Re: [9fans] Some questions...

Re: [9fans] Some questions...

[presotto]

[-- Attachment #1: Type: text/plain, Size: 723 bytes --]

No, all of our stand alone systems have IDE drives.  We just don't
do that much writing to the disks, only log traffic, and the UPS's
make sure the system doesn't get shut off in surprising ways.  Our
auth server has been up since Dec 1, which is when I took it down to
put a new kernel on it.  I'm not pretending that kfs is robust, just
that it is sufficient in this situation, i.e, as a pretty much read
only fs.  I do use it for weeks at a time on my laptop when travelling,
but I've never used it as a real production system.  It was originally
cloned from a much older version of our file server code and though
we fix bugs when we experience them, we don't work it hard enough to
experience many of them.

[-- Attachment #2: Type: message/rfc822, Size: 1767 bytes --]

From: okamoto@granite.cias.osakafu-u.ac.jp
To: 9fans@cse.psu.edu
Subject: Re: [9fans] Some questions...
Date: Wed, 31 Jan 2001 12:50:17 0900
Message-ID: <20010131035018.A3B36199FF@mail.cse.psu.edu>

>  While kfs isn't
>the most robust software, it's stable enough for something with a limited

Isn't your kfs residing on scsi HDD but not IDE?

I remind now that I had no such trouble on kfs when I met this new release
first time.  I compiled fileserver, auth server kernel on the machine with 48MB
ram using kfs local filesystem.  Only difference between that and present 
notebook is that the notebook uses IDE disk, however, the machine last year 
used scsi (sd53c8xx) local disk.  I may be wrong though, because I didn't use 
kfs so efficiently at that time.

Kenji

Re: [9fans] Some questions

[YAMANASHI Takeshi]

On Fri Jan 30 12:33:45 JST 2004, Geoff Collyer wrote:
> Plan 9's scroll bars let one scroll a precise number of lines, forward
> or backward, or move to a relative position within the window.  I've
> liked this since I first encountered it on the Blit.

I totally agree wtih you about the behaviour of scroll bar in Plan 9.
Once the cursor is placed within the bar, you can go on back and forth
in precisely predictable amount *WITHOUT MOVING THE MOUSE AGAIN*.
This is my most favourite feature of Plan 9 scroll bar.

But still, I use wheel as often as scroll bar.  :)
--

Re: [9fans] Some questions

[Scott Schwartz]

| > Plan 9's scroll bars let one scroll a precise number of lines, forward
| > or backward, or move to a relative position within the window.  I've
| > liked this since I first encountered it on the Blit.

That's not true in sam (last I checked), if lines are long enough to wrap.
In that case, scrolling forward n clicks and then backwards n clicks
doesn't land you where you started.

Re: [9fans] Some questions

[YAMANASHI Takeshi]

On Fri Jan 30 12:11:40 JST 2004, okamoto@granite.cias.osakafu-u.ac.jp wrote:
> In the same logic, we need horizontal scroll button. ?

Now that you mentioned it, I confess that I have desired
a track ball on the mouse instead of the wheel for a couple
of years.

It would be a horrible thing to you all. :)
--

Re: [9fans] Some questions

[mirtchov]

> On Fri Jan 30 12:11:40 JST 2004, okamoto@granite.cias.osakafu-u.ac.jp wrote:
>> In the same logic, we need horizontal scroll button. ?
>
> Now that you mentioned it, I confess that I have desired
> a track ball on the mouse instead of the wheel for a couple
> of years.
>
> It would be a horrible thing to you all. :)
> --

the ibm mouse I posted a link to earlier has a trackpoint, which does
the same thing as a trackball, with less movement

Re: [9fans] Some questions

[YAMANASHI Takeshi]

On Fri Jan 30 11:46:44 JST 2004, boyd, rounin wrote:
> 1) to click you need to waste neruons on not scrolling

oh.  But in some mouse, you can use other buttons instead
of the wheel.  My MS Intellimouse has 4 buttons on it
beside the wheel.

What if there is a mouse with 3 buttons and a wheel on it?
The wheel is only for rolling and not used ever as a button.

> 2) there is a perfectly good interface to scroll (no wheel)

Sometimes, I am too lazy to move cursor to scroll bar.
I just roll the wheel instead.
--

Re: [9fans] Some questions

[okamoto]

> What if there is a mouse with 3 buttons and a wheel on it?
> The wheel is only for rolling and not used ever as a button.

In the same logic, we need horizontal scroll button. ☺

Kenji

Re: [9fans] Some questions

[George Michaelson]

On Fri, 30 Jan 2004 12:11:02 +0900 okamoto@granite.cias.osakafu-u.ac.jp
wrote:

>> What if there is a mouse with 3 buttons and a wheel on it?
>> The wheel is only for rolling and not used ever as a button.
>
>In the same logic, we need horizontal scroll button. ___
>
>Kenji

If we were Apes, we might be able to exploit mice with more degrees of
freedom, but alas my pre-arthritic (not really, but it does feel like it
sometimes) fingers really only have two apparent motions, not three.

thus pressing, and front-back scrolling by a finger is possible, while
side to side is not.

side to side and whole-of-mouse moving is a *HAND* motion. Personally I
prefer to give my computer the finger.

