Re: Should setup suggests to downgrade? [WAS: Lillypond for cygwin]

["David A. Cobb" <superbiskit at cox dot net>]

Robert Collins wrote:

>>-----Original Message-----
>>From: Christopher Faylor [mailto:cgf@redhat.com] 
>>Sent: Saturday, April 06, 2002 4:20 PM
>>
>>I'm not sure how.  If you're talking about my suggestion, if 
>>you switched to a test view and installed things, the next 
>>time you ran setup you'd still default to current so, I 
>>presume that setup would still suggest a downgrade.
>>
>
>Sorry, I was taking as read that if we made the three things into three
>effectively different distributions that setup would remember the last
>choice of prev/curr/test.
> 
>
>>setup should probably *never* suggest either going backwards 
>>or uninstalling.
>>
>
>I think it should suggest uninstalling when a requirement is uninstalled
>- but in a very visible fashion. As for going backwards, I haven't
>thought that completely through yet.
>
>Rob
>
Even then, it isn't spontaneously suggesting going backwards: the user 
selected to uninstall something and this "suggestion" is simply the 
inverse or adding dependencies to install.

On the same general topic: IMHO it would be much friendlier to default 
to *install* packages that are new, rather than skipping them.  If 
feasible, clearly mark them as "new."

-- 
David A. Cobb, Software Engineer, Public Access Advocate
"By God's Grace I am a Christian man, by my actions a great sinner." -- The Way of a Pilgrim; R. M. French, tr.
Life is too short to tolerate crappy software.
.