Hello David, On Sun, 2016-02-21 at 23:14 +0100, David Kalnischkies wrote: > Hi > > (I don't understand the first message in the bug, so I refer only to > the > second in the hope that the second is a/the example for the "bug") > The problem is just -qq vs -qqq. I was just hoping to have a consistency in behavior when using the option with apt commands. For example; if you use the same option for the upgrade/install target, it is not quiet but gives you a prompt; so you have to use an extra 'q' rrs@learner:~/.rrs-home/devel/apt-offline/apt-offline (master)$ apt -qq --print-uris install emacs The following additional packages will be installed: emacs24 emacs24-bin-common emacs24-common libm17n-0 libotf0 m17n-db Suggested packages: emacs24-common-non-dfsg m17n-docs Recommended packages: emacs24-el The following NEW packages will be installed: emacs emacs24 emacs24-bin-common emacs24-common libm17n-0 libotf0 m17n-db 0 upgraded, 7 newly installed, 0 to remove and 12 not upgraded. Need to get 16.9 MB/18.2 MB of archives. After this operation, 91.6 MB of additional disk space will be used. Do you want to continue? [Y/n] n Abort. 2016-02-22 / 15:07:35 ♒♒♒ ☹ => 1 vs rrs@learner:~/.rrs-home/devel/apt-offline/apt-offline (master)$ apt -qqq --print-uris install emacs 'http://ftp.debian.org/debian/pool/main/e/emacs24/emacs24-common_24.5+1 -6_all.deb' emacs24-common_24.5+1-6_all.deb 12825770 MD5Sum:fb43f4fd59052a28cc3ed1b5058b04d7 'http://ftp.debian.org/debian/pool/main/e/emacs24/emacs24-bin-common_24 .5+1-6+b1_amd64.deb' emacs24-bin-common_24.5+1-6+b1_amd64.deb 256632 MD5Sum:62e8c7b5c3c433ad7f9f9796885662db 'http://ftp.debian.org/debian/pool/main/libo/libotf/libotf0_0.9.13-3_am d64.deb' libotf0_0.9.13-3_amd64.deb 52794 MD5Sum:a2593897a324bad727ce5d848426fb32 'http://ftp.debian.org/debian/pool/main/m/m17n-lib/libm17n-0_1.7.0-3_am d64.deb' libm17n-0_1.7.0-3_amd64.deb 249092 MD5Sum:8e944e96b488f5d1fe119aeb4f8cf98a 'http://ftp.debian.org/debian/pool/main/e/emacs24/emacs24_24.5+1-6+b1_a md64.deb' emacs24_24.5+1-6+b1_amd64.deb 3495008 MD5Sum:e05e15580c68aa77b15ce9b370778c54 'http://ftp.debian.org/debian/pool/main/e/emacs-defaults/emacs_46.1_all .deb' emacs_46.1_all.deb 1634 MD5Sum:1f115942065ac452467e02377368ee22 2016-02-22 / 15:07:40 ♒♒♒ ☺ Now if I do the same with update (my previous email has that example), I need to lower 1 level. > > On Sun, Feb 21, 2016 at 11:20:42PM +0530, Ritesh Raj Sarraf wrote: > > When triggering update with -qqq, it goes ultra quiet. > > > > > > rrs@learner:~$ apt -qqq --print-uris update > > 2016-02-21 / 23:19:10 ♒♒♒ ☺ > > rrs@learner:~$ > > And this is a problem exactly why? > It is just not consistent in the update vs upgrade/install use case. > You might have noticed that this is the case with apt-get -qq > already, > so apt actually takes the liberty of requiring more q's for the same > effect Yes. but in case of apt-get, the single '-qq' behavior is consistent across update, upgrade etc. > (I never quite liked the "ultra quiet" directly after "slightly > less chatty", but changing that while staying compatible in the code > is > hard, so for now I introduced 3 levels only… and the new level-2 is > a small step up from level-1 I found useful for testing…). > > I realize now that our recent redesign of the --help messages removed > a very relevant information tidbit for this discussion: > -qq No output except for errors > > We should probably move that to the manpage (although it isn't > entirely > correct, but relatively close to reality). > Oh Yes. These option seem to have gone away from the documentation. I reported the bug because I've been using it in apt-offline. > So, works as expected or what exactly do you expect? > > > Maybe the expectation is that "apt" doesn't have such options as they > aren't really suited for interactive use, but then again I sometimes > use > them "interactively" – if only to test things and I don't think there > is > a lot of point in discriminating options based on the command. > > > Oh, and as your message indicates a misunderstanding: "apt" will > NEVER > guarantee a stable interface – that is actually the point of > introducing > apt in the first place. We have a stable interface which can be used > in > scripts with apt-get and co as they will neither disappear nor do > they > change default behaviour.¹ "apt" is just another binary which comes > with > a slightly different set of default options we deem more reasonable > for > interactive usage, but uses the very same code as apt-get and co. > > So if you have a script, please continue to use 'apt-get update > --print-uris' in it. > Thanks for explaining this. I added apt support because it is much faster and did slightly better dependency resolving (I guess which is courtesy of the different set of defaults). -- Ritesh Raj Sarraf | http://people.debian.org/~rrs Debian - The Universal Operating System
Attachment:
signature.asc
Description: This is a digitally signed message part