Re: [9fans] Plan 9 versus CORBA?

Re: [9fans] Plan 9 versus CORBA?

[presotto]

802.11's WEP was partially Lucent's fault, partially a bunch of other companies
that were also on the committee.  WEP stands for wired equivalent privacy.
The intention was to guard against casual snooping.  It was known at its
inception that it was not a wonderful solution.  Unfortunately, it
was sold as much more than that.  And even that fails after the Shamir
and Stubblefield work.  IEEE has a committee trying to address the situation
with better encryption.  After leaving Lucent, I'm no longer in the loop.
The early proposals were flawed and shot down, which means that this
time they're getting better review.

Putting encryption in the network is like firewalls.  You reject stuff
before routing it around.  It also provides some minimal protection
for all the uncontrolled/unprotected systems on the network.  However,
its not a substitute for end-to-end encryption.  It's also painfully
hard to change when flaws are found.  If they upgraded the current
standard to include key distribution (there's already a Lucent product
that does that over WaveLAN) then I'ld be happy to continue using it.
However, I'ld also continue doing my own encryption.  I don't trust
the systems authorized to use our network any more than I trust the
ones that aren't authorized.

The upcoming Plan 9 version, tries to solve some data privacy/integrity
problems by

(1) taking all authentication out of the kernel and the applications
   and centralizing it in an authentication server on every cpu.  The
   intent is to be able to switch algorithms with ease and to allow
   the simultaneous use of multiple algs and protocols.  Also, you can
   export the server to machines you CPU to so that you don't have to
   depend on the current 'speaks for' relation if you don't want to.
(2) encrypt all connections.  At the moment the encryption keys are
   part of the authentication process.  However, we'll probably move
   to just using TLS to encrypt all connections before authentication
   occurs.
(3) secure servers using PAK or SRP to store a user's secrets.  The
   server can be centralized or it could be personalized like a bitsy
   that you keep in your pocket.  It's the user's choice where to
   keep the info; you could type it in every time if you want.

We've got most of it working.  All of our CPU connections have been enrypted
for a while.  Soon the file server ones will also.  We've got to see what
that does to throughput.  We've already noticed that it really slows
down modems that currently get some of their speed by compressing.  That
means we'll probably have to compress also.  Luckily processors keep
getting faster.  All the extra cycles that Windows requires to have drop
alpha blended menus can serve our encryption needs.

Rob and jmk are just now converting our machines to run the new system.
There's a lot of change between 9P2000 and the new security infrastructure
so it'll be a while before we release it.

We (ynl & ehg) have also added IPv6 to the stack.  That hasn't been
merged in yet but we'll do it before the release.

We're currently considering what to do about IPsec.  I really don't like
it as a solution, partially for the same reason I don't like WEP:
its done too low.  However, we'll probably have to do it to
be compatible.  Security associations just bother me because they
represent such a violation of layering.  The advantage is the same
as that of WEP, programs don't have to take part in the security, it
can be done around them.  Also, headers are also encrypted and mac'd
so that a number of attacks, like the SYN one, become much harder.

Anyways, that's the current plan.  Rsc is doing a lot of it from
MIT, I'm doing some from Avaya, and the usual suspects are working
on it from Lucent.

Sorry for the long message.

Re: [9fans] Plan 9 versus CORBA?

[jmk]

On Wed Sep 26 05:05:30 EDT 2001, boyd@fr.inter.net wrote:
> > I've been doing a fair amount of thinking about capabilities,
> > mandatory access control (MAC), and security labels, lately.
>
> i recall that jim reeds and ? put some simple, but powerful
> scheme into unix, back in the late 80s.
>

? would be Doug McIlroy and 'The IX Multilevel Secure Operating System'
CSTR #163 is available in pieces at http://cm.bell-labs.com/cm/cs/cstr.html.

Re: [9fans] Plan 9 versus CORBA?

[Boyd Roberts]

> ? would be Doug McIlroy and 'The IX Multilevel Secure Operating System'
> CSTR #163 is available in pieces at http://cm.bell-labs.com/cm/cs/cstr.html.

that'd be it.  iirc they added a chunk of bit vectors to just about
everything and used simple bitwise operations to implement the security
with a tiny performance hit.

Re: [9fans] Plan 9 versus CORBA?

[okamoto]

Sorry, just after I posted previous mail, I noticed that I'm taking wrong way.
If someone (uninvited) intruded our system, and resources have world-wide
readable permission, s/he can read that.  :-)

Kenji

Re: [9fans] Plan 9 versus CORBA?

