From owner-freebsd-hackers Mon Apr 24 19:10:37 1995 Return-Path: hackers-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id TAA05911 for hackers-outgoing; Mon, 24 Apr 1995 19:10:37 -0700 Received: from ref.tfs.com (ref.tfs.com [140.145.254.251]) by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id TAA05904 for ; Mon, 24 Apr 1995 19:10:31 -0700 Received: (from julian@localhost) by ref.tfs.com (8.6.8/8.6.6) id TAA03705; Mon, 24 Apr 1995 19:10:24 -0700 From: Julian Elischer Message-Id: <199504250210.TAA03705@ref.tfs.com> Subject: [to Install people] Re: comments on an attempted install To: dgaudet@cs.ubc.ca (Dean Gaudet) Date: Mon, 24 Apr 1995 19:10:24 -0700 (PDT) Cc: hackers@FreeBSD.org In-Reply-To: <199504242325.QAA17567@grolsch.cs.ubc.ca> from "Dean Gaudet" at Apr 24, 95 04:25:41 pm Content-Type: text Content-Length: 2854 Sender: hackers-owner@FreeBSD.org Precedence: bulk Well thank you for this... firstly a note to say that there is work underway to fix many of these things.. I have to agree that the install for 2.0 was the worst that I have ever seen, (or written) mainly because it wouldn't let experienced users over-ride it when it wanted to do the wrong thing. The install with the SNAPs is basically the 2.0 install.. 2.1 should be better anyway, this is the subject of on-going improvement, but it seems that you fell foul of just about ALL the problems I've seen at various times.. To recap: Ok, so Who is doing the new install....? Can some-one who knows, answer these points as to whether they are addressed in the new install.. if they are not, I may try to look at them myself..... julian > 1/ Bindist is FAR too big, and should be broken down to a Base, Net, Devel, and Extended set. I will say however that it's AMAZING how usable the system is after installing just the second floppy.. 2/ Boot floppy wouldn't work with remmapped bios.. not sure where we should look for that.. I don't understand the remapping very well. It is possible that FBSD could use the remapped geometry if you'd just typed it in at teh right time.. {?} 3/ we need to be able to create other partition types. 4/ The disklabel part of the install is totally cryptic and you have to know exactly what the program is doing before you have a prayer of running it.. e.g. write disklabel before assign. 5/ disklabel had old info from before teh new fdisk (I understand how this happens, but we should look at how to fix that when we re-write the MBD...(invalidate the label?)) 6/ It should be possible to proceed without re-disklabelling your disks DAMMIT! I understand that you may need to 'assign' them again as this info is not written to disk yet. You should have booted the kernel using the -c option which would have dynamically reconfigured the kernel to suit your cards. 7/ In the initial install boot, the -c option should be COMPULSORY or at least better publisised (It is, but not well) 8/ better handling of domain name and host name. 9/ Net values should default to previous values. 10/ sysinstall should be runnable after setup, so that you can install other packages as needed. 11/ packages stuff is totally cryptic. there is no good documentation IN THE NORMAL PLACES as to what the f*ck to DO with the packages This information needs to be plastered EVERYWHERE. > 12/ screen failed (separate problem) > [...] user goes and installs LINUX > > [continued] > 13/ Slice code is pooly documented .. docs still refer to disklabel for DOS partitions. > in a disklabel. This seems like a really arbitrary restriction to me. > Isn't /dev/wd0s2 the second slice on disk 0? No, wd0s1 is..... 0,1,2,3 right? > 14/ swap-files are real good when the install is wrong..... julian