World: r3wp

If you're using a HP/UX from, say, 1990, you'll probably be out of 
luck, though.

I have all those systems, minus the word "recent".

I try to stay about 3-5 years behind "recent" in order to make the 
code work over the greatest range of systems.

Not a good strategy on some forward-lookin platforms, though.

It's funny how few operating systems are designed to just run software, 
not to become continuous IT management projects.

Well, forward-looking platforms either run prior code, or are totally 
new, so require some degree of time investment to get started on 
them.

Well, just as REBOL deprecates "disarm", other projects deprecate 
what they deem dangerous legacy.

REBOL uses so little of the OS libs, we could almost have our own 
libs to replace them.  Exceptions are the TCP stack and stdio/filesystem.

heh.. I thought Carl was going to say he was using Amiga System V 
unix

No, but I do have an Amiga here with 4.1. I should check if it's 
running git.

Steve said it is not.

ditto

(ssolie)

Let's get back to the issue of getting Git to run on Ubuntu Breezy 
:) ?

Ok. So much for distributed automated build, test, release.

We can start by concentrating on the "big three".

Let me see if I can find a curl-config that runs... then I can try 
the build again.

apt-get install libcurl3-gnutls-dev

(that contains curl-config)

Is there an apt-get option to ignore errors in deb urls?

You'll probably need to adapt your apt source, since breezy has been 
archived for a while.

Instead of 
  deb http://us.archive.ubuntu.com/ubuntu/breezy main restricted
use

  deb http://old-releases.ubuntu.com/ubuntu/breezy main restricted
(and change all other URLs accordingly.)

In other words, replace all *.ubuntu.com URLs by old-releases.ubuntu.com

Then run apt-get update, twice.

ok, that exists. progress!

those are the git build dependencies as marked in ubuntu packages:
https://gist.github.com/9cb86c37385fb66f9016

(all ubuntu package names)

compile goes a long way... blows out at http fetch.

try `make NO_CURL=1`

is git-http-fetch necessary?

Not really. Only if you want to access repositories via http:// or 
https://.You can live fine without that.

rebuilding with NO_CURL=1

This build is much bigger than I would have guessed.Linus and I have 
different definitions of "simple".

ok build complete, let's see if it runs...

The system itself is simple. Comes with a load of features, though.

(For many of which Linus' not the one to blame :)

ok, time to try the git on your url... scanning back.

git clone git://github.com/rebolsource/r3-hostkit.git

worked

lovely

now how old is your freebsd :) ?

based on your ubuntu, it's probably 5.4 :) ?

So, that clone gave me a full local repository, right? So, I can 
mess around with it... to get to know it?

Exactly!

No, I can't get away with that on freebsd. I think it's newer.

Probably updated it about 2 months ago.

I'd really suggest updating your Ubuntu at least to the previous 
LTS (8.04).

But that's another story :)

Yes, we use 8.xx on our servers.

So, what are git templates?