-george

Re: [9fans] Some questions

[ron minnich]

On Fri, 30 Jan 2004, YAMANASHI Takeshi wrote:

> Sometimes, I am too lazy to move cursor to scroll bar.
> I just roll the wheel instead.


My only observation is that people who have the wheel really like it. Yes,
there are other ways to get what the wheel does, and yes, folks who like
wheels are aware of those ways. But, if the wheel is there, they use it
and like it.

ron

Re: [9fans] Some questions

[okamoto]

>But, if the wheel is there, they use it 
> and like it.

I have one for my home Win2k, but I don't use that wheel.
Probably because I'm "peculiar man". ☺

Kenji

Re: [9fans] Some questions

[ron minnich]

On Fri, 30 Jan 2004 okamoto@granite.cias.osakafu-u.ac.jp wrote:

> I have one for my home Win2k, but I don't use that wheel.
> Probably because I'm "peculiar man". ☺
> 

well let's put it this way. My 83-year-old dad not only adopted to the
scroll wheel but he bought a mouse with a few extra buttons, then he
programmed the button functions, and gets along fine. I find all those
buttons annoying. It's all personal preference.  We're all peculiar.

ron

Re: [9fans] Some questions

[okamoto]

> buttons annoying. It's all personal preference.  We're all peculiar.

Let's remember again, mouse has been called "pointing" device,
and then, it has a ball and some buttons which is considered to
be all for that function.   If something must be 'INNOVATED', we have
to recognise and accept the falts of that concept of pointing device.
Yes, I'm a fundametarist (not in the meaning of rerigous one of course).

I should not accept anything only by the reason it's convenient! ☺

Kenji

Re: [9fans] Some questions

[George Michaelson]

>Yes, I'm a fundametarist (not in the meaning of rerigous one of course).
>

gosh. does a fundamentarist get to study... fundaments? I hope we get to
the bottom of this shortly...

-George

Re: [9fans] Some questions

[Geoff Collyer]

What presotto said.

Also, unfortunately the wheel gives imprecise control of scrolling.
You move the wheel, and the screen jumps by some unpredictable amount.
If you have to scroll through a lot of text, it can take a long time
to get there.  So I've used the wheel and don't like it.

Plan 9's scroll bars let one scroll a precise number of lines, forward
or backward, or move to a relative position within the window.  I've
liked this since I first encountered it on the Blit.

Re: [9fans] Some questions

[okamoto]

> Plan 9's scroll bars let one scroll a precise number of lines, forward
> or backward, or move to a relative position within the window.  I've
> liked this since I first encountered it on the Blit.

I have been always having some doubts why that scroll mouse appeared.
It must be done because industry needs 'INNOVATION" all the time to feed
some foods to them...  Too much of 'INNOVATIONS' these days!

Kenji

Re: [9fans] Some questions

[Geoff Collyer]

I'm just surprised that they didn't put the wheel on the keyboard, so
that all those people who hate using a mouse could scroll quickly
without taking their hands off their beloved keyboard.

Re: [9fans] Some questions

[Lyndon Nerenberg]

--On 2004-1-29 7:44 PM -0800 Geoff Collyer <geoff@collyer.net> wrote:

> I'm just surprised that they didn't put the wheel on the keyboard, so
> that all those people who hate using a mouse could scroll quickly
> without taking their hands off their beloved keyboard.

Actually, having the keyboard nipple do scrolling instead of mouse
movement seems appealing.

--lyndon

Re: [9fans] Some questions

[Fco.J.Ballesteros]

> I'm just surprised that they didn't put the wheel on the keyboard, so
> that all those people who hate using a mouse could scroll quickly
> without taking their hands off their beloved keyboard.

Are you sure? Mi keyboard has a wheel (logitech bluetooth kbd).

The thing is that once I have my hand over the mouse, it's easier to use
that wheel than it is to move the keyboard hand out of its typing position
just to reach the wheel.

Re: [9fans] Some questions

[Geoff Collyer]

[-- Attachment #1: Type: text/plain, Size: 88 bytes --]

Sigh.  Just demonstrates that reality has become so twisted that it's
hard to parody.

[-- Attachment #2: Type: message/rfc822, Size: 2188 bytes --]

From: Fco.J.Ballesteros <nemo@plan9.escet.urjc.es>
To: 9fans@cse.psu.edu
Subject: Re: [9fans] Some questions
Date: Fri, 30 Jan 2004 08:28:46 +0100
Message-ID: <aae12744d80fc572f008c0c4d27874c7@plan9.escet.urjc.es>

> I'm just surprised that they didn't put the wheel on the keyboard, so
> that all those people who hate using a mouse could scroll quickly
> without taking their hands off their beloved keyboard.

Are you sure? Mi keyboard has a wheel (logitech bluetooth kbd).

The thing is that once I have my hand over the mouse, it's easier to use
that wheel than it is to move the keyboard hand out of its typing position
just to reach the wheel.

Re: [9fans] Some questions

[Fco.J.Ballesteros]

> Sigh.  Just demonstrates that reality has become so twisted that it's
> hard to parody.

Yep. It's even more amazing, there is a key with label "buying". I didn't
try it since I couldn't find the card reader for the VISA. The last is a joke,
but the former is not.

