Linux-Networking Digest #763, Volume #11 Fri, 2 Jul 99 18:14:22 EDT
Contents:
Re: Maximus/opus or clone of them for Linux (B'ichela)
Re: Can I use Linux as a server to a Windows based calendar/scheduling application?
("Steve Hiner")
Linux-to-Linux and IP Masq ("Scotty Perkins")
Re: Why not C++ (Nathan Myers)
Bootpd failing on Redhat 5.2 (Chuck Bolvin)
3Com 3c503/16 etherlink II problem. (Jason Martinez)
Cause of many RX errors, 3c595? (Mike Klein)
Re: php3 (Bill Pitz)
eth0: transmit timeout ([EMAIL PROTECTED])
Re: C++ templates: More than Turing Complete? (Nathan Myers)
Re: Ethernet to Token-Ring Bridge ("Andrey Smirnov")
Re: Office2K install from Samba fails (Bill Pitz)
Re: Problems running PPP scripts.... (Clifford Kite)
Re: aol under linux? (Clifford Kite)
Re: Installing RH linux onto P166 laptop (Bill Pitz)
Re: IP Masquerading problem ("chris")
Re: Private DNS useless?? (Bernd Eckenfels)
d-link card ("Jim Ingram")
aol under linux? (LikeFUN024)
Re: Good docs for PGP/Linux? (Bill Pitz)
Re: Why not C++ (Johan Kullstam)
Re: C++ templates: More than Turing Complete? (Johan Kullstam)
Re: C++ templates: More than Turing Complete? (Greg Comeau)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (B'ichela)
Subject: Re: Maximus/opus or clone of them for Linux
Date: Fri, 2 Jul 1999 13:27:57 -0400
Reply-To: [EMAIL PROTECTED]
In article <7lhe1i$auk$[EMAIL PROTECTED]>, Jonathan Guthrie wrote:
>That is correct. I spoke with Scott Dudley (the guy who wrote Maximus)
>about it about a year ago, and told him that I wanted to do something like
>a Linux port of Maximus. Scott has essentially stopped working on Max
>because he feels that the world has gone beyond BBSes.
Same thing happened with Opus development. Amazingly there is ONE
bbs left in my area that I actually know of. Opus 1.79 system called
The Anxiety Closet 860-423-6099. I used to pass mail to/from him when I was
in fidonet.
I myself do not feel bbs's are dead as there is still a small nitch
where they still can beat the pants off of the internet, local area
discussions and community feel. With the internet one does not get that
type of local feel.
(I just reactivated my bbs line for a bbs system for linux if I can find
one. you can call it at 860-423-2782). The thing is with Linux I can set up
both a local dialup and a internet telenetable site. (I just ordered a 4 com
card that is similar to the AST fourport card so I can plug in more terminals
and modems here).
> Scott had, at one
>point some time around 1992 or 1993, offered me the source code for Max so
>that I could port it to Linux. I didn't take it at that time, although I
>wish that I had.
>
Bummer, I really would like to see a Maximus or Opus linux software package
for those of us who understood that software and wanted to use it with the
internet/dialup capability of linux.
>Bob Juge, who has Scott's ear and who happens to be a customer of mine,
>has told me that he has suggested that Scott declare the product
>commercially dead and release the source for Max to everyone. So far, it
>hasn't happened. We shall see. (Who knows. Perhaps I shall respond to
>the email he wrote me.)
What email? No matter. if you could find out more I would
appreciate any leads or compatible bbs's for linux that are set up similar
to opus/maximus. BTW if maximus/opus can use a linux version of squish that
would help for fidonet echomail too. I really am trying to get "Planet
Maca's Opus" back online again.
--
A pearl of wisdom from the y2K newsgroups:
=========================================================================
Y2K appears to be the Baby Boomers mid-life crisis, and it has the
potential to be a dandy.
-- Anonymnous --
==========================================================================
B'ichela
------------------------------
From: "Steve Hiner" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc,comp.os.linux.questions
Subject: Re: Can I use Linux as a server to a Windows based calendar/scheduling
application?
Date: Fri, 2 Jul 1999 11:04:30 -0700
Karl-Heinz,
I knew it would be a good idea to post my question. I didn't know I would
get this good a suggestion.
I hadn't even considered StarOffice. I can save on download time by using
that copy sitting on my hard drive (I knew there was some reason I hadn't
deleted it even though I wasn't using it). I have the Linux and Windows
versions of the original 5.0 release of SO (I think there have been bug
fixes since I picked it up).
How does the server side work?
Does the Linux version have the server software included with it?
