Re: Small MAKEDEV bug

2000-05-09 Thread Sheldon Hearn



On Mon, 08 May 2000 18:56:03 -0400, "Jeroen C. van Gelderen" wrote:

 Now it's a question of "the UNIX way" vs. convenience/userfriendlyness
 :-)

I don't agree.  I think this is an issue of avoiding changes that
unnecessarily astonish existing users.  If you can find ways to improve
MAKEDEV that don't inconvenience those already familiar with it, great.
If your improvements astonish these people, you need to put a bit more
thought into what you're actually scoring.

Lately, we're seeing a lot of vanity changes; this is a trend that can
only end in tears.

Ciao,
Sheldon.


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message



Re: Small MAKEDEV bug

2000-05-09 Thread Sheldon Hearn



On Tue, 09 May 2000 10:26:05 +0200, Sheldon Hearn wrote:

 I don't agree.  I think this is an issue of avoiding changes that
 unnecessarily astonish existing users.  If you can find ways to improve
 MAKEDEV that don't inconvenience those already familiar with it, great.
 If your improvements astonish these people, you need to put a bit more
 thought into what you're actually scoring.

I'd like to follow up on myself with a comment that takes a slightly
broader view on the problem.

As FreeBSD's user base grows, we're going to see two different demands
in increasing contention with each other.

1) The demand for an intuitive, consistent interface.
2) The demand for a constant interface.

The former represents the need for POLA to be upheld throughout the
system at a given moment in time.  The latter represents the need for
POLA to be upheld from one release to the next, in the arena of
"backward compatibility".

Both are valid issues that deserve careful thought.  In many situations,
we can fix inconsistencies "in the moment" without forcing folks to
change the way they do things from on release to the next.  In those
situations where we have to make a decision one way or another, the
decision is going to become more and more difficult as the project moves
forward.

I don't think that the only rule of thumb that can be applied where
the two demands are in contention with one another, is that it's worth
thinking carefully about what we gain for the change.  But we can
_certainly_ agree that it's always worth thinking carefully about ways
to satisfy _both_ demands wherever possible.

Ciao,
Sheldon.


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message



Re: Small MAKEDEV bug

2000-05-09 Thread Bruce Evans

On Mon, 8 May 2000, Tim Vanderhoek wrote:

 On Mon, May 08, 2000 at 06:56:03PM -0400, Jeroen C. van Gelderen wrote:
   
   I don't buy it :-).  This syntax is similar to a special case of the syntax
   of jot(1).  It's better to use jot(1) directly, e.g.:
   
   MAKEDEV $(jot -w da 2 0)# make 2 acd devices beginning at acd0
 
 b$ which jot
 /usr/bin/jot
 b$
 
 The jot utility doesn't appear to be in /bin.

You can just type all the device names or use a shell loop when /usr/bin
is not mounted.

 b$ echo '$(jot -w da 2 0)' | wc
1   5  17
 b$ echo $(jot -w da 2 0) | wc
1   2   8
 b$
 
 Heh.

Yes, it is much faster to type all the device names than to even type
the command to generate them when there are a small number of device
names.

 /me mumbles something about the prototypical UNIX hacker...  :-)

I wouldn't use jot for MAKEDEV'ing disks in practice :-).  Part of
my point is the new syntax for MAKEDEV is just as hard to remember
as the syntax for jot.  You would use it once or twice per millenium
after install more than about 8 drives at once.

Bruce



To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message



Re: Small MAKEDEV bug

2000-05-08 Thread David O'Brien

On Sun, May 07, 2000 at 03:27:07PM -0400, Jeroen C. van Gelderen wrote:
 Or just settle for a more intuitive solution: 
  MAKEDEV acd2   creates /dev/acd2
  MAKEDEV 2 acd  creates /dev/acd[01]
 which would allow for "MAKEDEV 64 da" and "MAKEDEV 256 pty"

I agree with this syntax and after sending my message to you, was sitting
there thinking "MAKEDEV num_of_devs dev_name" would make a really
nice clear syntax.  If you can get BDE's buy-in and other BSD
traditionalists I think this would be great.
 