Re: [9fans] Some questions

[matt]

> I'm just surprised that they didn't put the wheel on the keyboard

newer super duper Microsoft keyboards have a wheel on the left of the keyboard near esc

presumably others will follow

Re: [9fans] Some questions

[boyd, rounin]

> newer super duper Microsoft keyboards have a wheel on the left of the
keyboard near esc

sony seem to be fascinated by 'jog dials'. i have one on soma, but _never_
use it.

Re: [9fans] Some questions

[ron minnich]

On Sun, 8 Feb 2004 matt@proweb.co.uk wrote:

> > I'm just surprised that they didn't put the wheel on the keyboard
>
> newer super duper Microsoft keyboards have a wheel on the left of the
> keyboard near esc
>
> presumably others will follow


not to mention the keyboard with the optical mouse sensor on the bottom.
You can move the keyboard around and it replaces your mouse!

ron

Re: [9fans] Some questions

[boyd, rounin]

> Plan 9's scroll bars let one scroll a precise number of lines, forward
> or backward, or move to a relative position within the window.  I've
> liked this since I first encountered it on the Blit.

yup

Re: [9fans] Some questions

[Jack Johnson]

On Jan 29, 2004, at 7:32 PM, Geoff Collyer wrote:
> Plan 9's scroll bars let one scroll a precise number of lines, forward
> or backward, or move to a relative position within the window.

I've always liked page's interface: just drag it where you want it
(though now I'm even more partial to 1280x1024).  The
click-to-next-page is nice, too.

I've often thought it would be an effective way to ditch scrollbars
altogether, maybe using a key chord (or additional button?) to signify
that you wanted to scroll rather than click/select/cut/snarf/paste.

-Jack

Re: [9fans] Some questions

[Geoff Collyer]

page's drag-it-where-you-want-it interface is fine as far as it goes,
but it doesn't replace the middle button on the scrollbar for relative
positioning.  It works in page because you are only ever positioning a
single page on the screen, and the right button is used to position
within the document.

Re: [9fans] Some questions

[a]

// I've often thought it would be an effective way to ditch scrollbars 
// altogether, maybe using a key chord (or additional button?) to signify 
// that you wanted to scroll rather than click/select/cut/snarf/paste.

um, like PageUp/PageDown? :-)

i will say that MS has some feature in at least some of their apps where
pressing and holding a button (middle? right?) and then moving the mouse
will cause the page to scroll in that direction proportionally to the
distance from the original pressing point, and that's kinda nice. i've
used it several times, and while it was awkward for me personally, i got
the impression that said awkwardness was mostly because i wasn't used to
it; i've got a friend who i observe using it frequently without issue.
wheel mice, however, i've simply detested every time i've used one. they
get in the way, are easy to hit as a button accidentally, and hard to use
as a button intentionally. i'm not tied to the three-button model (i've
had a four-button logitech a few years back that i really liked), but the
wheel just feels like a kludgy "feature".
ア

Re: [9fans] Some questions

[Charles Forsyth]

[-- Attachment #1: Type: text/plain, Size: 63 bytes --]

i suspect they don't use middle-button chording much, though.

[-- Attachment #2: Type: message/rfc822, Size: 2686 bytes --]

From: ron minnich <rminnich@lanl.gov>
To: 9fans@cse.psu.edu
Subject: Re: [9fans] Some questions
Date: Thu, 29 Jan 2004 20:17:38 -0700 (MST)
Message-ID: <Pine.LNX.4.44.0401292016330.28001-100000@maxroach.lanl.gov>

On Fri, 30 Jan 2004, YAMANASHI Takeshi wrote:

> Sometimes, I am too lazy to move cursor to scroll bar.
> I just roll the wheel instead.


My only observation is that people who have the wheel really like it. Yes,
there are other ways to get what the wheel does, and yes, folks who like
wheels are aware of those ways. But, if the wheel is there, they use it
and like it.

ron

Re: [9fans] Some questions

[David Presotto]

[-- Attachment #1: Type: text/plain, Size: 735 bytes --]

Some of us are more dexterous than others (or sinister if
we use left hand mice?).  If you can handle a 6 button mouse
with a scroll wheel attached (or a one handed keyboard  for
that matter), more power to you.  I just don't do well
with more than 3 and want to make sure they don't all
disappear.  For that matter, I can just barely handle chording on
acme and I know lots of old mac users that were apalled at
3 button mice.

I had a mouse with 4 for a while and ended up sticking a straw
under the 4th one so I wouldn't keep accidentally clicking it.
The mice with wheels doubling as a button 2, I find even more
annoying since it makes button 2 harder to use than the other
two buttons.  I tend to hit it by accident.

[-- Attachment #2: Type: message/rfc822, Size: 2674 bytes --]

From: YAMANASHI Takeshi <uncover@beat.cc.titech.ac.jp>
To: 9fans@cse.psu.edu
Subject: Re: [9fans] Some questions
Date: Fri, 30 Jan 2004 12:04:45 +0900
Message-ID: <96f058ee1bda3356f6c3794a1ee47eaf@orthanc.cc.titech.ac.jp>

On Fri Jan 30 11:46:44 JST 2004, boyd, rounin wrote:
> 1) to click you need to waste neruons on not scrolling

oh.  But in some mouse, you can use other buttons instead
of the wheel.  My MS Intellimouse has 4 buttons on it
beside the wheel.

