Re: [9fans] Re: ideal plan9 laptop

Re: [9fans] Re: ideal plan9 laptop

[forsyth]

>>Also my 9load in (/n/9fat) has the 'a' [append] bit on.  This must be a bug.
          The file attribute flags used by the DOS file system do not
          map directly to those used by Plan 9.  Since there is no
          concept of user or group, permission changes via wstat (see
          stat(2)) will fail unless the same (read, write, execute)
          permissions are specified for user, group, and other.  For
          example, removing write permission in Plan 9 corresponds to
          setting the read-only attribute in the DOS file system.
          Most of the other DOS attributes are not accessible.

          Setting the exclusive use flag (CHEXCL) in Plan 9 corre-
          sponds to setting the system use attribute in the DOS file
          system.  Such files are not actually restricted to exclusive
          use, but do merit special treatment that helps in the cre-
          ation of boot disks: when dossrv allocates a new block for
          such a file (caused, say, by a write that fills the file's
          last allocated block), it succeeds only if it can arrange
          for the file to be stored contiguously on disk.

          Since other operating systems do not guarantee that system
          files are laid out contiguously, the CHAPPEND mode bit is
          set in file stat information only when the file is currently
          contiguous.  Attempts to set the CHAPPEND mode bit explic-
          itly will cause dossrv to try to make the file contiguous,
          succeeding only if this is possible.

Re: [9fans] Re: ideal plan9 laptop

[Matt H]

On Wed, 30 Jan 2002 16:31:53 0000
"forsyth@caldo.demon.co.uk" <forsyth@caldo.demon.co.uk> wrote:

> >>Also my 9load in (/n/9fat) has the 'a' [append] bit on.  This must be a bug.
>           The file attribute flags used by the DOS file system do not
>           map directly to those used by Plan 9.

so is the 'a' mapped onto DOS's Archive bit?

Re: [9fans] Re: ideal plan9 laptop

[Boyd Roberts]

forsyth@caldo.demon.co.uk wrote:
>           Since other operating systems do not guarantee that system
>           files are laid out contiguously, the CHAPPEND mode bit is
>           set in file stat information only when the file is currently
>           contiguous.  Attempts to set the CHAPPEND mode bit explic-
>           itly will cause dossrv to try to make the file contiguous,
>           succeeding only if this is possible.

Thanks for the info.

That is awful.  It also explains why my 9load crashed and burned; it not
being contiguous, perchance?

I shall try again.

Those VAIOs will come 'round to 'my way of thinking'.

Re: [9fans] Re: ideal plan9 laptop

[Boyd Roberts]

Russ Cox wrote:
> The whole thing is a crock.

You said it.

In some mad hack last night I put latin-1 support into the kernel
so it's a bit less painful to use [weird LCD xga card stuff so
revert to raw typing qwerty on azerty at rc and ed].

Last trick was [kfs stuff omitted for brevity]:

    9fat:
    cd /n/9fat
    > d9load
    chmod +al d9load
    cp /sys/src/boot/pc/9load d9load
    mv 9load o9load
    mv d9load 9load

and reboot.

It then goes trying to boot a gz'd kernel [thinking it's a floppy
install?] and fails.

To fix the mess I boot from:

    sdCO!9fat!9pcdisk

Enough already :(

I've already written the Makefile from hell on linux today :( :(

Re: [9fans] Re: ideal plan9 laptop

[Russ Cox]

>     chmod +al 9load
>
> Leave off the l and you get EPERM.

That was intentional, because it's so
magical.  You're only allowed to make
system files (the l bit) contiguous.

The whole thing is a crock.

Russ

Re: [9fans] Re: ideal plan9 laptop

[Boyd Roberts]

Russ Cox wrote:
> Umm... But you said it had the 'a' bit set.
> It only has the 'a' bit set when it is contiguous.

    chmod +al 9load

Leave off the l and you get EPERM.

Re: [9fans] Re: ideal plan9 laptop

[Boyd Roberts]

Russ Cox wrote:
> Umm... But you said it had the 'a' bit set.
> It only has the 'a' bit set when it is contiguous.

