When I was watching that Lunduke's video a couple of days ago initially I was thinking he's just making a joke of that Vendefoul Wolf distro on 200MB box. I recalled using FreeBSD as access server with lots of modems (PPP/SLIP), Apache, Samba and QuakeWorld server running on a box with just 32MB of RAM. That was also my daily working machine with XF86 and Enlightenment desktop manager, circa 2000. So, 200MB is a whole lot of memory!
That's interesting, thanks. I feel a need for simple multitasking/networking OS for synthesizable RV32I core (not RTOS like, but more like Unix or CP/M). Would be nice to try Plan9 on it once port is out.
Get two FTDI FT232RL chips, connect them together on serial side (RXD->TXD, TXD->RXD, GND<->GND). Plug into USB ports of your computers, run terminals (or any other software that supports serial I/O), send/receive data. Can use XYZModem to send files, PPP for TCP/IP networking, etc. No programming involved. Cheap as hell.
Thanks for the idea. I think it's a great one, but I have a few concerns:
1) The two computers are on 2 different grounds, so I believe it could damage the computers. So I would like to isolate the ground somehow, I am not sure what the options are. I tried to look for opto-isolated options, but didn't really find something clear. Do you have any suggestion?
2) This specific chip is marked NRND (not recommended for new designs) on ftdi's website, So it feels slightly wrong to use it?
"On Unix, I used RCS fairly happily. It was too verbose, but I fixed
that. RCS, like most version control systems, also believed that the
RCS archive or repository was authoritative and I don't accept that, I
think that the checked-out source is authoritative. If you are
grepping source, you really don't want to grep the archive files, and
you shouldn't have to check out source just to compile. I view the
archive as just history. So I added options to do things my way.
For my purposes, the Plan 9 dump, cp, diff and idiff work well. I
don't miss a formal version control system.
Git is a horrific botch of a version control system: wasteful, slow,
stupid, and unpredictable. When I check in source file(s), I want a
snapshot taken always, and I don't want to be told, no, you can't do
that, you have to rebase. The fundamental problem is that companies
want to believe that multiple programmers can edit the same program at
the same time. That's a great way to introduce bugs and break things.
When editing a program, you want to have a stable view of it, so you
know what you can rely on, you don't want it to be shifting underfoot.
The idea that you can apply an arbitrary set of diffs, or subset, to
some version of that program, and get a meaningful and correct program
out is absurdly optimistic; I wouldn't trust it."
reply