This actually solves another big problem over there. They are currently
using WordStar 2000 (ick). I grew up using WordStar on my old 8088 system
but by today's standards it is way out of date. Remember the days when you
could fit your OS, wordprocessor and all your documents on one 5.25" 360k
floppy?
I am going to have to go check out the pricing. Knowing you work over
there - do they give discounts to non-profit organizations? (I realize that
the normal selling price is basically a discount, especially compared to MS
Office, but I thought I'd check)
Wow, I could do so much with StarOffice, relatively cheaply too. You may
have just sold 30 copies of your suite! (If we do buy it will it do you any
good for me to tell them that you are responsible? If you write back it
would be beneficial if you would email me directly as well since my ISPs
news server tends to kill messages quickly and this is a long weekend here
in the US.)
A very happy,
Steve Hiner
developer {AT} isiaz {DOT} com
------------------------------
From: "Scotty Perkins" <[EMAIL PROTECTED]>
Subject: Linux-to-Linux and IP Masq
Date: Fri, 2 Jul 1999 12:12:32 -0700
Here's an easy for you, hopefully:
I have a RH 5.2 box that has IP Masq configured to masquerade any device on
my private subnet (192.168.x.x) as the modem when I dial up to my ISP. I
have three computers on my little network using this setup without any
issues, a WinNT/Win98 dual-boot box, a Win95 laptop, and a Mac.
However, I added RH 6.0 to the dual-boot box, and when I boot into it, I
cannot get any packets, even though I configured the network/subnet
mask/gateway the same as on the other OS configurations. I also used an IP
one up from the Windows configuration (.6 versus .5) just so there would not
be a hostname-versus-IP conflict when I boot to a different hostname in
Linux as in NT/98.
Is there something trivial about RH 6.0 or Linux-to-Linux connectivity
through IP Masq that I need to worry about?
Any suggestions are appreciated.
scotty
------------------------------
From: [EMAIL PROTECTED] (Nathan Myers)
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.development.system
Subject: Re: Why not C++
Date: 2 Jul 1999 13:44:21 -0700
Timo Tossavainen <[EMAIL PROTECTED]> wrote:
>Your Lisp-bashing has gone a bit too far.
I have not bashed Lisp. I have not seen any bashing of Lisp here.
The subject is C++, and I have posted correcting falsehoods
written about C++. It happens that Lisp doesn't meet my needs,
but it doesn't seem to have been meant to. That's OK.
--
Nathan Myers
[EMAIL PROTECTED] http://www.cantrip.org/
------------------------------
From: Chuck Bolvin <[EMAIL PROTECTED]>
Subject: Bootpd failing on Redhat 5.2
Date: Fri, 02 Jul 1999 15:52:43 -0500
I've just recently re-installed a machine with RedHat 5.2 that was
running 5.1. The install went very smoothly. Everything is running well
except booptd. I have used the same 'bootptab' and the entries in the
tftpboot supdirectory are the same. The install of bootp from the RedHat
CD also went ok. No errors when compiled and installed. The problem I'm
seeing is every time a bootp message is received, Linux sees it, but
does not respond to it. Instead I keep seeing messages in the log that
show the message coming in, but after about 30 entries or so, the kernel
shuts down the bootpd daemon with a message saying the bootp service is
failing (looping). The version of bootp I'm using is the one from the
RedHat CD, version 2.4.3-7 I believe. I can't seem to figure out how to
find the problem. I get no other diagnostic messages.
The devices sending in the bootp requests are Wireless LAN Access
Points.
Any ideas anyone????
Thanks In Advance!!!
Chuck Bolvin
Mgr, Technical Services
POS Information Systems
------------------------------
Date: Fri, 02 Jul 1999 13:16:09 -0700
From: Jason Martinez <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Subject: 3Com 3c503/16 etherlink II problem.
I have been having some problems with this 3c503 etherlink card (I
think).
I have SuSE 6.1 and it reconizes the card and all diagnostics indicate
the card is working but when I go to communicate with another machine
nothing happens. So, a couple of questions.
1. There appears to be a lamp near the interal transiever, is this
suppose to blink when there is signals coming in and out? (mine does not
do this).
2. Is there a way to test just the card? because using the loopback
works with or without the card. (so does pinging the local machine).
So before I throw money at some new cards I would like to know if there
is a way to tell if the card is physically working.
Thanks
Jason Martinez
[EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (Mike Klein)
Subject: Cause of many RX errors, 3c595?