-- 
-- David([EMAIL PROTECTED])


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message



Re: Small MAKEDEV bug

2000-05-08 Thread Poul-Henning Kamp

In message [EMAIL PROTECTED], "David O'Brien" writes:
On Sun, May 07, 2000 at 03:27:07PM -0400, Jeroen C. van Gelderen wrote:
 Or just settle for a more intuitive solution: 
  MAKEDEV acd2   creates /dev/acd2
  MAKEDEV 2 acd  creates /dev/acd[01]
 which would allow for "MAKEDEV 64 da" and "MAKEDEV 256 pty"

I agree with this syntax and after sending my message to you, was sitting
there thinking "MAKEDEV num_of_devs dev_name" would make a really
nice clear syntax.  If you can get BDE's buy-in and other BSD
traditionalists I think this would be great.

Make it
MAKEDEV -num_of_devs dev_name
and there will be no ambiguity.

--
Poul-Henning Kamp   | UNIX since Zilog Zeus 3.20
[EMAIL PROTECTED] | TCP/IP since RFC 956
FreeBSD coreteam member | BSD since 4.3-tahoe
Never attribute to malice what can adequately be explained by incompetence.


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message



Re: Small MAKEDEV bug

2000-05-08 Thread Jeroen C. van Gelderen

David O'Brien wrote:
 
 On Sun, May 07, 2000 at 03:27:07PM -0400, Jeroen C. van Gelderen wrote:
  Or just settle for a more intuitive solution:
   MAKEDEV acd2   creates /dev/acd2
   MAKEDEV 2 acd  creates /dev/acd[01]
  which would allow for "MAKEDEV 64 da" and "MAKEDEV 256 pty"
 
 I agree with this syntax and after sending my message to you, was sitting
 there thinking "MAKEDEV num_of_devs dev_name" would make a really
 nice clear syntax.  If you can get BDE's buy-in and other BSD
 traditionalists I think this would be great.

The good part of this solution is that it's backwards compatible.
If the first argument is an integer and the second a device name
without a suffix we do the new thing, if not we revert to the 
traditional behaviour.

bde: what do you think?

Cheers,
Jeroen


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message



Re: Small MAKEDEV bug

2000-05-08 Thread Jeroen C. van Gelderen

Poul-Henning Kamp wrote:
 
 In message [EMAIL PROTECTED], "David O'Brien" writes:
 On Sun, May 07, 2000 at 03:27:07PM -0400, Jeroen C. van Gelderen wrote:
  Or just settle for a more intuitive solution:
   MAKEDEV acd2   creates /dev/acd2
   MAKEDEV 2 acd  creates /dev/acd[01]
  which would allow for "MAKEDEV 64 da" and "MAKEDEV 256 pty"
 
 I agree with this syntax and after sending my message to you, was sitting
 there thinking "MAKEDEV num_of_devs dev_name" would make a really
 nice clear syntax.  If you can get BDE's buy-in and other BSD
 traditionalists I think this would be great.
 
 Make it
 MAKEDEV -num_of_devs dev_name
 and there will be no ambiguity.

I'm confused, what ambiguity does it remove exactly? There are no device 
names that parse as a valid integer, are there? The problem I see with 
your solution is that it makes the count look optional while it isn't.

If you really need to avoid ambuigity and accept the fact that a count
can look optional, I'd use "-n num_of_devs" as the count in your
suggestion looks like a negative number...

Does that make sense?

Cheers,
Jeroen


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message



Re: Small MAKEDEV bug

2000-05-08 Thread Bruce Evans

On Mon, 8 May 2000, David O'Brien wrote:

 On Sun, May 07, 2000 at 03:27:07PM -0400, Jeroen C. van Gelderen wrote:
  Or just settle for a more intuitive solution: 
   MAKEDEV acd2   creates /dev/acd2
   MAKEDEV 2 acd  creates /dev/acd[01]
  which would allow for "MAKEDEV 64 da" and "MAKEDEV 256 pty"
 
 I agree with this syntax and after sending my message to you, was sitting
 there thinking "MAKEDEV num_of_devs dev_name" would make a really
 nice clear syntax.  If you can get BDE's buy-in and other BSD
 traditionalists I think this would be great.