[okamoto]

I seemed to have written a not enough message.

'hidden' in my previous mail means (should be) hidden from the outside user
such as ftp user, html reader, and uninvited intruder.

For an example, if I have something should be public to the members of internal
Plan 9 users, which looks like (world-wide readable)

--rw-r--r-- M 4 okamoto okamoto     87 Apr  5 14:13 plumbing.

'none' user also can read this.  I think you mean, in that case we should have
some group permission to common to 'all' the members.   However, it's combersome
to have such group which includes all the internal members but outside ones.
This will be solved just if we don't have 'none' user.  This is the point I wanted to
write.  Of course I know Plan 9 solved much about the problem of super user by
making the file server standalone, and separated it from terminals etc.

Kenji

PS.  If we have confidence we will not be intruded, this is not a serious problem
of course.

Re: [9fans] Plan 9 versus CORBA?

[Boyd Roberts]

the 'none' login could always lead to a ches/bellovin style 'jail' :)

Re: [9fans] Plan 9 versus CORBA?

[geoff]

Kenji, I guess I don't understand your problem.  You say
``someone can enter as 'none' user and read hidden internal data
in a future...''.  If it's hidden internal data, it should be owned
by a group that "none" does not belong to, and world access should
be denied, which stops access by "none" and any other user not in
the group.  So it sounds like you've got ``hidden internal'' [sic] data
that is world-readable (thus neither hidden nor internal) and you're
trying to stop "none" from reading it, which sounds like you're trying
to solve the wrong problem.  Or I have I misunderstood your situation?

Re: [9fans] Plan 9 versus CORBA?

[Boyd Roberts]

'none' is a good idea.  it still could be arbitrarily tightened up.

Re: [9fans] Plan 9 versus CORBA?

[okamoto]

> Partticularly ion the popint
 >they have it per object, and have more variable security schem, or

s/ion the popoint/ in the point/
 sorry

Re: [9fans] Plan 9 versus CORBA?

[okamoto]

Thanks David.

I might have feared too much.   However, when we are considering the
security, we assume we may be intruded by someone with bad will someday.

In the present Plan 9, 'none' user cannot login our system from outside,
I mean other than terminals, which may be enough to prevent accidental
reading the data by such external user.

If the above statement is correct, someone can enter as 'none' user
and read hidden internal data in a future...

I wrote that thing because I've been touched Amoeba5.3 this month,
where they have different security scheme using capability which attracted
me somewhat (as far as I can understand it).  Partticularly ion the popint
 they have it per object, and have more variable security schem, or
in other words, capability.  Then, I thought we can make some change of
security scheme other than just file permission scheme...  Of course,
I don't know this may break the integrity of our system.

That is the story why I wrote the previous message.  I think I understand
your stance, and I agree with you now.

The 'noworld' user may be attractive to open a database for external use,
thanks.

Kenji

PS.  This is not serious but some kind of kidding.  Students will like to find
something interesting which is not similar to us sometime.  S/he can
append joke mail to someone, such as supervisor :-), where s/he has no
bad will though.  ^_^

Re: [9fans] Plan 9 versus CORBA?

[Christopher Nielsen]

On Wed, Sep 26, 2001 at 12:13:29PM +0900, okamoto@granite.cias.osakafu-u.ac.jp wrote:

[snip]
> I wrote that thing because I've been touched Amoeba5.3 this month,
> where they have different security scheme using capability which attracted
> me somewhat (as far as I can understand it).  Partticularly ion the popint
>  they have it per object, and have more variable security schem, or
> in other words, capability.  Then, I thought we can make some change of
> security scheme other than just file permission scheme...  Of course,
> I don't know this may break the integrity of our system.
[snip]

I've been doing a fair amount of thinking about capabilities,
mandatory access control (MAC), and security labels, lately.
Unfortunately, I haven't had time to explore these thoughts
in depth due to a rather demanding employer, but I like the
idea of capabilities and MAC.

That said, has anyone put any thought into implementing these
for plan9?

--
Christopher Nielsen - Metal-wielding pyro techie
cnielsen@pobox.com
"Any technology indistinguishable from magic is
insufficiently advanced." --unknown

Re: [9fans] Plan 9 versus CORBA?

[David Arnold]

-->"Christopher" == Christopher Nielsen <cnielsen@pobox.com> writes:

  Christopher> "Any technology indistinguishable from magic is
  Christopher> insufficiently advanced." --unknown

am i being dense, or should one of those `in's be removed?



d

Re: [9fans] Plan 9 versus CORBA?

