[Coco] nitros9 level3

Lothan lothan at newsguy.com
Sat Oct 20 20:08:08 EDT 2012


From: Gene Heskett

> And the first thing I run into is that ALL the defs have been renamed, so 
> a
> make dsk in level3/coco3_6309 goes completely stark raving blitherfousy.
> However it appears the level1 and level 2 coco_6309 stuff at least builds,
> using mamou.

This is what I've done to get a working build on Ubuntu 12:

https://sourceforge.net/projects/toolshed/files/ToolShed/ToolShed%202.0/

Download toolshed_2.0_linux.tar.gz from the link above and extract 
everything to a bin folder in your path, such as /bin, /usr/bin, ~/bin, or 
whatever. What this one, I can build the entire nitros9 tree without any 
errors. At least I've run make clean, make dsk, make dskcopy, etc. and 
haven't seen any errors yet.

I did downloaded the Toolshed source, but I can't get it build because it 
can't find fuse.h. I have no idea where to obtain that file at the moment. I 
thought it was from the Fuse stack (Filesystem in Userspace), but I already 
have that installed and it doesn't seem to have a fuse.h file. This is when 
I backed out to find the pre-built binaries and found they worked fine, 
although the pre-built binaries are about 4 years old.

> But it appears that copy -r may have a src date to target date problem.
>
> All the stuff in the MODULES tree, a tree I created not more than 10 days
> ago from scratch, including the z stuff Aaron has been working on, carry
> dates, which are correct in the dsk image, but as installed in my
> /dd/bootmaker-stf tree by dsave, have file creation dates just short of 2
> years old, or even older, like
>
>  0  2004/08/04 13:18  ----r-wr 718 2D1 lltc3.dr !
>
> My clock here is ntp set, milliseconds from WWV in Boulder Co., and my
> coco's clock, which has not been set in yonks, says
> {t2|08}/DD/bootmaker-stf:date -t
>
> October 20, 2012  14:26:46
>
> Nominally 10 minutes slow.  My TC^3 has an rtc I haven't set in at least a
> year.  I reset it just now, but that doesn't explain the old dates in view
> of the observation that this whole /opt/nitros9 tree on my drive carries a
> date -t (ls -l) of about an hour ago, when I pulled it.
>
> The dates in the just built and dw re-mounted .dsk are correct.
>
> This could make version tracking on my coco's drive into a cast iron
> nightmare.  How can we fix this?

I've never used Drivewire, so I'm not sure where the timestamp issue is 
coming from.




More information about the Coco mailing list