I don't buy it :-).  This syntax is similar to a special case of the syntax
of jot(1).  It's better to use jot(1) directly, e.g.:

MAKEDEV $(jot -w da 2 0)# make 2 acd devices beginning at acd0

Bruce



To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message



Re: Small MAKEDEV bug

2000-05-08 Thread Jeroen C. van Gelderen

Bruce Evans wrote:
 
 On Mon, 8 May 2000, David O'Brien wrote:
 
  On Sun, May 07, 2000 at 03:27:07PM -0400, Jeroen C. van Gelderen wrote:
   Or just settle for a more intuitive solution:
MAKEDEV acd2   creates /dev/acd2
MAKEDEV 2 acd  creates /dev/acd[01]
   which would allow for "MAKEDEV 64 da" and "MAKEDEV 256 pty"
 
  I agree with this syntax and after sending my message to you, was sitting
  there thinking "MAKEDEV num_of_devs dev_name" would make a really
  nice clear syntax.  If you can get BDE's buy-in and other BSD
  traditionalists I think this would be great.
 
 I don't buy it :-).  This syntax is similar to a special case of the syntax
 of jot(1).  It's better to use jot(1) directly, e.g.:
 
 MAKEDEV $(jot -w da 2 0)# make 2 acd devices beginning at acd0

From this it follows that MAKEDEV should be modified to create just it's 
argument: MAKEDEV dev8 creates just dev8, not dev0-dev7.
Otherwise
MAKEDEV $(jot -w da 6 4) wouldn't work or violate POLA. Agreed?

Now it's a question of "the UNIX way" vs. convenience/userfriendlyness
:-)
Is it acceptable to have all users juggle with jot(1) or can we build
in a convenience syntax that covers 95% of all uses? I'd think the
latter, 
otherwise we might as well force our users to use mknod(8) and chmod(1) 
directly instead of MAKEDEV; After all, MAKEDEV is just a convenient 
wrapper around those commands.

So I'd still propose:
  MAKEDEV count device_name_without_suffix
  MAKEDEV device_name_with_suffix ...

As a consolation, added such a special syntax can be added in a few
lines
at the top of MAKEDEV, after which it recursively calls MAKEDEV with the
appropriate jot(1)-expanded device list. So it doesn't clobber the code.

Thoughts?

Cheers,
Jeroen


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message



Re: Small MAKEDEV bug

2000-05-08 Thread Tim Vanderhoek

On Mon, May 08, 2000 at 06:56:03PM -0400, Jeroen C. van Gelderen wrote:
  
  I don't buy it :-).  This syntax is similar to a special case of the syntax
  of jot(1).  It's better to use jot(1) directly, e.g.:
  
  MAKEDEV $(jot -w da 2 0)# make 2 acd devices beginning at acd0

b$ which jot
/usr/bin/jot
b$

The jot utility doesn't appear to be in /bin.


b$ echo '$(jot -w da 2 0)' | wc
   1   5  17
b$ echo $(jot -w da 2 0) | wc
   1   2   8
b$

Heh.

/me mumbles something about the prototypical UNIX hacker...  :-)


-- 
Signature withheld by request of author.


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message



Re: Small MAKEDEV bug

2000-05-07 Thread Jeroen Ruigrok van der Werven

-On [2506 21:55], Bruce Evans ([EMAIL PROTECTED]) wrote:
On Sat, 6 May 2000, Maxim Sobolev wrote:

 I've just noticed that "sh MAKEDEV acd1" doesn't produce node for acd1 due to
 incorrect comparasion in the "while" loop. This affecting both 4.0-STABLE and
 5.0-CURRENT. With this message I'm attaching short patch which should solve
 this little problem.

This is the intended behaviour.  "sh MAKEDEV acdN" is supposed to create
N acd devices, numbered from 0 to N-1.  This broken behaviour was introduced
for cd*, mcd* and scd* in rev.1.171.  It has since spread to acd*.  Other
types of disks are handled correctly.