[Boyd Roberts]

> I've been doing a fair amount of thinking about capabilities,
> mandatory access control (MAC), and security labels, lately.

i recall that jim reeds and ? put some simple, but powerful
scheme into unix, back in the late 80s.

Re: [9fans] Plan 9 versus CORBA?

[okamoto]

>In fact, we continually reduce the power of none to make it
>less dangerous.

I don't know this is possible or not, however, I don't like to open all
the sources and data to the public from network access like now for
none user.   Can't we restrict the power of 'none' only to exec permition?

Kenji

Re: [9fans] Plan 9 versus CORBA?

[david presotto]

We've reduced none to only have world access to things.  We
can't make just exec access available because the file server can't
really distinguish execution from reading.

I take it removing world read access from files isn't good enough
for you?  We also invented a group called noworld.  When anyone
in that group tries to access a file, the permission bits are first anded
with 0x770 for files and 0x771 for directories.  We use this right now
to create sandbox'd users.  They can't access anything unless they own
it or are a member of its group.  Somewhere twixt that and what
we currently call none is probably the right solution.

Of course this takes a lot of forethought.  It's really easy to build
environments where noworld users can't even set up a namespace.
You're right though.  This would make the listeners a lot more
secure.

----- Original Message -----
From: <okamoto@granite.cias.osakafu-u.ac.jp>
To: <9fans@cse.psu.edu>
Sent: Tuesday, September 25, 2001 9:48 PM
Subject: Re: [9fans] Plan 9 versus CORBA?


> >In fact, we continually reduce the power of none to make it
> >less dangerous.
>
> I don't know this is possible or not, however, I don't like to open all
> the sources and data to the public from network access like now for
> none user.   Can't we restrict the power of 'none' only to exec permition?
>
> Kenji
>
>

Re: [9fans] Plan 9 versus CORBA?

[presotto]

[-- Attachment #1: Type: text/plain, Size: 205 bytes --]

I wouldn't get rid of none.  I still like the idea of running
the listener and services that don't need more access as none.
In fact, we continually reduce the power of none to make it
less dangerous.

[-- Attachment #2: Type: message/rfc822, Size: 1585 bytes --]

From: okamoto@granite.cias.osakafu-u.ac.jp
To: 9fans@cse.psu.edu
Subject: Re: [9fans] Plan 9 versus CORBA?
Date: Wed, 26 Sep 2001 10:26:22 +0900
Message-ID: <20010926012552.68EAE1998A@mail.cse.psu.edu>

Thanks Presotto for your elaborated message.

>   export the server to machines you CPU to so that you don't have to
>   depend on the current 'speaks for' relation if you don't want to.

Does this mean we can eliminate 'none' user from Plan 9 system?
I don't like that user, although he is not spade ace. :-)

Kenji

Re: [9fans] Plan 9 versus CORBA?

[okamoto]

Thanks Presotto for your elaborated message.

>   export the server to machines you CPU to so that you don't have to
>   depend on the current 'speaks for' relation if you don't want to.

Does this mean we can eliminate 'none' user from Plan 9 system?
I don't like that user, although he is not spade ace. :-)

Kenji

Re: [9fans] Plan 9 versus CORBA?

[forsyth]

>>However,
>>its not a substitute for end-to-end encryption.

especially when the bits aren't all on a single wireless network.

Re: [9fans] Plan 9 versus CORBA?

[rob pike]

I think you misrepresent the purpose of security.  Its role is to
prevent us getting work done.  If someone constructs a security
solution that is usable, experts will focus on it like a cat watching
a mouse hole until a fatal flaw is found.  This results in three
things: 1) The technology is disabled, making it impossible to work
again.  2) A solution is worked on, distracting people from getting
regular work done.  3) Finally, a new solution is deployed, requiring
people to spend time updating their systems and networks rather than
getting work done.  At this point, security has failed because people
are working, so the cat goes back to the hole and in a few days the
mouse emerges and is caught and life returns to normal.

So the rule of security is the following: if you are able to work on
something other than security, your system is insecure.

-rob

Re: [9fans] Plan 9 versus CORBA?

[Dan Cross]

In article <20010925141950.92B86199DD@mail.cse.psu.edu> you write:
>I think you misrepresent the purpose of security.  [...]
>
>So the rule of security is the following: if you are able to work on
>something other than security, your system is insecure.

Shhh!  I told you not to TELL anyone!  Way to go, Rob, letting the cat
out of the bag.  Shesh.

	- Dan C.

Re: [9fans] Plan 9 versus CORBA?

