I have collection of Debian 2.0 packages. Downloaded mainly 5.5.98. Some packages later, so I use 'scan' instead of Packages file. I started with clean partitions, because i wanted to protect old working Debian.
Selection of packages is normal, but installation ends: "internal error - no filename at -e line 12, <P> chunk 13. installation script returned error exit status 1. Press RETURN to continue." Sometimes if i enter selection after this dselect crashes (segmentation fault, core dump). Twice this created invisible file(s) of few MB. Only trace of that file(s) is that outputs of df and du don't match. Only 'e2fsck -f' frees that disk space. I tested with kernel from rescue disk and also 2.0.33 kernel i have compiled earlier. I have no hardware problems with old Debian installation. I guess some installation script crashes. Dselect does not report witch package is broken, not even with -D. How do i found witch one? Or at least i don't find this useful: getcurrentopt() cmethopt open getcurrentopt() cmethopt eof getcurrentopt() cmethopt read getcurrentopt() cmethopt len and newline getcurrentopt() cmethopt space getcurrentopt() cmethopt meth name `disk' getcurrentopt() cmethopt meth found; opt `mounted' getcurrentopt() cmethopt opt found I tried to found it by selecting only few packages, but that failed. Perhaps order in witch packages are installed is important. I think this is also dselect bug, because it should report, at least with -D, what it is going to try next. - Jukka -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