Reply-To: mike - at - kleinnet.com
Date: Fri, 02 Jul 1999 17:20:35 GMT
I have an RH5.1 system with 2.0.36 kernel and Donald Becker's updated
3com drivers (0.99H). There are 2 ethernet cards in the system, both
3c595. eth0 serves the internal network (small home network) and eth1
my ADSL connection. They are on different IRQs.
The problem: eth0 consistently has many RX errors, which appears to be a
significant performance factor. There is a large difference between
network bandwidth into and out of the Linux box, and slow performance
seems to be correlated with a large increase in reported RX errors.
I turned on debugging in the ethernet driver module and it reports that
every error consists of a CRC error, with a few other occasional other
types. My suspicion is once the CRC errors are solved, the others will
be also.
eth1, the card attached to the ADSL box, has no errors of any kind.
Debugging steps I have already tried:
- try a different 3c595. In fact I have tried 3 different
ones! They all have the same problem.
- assign a different IRQ. No change.
- remap priorities using irqtune. No change.
This problem has been the one stumping me the longest on this system,
which is otherwise a model of excellent performance and reliability. It
appears others have similar problems (but apparently rarely) based on a
Deja search.
Any help here would be greatly appreciated!
ifconfig and messages data follows. Please let me know if there is any
other data that would be useful.
Thank you,
-Mike Klein
mike - at - kleinnet.com
eth0 Link encap:Ethernet HWaddr XX:XX:XX:XX:XX:XX
inet addr:192.168.0.1 Bcast:192.168.0.255 Mask:255.255.255.0
EtherTalk Phase 2 addr:65280/81
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:1786793 errors:21874 dropped:21874 overruns:0
TX packets:1391613 errors:0 dropped:0 overruns:0
Interrupt:10 Base address:0xff80
kernel: eth0: 3Com 3c595 Vortex 100baseTX at0xff80, XX:XX:XX:XX:XX:XX,
IRQ 10
kernel: 64K word-wide RAM 3:1 Rx:Tx split, autoselect/10baseT
interface.
kernel: eth0: Overriding PCI latency timer (CFLT) setting of 64, new
value is 248.
------------------------------
From: Bill Pitz <[EMAIL PROTECTED]>
Subject: Re: php3
Date: Fri, 02 Jul 1999 03:23:48 GMT
Haaino Beljaars <[EMAIL PROTECTED]> gave us the interesting posting of:
http://www.amazon.com and search for "PHP"
They have two books (Core PHP Programming and Practical PHP) there, both
of which I would consider to be excellent resources. You're much better
off going with print resources...
-Bill
--
Bill Pitz [EMAIL PROTECTED]
Silicon Valley North, Inc. www.svn.net
Internet and World Wide Web Services (707) 781-9999
------------------------------
From: [EMAIL PROTECTED]
Subject: eth0: transmit timeout
Date: Fri, 02 Jul 1999 20:54:35 GMT
I recently did a workstation install of Redhat
6.0, on a machine that had been running 5.1 which
was connected to our LAN. The ethernet card it
uses is AMD PC/NET32, and the card was detected
during the installation. I get the following
error message in console mode: eth0: transmit
timeout status 0373... retransmitting. Then the
error message repeats itself. I am unable to
access the LAN... presumably because the card is
not working properly. Any suggestions?
-Gideon Caplovitz
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: [EMAIL PROTECTED] (Nathan Myers)
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.development.system
Subject: Re: C++ templates: More than Turing Complete?
Date: 2 Jul 1999 13:49:34 -0700
Davin McCall <[EMAIL PROTECTED]> wrote:
>[EMAIL PROTECTED] (Nathan Myers) wrote:
>
>>The convention in some communities is to ignore performance
>>and engineering rigor. Certainly anything can be achieved in
>>assembly language (or in T-code, if you like) that can be achieved
>>with a C++ _program_, given infinite time and coding effort.
>
>Infinite? Finite, surely!
"Indefinite time and coding effort", then, if you like.
But by quibbling you have missed the point.
>>Given the constraints of a sensible _library_ interface and
>>equivalent-to-machine-code performance, C++ achieves what
>>other languages haven't, yet.
>
>I would say "approaching sensible" and almost completely disagree with
>the performance statement. In most cases a compiled program written in
>a high level language performs very acceptably, but not nearly as well
>as an equivalent program hand coded in assembly language (or machine
>code).
This may have been true at one time. Modern CPUs are so complex
that hand-coding is unlikely to do better than a good compiler
on a significant program.
The point you missed by quibbling is that once you cleave the solution
space along a library boundary, you have left the domain of "computer
science" and are firmly in the domain of "engineering" where your
precious axioms are just obvious facts, and the hard problems
involve tradeoffs and organizational choices.