[presotto]

WEP works incredibly well now.  Imagine all the anguish avoided by no longer
having to arm wrestle the network password out of reluctant administrators
forced by company policy to not give it to you.  Now you can just say,
``Don't get up, I'll fetch it myself.''

Or just look at WEP as the free sample to hook people on the security drug.
Now that they're dependent and it doesn't work anymore they're all
willing to pay big bucks for the hard stuff; IPsec, pptp, TLS, ssh.
It's like a dream come true.

Re: [9fans] Plan 9 versus CORBA?

[rob pike]

> tree-borne reproductive organs wrapped in wrinkly wood.

Pressed flowers?

-rob

Re: [9fans] Plan 9 versus CORBA?

[Andrew Simmons]

So that's what the C++ Programming Language Special Edition was made for.
Thanks for the tip.

----------
In article <20010924224616.8A73919AAD@mail.cse.psu.edu>,
rob@plan9.bell-labs.com (rob pike) wrote:

>
> Pressed flowers?
>
> -rob

Re: [9fans] Plan 9 versus CORBA?

[Fco.J.Ballesteros]

:  I'll definitely try Plan 9 out, but may not be allowed to use it
:  because it is not Object Oriented and because the compiler doesn't
:  support const, both of which are Bad Things. This is completely off

You can tell your boss that Plan 9 is object based, but
tell him/her that

	"...those weird Plan 9 guys refer to their objects as `files',
and to their object servers as `file servers'..."

good luck

Re: [9fans] Plan 9 versus CORBA?

[Fco.J.Ballesteros]

:  You mean, you're actually at an _academic_ institution, not a
:  psychiatric one?!
:
Some times it looks like... most of the times, actually.

:  ++L
:
:  PS: Jokes asides, what the most recent URL for your kernel commentary,
:  it is high time I made a(nother) serious study of it?

You can find links at http://plan9.escet.urjc.es/
I have not work on it since long ago, though. So you already
may have the last one.

Hope to get some time to get back to it during fall.

Re: [9fans] Plan 9 versus CORBA?

[anothy]

// ...I'm finding it an increasing strain just to lift the books I need
// to consult - over 1000 pages each.

Aha! somebody call OSHA! i guess whatever plan 9 costs people
through RSI from chording in Acme is offset by savings in pulled
muscles from lifting those insane reference manuals.
-α.

Re: [9fans] Plan 9 versus CORBA?

[Dan Cross]

In article <20010921152953.3090719A9C@mail.cse.psu.edu> you write:
>// ...I'm finding it an increasing strain just to lift the books I need
>// to consult - over 1000 pages each.
>
>Aha! somebody call OSHA! i guess whatever plan 9 costs people
>through RSI from chording in Acme is offset by savings in pulled
>muscles from lifting those insane reference manuals.

Actually, the CORBA reference manual I have is pretty light.  I
think it was printed on light paper, though.  It is around 1000
pages.  Of course, the book I have on transaction processing
(G&R; not to be confused with a cheesy band) is thicker and heavier,
with small print and tiny margins.  It doesn't really contain any
code, though, and only glosses over discussing API's (``there is
this thing called CICS...IBM made it.  Learn PL/I.'').

	- Dan C.

Re: [9fans] Plan 9 versus CORBA?

[Fco.J.Ballesteros]

[-- Attachment #1: Type: text/plain, Size: 48 bytes --]

I think I should  resign as a professor ☺.


[-- Attachment #2: Type: message/rfc822, Size: 2114 bytes --]

From: Ronald G Minnich <rminnich@lanl.gov>
To: <9fans@cse.psu.edu>
Subject: Re: [9fans] Plan 9 versus CORBA?
Date: Fri, 21 Sep 2001 08:29:37 -0600 (MDT)
Message-ID: <Pine.LNX.4.33.0109210829240.11328-100000@snaresland.acl.lanl.gov>

On Fri, 21 Sep 2001, andrey mirtchovski wrote:

> when i did a small undergraduate presentation on plan9 (i tried to do
> distributed bioinformatics computations) one of my professors asked me the
> same question: "why not corba?"..

that's why we don't let professors write code.

ron

Re: [9fans] Plan 9 versus CORBA?

[Lucio De Re]

On Fri, Sep 21, 2001 at 04:54:38PM +0200, Fco.J.Ballesteros wrote:
>
> I think I should  resign as a professor ?.
>
You mean, you're actually at an _academic_ institution, not a
psychiatric one?!

++L

PS: Jokes asides, what the most recent URL for your kernel commentary,
it is high time I made a(nother) serious study of it?

Re: [9fans] Plan 9 versus CORBA?

