Hi again! On Nov 14, 2007 12:55 PM, Andreas Tille <[EMAIL PROTECTED]> wrote: > On Wed, 14 Nov 2007, Charles Plessy wrote: > > > although I do not expect users to decide to install one without the > > other. But now I wonder if this fits the renewed interpretation of what > > Recommends: means, now that front-ends to dpkg install them by default. > > > This is exactly what you want. While users might install dialign-t-data > perfectly without breaking there system or the functionality of dialign-t-data > (which is prividing data for dialign-t ... the user might run a local > copy for whatever reason), it is _Recommended_ to install dialign-t as > well which is reasonably done by apt-get by default.
That is right. > > I think that I would only Suggest: dialign-t, to avoir a relationship > > which would be too circular. > > No. Circular _Depends_ should be avioded. I never heard that > circular _Recommends_ would cause any problem. It just makes sense in > most use cases to not install dialign-t-data without dialign-t - so it > is _Recommended_. Exactly. > > Or how about: "This package contain the score matrices and probability > > distribution files that DIALIGN-T needs to align peptidic and nucleic > > sequences." > > I prefer this one. So we change to this one :-) > > Thanks a lot for the help ! I am running out of time today to apply the > > changes... > > But there is so much time left for today - I think Nelson will not even > had Lunch today. ;-)) Now I am back from lunch ;-) Best regards, Nelson -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]