2015-09-10 4:42 GMT+01:00 <[email protected]>: > MAFM> Anyway, in the end I changed it only slightly, "aptitude -v update": > > How about for plain aptitude, no -v. > I (happily) remember 10 years ago there was no such message.
As I see it, the problem is that if we remove the message after being there several years, we will also have bug reports asking it to put it back by default because "it was very useful, and typing an extra '-v' is a hassle". For example, in #458351 you were asking for an option to hide this, not to not show it by default, and there already is one -- '-q'. (But then it appears an extra "Reading package lists..." which doesn't appear by default or with verbose modes). > Isn't it bad to output something the user doesn't understand even if > they didn't use -v? I disagree with users not understanding that, or that being a big problem. The first message, "There are 0 new [-1]", is specially cryptic, and I agree that "[-1]" can be taken for an error. But if one is used to aptitude, and apt and Debian terminology in general, talking about "new", "upgrades" or "upgradable", and "obsoletes" should not be too alien. In the curses interface there are separate groups for those in the default view, and "broken" is mentioned all the time (if there is something broken, but if not that element doesn't appear in the status line anyway). In fact if the user attempts to upgrade after such message, the BROKEN message will appear listing the packages, as they will appear "upgradable but not upgraded", etc. So in context, if after issuing an "aptitude update" when using unstable/testing, or stable point or security updates, with the very intention to see if there are new or updated packages, I see this : ... Fetched 617 kB in 12s (51.3 kB/s) Current status: 287 (+4) updates, 4629 (-3) new. ...it is not a big stretch to imagine that there are 4 new updates available, and a total of 287 not updated (because I know that I haven't upgraded many in a while), and the other "new" is the huge number of packages that I know that I have not reviewed for months, so I didn't issue a "forget new" for ages. Similar for "full-upgrade" or other commands where it apepars. The result of seeing those numbers in the "current status stats", it is action-reaction, precisely why I issued the "aptitude update" command -- to see if there's anything new or updated to upgrade to. I can see how one cannot realise immediately about this the first time depending on the circumstances, but I cannot see how the numbers might be misinterpreted in that context long term, or cannot be seen as an effect of your actions. Even if one doesn't realise immedialy, and if the numbers are low (e.g. <20), when doing other commands (e.g. aptitude search ~U) or in the curses interface one can realise that the numbers match, and that there were those +4 new packages that weren't upgradable in the last update. And even if after that the user doesn't understand after a while what the numbers mean, they can be safely ignored. In any case, I don't think that there is now any chance that the sentence above can be mistaken for an error, which was mostly the point of the tiny change. And I am not keen on hiding it by default or changing it in more fundamental ways at this point, because of upcoming complaints in the other direction. > Or maybe there should be a way to figure out what they mean via a -v -v > doubled, sort of like perl: > > DESCRIPTION > The "diagnostics" Pragma > This module extends the terse diagnostics normally emitted by both the > perl compiler and the perl interpreter (from running perl with a -w > switch or "use warnings"), augmenting them with the more explicative > and endearing descriptions found in perldiag. Like the other pragmata, > it affects the compilation phase of your program rather than merely the I think that the whole quietness/verbose business in aptitude is very inconsistent. It starts with being possible to specify quiet+verbose independently and at the same time (yes, aptitude can be quiet and verbose at once), so one sees all kind of inconsistencies (like the example about "Reading package lists..." above). Also, one can specify "-vvvv" but not "-v=3", "-q" cannot be specified as "-qqq" and has to use numbers, "-q=3". So this area needs a complete review, rethink and perhaps overhaul, considering all messages shown in the different verbose levels as a whole. At that time suggestions like this could be implemented. But this is not going to happen immediately or in the short term, I think. Cheers. -- Manuel A. Fernandez Montecelo <[email protected]> _______________________________________________ Aptitude-devel mailing list [email protected] http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/aptitude-devel