What if there is a mouse with 3 buttons and a wheel on it?
The wheel is only for rolling and not used ever as a button.

> 2) there is a perfectly good interface to scroll (no wheel)

Sometimes, I am too lazy to move cursor to scroll bar.
I just roll the wheel instead.
--

Re: [9fans] Some questions

[YAMANASHI Takeshi]

On Fri Jan 30 00:35:39 JST 2004, boyd, rounin wrote:
> better.  damn, why don't they just use plain ol' buttons? :(

Just for my curiosity, why you guys regard the wheel
distasteful?
--

Re: [9fans] Some questions

[boyd, rounin]

> Just for my curiosity, why you guys regard the wheel
> distasteful?

1) to click you need to waste neruons on not scrolling

2) there is a perfectly good interface to scroll (no wheel)

Re: [9fans] Some questions

[David Presotto]

[-- Attachment #1: Type: text/plain, Size: 24 bytes --]

We enjoy the bumpy ride.

[-- Attachment #2: Type: message/rfc822, Size: 2354 bytes --]

From: YAMANASHI Takeshi <uncover@beat.cc.titech.ac.jp>
To: 9fans@cse.psu.edu
Subject: Re: [9fans] Some questions
Date: Fri, 30 Jan 2004 11:36:50 +0900
Message-ID: <789e0ff6967657d899173069a104dbc5@orthanc.cc.titech.ac.jp>

On Fri Jan 30 00:35:39 JST 2004, boyd, rounin wrote:
> better.  damn, why don't they just use plain ol' buttons? :(

Just for my curiosity, why you guys regard the wheel
distasteful?
--

Re: [9fans] Some questions

[davide+p9]

> The Sun type-6 mice aren't bad:

...except for not being optical (yet another
case of "What *were* they thinking over at
the SunMouse division?  Haven't they realized
that ball mice are 20th-century?").

I defenestrated mine on day 2 in favor of
a Compaq PS2/USB optical on sale at Radio
Shack.

Dave Eckhardt

Re: [9fans] Some questions

[David Presotto]

On Wed Jan 28 08:49:47 EST 2004, anastas@ceid.upatras.gr wrote:
> 1) How often is plan9.iso.bz2 regenerated(both binaries and iso image)?

Every morning at 4:45AM EST if the source tree has changed.

> 2) What is the status of libcontrol? I am not asking about stability but
> about its design. Are you satisfied with it or you lack ( time |
> resources | better ideas )* to build somesting else? Are there any plan9
> applications that make heavy use of it like EFIS
> (http://wwwhome.cs.utwente.nl/~sape/gos/plan9.ps)?

We use it for the factotum gui and for an audio juke box.

> 3) The first edition of the system contained a chess database chdb(7). Is
> the source still available?

Ask ken@entrisphere.com

[9fans] Some questions

[Anastasopoulos S]

Hi

I have three easy to answer questions

1) How often is plan9.iso.bz2 regenerated(both binaries and iso image)?

2) What is the status of libcontrol? I am not asking about stability but
about its design. Are you satisfied with it or you lack ( time |
resources | better ideas )* to build somesting else? Are there any plan9
applications that make heavy use of it like EFIS
(http://wwwhome.cs.utwente.nl/~sape/gos/plan9.ps)?
By the way was there a GUI for the blit terminal?

3) The first edition of the system contained a chess database chdb(7). Is
the source still available?

Thanks
Spyros

Re: [9fans] Some questions

[boyd, rounin]

> By the way was there a GUI for the blit terminal?

sure was.  that was the whole point.

now if i could just get a depraz mouse ...

Re: [9fans] Some questions

[jmk]

On Wed Jan 28 09:01:40 EST 2004, boyd@insultant.net wrote:
> > By the way was there a GUI for the blit terminal?
>
> sure was.  that was the whole point.
>
> now if i could just get a depraz mouse ...
>

On Mon Sep 24 18:40:32 EDT 2001, jmk@plan9.bell-labs.com wrote:
> I'm afraid you'd be disappointed if you used a Depraz mouse today.
> A couple of years ago we had some converted to USB only to discover
> that even the cheapest modern mice are better mechanically, the
> buttons are incredibly clunky and the mice were not very accurate in
> positioning (I'm not talking about the mouse shape here, I still
> think the Depraz has the others beat hands down).
>
> On Mon Sep 24 18:30:33 EDT 2001, boyd@fr.inter.net wrote:
> > gimme a 5620 keyboard and a dupraz (sp?) mouse anyday.
> >

Re: [9fans] Some questions

[splite]

On Mon Sep 24 18:40:32 EDT 2001, jmk@plan9.bell-labs.com wrote:
> I'm afraid you'd be disappointed if you used a Depraz mouse today.
> A couple of years ago we had some converted to USB only to discover
> that even the cheapest modern mice are better mechanically, the
> buttons are incredibly clunky and the mice were not very accurate in
> positioning (I'm not talking about the mouse shape here, I still
> think the Depraz has the others beat hands down).

No attempt to put a modern mouse's guts in a Depraz shell?


