[Oberon] Re. BB, LNO + Linux

[easlab at absamail.co.za]

>  > >  > Is it possible to get a graphical display like NO in a VT ?
>  > >
> Jan Verhoeven wrote:
>  > > Yes, if you use SVGAlib.
>  >
>  > How would I do a minimal test/confirmation ?
> 
> On my website http://fruttenboel.verhoeven272.nl navigate to 
> Languages -> Mocka -> SVGAlib. I ported SVGAlib to Modula-2. 
> I show how to use the library and the nice side effects it offers.
> http://fruttenboel.verhoeven272.nl/mocka/svgalib.html

Ok. But I can't afford a big detour. 
I just wanted to click a few already existing commands.
BTW If you ported C/C++ SVGA-utils to Modula, did you 
manually 'translate' some/all of the C/C++ code ?

How about a 'tool/utility/set-of-captured-heuristics'
to assist in such a task ?
I.e. instead of 10 efforts to port drivers to Oberon, all
using weak [mostly manual] tools, invest some effort 
to make tool[s] which will facilitate ALL future porting
efforts  - move up the productivety heirarchy.

Such a 'tool' doesn't have to be 'determistically
automagical': like your pocket calculator which
gurantees to give the 7th root of 17 exact to
N decimals; to be immediately productive.

NO's transparent/open-ended/extendability
allows such projects to be in use/testing, while the
next version is evolving.

An example of what is possible and what should be
avoided: how many user have used [or even know of]
Decoder.*  ?  The disassembling capabilities seem to
offer massive analytic power but seem buggy/incomplete.
Perhaps that's why it's not documented ?
Another project which already has much value for potential
users, and which could/should be refined.  
X86-chips will be here still a long time still ?

A method to capture [and refine as required] the ETH-student
project's value is needed.  Don't just let them be wasted ?
----------

Edgar wrote:
> > ] It uses the framebuffer, so /dev/fbdev should be there.
> > I've only got /dev/fb0 ...30
> >   what are these ?
> > How/where do I get /dev/fbdev ?
> 
> Qick fix: "cd /dev; ln -s fb0 fbdev" should do the trick.

Once I've sym-linked, how would I make minimal tests
of the 'frame buffer' facility ?

> > Does that give you the graphical capability instead of using X-windows ?
> > If so, why not just use normal X-windows ?
> 
> Why use another SW layer like X when LNO can just use the framebuffer 
> directly instead of using X which then writes to the framebuffer ?

OK, but since you talk about 'Ctrl-Alt-F1' you were apparently
already in X-window ?

The LNO which I manged to get running again is V 2.3.6,
which runs under X.

Although I'll always prefer the minimalist approach, I see
no point in [contrary to my V 2.3.6] having LNO run directly
in a VT using a frame buffer, instead of in X-window.
Because the main reason for using LNO instead of NO, would 
be eg. to have linux application simultaneously available
with LNO on the same box, eg. viewing a *.pdf/*.ps 
while making notes and doing stuff under LNO & linux 
in the same boot-session.

But OK, you might want to use LNO only with some 
non-X linux application.

I've read of V2.3.7.
Q- did it fix the inability to delete with mouse-cording,
    or does this problem originate from my X-setup ?

LNO is great in that it allows you to handle linux, NO & FAT
all together.

== Chris Glur.