On Tue, 2010-04-20 at 08:36 -0700, Adam Williamson wrote: > Hi, all. At the last blocker meeting, we realized there are no release > criteria regarding remote-accessible installs. If you can't install > Fedora to be remote-accessible out of the box, some sysadmins will have > a lot of trouble. So I propose we add this to the criteria: > > "With the correct install configuration, it should be possible to make > an installed system immediately remotely accessible (you should be able > to install with a secure remote access service active and unblocked by > firewall-type mechanisms)" > > Does that seem reasonable? Should it be a Beta or Final criterion, do > you think? Thanks for drafting Adam. For those not in the blocker meeting, this came out of a discussion around https://bugzilla.redhat.com/show_bug.cgi?id=568528 (firewall --disabled still produces a blocking /etc/sysconfig/iptable). I like the proposed criterion above. Originally, I was thinking of a more general statement along the lines of "the install must honor all kickstart commands provided". While it should, it's not something we exhaustively test now and not every failure of this nature would have the same impact to users (and the release). I appreciate how you've limited to the scope to specific kickstart commands [2] (rootpw, firewall, network, potentially selinux and authconfig). Should we be able to report on the health of each kickstart command in the future ... we can certainly consider extending the criteria. NOTE: you're wording doesn't limit this to kickstart installs. The kickstart installs only serve to help me capture the failure cases this would cover. Thanks, James [1] It would find bugs sure, but we don't want to further bury ourselves in bugs [2] https://fedoraproject.org/wiki/Anaconda/Kickstart
Attachment:
signature.asc
Description: This is a digitally signed message part
-- test mailing list test@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe: https://admin.fedoraproject.org/mailman/listinfo/test