The stage 1 and stage 2 instalation instructions have been removed from the handbook. I personally remember my initial confusion as to the stage 1, stage 2, and stage 3 decicion when i read the handbook for the first time, and I'm guessing this is why stage 1 and 2 were removed, but it allows the following situation to happen: I went to install (planning to do a stage 1 or stage 2, (for the coolness factor)) The new directions only contained a stage 3 install, so I just went with that and downloaded the stage-3 athlon-xp tarball. I unpacked it via the handbook. I went to edit make.conf and make.conf.example said that binaries compiled for the athlon-xp will break on the athlon thunderbird (which I have). The handbook also states that only stage 1 users should change their CHOST setting. Which imples that grabbing the standard x86 stage-3 (which is 486 optimized?) and changing optimization to athlon (and CHOST acordingly) would be a bad thing. At this point the user is stranded. There is a small blurb about stage1/stage2 in the faq, but there is something very unsettling about it. Perhaps either how terse it is, or that the instructions are no longer in sequence. I personally am fine as I have a old copy of the handbook that has stage 1 and stage 2 information in it. We definately don't want to revert the handbook to the old "there's stage 1, stage 2, and stage 3 types of installs, but were not going to recomend one method over the other, you have to figure that out on your own" kind of attitude, but what if we add the stage1/stage2 instructions back in as a appendix, and adding astricks, or daggars, or something to mark the points in the handbook instructions at which you should reference the stage1/stage2 appendix? We should still be able to have a "don't do stage 1 or stage 2 installs, it's for people other then *you*" attitude, come across so that people know to pick stage 3 by default.
A bootstrapping guide is in progress. BTW, you could have used the i686 stage3 *** This bug has been marked as a duplicate of 112661 ***
Created attachment 73992 [details] Proposed fix running this should fix the problem
(In reply to comment #2) > Created an attachment (id=73992) [edit] > Proposed fix There's no point in attaching 1MB .tar files to the bugzilla, especially to resolved bugreports...
A bootstrapping guide sounds great and will resolve the "missing stage1 and stage2" issue. It appears from what he said that I could have used the pentium pro (i686) stage3. I found this to be non-obvious, and I assume that other people will have the same confusion. This is what my fix addresses. Should I have opened it as a new bug report? And it's one meg because I left a executable in it just in case people were too lazy to emerge the dependancy.