Couple of good color pictures of the red Depraz mouse in:

http://www.billbuxton.com/input06.ChordKeyboards.pdf

Re: [9fans] Some questions

[boyd, rounin]

> No attempt to put a modern mouse's guts in a Depraz shell?

story i heard is that they cost $US 500 a pop in the mid 80's -- ouch.

Re: [9fans] Some questions

[Martin C.Atkins]

On Wed, 28 Jan 2004 10:52:42 -0500 jmk@plan9.bell-labs.com wrote:
> On Wed Jan 28 09:01:40 EST 2004, boyd@insultant.net wrote:
>...
> > now if i could just get a depraz mouse ...
> On Mon Sep 24 18:40:32 EDT 2001, jmk@plan9.bell-labs.com wrote:
> > I'm afraid you'd be disappointed if you used a Depraz mouse today.
>...
> >
> > On Mon Sep 24 18:30:33 EDT 2001, boyd@fr.inter.net wrote:
> > > gimme a 5620 keyboard and a dupraz (sp?) mouse anyday.

Having downloaded:
	http://www.billbuxton.com/input06.ChordKeyboards.pdf
I see that the Depraz mouse is indeed what I thought it was. I think
I used one on a Whitechapel workstation (anyone remember them! :-),
way back.

I thought that the idea, while attractive and initially very
comfortable, was flawed because it forced you to use whole arm
movements (keeping the wrist off the table), rather than wrist and
finger movements with the arm/wrist resting on the table.
Consequently I found it much more tiring to use than a 'normal'
mouse, and also much more difficult to position precisely.

Was others' experience more positive?

Martin

--
Martin C. Atkins			martin@parvat.com
Parvat Infotech Private Limited		http://www.parvat.com{/,/martin}

Re: [9fans] Some questions

[boyd, rounin]

this thing has a wheel (vomit), but is optical.  it's nice to
use, althought a little small:

    http://www.mitsubishi-electric.com.au/PRODUCTS/COMPP/mice/DT210M.htm

the depraz were clunky, but the only game in town back in '85;  coupla blits
and one jerq.

Re: [9fans] Some questions

[mirtchov]

I use this one daily (for the past year and a half):

http://www-132.ibm.com/webapp/wcs/stores/servlet/ProductDisplay?catalogId=-840&langId=-1&partNumber=31P7405&storeId=1

it's optical, has no wheel (the blue thing in the middle is analogous
to the little 'nipple' on the ibm laptops, but doesn't work under Plan
9 or X) and is large and sturdy.

the middle button is somewhat small, but I've never had a problem
missing chords with it..

my $0.02 :)

andrey

Re: [9fans] Some questions

[boyd, rounin]

>
http://www-132.ibm.com/webapp/wcs/stores/servlet/ProductDisplay?catalogId=-840&langId=-1&partNumber=31P7405&storeId=1

better.  damn, why don't they just use plain ol' buttons? :(

Re: [9fans] Some questions

[a]

my experience was quite positive. i found it more comfortable to
use for long periods of time that most other mice i've used, and
the "clunky" buttons nice (reliable, obvious when you've made
contact, and hard to trigger inadvertantly). i got two of the
usb-converted mice jmk mentioned, and (except for the busted one
which moved the cursor down about quarter inch for every inch it
moved up on the screen) didn't find them more dificult to position
than average (not the best, but certainly not the worst). the
mechanics are worse than (modern) average, though. but then, good
optical mice have me spoiled.

bah. all mice suck today. i've got a running thread with boyd
trying to identify a usb three-button (wheels aren't buttons)
mouse with either optics or high-quality mechanics and a shape
that isn't plain stupid. it's not been particularly fruitful.
ア

Re: [9fans] Some questions

[boyd, rounin]

> bah. all mice suck today. i've got a running thread with boyd
> trying to identify a usb three-button (wheels aren't buttons)
> mouse with either optics or high-quality mechanics and a shape
> that isn't plain stupid. it's not been particularly fruitful.

no.  piers was lamenting the same folly, when caught up
with him in sydney.  as Valen Halen said:

    you can't get this stuff no more

i have a bunch of 3 button mice (no wheels), but they're
not optical.  i'm sold on optical (now) but the price you
pay is that damn wheel.

it would cost _less_ to make 'em without the wheel, but ...

Re: [9fans] Some questions

[splite]

On Thu, Jan 29, 2004 at 04:52:02AM -0500, a@9srv.net wrote:
>
> bah. all mice suck today. i've got a running thread with boyd
> trying to identify a usb three-button (wheels aren't buttons)
> mouse with either optics or high-quality mechanics and a shape
> that isn't plain stupid. it's not been particularly fruitful.

The Belkin 3-button USB ball mouse isn't too terrible.  It's similar in
shape to the current 3-button Logitechs, if you like that sort of thing.
Mechanics aren't the greatest, but probably better that a Depraz.  Nice
part is that they're often free after rebate from OfficeMax.  (And no
damn wheel, either.)

A few years ago I saw a three-button mouse with two wheels, one on each
side of the center button.  It was a cheap-looking no-brand job so I didn't
buy it, but I thought the idea was interesting.

Re: [9fans] Some questions

[boyd, rounin]

> The Belkin 3-button USB ball mouse isn't too terrible.

the logitec gaming mouse is great, but a) it's not optical
and b) you can't get this stuff no more.  i have one.

