Linux-Misc Digest #664, Volume #20 Wed, 16 Jun 99 18:13:17 EDT
Contents:
SUID programs: are they normal? (Ding-Jung Han)
Dynamic protocols ("y")
Re: Could Microsoft Cheat On The New Mindcraft Benchmark? (was: Mindcraft Retest
News ("Otto")
Re: Could Microsoft Cheat On The New Mindcraft Benchmark? (was: Mindcraft Retest
News (Mark S. Bilk)
URGENT--pppd and chat ([EMAIL PROTECTED])
Re: editorial: Stupid Linux Tricks (Gene Wilburn)
Re: kernel mode programs (yes, this is on topic) (Scott Lanning)
Re: DOS command line replacement ("Art S. Kagel")
Re: What are the differences between mySQL and mSQL? (bryan)
Re: Linux systems- Poor security (Erik de Castro Lopo)
Re: What are the differences between mySQL and mSQL? (bryan)
Linux&PHP+NT&Oracle ([EMAIL PROTECTED])
Re: SUID programs: are they normal? (Mike Khalili)
Re: Repartition EXT2 without data loss? (Stephen George)
undefined reference to ct_bind (Stephen George)
Re: A Capitalists view of freedom
Re: Could Microsoft Cheat On The New Mindcraft Benchmark? (was: Mindcraft Retest News
Re: kernel mode programs (yes, this is on topic) (Craig Dickson)
Re: tell modem "not to wait for dial tone"? ("Matthew Haley")
----------------------------------------------------------------------------
From: Ding-Jung Han <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.security
Subject: SUID programs: are they normal?
Date: Wed, 16 Jun 1999 16:16:17 -0400
Hi,
I'm a newbie to all this security stuff. Just checked if there's any
suid programs on my linux box, and I found more than I expected (see
below). Are they normal? Is there any one that is unnecessarily suid'd?
TIA
Ben
---
[root@Legarto /]# find / -fstype ext2 -perm -04000 -user root -print
/usr/X11R6/bin/Xwrapper
/usr/X11R6/bin/crxvt
/usr/X11R6/lib/X11/xmcd/bin-Linux-i686/xmcd
/usr/X11R6/lib/X11/xmcd/bin-Linux-i686/cda
/usr/bin/at
/usr/bin/chage
/usr/bin/gpasswd
/usr/bin/zgv
/usr/bin/chfn
/usr/bin/chsh
/usr/bin/disable-paste
/usr/bin/passwd
/usr/bin/lpq
/usr/bin/lpr
/usr/bin/lprm
/usr/bin/suidperl
/usr/bin/procmail
/usr/bin/sudo
/usr/bin/rcp
/usr/bin/rlogin
/usr/bin/rsh
/usr/bin/newgrp
/usr/bin/crontab
/usr/bin/sperl5.00503
/usr/sbin/usernetctl
/usr/sbin/sendmail
/usr/sbin/traceroute
/usr/sbin/userhelper
/usr/libexec/pt_chown
/bin/mount
/bin/su
/bin/umount
/bin/login
/bin/ping
find: /proc/17817/fd/4: No such file or directory
/sbin/pwdb_chkpwd
/opt/netscape/movemail
------------------------------
From: "y" <[EMAIL PROTECTED]>
Subject: Dynamic protocols
Date: Wed, 16 Jun 1999 17:16:15 -0300
Hi,
Does anybody know what dynamic routing protocols are supported by RedHat Linux
5.2?
Thanks in advance for any help.
Ars
------------------------------
From: "Otto" <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.linux.networking,comp.os.ms-windows.nt.advocacy,comp.os.linux.advocacy,comp.infosystems.www.servers.unix
Subject: Re: Could Microsoft Cheat On The New Mindcraft Benchmark? (was: Mindcraft
Retest News
Date: Wed, 16 Jun 1999 20:35:23 GMT
Mark S. Bilk <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> Microsoft has a history of cheating on benchmarks and
> rigging software to prevent competitive products from
> functioning. Could they do that in the Mindcraft retest
> that's now taking place?
