 24:00). Using times like "13th of February 2012 at > 23:59:59 UTC" or similar would be unambiguous. If only I could be able to respect delays as exactly as this...:-) I should once show you how many packages I'm working on at the same time, really...:-) And to conclude this mail, I hope that my answer brought you more light about how my "little thing" is working and that the next package of yours I will fall upon, will not bring us into writing too grumpy mails..:-) And I hope I see you at DC12, of course!
signature.asc
Description: Digital signature