Re: [9fans] Some questions

[Scott Schwartz]

| trying to identify a usb three-button (wheels aren't buttons)
| mouse with either optics or high-quality mechanics and a shape
| that isn't plain stupid.

The Sun type-6 mice aren't bad:

	http://www.memoryxsun.com/3703632.html

Re: [9fans] Some questions

[jj]

a@9srv.net wrote in message 
> bah. all mice suck today. i've got a running thread with boyd
> trying to identify a usb three-button (wheels aren't buttons)
> mouse with either optics or high-quality mechanics and a shape
> that isn't plain stupid. it's not been particularly fruitful.

I used this one once (non-optical version) and I liked it a lot.
It fits nicely on the hand, there is 3 buttons and the wheel does not
seem to be a problem if you plan on ignoring it...now the price is
another issue

    http://www.contourdesign.com/pmo/


Did anybody tried it ? 

jj

Re: [9fans] Some questions

[boyd, rounin]

>     http://www.contourdesign.com/pmo/

sure is a strange looking thing.  beware:

    http://www.contourdesign.com/cgi-bin/webstore/web_store.pl

    This form is for orders shipping to USA, Canada and Mexico ONLY.

    If you are located outside the USA, Canada or Mexico and would like
     to order one of our products, please visit the appropriate reseller
     link below for a list of dealers in your region. If you do not see a
reseller
     in your area, please contact us:

        http://www.contourdesign.com/contact_info.htm

Re: [9fans] Some questions

[boyd, rounin]

> 3) The first edition of the system contained a chess database chdb(7). Is
> the source still available?

doubtful, but presotto should know.

Re: [9fans] Some questions

[mirtchov]

> 2) What is the status of libcontrol? I am not asking about stability but
> about its design. Are you satisfied with it or you lack ( time |
> resources | better ideas )* to build somesting else?

I had modified Charles Forsyth's 2e MineSweeper for 3e to use the
control(2) library heavily -- each square on the grid was a control
widget, all sharing the same channel on which the main chunk of the
code sent messages around.  the control library also took care of
displaying the bitmapped images.

	http://pages.cpsc.ucalgary.ca/~mirtchov/lanlp9/minesweeper/

I found libcontrol perfectly capable of standing on its own.  At the
time I was unhappy how some of the components behaved, but changing
them also proved painless.

It broke when 4e came out (I just found out recently) due to a change
in the control message convention.  I haven't bothered to rewrite it
(it was just an exercise in using libcontrol) because it was
superseded by a more colourful and correct interpretation of windows'
MineSweeper:

	http://www.vitanuova.com/plan9/co/mines

andrey

Re: [9fans] Some questions

[David Presotto]

[-- Attachment #1: Type: text/plain, Size: 93 bytes --]

I think we're talking about different libcontrols.  The current one is
new, written by sape.

[-- Attachment #2: Type: message/rfc822, Size: 2825 bytes --]

From: mirtchov@cpsc.ucalgary.ca
To: 9fans@cse.psu.edu
Subject: Re: [9fans] Some questions
Date: Wed, 28 Jan 2004 08:32:51 -0700
Message-ID: <7a0d071bb28c1d5190453f39024cc60b@plan9.ucalgary.ca>

> 2) What is the status of libcontrol? I am not asking about stability but
> about its design. Are you satisfied with it or you lack ( time |
> resources | better ideas )* to build somesting else?

I had modified Charles Forsyth's 2e MineSweeper for 3e to use the
control(2) library heavily -- each square on the grid was a control
widget, all sharing the same channel on which the main chunk of the
code sent messages around.  the control library also took care of
displaying the bitmapped images.

	http://pages.cpsc.ucalgary.ca/~mirtchov/lanlp9/minesweeper/

I found libcontrol perfectly capable of standing on its own.  At the
time I was unhappy how some of the components behaved, but changing
them also proved painless.

It broke when 4e came out (I just found out recently) due to a change
in the control message convention.  I haven't bothered to rewrite it
(it was just an exercise in using libcontrol) because it was
superseded by a more colourful and correct interpretation of windows'
MineSweeper:

	http://www.vitanuova.com/plan9/co/mines

andrey

Re: [9fans] Some questions

[mirtchov]

> I think we're talking about different libcontrols.  The current one is
> new, written by sape.

my mistake...  they looked sufficiently similar to assume the old one
has been modified.

Re: [9fans] Some questions

[Sape Mullender]

>> I think we're talking about different libcontrols.  The current one is
>> new, written by sape.
>
> my mistake...  they looked sufficiently similar to assume the old one
> has been modified.

That's right.  The bulk is rob's; the composition of controls is mine.

	Sape

Re: [9fans] Some questions

[David Presotto]

[-- Attachment #1: Type: text/plain, Size: 78 bytes --]

Ignore that last message, I misread andrey's message.  I read
2e and swooned.

[-- Attachment #2: Type: message/rfc822, Size: 4984 bytes --]

[-- Attachment #2.1.1: Type: text/plain, Size: 93 bytes --]

I think we're talking about different libcontrols.  The current one is
new, written by sape.