Me, not knowing about the 'a' bit just used cp with my new 9load.

Now I know about it, so I'll retry it.

When I saw the 'a' bit set I thought 'hmm... better use mv'.

Re: [9fans] Re: ideal plan9 laptop

[Russ Cox]

> That is awful.  It also explains why my 9load crashed and burned; it not
> being contiguous, perchance?

Umm... But you said it had the 'a' bit set.
It only has the 'a' bit set when it is contiguous.

Re: [9fans] Re: ideal plan9 laptop

[Russ Cox]

matt@proweb.co.uk:
> so is the 'a' mapped onto DOS's Archive bit?

No, the 'a' means "this file is currently contiguous on disk,
as 9load must be in order to be the target of a single
sector bootstrap loader".  It's a clumsy hack and I'm sorry
I did it, but there's really no good alternative that
doesn't require writing custom tools.  At least this
way I can use chmod and ls.

Russ

Re: [9fans] Re: ideal plan9 laptop

[Boyd Roberts]

jmk@plan9.bell-labs.com wrote:
>
> Although they function the same, Intel have recently started giving new
> PCI Device IDs to the 82557s connected to (or embedded in) their 'hub'
> chipsets. You may have to alter the detection code in the driver (and 9load)
> to recognise them:

I don't have this code in my driver(s), but I do have a 9load (on a floppy
that will recognise the card:

    ether0: i82557: port 0xFCC0 irq 9: 0800460EB796

Also my 9load in (/n/9fat) has the 'a' [append] bit on.  This must be a bug.

>         i82557pci(void)
>         {
>                 int port;
>                 Pcidev *p;
>                 Ctlr *ctlr;
>
>                 p = nil;
>                 while(p = pcimatch(p, 0x8086, 0)){
>                         switch(p->did){
>                         default:
>                                 continue;
>                         case 0x1209:            /* Intel 82559ER */
>                         case 0x1229:            /* Intel 8255[789] */
>                         case 0x1031:            /* Intel 82562EM */
>                         case 0x2449:            /* Intel 82562ET */
>                                 break;
>                         }
>                         ...

Re: [9fans] Re: ideal plan9 laptop

[jmk]

Although they function the same, Intel have recently started giving new
PCI Device IDs to the 82557s connected to (or embedded in) their 'hub'
chipsets. You may have to alter the detection code in the driver (and 9load)
to recognise them:

	i82557pci(void)
	{
		int port;
		Pcidev *p;
		Ctlr *ctlr;

		p = nil;
		while(p = pcimatch(p, 0x8086, 0)){
			switch(p->did){
			default:
				continue;
			case 0x1209:		/* Intel 82559ER */
			case 0x1229:		/* Intel 8255[789] */
			case 0x1031:		/* Intel 82562EM */
			case 0x2449:		/* Intel 82562ET */
				break;
			}
			...

On Wed Jan 30 06:35:34 EST 2002, boyd@strakt.com wrote:
> > Sony products should be avoided ...
>
> Hmm, I hope this is just an id string thing, but it doesn't
> find my on board:
>
>     i8255x PCI 10/100Base-T
>     PCI Bus 0 ctrl 11 func 0
>     IRQ 9
>     Mem FECFF000-FECFFFFF
>         FCCO-FCFF
>         FED0000-FEDFFFFF

Re: [9fans] Re: ideal plan9 laptop

[Boyd Roberts]

> Sony products should be avoided ...

Hmm, I hope this is just an id string thing, but it doesn't
find my on board:

    i8255x PCI 10/100Base-T
    PCI Bus 0 ctrl 11 func 0
    IRQ 9
    Mem FECFF000-FECFFFFF
        FCCO-FCFF
        FED0000-FEDFFFFF

Re: [9fans] Re: ideal plan9 laptop

[Thomas Bushnell, BSG]

ravage@einstein.ssz.com (Jim Choate) writes:

> On Mon, 28 Jan 2002, Thomas Bushnell, BSG wrote:
>
> > ravage@einstein.ssz.com (Jim Choate) writes:
> >
> > > Sony products should be avoided until they change their tune toward Aibo
> > > hackers...
> >
> > Say more--I don't know what the issue is to which you refer, but I'm
> > interested.
>
> Google Aibo, it's current news.

