From mboxrd@z Thu Jan 1 00:00:00 1970 From: dugo@xs4all.nl (Jacob Goense) Date: Wed, 08 Feb 2017 15:10:24 +0100 Subject: [TUHS] Code bloat In-Reply-To: <4b06386d-6000-4d22-bbb1-2719479b5320@SG2APC01FT060.eop-APC01.prod.protection.outlook.com> References: <930c52a0c279cdd7d44953aa403a504a8622bb83@webmail.yaccman.com> <20170208025538.GE65698@eureka.lemis.com> <90190d89aaeefbf0b540a28436468835@xs4all.nl> <4b06386d-6000-4d22-bbb1-2719479b5320@SG2APC01FT060.eop-APC01.prod.protection.outlook.com> Message-ID: <72f15b962c85b6e584ef1f15965af55b@xs4all.nl> On 2017-02-08 14:10, jsteve at superglobalmegacorp.com wrote: > After I had pasted a bunch of 386BSD pl0.24 + a CVS export of 0.8 I > did get a booting system. I remember that Victor Frankenstein ;) Proved how close these 2 were. > Then I found an old ftp site that had 0.8 > I made a mirror of it, then it disappeared. Very glad that piece of history got unearthed. > I did some minor work on installing it on Bochs years ago, and VMware, > from what I recall, NetBSD 1.0/1.1 can boot 386BSD’s kernel, while > the 386BSD boot diskette didn’t work under emulation, NetBSD’s > does, and I used that to kickstart an installation. Same with the > boot blocks on the harddisk image. 386BSD is now bootable in Bochs, with very, very specific settings. One that works at: https://raw.githubusercontent.com/dugoh/tobochs/master/bochsrc Back then it required a patch against bochs too as the boot blocks do some truly weird stuff with the PIC (polling OCW3?), something most emulators don't implement or even barf on. These 2 little marvels didn't have much bloat, but the Bostification had already set in. My idea of a true diet x86 UNIX system would be a report of Tahoe without resorting to gcc/gas or anything else that smells like RMS.