[-- Attachment #2.1.2: Type: message/rfc822, Size: 2825 bytes --]

From: mirtchov@cpsc.ucalgary.ca
To: 9fans@cse.psu.edu
Subject: Re: [9fans] Some questions
Date: Wed, 28 Jan 2004 08:32:51 -0700
Message-ID: <7a0d071bb28c1d5190453f39024cc60b@plan9.ucalgary.ca>

> 2) What is the status of libcontrol? I am not asking about stability but
> about its design. Are you satisfied with it or you lack ( time |
> resources | better ideas )* to build somesting else?

I had modified Charles Forsyth's 2e MineSweeper for 3e to use the
control(2) library heavily -- each square on the grid was a control
widget, all sharing the same channel on which the main chunk of the
code sent messages around.  the control library also took care of
displaying the bitmapped images.

	http://pages.cpsc.ucalgary.ca/~mirtchov/lanlp9/minesweeper/

I found libcontrol perfectly capable of standing on its own.  At the
time I was unhappy how some of the components behaved, but changing
them also proved painless.

It broke when 4e came out (I just found out recently) due to a change
in the control message convention.  I haven't bothered to rewrite it
(it was just an exercise in using libcontrol) because it was
superseded by a more colourful and correct interpretation of windows'
MineSweeper:

	http://www.vitanuova.com/plan9/co/mines

andrey

Re: [9fans] Some questions

[okamoto]

> 2) What is the status of libcontrol? I am not asking about stability but
> about its design. Are you satisfied with it or you lack ( time |
> resources | better ideas )* to build somesting else?

We used its older version for our application, which you can get from
http://basalt.cias.osakafu-u.ac.jp/plan9/marsv.tgz.   However, I'm afraid
you cannot compile it correctly on your latest Plan 9 distribution.
This is because we used ourown layer structures in it.  I know it's
better to rewrite it, however, it is still working here, and I have no
enough time to do so now...

Kenji

Re: [9fans] Some questions...

[tad]

I've got a CPU server running KFS that's been up since August:

http://vampira.csh.rit.edu

	vampira% ls -l '#c'
	--rw-rw-rw- c 0 bootes bootes  0 Aug 11 01:51 #c/authcheck
	...

-Tad

Re: [9fans] Some questions...

[okamoto]

>  While kfs isn't
>the most robust software, it's stable enough for something with a limited

Isn't your kfs residing on scsi HDD but not IDE?

I remind now that I had no such trouble on kfs when I met this new release
first time.  I compiled fileserver, auth server kernel on the machine with 48MB
ram using kfs local filesystem.  Only difference between that and present 
notebook is that the notebook uses IDE disk, however, the machine last year 
used scsi (sd53c8xx) local disk.  I may be wrong though, because I didn't use 
kfs so efficiently at that time.

Kenji

Re: [9fans] Some questions...

[forsyth]

>>I'm enthused :)

it's worth it.

Re: [9fans] Some questions...

[presotto]

[-- Attachment #1: Type: text/plain, Size: 557 bytes --]

We have a number of CPU servers and file servers.  We run the auth server
on a kfs machine with a UPS.  That machine serves bootp/dhcp/tftp, authentication, and
consolefs.  It gives us a stable point in the network.  While kfs isn't
the most robust software, it's stable enough for something with a limited
function.  It gives us someplace to stand while repairing the world when
catastrophe strikes.  I also keep a copy of the the key files and kernels
and /lib/ndb on the file server so that we can rebuild the auth server
should it ever blow up.

[-- Attachment #2: Type: message/rfc822, Size: 1739 bytes --]

From: okamoto@granite.cias.osakafu-u.ac.jp
To: 9fans@cse.psu.edu
Subject: Re: [9fans] Some questions...
Date: Tue, 30 Jan 2001 10:45:48 0900
Message-ID: <20010130014611.DDE6B199E6@mail.cse.psu.edu>

>So I'm lead to wonder: Is it usually the case that the auth/CPU server is
>diskless? Or do you typically have an auth/bootp server with its own disk which
>is used to bring up the whole network? 

I don't see any reason to have kfs (even if it'd be robust enough) in auth/cpu
server.  Furthermore, diskless auth server can make the system more stable,
if we could assume that one of the most weak points in the present day PS 
hardware would be harddisk drive.  I suppose here we don't change file server 
so often.  ^_^
Any other idea?

Kenji

Re: [9fans] Some questions...

[Mike Acar]

presotto@plan9.bell-labs.com wrote:
> 
> We have a number of CPU servers and file servers.
[snip]

Ok, thanks to all who responded. I reconfigured things along the lines of
Charles Forsyth's example; now I just have to find another machine to try as a
CPU server to test things out a bit... And some terminals... Then maybe to try
hacking in some support for newer video cards...

I'm enthused :)

-- 
Brilliance and gorgeousness                        |   Mike Acar
And we tell ourselves we don't want the treasures  |   mike@trolltech.com
But we hate the glass anyway                       |

Re: [9fans] Some questions...

[forsyth]

>>I really hope the bootargs method works; that'll make me very
>>happy....

bootargs=il -g GATEWAY-IP ether /net/ether0 IP-ADDRESS IP-MASK 0 FILESERVER-IP IP-ADDRESS

Re: [9fans] Some questions...

[okamoto]

>sans a DHCP server, how does a ``diskless'' machine figure out it's own
>IP address?