I tried, but it doesn't get me far.  The first gajillion pages are all
Sony or resellers, and wouldn't say much anyway.  I went to the
www.aibosite.com page (which bills itself as for Aibo hackers), but I
can't find any info there...

Re: [9fans] Re: ideal plan9 laptop

[Jim Choate]

On Mon, 28 Jan 2002, Thomas Bushnell, BSG wrote:

> ravage@einstein.ssz.com (Jim Choate) writes:
>
> > Sony products should be avoided until they change their tune toward Aibo
> > hackers...
>
> Say more--I don't know what the issue is to which you refer, but I'm
> interested.

Google Aibo, it's current news.


 --
    ____________________________________________________________________

             Day by day the Penguins are making me lose my mind.

                                             Bumper Sticker

       The Armadillo Group       ,::////;::-.          James Choate
       Austin, Tx               /:'///// ``::>/|/      ravage@ssz.com
       www.ssz.com            .',  ||||    `/( e\      512-451-7087
                           -====~~mm-'`-```-mm --'-
    --------------------------------------------------------------------

Re: [9fans] Re: ideal plan9 laptop

[Matt H]

> too primitive for much else except pointing and grunting so one has to have
> a mouse anyway. [ask any graphics artist, or a long time illustrator user
> like me]

drawing, mouse, ugh!

Re: [9fans] Re: ideal plan9 laptop

[ozan s yigit]

ravage@einstein.ssz.com (Jim Choate) writes about the little keyboard
pointing thing:

> Yes it is, in fact it's the optimal pointing device available today.

too primitive for much else except pointing and grunting so one has to have
a mouse anyway. [ask any graphics artist, or a long time illustrator user
like me]

oz
---
www.cs.yorku.ca/~oz	 | if you couldn't find any weirdness, maybe
york u. computer science | we'll just have to make some!   -- hobbes

Re: [9fans] Re: ideal plan9 laptop

[Thomas Bushnell, BSG]

ravage@einstein.ssz.com (Jim Choate) writes:

> Sony products should be avoided until they change their tune toward Aibo
> hackers...

Say more--I don't know what the issue is to which you refer, but I'm
interested.

[9fans] Re: ideal plan9 laptop

[Jim Choate]

On Fri, 25 Jan 2002, Boyd Roberts wrote:

> Ronald G Minnich wrote:
> > Any recommendations?
>
> I think Sony VAIOs are great:

Sony products should be avoided until they change their tune toward Aibo
hackers...


 --
    ____________________________________________________________________

             Day by day the Penguins are making me lose my mind.

                                             Bumper Sticker

       The Armadillo Group       ,::////;::-.          James Choate
       Austin, Tx               /:'///// ``::>/|/      ravage@ssz.com
       www.ssz.com            .',  ||||    `/( e\      512-451-7087
                           -====~~mm-'`-```-mm --'-
    --------------------------------------------------------------------

[9fans] Re: ideal plan9 laptop

[Jim Choate]

On Fri, 25 Jan 2002, andrey mirtchovski wrote:

> > I don't like the little "keyboard eraser" pointing device anymore.
>
> in the more unfashionable northern corners of canada they call it
> 'the nipple'..  probably trying to pass it as an intuitive interface.
>
> which it isn't..

Yes it is, in fact it's the optimal pointing device available today.

AccuPoint's are the cats meow.


 --
    ____________________________________________________________________

             Day by day the Penguins are making me lose my mind.

                                             Bumper Sticker

       The Armadillo Group       ,::////;::-.          James Choate
       Austin, Tx               /:'///// ``::>/|/      ravage@ssz.com
       www.ssz.com            .',  ||||    `/( e\      512-451-7087
                           -====~~mm-'`-```-mm --'-
    --------------------------------------------------------------------