Kyle McDonald wrote: > Bart Smaalders wrote: >> Mike Gerdts wrote: >> >>> True enough. I suspect at some time there will be someone that says >>> "why does every machine have a user named jack?" and no matter what >>> approach is used a change will be needed. Or is this creating a user >>> that is specified by the person performing the installation? >>> >>> >> >> The fact that there's a jack user on every Indiana machine is a bug, plain >> and simple. >> >> > I know there isn't always room on the media for the live version, and a > full repository. > > But this is one of the reasons that I don't get a warm fuzzy feeling > when I hear talk of 'copying the livecd to the disk' as the main part of > the install. > > There will always be customizations like 'jack' on the livecd. I'm not > comfortable that the list of things to 'undo' after copying the files to > the drive will always be 100% up to date. > > This leaves me wanting to always install directly from the set of > packages. It's also very similar to the reason I don't trust upgrades. >
The set of customizations is actually pretty small, and will get smaller as we work our way towards integrating the live CD technology as an architectural element of the system, rather than as a set of add-ons as it is right now. Ultimately, there shouldn't be a difference in the software that results. We're already doing pretty well, as "pkg verify" doesn't show much of anything that we control as being amiss as of preview 2. BTW, you don't end up with 'jack' on every system, it gets filtered out at installation. The fact that it ends up in the ipkg branded zones is a bug that I think should be easily fixable now that user actions are ready, as we ought to be able to move this passwd entry to the live CD-specific package. Dave