I hardcoded it into our auth/cpu server kernel.  :-)
I suppose some more elegant way must be there now, however,
I'm keeping the words that if you have no problem in something, 
don't touch it.  ^_^

Kenji

Re: [9fans] Some questions...

[Dan Cross]

In article <20010130064701.60BC8199E6@mail.cse.psu.edu> you write:
>>sans a DHCP server, how does a ``diskless'' machine figure out it's own
>>IP address?
>
>I hardcoded it into our auth/cpu server kernel.  :-)

Ha!  :-)

>I suppose some more elegant way must be there now, however,

I really hope the bootargs method works; that'll make me very
happy....

>I'm keeping the words that if you have no problem in something, 
>don't touch it.  ^_^

Here here!  btw- if you want the real American slang effect,
go back to the original: ``if it ain't broke, don't fix it!''  :-)

	- Dan C.

Re: [9fans] Some questions...

[rsc]

You'd have to put the kernel on a floppy
disk and set fs= and auth= in plan9.ini.
It only consults the IP address entries when
ipconfig can't find them from DHCP.

Russ

Re: [9fans] Some questions...

[Dan Cross]

In article <200101300350.WAA16390@smtp2.fas.harvard.edu> you write:
>You'd have to put the kernel on a floppy
>disk and set fs= and auth= in plan9.ini.
>It only consults the IP address entries when
>ipconfig can't find them from DHCP.

Thanks for the note, Russ.  That aspect of things is documented in
plan9.ini(8); I was confused about a more basic issue: given a network
sans a DHCP server, how does a ``diskless'' machine figure out it's own
IP address?

As usual, I'm guilty of not reading enough before posting.  :-)  From
reading in the 9fans archives, it would seem that one can specify
ipconfig style arguments on the bootargs line when selecting il in
plan9.ini.  That solves that problem, and I can't wait to try it out on
the Plan 9 network when I get into the office tomorrow....

Hmmm....  I seem to be getting confused a lot.  Guilliani must be
putting something in the water.  ;-)

	- Dan C.

Re: [9fans] Some questions...

[okamoto]

>So I'm lead to wonder: Is it usually the case that the auth/CPU server is
>diskless? Or do you typically have an auth/bootp server with its own disk which
>is used to bring up the whole network? 

I don't see any reason to have kfs (even if it'd be robust enough) in auth/cpu
server.  Furthermore, diskless auth server can make the system more stable,
if we could assume that one of the most weak points in the present day PS 
hardware would be harddisk drive.  I suppose here we don't change file server 
so often.  ^_^
Any other idea?

Kenji

Re: [9fans] Some questions...

[Dan Cross]

In article <20010130014611.DDE6B199E6@mail.cse.psu.edu> you write:
>Any other idea?

Well, here's my naive question for the day: without a diskful boot
server machine, how does one boot the boot server (since it has no
one to bootp off of)?  My PC CPU server doesn't seem to ask for an
IP address when it boots, and I can't seem to find a way to set it
from plan9.ini....

I'm curious about this, too....

	- Dan C.

Re: [9fans] Some questions...

[forsyth]

in the past i've run a standalone auth/boot
server with a local (private) file system, as
some others do.

i didn't want to do that this time.
currently, at a suitable point in our /rc/bin/cpurc at work i do:

switch($sysname) {
case doppio
	ip/dhcpd
	ip/tftpd

	auth/keyfs -m /mnt/keys /adm/keys >/dev/null >[2=1]
	aux/listen -q -d /rc/bin/service -t /rc/bin/service.auth il
	aux/listen -q -d /rc/bin/service -t /rc/bin/service.auth tcp
	auth/cron >>/sys/log/cron >[2=1] &
	sleep 3	# give auth a chance

case *
	aux/listen -q -d /rc/bin/service il
	aux/listen -q -d /rc/bin/service tcp
}

which starts authentication and booting services on only one server (doppio).
it shares the same file server as everything else.

[9fans] Some questions...

[Mike Acar]

So, after some interesting experiments and lots of digging through DejaNews
we've gotten a Plan 9 file and CPU/authentication server working. But I have
some questions regarding the normal network setup...

I installed a standalone CPU/auth server, following the instructions in the
Getting Started document. Then I built a plan9pc fs kernel, booted it on our
fileserver, configured the filesystems, mounted them on the CPU/auth server,
and used wrap/inst install the distribution on them.

At this point, though, I realized that people logging into the CPU/auth server
from terminals won't get the filesystems exported by the file server, but
instead the CPU/auth server's local disk. So I again followed the instructions
from the Getting Started document to get a CPU/auth server (i.e. enable the
authentication services, run them from the cpurc, configure /lib/ndb/local etc)
on the file server filesystem. Life is good now; terminals and the CPU/auth
server have the same root filesystem.

However, all CPU servers booted off this file server will also act as
authentication servers. This seems suboptimal, to say the least :)

So I'm lead to wonder: Is it usually the case that the auth/CPU server is
diskless? Or do you typically have an auth/bootp server with its own disk which
is used to bring up the whole network? The couple of network descriptions I've
found (Charles Forsyth's and Nigel Roles's), while pretty interesting, don't
really answer this question for me.

--
Brilliance and gorgeousness                        |   Mike Acar
And we tell ourselves we don't want the treasures  |   mike@trolltech.com
But we hate the glass anyway                       |