Bah, bah, bah.  I am really starting to wonder about this sunburn thing.

Can we settle this once and for all in a slightly sane manner?

I committed the change so that MAKEDEV acd1 creates acd1 and not just
acd0.

I personally think this is more consistent with the wd/sa/da/ad
numbering scheme and would propose to fix the other cd* entries
likewise.  Because otherwise somebody other than me will make the same
(commit) mistake x days/weeks/months/years into the future.

Opinions?

-- 
Jeroen Ruigrok van der Werven  Network- and systemadministrator
[EMAIL PROTECTED]VIA Net.Works The Netherlands
BSD: Technical excellence at its best  http://www.via-net-works.nl
In this short time of promise, you're a memory...


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message



Re: Small MAKEDEV bug

2000-05-07 Thread David O'Brien

On Sun, May 07, 2000 at 04:59:46PM +0200, Jeroen Ruigrok van der Werven wrote:
 Can we settle this once and for all in a slightly sane manner?
 
 I committed the change so that MAKEDEV acd1 creates acd1 and not just
 acd0.

This is wrong.  ``MAKEDEV acd2'' should either create only /dev/acd2*, or
/dev/acd[01]*.

It would be nice to fix our inconsistency problem.  Looking at the
4.4Lite vendor import to find the BSD way would be a good start.

-- 
-- David([EMAIL PROTECTED])


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message



Re: Small MAKEDEV bug

2000-05-07 Thread Jeroen C. van Gelderen

[CC culled, -stable removed]

David O'Brien wrote:
 
 On Sun, May 07, 2000 at 04:59:46PM +0200, Jeroen Ruigrok van der Werven wrote:
  Can we settle this once and for all in a slightly sane manner?
 
  I committed the change so that MAKEDEV acd1 creates acd1 and not just
  acd0.
 
 This is wrong.  ``MAKEDEV acd2'' should either create only /dev/acd2*, or
 /dev/acd[01]*.
 
 It would be nice to fix our inconsistency problem.  Looking at the
 4.4Lite vendor import to find the BSD way would be a good start.

Or just settle for a more intuitive solution: 
 MAKEDEV acd2   creates /dev/acd2
 MAKEDEV 2 acd  creates /dev/acd[01]
which would allow for "MAKEDEV 64 da" and "MAKEDEV 256 pty"

 -- David([EMAIL PROTECTED])

Jeroen


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message



Re: Small MAKEDEV bug

2000-05-06 Thread Bruce Evans

On Sat, 6 May 2000, Maxim Sobolev wrote:

 I've just noticed that "sh MAKEDEV acd1" doesn't produce node for acd1 due to
 incorrect comparasion in the "while" loop. This affecting both 4.0-STABLE and
 5.0-CURRENT. With this message I'm attaching short patch which should solve
 this little problem.

This is the intended behaviour.  "sh MAKEDEV acdN" is supposed to create
N acd devices, numbered from 0 to N-1.  This broken behaviour was introduced
for cd*, mcd* and scd* in rev.1.171.  It has since spread to acd*.  Other
types of disks are handled correctly.

Bruce



To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message



Re: Small MAKEDEV bug

2000-05-06 Thread Maxim Sobolev

Bruce Evans wrote:

 On Sat, 6 May 2000, Maxim Sobolev wrote:

  I've just noticed that "sh MAKEDEV acd1" doesn't produce node for acd1 due to
  incorrect comparasion in the "while" loop. This affecting both 4.0-STABLE and
  5.0-CURRENT. With this message I'm attaching short patch which should solve
  this little problem.

 This is the intended behaviour.  "sh MAKEDEV acdN" is supposed to create
 N acd devices, numbered from 0 to N-1.  This broken behaviour was introduced
 for cd*, mcd* and scd* in rev.1.171.  It has since spread to acd*.  Other
 types of disks are handled correctly.

How broken behaviour could be "intended"? It is unclear why *cd* devices should be
different from all others types of disk devices. Users usually being confused when
dealing with special cases like that. I think that there will be no problem if it
would create N+1 devices at least until someone will reimplement it correctly.

-Maxim



To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message