[Unix-jun72] boot_080511.tar.gz -- UNIX V1 bootstrap

[Unix-jun72] boot_080511.tar.gz -- UNIX V1 bootstrap

[James A. Markevitch]

A non-text attachment was scrubbed...
Name: boot_080511.tar.gz
Type: application/octet-stream
Size: 10084 bytes
Desc: not available
URL: <http://minnie.tuhs.org/pipermail/tuhs/attachments/20080511/cdc2d155/attachment.obj>

[Unix-jun72] boot_080511.tar.gz -- UNIX V1 bootstrap

[Tim Newsham]

Thank you.  I incorporated this into the svn tree and the image building 
process.  We're now using the bos boot method instead of loading the 
kernel with a simh loadfile.

Tim Newsham
http://www.thenewsh.com/~newsham/

[Unix-jun72] boot_080511.tar.gz -- UNIX V1 bootstrap

[Warren Toomey]

On Sun, May 11, 2008 at 01:31:25PM -1000, Tim Newsham wrote:
> Thank you.  I incorporated this into the svn tree and the image building 
> process.  We're now using the bos boot method instead of loading the 
> kernel with a simh loadfile.

Tim, the new boot method seems to make the output from the running kernel
very sluggish, e.g. ls -l /bin takes ages to arrive. Can you do a quick
compare old vs. new?

	Warren

[Unix-jun72] boot_080511.tar.gz -- UNIX V1 bootstrap

[Warren Toomey]

On Mon, May 12, 2008 at 11:47:21AM +1000, Warren Toomey wrote:
> Tim, the new boot method seems to make the output from the running kernel
> very sluggish, e.g. ls -l /bin takes ages to arrive. Can you do a quick
> compare old vs. new?

I take that back. I was ssh'd into my home machine and running simh over
the link. I think the output from the 1st Edition kernel comes out in
clumps (i.e. a small internal buffer's worth), and each clump was being
shipped over the link as a single TCP packet; hence I was seeing latency
between the clumps. My apologies.

	Warren