test please delete

[BrowneIssue]

it is hips brownei hips.

What is?

[BrowneIssue]

Ubuntu

Currently at:

"Installing system
Configuring apt
Scanning the mirror...
82%"

I could have written this opsys to disk with a magnet on a lolly stick by now.

What on earth can it be doing?

[BrowneIssue]

Ubuntu

Here's something else for me to worry about:

Cannot access security updates

The security updates on security.ubuntu.com couldn't be blah blah blah for three lines blah blah later.

Commented out entries for security.ubuntu.com have been added to the /etc/apt/sources.list file


Well, that's alright then. Thank heavens for short, meaningful messages.

[BrowneIssue]

Ubuntu

Oh goody.

Having dealt with that, back to 83%.

[BrowneIssue]

For some reason, my last bottle of beer tasted just like canned McEwan's used to in the 1980s.

Weird.

[BrowneIssue]

Ubuntu

"Configuring boot loader"! We can't be far off, surely?

[BrowneIssue]

Ubuntu

And then, all of a sudden, it's done.

Pop CD out, reboot from USB stick and it's up.

No network yet - that's probably next.

Moral: when trying to install Ubuntu to an external drive (eg USB stick) you must partition the drive in gparted BEFORE installing Ubunto onto it. It doesn't like a Windows-created partition.

[NickFitz]

That sort of activity made sense when Honeywell built a brand new mainframe and the first assembler is written in machine code with just enough code to produce a slightly better assembler.

When I implemented Forth (or FORTH if you insist) on the Atari ST I wrote the nucleus in assembler, including defining the word headers complete with link field and code field. Once I'd got that written I could then write the rest of the system and development environment in Forth itself. Then I implemented a Forth assembler (Reverse Polish 68000 assembly language FTW!) and rewrote the original assembler source in RP... and then it could assemble itself

From that point on I could implement Forth anywhere else by first writing an assembler for the target processor, re-implementing a very small number of low-level words, and compiling a target image which I could then send to the target (via RS-232, naturally) and lo! it worked

Given that my implementation strategy was the same as that used by Chuck Moore (who invented the language), the nucleus actually included the OS - the only existing code it used was the firmware BIOS of whatever system it might be on. Other than that it did all its own disc handling and so on.

IIRC the ST nucleus - the absolute bare minimum of code to enable everything else to happen - came out to about 14K, although with the development environment loaded (including an editor and a debugger) it rose to the dizzy heights of about 22K

[NickFitz]

My intention was to show just how easy this is and to show just how clever I am by being able to install and configure Linux for the first time and get it done before lunch.

Dream on...

[NickFitz]

You won't see this until tonight at least but the thought counts, I hope:

Cheers BI