Why would they need to cheat? As it is NT can beat the crap out of Linux on
the high end hardware, so what?Live with it.
It boils down to matter of trust, which no matter what's being done the
Linux community will never be satisfied. The test is not even over yet, no
results are available, you already crying foul play. Just make sure you'll
have plenty of tissues at hand when the results come out :).
Dislike the way the test is performed? Do your own using the same hardware.
------------------------------
Crossposted-To:
comp.os.linux.networking,comp.os.linux.advocacy,comp.infosystems.www.servers.unix
From: [EMAIL PROTECTED] (Mark S. Bilk)
Subject: Re: Could Microsoft Cheat On The New Mindcraft Benchmark? (was: Mindcraft
Retest News
Date: Wed, 16 Jun 1999 21:25:09 GMT
In article <[EMAIL PROTECTED]>, Paul D. Smith <[EMAIL PROTECTED]>
wrote:
[Mark Bilk wrote:]
>>> The tests would have to be repeated on hardware that is
>>> known to have unmodified microcode, and with software and
>>> configuration data that are also known not to have been
>>> meddled with.
>Well, the changes may be specifically designed to allow the system to
>perform much better on that particular benchmark, but are useless at
>best or very bad at worst for normal, day-to-day usage.
>
>Compiler writers have a long and storied history of this kind of thing:
>they like to write specific optimizations for well-known benchmarks so
>their compilers look better in comparisons--but most benchmarks are
>fairly poor at emulating real-life usage so these kinds of targeted
>optimizations almost never result in any real code running faster.
Exactly. I think there was a compiler in the early PC era
that emitted a hand-optimized routine for the "Sieve of
Eratosthenes" prime number algorithm, since that was used
as a benchmark by Byte magazine and others.
>Anyway, changing hardware microcode has nothing to do with the
>performance of the OS: that gives a skewed result for the
>benchmark... it's like running benchmarks on different hardware. The
>comparison you want to make is between the operating systems themselves,
>so everything else has to be as invariant as possible.
What I had in mind here is that the microcode in, e.g., the
NICs, could be hacked to detect whether IIS or Samba was
running. Since web pages are sent out with a header that
includes the name of the server software, the network card
would simply have to read that header. Then, if it detected
Samba, it would impede packet transmission (by telling the
CPU it was "busy"), and possibly drop some packets going in
either direction. It could include a timer that would
disable it after, say, 24 hours, so it wouldn't be found
later if there were an investigation.
If that seems farfetched, this article describes the stealthed,
encrypted code that Microsoft put into beta versions of Windows
3.1 to detect DR-DOS, put up an error message, and fail by
default:
<a href="http://www.ddj.com/articles/1993/9309/9309d/9309d.htm">MS Code to Kill DR-DOS
-- SEP93: Examining the Windows AARD Detection Code</a>
------------------------------
From: [EMAIL PROTECTED]
Subject: URGENT--pppd and chat
Date: Wed, 16 Jun 1999 19:44:25 GMT
I wrote a Perl script that brings up multiple ppp sessions
consequtively or at once. From the script, I issue a pppd command that
looks like this:
pppd /dev/ttyDx lock modem nodefaultip.....connect "chat -v -f
'/etc/ppp/ppp.chatscript'"
The problem is that this command is issued many many times...in a loop
basically (this is my intention...). Some calls are not successful and
that's fine, but numerous instances of pppd and chat continue to run,
even after the phones hang up. Also, before restarting the cycle of
issuing pppd commands, I use the ppp-off script to make sure that no
connections are up.
I'm wondering how this could be resolved. My memory gets eaten up
quickly, and the system eventually falls into segmentation fault, and
crashes. I need to be able to leave this script running, unattended,
over a weekend, if I wanted to!
Thanks,
Sherif
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: Gene Wilburn <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux,comp.os.linux.x,comp.os.linux.development
Subject: Re: editorial: Stupid Linux Tricks
Date: Wed, 16 Jun 1999 20:32:49 GMT
Mike Bartman wrote:
[snip]
> there they asked if he wanted to do the OS too. He said he knew
> nothing about writing an OS (the truth, as history records...;^), but
> they said they had lots of folks who did, and they'd help him get
> started. Then he remembered a local guy who'd done his own port of
> CP/M to the 8086 chip. I forget the guy, but seem to recall the name
> "Seattle Computers" vaguely in relation to this...it's been a while!
> :^). Anyway, Billy calls him up and asks if he'd be interested in
> selling his CP/M port for the 8086...the guy (knowing nothing about
> IBM's interest) says he guesses so, how much? Bill offers him about
> $50,000, and the rest is history.
His first name was Tim and I seem to recall his last name was Paterson.
I believe his product was called Seattle DOS, and it was essentially an
8086 clone of CP/M. For years folks at MS used to talk about Timcode --
bits of assembly language code in the original DOS that none of them
knew what it did. It wasn't until DOS 5 or 6 that all the Timcode was
removed.
--
===================================================================
Gene Wilburn, Northern Journey Online, http://www.interlog.com/~njo
===================================================================
------------------------------
From: [EMAIL PROTECTED] (Scott Lanning)
Crossposted-To: comp.lang.functional
Subject: Re: kernel mode programs (yes, this is on topic)
Date: 16 Jun 1999 20:59:04 GMT
Jason Stokes ([EMAIL PROTECTED]) wrote:
: I was just thinking it's so nice that C programmers are
: considerate enough t strp the lngage of all thse rdndnt vwls.
: Gotta save the five MicroScns it-d take to type out vars
: in_fll, yKnw?
Apparently you haven't run into FORTRAN programmers... :)
================================================================
SUBROUTINE QSORTI (ORD,N,A)
C
C==============SORTS THE ARRAY A(I),I=1,2,...,N BY PUTTING THE
C ASCENDING ORDER VECTOR IN ORD. THAT IS ASCENDING ORDERED A
C IS A(ORD(I)),I=1,2,...,N; DESCENDING ORDER A IS A(ORD(N-I+1)),
C I=1,2,...,N . THIS SORT RUNS IN TIME PROPORTIONAL TO N LOG N .
C
C
C ACM QUICKSORT - ALGORITHM #402 - IMPLEMENTED IN FORTRAN 66 BY
C WILLIAM H. VERITY, [EMAIL PROTECTED]
C CENTER FOR ACADEMIC COMPUTING
C THE PENNSYLVANIA STATE UNIVERSITY
C UNIVERSITY PARK, PA. 16802
C
IMPLICIT INTEGER (A-Z)
C
DIMENSION ORD(N),POPLST(2,20)
INTEGER X,XX,Z,ZZ,Y
C
C TO SORT DIFFERENT INPUT TYPES, CHANGE THE FOLLOWING
C SPECIFICATION STATEMENTS; FOR EXAMPLE, FOR FORTRAN CHARACTER
C USE THE FOLLOWING: CHARACTER *(*) A(N)
C
INTEGER A(N)
C
NDEEP=0
U1=N
L1=1
DO 1 I=1,N
1 ORD(I)=I
2 IF (U1.LE.L1) RETURN
C
3 L=L1
U=U1
C
C PART
C
4 P=L
Q=U
C FOR CHARACTER SORTS, THE FOLLOWING 3 STATEMENTS WOULD BECOME
C X = ORD(P)
C Z = ORD(Q)
C IF (A(X) .LE. A(Z)) GO TO 2
C
C WHERE "CLE" IS A LOGICAL FUNCTION WHICH RETURNS "TRUE" IF THE
C FIRST ARGUMENT IS LESS THAN OR EQUAL TO THE SECOND, BASED ON "LEN"
C CHARACTERS.
C
X=A(ORD(P))
Z=A(ORD(Q))
IF (X.LE.Z) GO TO 5
Y=X
X=Z
Z=Y
YP=ORD(P)
ORD(P)=ORD(Q)
ORD(Q)=YP
5 IF (U-L.LE.1) GO TO 15
XX=X
IX=P
ZZ=Z
IZ=Q
C
C LEFT
C
6 P=P+1
IF (P.GE.Q) GO TO 7
X=A(ORD(P))
IF (X.GE.XX) GO TO 8
GO TO 6
7 P=Q-1
GO TO 13
C
C RIGHT
C
8 Q=Q-1
IF (Q.LE.P) GO TO 9
Z=A(ORD(Q))
IF (Z.LE.ZZ) GO TO 10
GO TO 8
9 Q=P
P=P-1
Z=X
X=A(ORD(P))
C
C DIST
C
10 IF (X.LE.Z) GO TO 11
Y=X
X=Z
Z=Y
IP=ORD(P)
ORD(P)=ORD(Q)
ORD(Q)=IP
11 IF (X.LE.XX) GO TO 12
XX=X
IX=P
12 IF (Z.GE.ZZ) GO TO 6
ZZ=Z
IZ=Q
GO TO 6
C
C OUT
C
13 CONTINUE
IF (.NOT.(P.NE.IX.AND.X.NE.XX)) GO TO 14
IP=ORD(P)
ORD(P)=ORD(IX)
ORD(IX)=IP
14 CONTINUE
IF (.NOT.(Q.NE.IZ.AND.Z.NE.ZZ)) GO TO 15
IQ=ORD(Q)
ORD(Q)=ORD(IZ)
ORD(IZ)=IQ
15 CONTINUE
IF (U-Q.LE.P-L) GO TO 16
L1=L
U1=P-1
L=Q+1
GO TO 17
16 U1=U
L1=Q+1
U=P-1
17 CONTINUE
IF (U1.LE.L1) GO TO 18
C
C START RECURSIVE CALL
C
NDEEP=NDEEP+1
POPLST(1,NDEEP)=U
POPLST(2,NDEEP)=L
GO TO 3
18 IF (U.GT.L) GO TO 4
C
C POP BACK UP IN THE RECURSION LIST
C
IF (NDEEP.EQ.0) GO TO 2
U=POPLST(1,NDEEP)
L=POPLST(2,NDEEP)
NDEEP=NDEEP-1
GO TO 18
C
C END SORT
C END QSORT
C
END
--
Scott Lanning: [EMAIL PROTECTED], http://physics.bu.edu/~slanning
"Windows was designed to keep the idiots away from Unix so we could
hack in peace. Let's not break that." --Tom Christiansen
------------------------------
From: "Art S. Kagel" <[EMAIL PROTECTED]>
Subject: Re: DOS command line replacement
Date: Wed, 16 Jun 1999 17:06:39 -0400
Reply-To: [EMAIL PROTECTED]
Kaya Imre wrote:
>
> Would someone please tell me how to implemet the DOS command
>
> copy a + b > prn
>
> in Linux
cat a b > prn
If you mean "concatenate files a & b into a file named prn". If you
mean to print the contents of files a & b to the standard printer try
the following:
cat a b >>/dev/lp0
But either of the following is preferred since it uses the spooler
queue:
lpr a b
cat a b | lpr
Art S. Kagel
------------------------------
From: bryan <[EMAIL PROTECTED]>
Subject: Re: What are the differences between mySQL and mSQL?
Crossposted-To:
comp.os.linux.development.system,comp.os.linux.development.apps,comp.os.linux.setup,comp.lang.java.databases
Date: Wed, 16 Jun 1999 21:11:49 GMT
In comp.os.linux.development.apps Don Baccus <[EMAIL PROTECTED]> wrote:
: In article <_FA93.273$[EMAIL PROTECTED]>,
: bryan <[EMAIL PROTECTED]> wrote:
: >I like the trans. of postgresql, but it has some bugs that make me
: >want to avoid postgresql.
: >since I can structure my code (pretty much) so that transactions
: >aren't a MUST-HAVE, I choose mysql. for now, at least.
: Your criticism of postgres is valid, indeed I was going to
: use Sybase at one point (which is available at no cost for
: Linux, even deployed to back up a web site, and which is an
: enterprise-class db engine).
: However ... my experience with 6.5 betas over the last several
: weeks have convinced me that the current development crew has
: got a handle on fixing bugs and fixing many of the shortcomings
: that have caused many (including myself, in the past) to avoid
: Postgres.
: 6.5's a vast improvement over 6.4, particularly for
: high-concurrency environments like web sites.
: They still have a ways to go, but seem to have a handle on
: what's important and are working hard to make the thing
: even more solid. If 6.6 is as big an improvement over
: 6.5 as 6.5 is over 6.4, there will be little left to complain
: about.
agreed - and I fully intend to watch the stability of PGsql. and when
its ready, I'll be ready to integrate with them ;-)
as an exercise, I wrote a thin abstraction layer (regular C) that sits
above mysql and postgres. it allowed me to test both by doing:
make DB=PGRES
or
make DB=MYSQL
and then run the app. very handy having multiple sql engines to help
sort the app bugs out ;-)
I almost wish those guys would agree on a common api and that way, no
apps would have to be re-written (in theory).
--
Bryan [at] Grateful.Net
http://www.Grateful.Net
------------------------------
From: Erik de Castro Lopo <[EMAIL PROTECTED]>
Subject: Re: Linux systems- Poor security
Date: Thu, 17 Jun 1999 07:06:39 +1000
Tim Philip Williams wrote:
>
> A while ago we had a security breach involving multiple linux boxes and
> as a consequence, our IT staff will probably be implementing a ban of
> the use of Linux! I use Linux as a development workstation (although
> I'm not a UNIX expert) ... does anyone know why the security of Linux is
> so bad? I assume that not all distributions are bad, but the ones with
> poor security give Linux a very bad name ... infact they give all free
> software a bad name .. I doubt if the IT staff will let me install
> FreeBSD instead of Linux.
> A very annoyed Tim (who will probably be forced to use Windows NT)
Most linux distributions are not very secure out of the box. The same
can be said for Solaris, HP-UX, SCO and most other unices. All of these
systems can however be secured to be more secure than NT.
There is currently two separate efforts to produce maximal security
Linux distributions of Linux. You can probably read about them on
comp.os.linux.security.
Cheers,
Erik
--
+-------------------------------------------------+
Erik de Castro Lopo [EMAIL PROTECTED]
+-------------------------------------------------+
"Only wimps use tape backup: *real* men just upload their
important stuff on FTP, and let the rest of the world
mirror it ;)" -- Linus Torvalds
------------------------------
From: bryan <[EMAIL PROTECTED]>
Subject: Re: What are the differences between mySQL and mSQL?
Crossposted-To:
comp.os.linux.development.system,comp.os.linux.development.apps,comp.os.linux.setup,comp.lang.java.databases
Date: Wed, 16 Jun 1999 21:13:09 GMT
In comp.os.linux.development.apps Don Baccus <[EMAIL PROTECTED]> wrote:
: In article <8Gs93.60$[EMAIL PROTECTED]>,
: bryan <[EMAIL PROTECTED]> wrote:
: >simple - cause not all the linux db's HAVE transactions. that's why!
: Arguing that you don't need transactions because you don't use
: a db that does them is a bit like arguing that backups are useless
: because you personally don't own a backup device...
I NEVER (!) used the word or concept of 'useless'. of COURSE xactions
aren't useless. BUT not every app needs them - and some apps can be
written in such a way that xactions aren't essential.
--
Bryan [at] Grateful.Net
http://www.Grateful.Net
------------------------------
From: [EMAIL PROTECTED]
Subject: Linux&PHP+NT&Oracle
Date: Wed, 16 Jun 1999 20:10:26 GMT
Hello,
We use a Linux server with Apache & PHP3 and an NT server with Oracle8.
Is there any wise soul who could tell us what we need to do in order to
get the data from the Oracle database to the php-generated pages.
-Chad
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: [EMAIL PROTECTED] (Mike Khalili)
Crossposted-To: comp.os.linux.security
Subject: Re: SUID programs: are they normal?
Date: 16 Jun 1999 22:00:55 GMT
Reply-To: [EMAIL PROTECTED]
On Wed, 16 Jun 1999 21:28:47 GMT, Juergen Heinzl <[EMAIL PROTECTED]> wrote:
>In article <[EMAIL PROTECTED]>, Mike Khalili wrote:
>>On Wed, 16 Jun 1999 16:16:17 -0400, Ding-Jung Han <[EMAIL PROTECTED]> wrote:
>>>Hi,
>>>
>>>I'm a newbie to all this security stuff. Just checked if there's any
>>>suid programs on my linux box, and I found more than I expected (see
>>>below). Are they normal? Is there any one that is unnecessarily suid'd?
>>
>>Depends what you call normal. Depends what you call necessary. I'll put
>>some quick explanations for the stuff I'm familiar with.
>
>I'll join in and try to fill out the rest ...
>[...]
>>>/usr/X11R6/bin/Xwrapper
>>>/usr/bin/chage
>>
>>Not sure. Don't think so.
>Yes, it must be able to read shadow but "normal" users can only
>use -l.
Things that need read-only privledge to the shadow file should *NOT* be
setuid(). That's too much privledge. Create a groups shadow. chgrp to
group shadow on the shadow file. set the perms to 640. Then make all the
apps that need read permission to the shadow file sgid shadow. This means
if something's wrong, the worst that can happen is someone gets to read (but
not write to) your shadow file, rather than having them have write access to
the shadow file as well as having all root perms.
>>>/usr/libexec/pt_chown
>>
>>Not sure. Depends if something uses it.
>Yes, but ... the but is some suid programmes do not require it
>like xterm but others do, leave it.
>[...]
The best way to implement this I think is to create a term daemon. The term
daemon runs as root. The only thing it does is allocate ttys to the
terminals. The term (xterm, rxvt) ect. should use ipc to tell the daemon
"Hey I could use a terminal", and if one is available the term daemon could
allocate one, and take care of writing to utmp. The term daemon could also
keep track of what terms have died, and don't need a tty anymore, thus
eliminating the possibility of term bugs leading to running out of ttys
(which could be a DOS attack). The pt_chown approach sounds like an
improvement over the traditional way of doing things, but still not quite as
good as this. Note: me saying this would be a good idea, doesn't mean I'm
going to sit down and write it. :)
------------------------------
From: Stephen George <[EMAIL PROTECTED]>
Subject: Re: Repartition EXT2 without data loss?
Date: Wed, 16 Jun 1999 16:29:58 -0400
Hi
Dont know if this will help but I have used a product called
partion magic to shrink partions. It says it supports linux
Stephen
------------------------------
From: Stephen George <[EMAIL PROTECTED]>
Subject: undefined reference to ct_bind
Date: Wed, 16 Jun 1999 16:43:17 -0400
Dear anyone
I am trying to compile with GNU c the sybase open client
programs.
I must have spent 30hrs on this and all I get is undefinded
reference to ct_describe + ct_cancel and lots more. Has anyone
any idea what I am doing wrong. As far as I can tell the code
gets to the link stage and it falls over in a heap.
I am typing in the $SYBASE/sample/ctlibrary directory
make all
or
make rpc
or
cc rpc.c -I. -I$SYBASE/include -L. -L$SYBASE/lib
Its still does not work and is driving me completely bananas
As far as I can see I have set all parameters corectly.
Any advice no matter how daft sounding will be gratefully
received.
------------------------------
From: [EMAIL PROTECTED] ()
Crossposted-To: comp.os.ms-windows.advocacy,comp.os.linux.advocacy,gnu.misc.discuss
Subject: Re: A Capitalists view of freedom
Date: Wed, 16 Jun 1999 13:36:21 -0700
On Wed, 16 Jun 1999 19:20:26 +0200, Matthias Warkus <[EMAIL PROTECTED]> wrote:
>It was the Wed, 16 Jun 1999 09:04:55 -0700...
>...and [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
>> On Tue, 15 Jun 1999 23:08:33 +0200, Matthias Warkus <[EMAIL PROTECTED]> wrote:
>> >It was the 15 Jun 1999 17:44:20 GMT...
>> >...and Aaron M. Renn <[EMAIL PROTECTED]> wrote:
>> >> On Tue, 15 Jun 1999 18:46:26 +0200, Richard Hickling <[EMAIL PROTECTED]>
>wrote:
[deletia]
>>
>> We're not the ones spouting righteous indignation about
>> our nation being more 'orderly' because we treat the
>> general population as if they cannot be trusted.
>>
>> This kind of situation rather supports our quaint notions
>> regarding liberty and immovable laws.
>
>Oh really.
Obviously, since you don't seem to have any problem
with the fundementally anti-democratic notion that
the general population can't be trusted.
Today the gun, tomorrow the ballot box. It's the
same justification either way. Logic good for one
is just as good for the other.
--
bash: the power to toast your registry in style... |||
/ | \
Seeking sane PPP Docs? Try http://penguin.lvcm.com
------------------------------
From: [EMAIL PROTECTED] ()
Crossposted-To:
comp.os.linux.networking,omp.os.ms-windows.nt.advocacy,comp.os.linux.advocacy,comp.infosystems.www.servers.unix
Subject: Re: Could Microsoft Cheat On The New Mindcraft Benchmark? (was: Mindcraft
Retest News
Date: Wed, 16 Jun 1999 13:37:18 -0700
On 16 Jun 1999 18:10:47 GMT, I R A Aggie <[EMAIL PROTECTED]> wrote:
>On 16 Jun 1999 12:51:36 -0400, Paul D. Smith <[EMAIL PROTECTED]>, in
><[EMAIL PROTECTED]> wrote:
>
>[about IE under Solaris]
>
>+ - It doesn't have any features except the browser (no mail, no news
>+ client, no editor, etc.)
>
>You say that like it is a bad thing.
For as big as it is, it is a bad thing.
--
bash: the power to toast your registry in style... |||
/ | \
Seeking sane PPP Docs? Try http://penguin.lvcm.com
------------------------------
From: [EMAIL PROTECTED] (Craig Dickson)
Crossposted-To: comp.lang.functional
Subject: Re: kernel mode programs (yes, this is on topic)
Date: 16 Jun 1999 14:05:31 -0700
Matt Harden <[EMAIL PROTECTED]> wrote:
>Craig Dickson wrote:
>
>> FP doesn't necessarily involve GC. The ML Kit uses regional analysis of the
>> code to infer at compile time when memory needs to be allocated and when it
>> can safely be freed.
>
>That sounds very nice. But since ML is not pure, it's harder to verify
>code safety. Does laziness make the regional analysis more difficult?
I don't have an answer for that one. It might.
>Does this analysis always result in code that doesn't need GC? It would
>be very nice to have this feature in a lazy language like Haskell.
I'm not sure if the ML Kit is completely GC-free, but my impression, from
what I've read, is that it is.
Leaving regional analysis aside, though, there are other techniques that
can be used to avoid freezing the whole system for a GC sweep. GC based on
reference counts makes it possible to delete things immediately when they
are no longer needed, just as one would do manually in a language like C.
This sort of GC could be implemented for a lazy language pretty easily, I
think.
Craig
------------------------------
From: "Matthew Haley" <[EMAIL PROTECTED]>
Subject: Re: tell modem "not to wait for dial tone"?
Date: Wed, 16 Jun 1999 14:04:54 -0700
Also at http://www.modem.com is a reference of the AT command set.
kokin wrote in message <7jvo2h$bho$[EMAIL PROTECTED]>...
>This is a option to tell the modem not to wait for dial tone before
>dialling out. How can i do this under Linux? Without this my tel
>line doesn't co-operate with my modem.
>
>Thanks in advance.
>
>kokin @ hong kong
------------------------------
** FOR YOUR REFERENCE **
The service address, to which questions about the list itself and requests
to be added to or deleted from it should be directed, is:
Internet: [EMAIL PROTECTED]
You can send mail to the entire list (and comp.os.linux.misc) via:
Internet: [EMAIL PROTECTED]
Linux may be obtained via one of these FTP sites:
ftp.funet.fi pub/Linux
tsx-11.mit.edu pub/linux
sunsite.unc.edu pub/Linux
End of Linux-Misc Digest
******************************