[Sape Mullender]

[-- Attachment #1: Type: text/plain, Size: 92 bytes --]

> Rob's heritage is Balkan, he's still getting used to vowels.

His name used to be pk


[-- Attachment #2: Type: message/rfc822, Size: 1572 bytes --]

From: jmk@plan9.bell-labs.com
To: 9fans@cse.psu.edu
Subject: Re: [9fans] Plan 9 versus CORBA?
Date: Fri, 21 Sep 2001 10:26:32 -0400
Message-ID: <20010921142640.88BF019AA5@mail.cse.psu.edu>

On Fri Sep 21 10:09:35 EDT 2001, andrew@mbmnz.co.nz wrote:
> ...
> On a totally unrelated note, I'd be interested to find out why rob
> pike spells his name in lower case. Is this a literary device, like ee
> cummings, or does Plan 9 not support upper case?

Rob's heritage is Balkan, he's still getting used to vowels.
Give hime some time and he'll get round to upper case.

Re: [9fans] Plan 9 versus CORBA?

[jmk]

On Fri Sep 21 10:09:35 EDT 2001, andrew@mbmnz.co.nz wrote:
> ...
> On a totally unrelated note, I'd be interested to find out why rob
> pike spells his name in lower case. Is this a literary device, like ee
> cummings, or does Plan 9 not support upper case?

Rob's heritage is Balkan, he's still getting used to vowels.
Give hime some time and he'll get round to upper case.

Re: [9fans] Plan 9 versus CORBA?

[suspect]

On Fri, 21 Sep 2001 jmk@plan9.bell-labs.com wrote:
> Rob's heritage is Balkan, he's still getting used to vowels.
> Give hime some time and he'll get round to upper case.
       ^^^^
I guess you got him covered on the extra vowels.
-

[9fans] Plan 9 versus CORBA?

[Andrew Simmons]

Hi

I'm working on a distributed application using C++ and CORBA, and
apart from the sheer mind-numbing complexity of both, I'm finding it
an increasing strain just to lift the books I need to consult - over
1000 pages each. I was wondering if Plan 9 might be worth considering
as a simpler alternative, and I would be interested in the views of
the participants of this news group, especially those of anyone who
has experience of both Plan 9 and CORBA. I'd also be interested in
people's views on the suitability of Plan 9 as a platform for
commercial development - my management might be rather nervous of
using an operating system perceived as too far out of the mainstream.

On a totally unrelated note, I'd be interested to find out why rob
pike spells his name in lower case. Is this a literary device, like ee
cummings, or does Plan 9 not support upper case?

Thanks
Andrew Simmons

Re: [9fans] Plan 9 versus CORBA?

[andrey mirtchovski]

when i did a small undergraduate presentation on plan9 (i tried to do
distributed bioinformatics computations) one of my professors asked me the
same question: "why not corba?"..

unfortunately i couldn't answer him... i tried to find justification on
why i chose plan9 and the only thing i came up with was something in lines
of 'one who has seen and gotten used to the simplicity of p9 would sway
away from such monstrocities'...

i thought that's not a very good justification, so i simply added
'besides, the bell-labs people think corba sux'... :)

anyway, i didn't get the highest possible mark :)

andrey

Andrew Simmons wrote:

> Hi
>
> I'm working on a distributed application using C++ and CORBA, and
> apart from the sheer mind-numbing complexity of both, I'm finding it
> an increasing strain just to lift the books I need to consult - over
> 1000 pages each. I was wondering if Plan 9 might be worth considering
> as a simpler alternative, and I would be interested in the views of
> the participants of this news group, especially those of anyone who
> has experience of both Plan 9 and CORBA. I'd also be interested in
> people's views on the suitability of Plan 9 as a platform for
> commercial development - my management might be rather nervous of
> using an operating system perceived as too far out of the mainstream.
>
> On a totally unrelated note, I'd be interested to find out why rob
> pike spells his name in lower case. Is this a literary device, like ee
> cummings, or does Plan 9 not support upper case?
>
> Thanks
> Andrew Simmons

Re: [9fans] Plan 9 versus CORBA?

[Ronald G Minnich]

On Fri, 21 Sep 2001, andrey mirtchovski wrote:

> when i did a small undergraduate presentation on plan9 (i tried to do
> distributed bioinformatics computations) one of my professors asked me the
> same question: "why not corba?"..

that's why we don't let professors write code.

ron

Re: [9fans] Plan 9 versus CORBA?

[Scott Schwartz]

| when i did a small undergraduate presentation on plan9 (i tried to do
| distributed bioinformatics computations) one of my professors asked me the
| same question: "why not corba?"..

