[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: My recent bug's and continuing effort to debconf-ize Debian

>>"Julian" == Julian Gilbey <J.D.Gilbey@qmw.ac.uk> writes:

 Julian> No, this isn't what I was talking about.  There was a discussion
 Julian> recently (should I try to track down the message numbers?) about some
 Julian> packages which could not use the debconf database setup, for example
 Julian> because the answers were too sensitive (passwords).  For these, the
 Julian> only option was to use interactive questions during the postinst.

 Julian> What I am asking is whether we can devise a way to handle these
 Julian> special cases, by allowing these interactive questions to all be
 Julian> handled at the end.  I don't know whether there is any way to get rid
 Julian> of them entirely; we should look back at the above-mentioned thread to
 Julian> answer that one.

	There is an additional winkle that also needs to be addressed
 -- and that is the down time of critical daemons. One may, under the
 current interactive process, shutdown a daemon (perhaps in preinst),
 ask questions that needs runtime (post unpack) answers, and restart
 the daemon, in a short time. Waiting until all packages have been
 installed and then restarting add to the downtime of that daemon. 

	How important is minimizing the time one leaves pacages in an
 unconfigured state? apt espescially jumps through hoops to minimize
 this time. 

	Mind you, I am not against installing post-install hooks into
 dpkg, and having a means of asking questions at the end -- I am just
 saying that this may be sub optimal for some packages. 

	I would also like to say that decisions like this should be
 the users decision -- people should be given a choice, rather than a
 having us mandate one course or the other.

 Johnson's First Law: When any mechanical contrivance fails, it will
 do so at the most inconvenient possible time.
Manoj Srivastava   <srivasta@debian.org>  <http://www.debian.org/%7Esrivasta/>
1024R/C7261095 print CB D9 F4 12 68 07 E4 05  CC 2D 27 12 1D F5 E8 6E
1024D/BF24424C print 4966 F272 D093 B493 410B  924B 21BA DABB BF24 424C

To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org

Reply to: