Assuming the cdimage.fs is simply an FFS image in a file like
miniroot.fs is, one could always mount the image on your vax and copy
the syssrc tar ball into it and be done with it. I'm glad Matt checked
in the fixes to distrib/vax that I was complaining about here a while
ago. 

Now, with the exception of dhcp (did you get this to build without
change Matt or did you disable -Werror or something?) my source tree is
nearly identical to the cvs one (I've got the dssi driver stuff I'm
working on and some stub changes to the rpc code in libsa as
differences)

--Chuck

Matt Thomas wrote:
> 
> At 02:47 PM 5/25/2000 -0400, NetBSD Bob wrote:
> >GREAT!  KUDOS!  and other assorted niceties......(:+}}... on that cdimage.iso
> >thingie.  That is exactly what I have been looking for, since my home net
> >is not on the internet, and I have to ferry things back and forth to play.
> >When I get home, tonight, it will get shovelled into the first VAX I stumble
> >across....(:+}}...
> >
> > > Nothing.  The cd contains an INSTALL kernel with a ramdisk with sysinst on
> > > it.  it also includes all the subsets in my 20000517 snapshot (which also
> > > includes the X subsets).  It does not, however, contain any sources.
> >
> >Foot in mouth light-bulb flashing propellerheaded-beanie dunce mode now.....
> >
> >What about the possibility of rolling into it just sufficient sys.tar.gz
> >sources to allow fiddling with the kernel, in the same iso image.
> >The sys tarball would add 18 megs more to its size, but might be a
> >welcome addition, for some of us (would the config and top-level tarballs
> >also be needed?) that could use a coherent matched source/binary set,
> >along the development stream.
> 
> Note this is just I needed to build for someone else.  I thought of
> including more but I just wanted a cut image.  But this is just a stepping
> stone.  The goal for me is that when the real 1.5-release CD comes out, one
> of the architectures it will be bootable on will be VAX.  There's still a
> bit of work to be done to get that done.
> 
> >My reasoning behind querying this is that having at least the minimal
> >kernel tarball sources handy, would make an even playing field from
> >which to work the system up for play and testing.  In my case, for
> >example, my sources usually lag a week behind the binaries, which
> >can lead to a bit of instability across my machines.  Having a least
> >common denominator matched source/binary set all in one place, would
> >be a nice feature.
> 
> That would be nice.  But for that, I'd probably include a couple more
> architectures (alpha, arm32, i386, macppc, pmax, sparc, and vax) and
> make a more complete CD image including a full anoncvs snapshot of
> src, pkgsrc, and xsrc.
> 
> >Would that be an unreasonable expectation, perhaps, if we were already
> >going to the trouble to roll a CD or image.fs thing of 50mb size anyway?
> >Another 20mb should not make that much difference....
> 
> Not really, but I'd just rather do a more complete job of it since I
> deal with more than VAXen.
> 
> --
> Matt Thomas               Internet:   matt@3am-software.com
> 3am Software Foundry      WWW URL:    http://www.3am-software.com/bio/matt/
> Cupertino, CA             Disclaimer: I avow all knowledge of this message