In an abstract way I like the idea of component architectures, but I
like the idea of clarity and simplicity even more.  In the bioinformatics
research that I do, we mostly use flat text files and some SQL.

While we're on the topic of scientific computing, a disadvantage that
Plan 9 and Linux share is that they chop up the address space.  One nice
thing about Solaris is that you can malloc several contiguous gigabytes.

Re: [9fans] Plan 9 versus CORBA?

[Ronald G Minnich]

On Fri, 21 Sep 2001, Andrew Simmons wrote:

> I'm working on a distributed application using C++ and CORBA, and
> apart from the sheer mind-numbing complexity of both, I'm finding it
> an increasing strain just to lift the books I need to consult - over
> 1000 pages each.

Yeah, that stuff really sucks. Plus it is so complex how do you know if
it's working right and that it will continue to work right. Plus on any
given day on any given machine your C++ code can refuse to compile or
work, for reasons unknown. Plus, last time I looked, CORBA runs about as
fast as congealed oatmeal. True story: I once asked a CORBA fanatic how
fast his ORB could run a simple null operation. "Really fast", he said,
"so fast I can hardly see it run. Must be 2 or 3 per second."

Plan 9 by contrast looks like the Next Right Thing. I haven't seen
anything CORBA does that Plan 9 can't do as well, although in a pinch the
corba manual set can double as a jackstand. But you do have to change your
thinking a bit.

As for commercial use: on that score, Plan 9 in people's minds is kind of
where Linux was 10 years ago (save for a couple design-ins). We're still
in the "what's that" stage. Now that it is finally Open Source that should
improve. So remind your bosses that somebody had to take a chance with
Unix, then other people took a chance with Linux, and the risk-takers
can win big. We're just beginning that battle out here, and I don't expect
to reach the "Oh! I get it!" stage for 5 more years. But we're the
gummint, so things can be slow.

> On a totally unrelated note, I'd be interested to find out why rob
> pike spells his name in lower case. Is this a literary device, like ee
> cummings, or does Plan 9 not support upper case?

so who remembers (if you do you're old) when unix used to be called the
"ee cummings operating system"

I remember it but refuse to admit I'm old. I still take @@@@ occasionaly
about failure to capitalize.

ron

Re: [9fans] Plan 9 versus CORBA?

[Andrew Simmons]

Thanks to all who replied. I had always assumed that mr pike was of
Dutch extraction, and that "pike" was an anglicised version of
"pijkstra".

On the question of manual weight, I'm using "Advanced CORBA
Programming in C++" by Henning & Vinoski - it's not quite as heavy as
Stroustrup's special edition. It's an excellent book in many ways, but
I feel rather as if I was calculating planetary orbits with the aid of
a 1000 page manual on epicycles. There must be a better way.

I'll definitely try Plan 9 out, but may not be allowed to use it
because it is not Object Oriented and because the compiler doesn't
support const, both of which are Bad Things. This is completely off
topic, but I've just been looking at an OO implementation of a CRC
calculation. In the bad old days you'd just write a five line function
to do this. In the good new days, you declare a CRC class with at
least three constructors, a destructor, a copy constructor, an
assignment operator, a Calculate method, and then you make the
calculated value private because God forbid people should be allowed
to access it directly and then you need an accessor method, or why not
have several such as GetCRCAsFormattedString I think I'll go and lie
down now it must be time for my medication.

Re: [9fans] Plan 9 versus CORBA?

[Boyd Roberts]

> On the question of manual weight, I'm using "Advanced CORBA
> Programming in C++" by Henning & Vinoski - it's not quite as heavy as
> Stroustrup's special edition.

this phenomena was investigated many years ago:

    http://chunder.com/stuff95/walnut.html

> or why not have several such as GetCRCAsFormattedString I think
> I'll go and lie down now it must be time for my medication.

check out the python PEP for MD5:

    http://python.sourceforge.net/peps/pep-0247.html

as far as medication goes, i'm all for:

    when in doubt, double the dose

Re: [9fans] Plan 9 versus CORBA?

[George Michaelson]

> > On the question of manual weight, I'm using "Advanced CORBA
> > Programming in C++" by Henning & Vinoski - it's not quite as heavy as
> > Stroustrup's special edition.
>
> this phenomena was investigated many years ago:
>
>     http://chunder.com/stuff95/walnut.html

Having just been around the national museum in Taiwan and admired the
polished stone artifacts from 5000BC, lovingly preserved by future generations
it has to be said stone age s/w (MH) seems to be lasting a lot longer than
tree-borne reproductive organs wrapped in wrinkly wood.

