Dave Miner wrote: >> >>> I'm not sure what ICT does at hard-disk install time, but by the >>> time it would run, it could just interrogate the liveCD's menu.lst >>> for the release string. If you are worried about which menu.lst >>> string it would parse, then how about adding a commented, keyed line >>> in menu.lst which would have a magic string ICT can look for, >>> followed by the release string it should use. >> ICT uses bootadm to generate the installed system's grub menu. >> Whatever that's on the liveCD's menu.lst is ignored. >> bootadm takes the title from /etc/release. So, we want to be able to >> retrieve the special title string somehow. >> My suggestion was to store it in .image_info, since that's where >> everything about the live CD image should be stored. >> Your suggestion is to store it as a comment in the live CD's >> menu.lst. Both approach would work. >> For me, I think it is easier to retrieve it from the .image_info. >> >> Anybody else have an opinion on this? >> > > I would put it in the .image_info file. > > Dave Thanks Dave for your input.
I was coding up the changes, and realized that AI will also need something similar for the installed system's grub menu. Otherwise, the installed system's grub menu will show the problem reported on 5416. During AI install, there's no way to access the .image_info file. So, how should we handle this AI specific situation? What do you think about the following approach? Instead of storing the special string in .image_info, DC will store it in the root_archive or one of the zlib somewhere. AI downloads those to its clients, and the ICT can look up this file and modify the grub menu accordingly. The adv to this approach is that DC and AI can share the same ICT for making this change. The disadvantage is that we are introducing yet another file. Alternatively, we can move the .image_info to the root archive. Thoughts? --Karen