Written in a TCL/TK gui on top of MH of course...

-George

PS Mind you, pickled MH probably heads towards bad things but pickled walnuts
   have nothing to do with Python.

PPS boxing and racing have the concept of weight for age don't they? isn't MH
    allowed to get a little fat around the middle now its past middle age?

--
George Michaelson       |  APNIC
Email: ggm@apnic.net    |  PO Box 2131 Milton QLD 4064
Phone: +61 7 3367 0490  |  Australia
  Fax: +61 7 3367 0482  |  http://www.apnic.net

Re: [9fans] Plan 9 versus CORBA?

[Boyd Roberts]

> it has to be said stone age s/w (MH) seems to be lasting a lot longer than
> tree-borne reproductive organs wrapped in wrinkly wood.

where were the bitmapped display, the ethernet, the laser printer invented
and integrated with the mouse?

Xerox PARC.

Re: [9fans] Plan 9 versus CORBA?

[George Michaelson]

> > it has to be said stone age s/w (MH) seems to be lasting a lot longer than
> > tree-borne reproductive organs wrapped in wrinkly wood.
>
> where were the bitmapped display, the ethernet, the laser printer invented
> and integrated with the mouse?
>
> Xerox PARC.

Boyd, that is *such* a non-sequiteur. Like, (a) bitmapped displays, ethernet
mice and printers are not email systems and (b) Xerox was notorious for
leading the way in developing rilly neat ideas that died in a ditch, such
as smalltalk.

You can't point to successes and say that makes the failures a success. Walnut
didn't fly. MH, shitfully concreted with #ifdef RAND as it is, persists.

Lets go sideways instead. if PARC was such a good idea, why did Xerox kill it?
Does Lucent share Xerox's ability to turn success into failure? I hope not!

cheers
	-George

Re: [9fans] Plan 9 versus CORBA?

[Boyd Roberts]

> Lets go sideways instead. if PARC was such a good idea, why did Xerox kill it?

'cos xerox wanted photocopies.  have you read _fumbling the future_?

Re: [9fans] Plan 9 versus CORBA?

[George Michaelson]

> > Lets go sideways instead. if PARC was such a good idea, why did Xerox kil
l it?
>
> 'cos xerox wanted photocopies.  have you read _fumbling the future_?
>
>

Nope, just talk to ex-PARC people. But thats a damn good reference and I'm
ordering a copy asap!

_Insanely Great_ lies in another dimension. As does _Accidental Empires_

_The Death of IBM_ is another good read, if somewhat dated. I suppose we
can look forward to _The Death of [Digital|Tandem|Compaq]_ next.

-George

Re: [9fans] Plan 9 versus CORBA?

[Boyd Roberts]

> Nope, just talk to ex-PARC people. But thats a damn good reference and I'm
> ordering a copy asap!

i've worked with ex-parc people and met taylor and seen him in action in
the SRC center meetings -- clever guy.

no doubt, so have a buncha other people on this list, so i'm no exception.

Re: [9fans] Plan 9 versus CORBA?

[Boyd Roberts]

> Does Lucent share Xerox's ability to turn success into failure? I hope not!

they would seem to be already seriouslt into failure.

i think they'll need roy 'n hg to dig 'em outa this one.

Re: [9fans] Plan 9 versus CORBA?

[George Michaelson]

> > Does Lucent share Xerox's ability to turn success into failure? I hope not!
>
> they would seem to be already seriouslt into failure.

WaveLAN stands out. well packaged solution that.

>
> i think they'll need roy 'n hg to dig 'em outa this one.

Roy is now into scriptwriting drama for ABC. He's gone all serious.

Never grow up.

-George

Re: [9fans] Plan 9 versus CORBA?

[Boyd Roberts]

> WaveLAN stands out. well packaged solution that.

doesn't the crypto suck?

Re: [9fans] Plan 9 versus CORBA?

[Scott Schwartz]

| > WaveLAN stands out. well packaged solution that.
|
| doesn't the crypto suck?

Link level encryption of any sort sucks, because it serves as an excuse
to not insure proper end-to-end integrity.  Easily sniffable wireless
ethernet focuses people's attention in a beautiful way.

Re: [9fans] Plan 9 versus CORBA?

[Dan Cross]

In article <20010925012306.16242.qmail@g.bio.cse.psu.edu> you write:
>Link level encryption of any sort sucks, because it serves as an excuse
>to not insure proper end-to-end integrity.  Easily sniffable wireless
>ethernet focuses people's attention in a beautiful way.

Unfortunately, that's just not the case, though.  802.11 encryption
was, as you say, a bandaid.  I think it's intention was largely to put
the barrier to entry for sniffing wireless Ethernet on par with that
required for sniffing ``normal'' Ethernet (where, obviously, you'd need
a wire or sensative equipment to pick up latent radiated energy from a
wire).  Now, the response isn't to focus on the problem, but to try and
``fix'' 802.11.  A lot of people who are putting in, eg, end-to-end
crypto are doing so ``temporarily'' until the problems with the
wireless LAN are ``fixed.''

The real problem is that too many people hear a word containing the
letters ``crypto'' and automatically assume that word is equivalent to
``security.''  As we all know, and has history and the world in general
have painfully demonstrated time and time again, reliance on
cryptography alone only gives a hollow sense of false security.
Attacks on crypto are rare in comparison to attacks against, eg, the
reliability of software and the vulnerabilities inherent in code
generated by lazy programmers.

What's really needed is a holistic approach, that takes into account
the ``big picture'' of security, and which emphasizes that there is no
magic pill that one can swallow to provide blanket security, and that
true security can only be achieved through a combination of
complementary techniques.

But, good luck selling that one.  :-(

	- Dan C.

Re: [9fans] Plan 9 versus CORBA?

[Boyd Roberts]

> What's really needed is a holistic approach, that takes
> into account the ``big picture'' of security ...

when i hear the word 'holistic', i reach for my 92FS ...

Re: [9fans] Plan 9 versus CORBA?

[Dan Cross]

In article <01bc01c1455d$7c4ef930$a2b9c6d4@SOMA> you write:
>> WaveLAN stands out. well packaged solution that.
>
>doesn't the crypto suck?

Yeah, but I'm not sure that's Lucent's fault.  btw- it's not so
much the crypto itself (unlike DVD), as the implementation of
the crypto.

	- Dan C.

Re: [9fans] Plan 9 versus CORBA?

[William Josephson]

On Mon, Sep 24, 2001 at 10:12:11PM -0400, Dan Cross wrote:
> >doesn't the crypto suck?
>
> Yeah, but I'm not sure that's Lucent's fault.  btw- it's not so
> much the crypto itself (unlike DVD), as the implementation of
> the crypto.

Somewhat more precisely, WEP is based on alleged RC4, but suffers from
poor handling of the initialization vectors.  A recent paper by Shamir
et al. gives a practical online, known plaintext only attack.

 -WJ

Re: [9fans] Plan 9 versus CORBA?

[Mike Warner]

This boils it down nicely.

-m

Andrew Simmons wrote:

> Thanks to all who replied. I had always assumed that mr pike was of
> Dutch extraction, and that "pike" was an anglicised version of
> "pijkstra".
> 
> On the question of manual weight, I'm using "Advanced CORBA
> Programming in C++" by Henning & Vinoski - it's not quite as heavy as
> Stroustrup's special edition. It's an excellent book in many ways, but
> I feel rather as if I was calculating planetary orbits with the aid of
> a 1000 page manual on epicycles. There must be a better way.
> 
> I'll definitely try Plan 9 out, but may not be allowed to use it
> because it is not Object Oriented and because the compiler doesn't
> support const, both of which are Bad Things. This is completely off
> topic, but I've just been looking at an OO implementation of a CRC
> calculation. In the bad old days you'd just write a five line function
> to do this. In the good new days, you declare a CRC class with at
> least three constructors, a destructor, a copy constructor, an
> assignment operator, a Calculate method, and then you make the
> calculated value private because God forbid people should be allowed
> to access it directly and then you need an accessor method, or why not
> have several such as GetCRCAsFormattedString I think I'll go and lie
> down now it must be time for my medication.

Re: [9fans] Plan 9 versus CORBA?

[Alexander Viro]

On Fri, 21 Sep 2001, Andrew Simmons wrote:

> Hi
>
> I'm working on a distributed application using C++ and CORBA, and
> apart from the sheer mind-numbing complexity of both, I'm finding it
> an increasing strain just to lift the books I need to consult - over
> 1000 pages each. I was wondering if Plan 9 might be worth considering
> as a simpler alternative, and I would be interested in the views of

Yes.  CORBA is an epitome of "clean APIs are hard, let's go shopping"
school of design.  In other words, it actively encourages API bloat and
considers that as a feature.  How hard it will be to make your code
use Plan 9 model for RPC depends on what you've already got, obviously,
but yes, result is very likely to be simpler and cleaner.