Language features which help transport these choices back and
forth across that boundary are essential engineering tools that
are often lacking in academic wonder-langauges.
>>Of course some other languages will achieve parity, someday, but
>>only those designed by people who fully understand the strengths of
>>standard C++. Maybe such a language will even succeed in avoiding
>>its weaknesses.
>
>I take it that you mean they must understand the principles, although
>not necessarily how they are applied in C++.
No, absolutely the opposite! Real, useful programs are written
using real language features. To understand principles you must
first understand the specific application. All valid principles
are derived from experience, however they may be dressed up after
the fact.
--
Nathan Myers
[EMAIL PROTECTED] http://www.cantrip.org/
------------------------------
From: "Andrey Smirnov" <[EMAIL PROTECTED]>
Subject: Re: Ethernet to Token-Ring Bridge
Date: Fri, 2 Jul 1999 14:03:35 -0700
Hello,
Unfortunately I can't give you an advise about bridging, but bear with me
for a second.
If you want to communicate between token ring network and ethernet network
(two separate TCP/IP networks!), you can use Linux machine with ethernet and
token ring cards as a router.
Good luck!
[EMAIL PROTECTED] wrote in message <7lj7q3$m2k$[EMAIL PROTECTED]>...
>Is there anyone in the Linux Community that has
>successfully setup a Ethernet to Token-Ring
>Bridge? If so could you please steer me in the
>right direction.
>
>ri
>
>
>Sent via Deja.com http://www.deja.com/
>Share what you know. Learn what you don't.
------------------------------
From: Bill Pitz <[EMAIL PROTECTED]>
Subject: Re: Office2K install from Samba fails
Crossposted-To: linux.samba,linux.redhat.misc,comp.protocols.smb
Date: Fri, 02 Jul 1999 21:14:50 GMT
In comp.os.linux.networking Greg Leblanc <[EMAIL PROTECTED]> babbled:
> Oops, guess I left that out of my post. I've tried it via both UNC
> (\\server\share\) and via a mapped drive letter. It really annoys me
> that I can't run it from my Samba server, but it works from my POS NT
> server. :-) Later,
> Greg
It's probably not going to let you since you're running Samba. The
installer has a way to detect it, I'm sure... But it is interesting
that you are able to run it from an NT server ... accessing it the
same way via \\server\share and a mapped drive? That makes it even
more strange just because I would guess something like that would be
implemented for "copy protection"
It could also be something to do with the way that UNIX handles files
as being case-sensetive... You might check some of the samba options
about that...
-Bill
--
Bill Pitz [EMAIL PROTECTED]
Silicon Valley North, Inc. www.svn.net
Internet and World Wide Web Services (707) 781-9999
------------------------------
From: kite@NoSpam.%inetport.com (Clifford Kite)
Subject: Re: Problems running PPP scripts....
Date: 2 Jul 1999 16:13:01 -0500
Cheesewizz ([EMAIL PROTECTED]) wrote:
: Here's my problem: I use a script 'ppp-on,' which connects just fine if
: the root user executes it, but if I execute that same script as a normal
: user, the script fails with these messages:
: Jul 2 00:57:34 blah pppd[296]: pppd 2.3.5 started by bob, uid 500
: Jul 2 00:57:35 blah pppd[296]: Connect script failed
: Jul 2 00:57:36 blah pppd[296]: Exit.
: I set pppd to run suid root, so that is not the problem, and I dont know
: what could be causing it. I have a feeling that it has something to do
: with permissions, but I don't know what file is causing this.
: This is what my ppp-on script looks like, if that will help:
: #!/bin/sh
: exec /usr/sbin/pppd connect 'chat -v "" ATL1DT8228836 CONNECT \
: "" ogin: username word: password' /dev/ttyS0 115200 noipdefault crtscts
: \
: modem defaultroute
Try specifing /usr/sbin/chat, the paths defined for the ordinary user
seldom include the system bin directories.
--
Clifford Kite <kite@inet%port.com> Not a guru. (tm)
/* Those who can't write, write manuals. */
------------------------------
From: kite@NoSpam.%inetport.com (Clifford Kite)
Subject: Re: aol under linux?
Date: 2 Jul 1999 15:02:58 -0500
LikeFUN024 ([EMAIL PROTECTED]) wrote:
: i've heard of some people getting aol to work under linux and i'm trying to
: find a page that might explain it to me in detail. any information would be
: greatly appreciated. [EMAIL PROTECTED]
Good Luck! AFAIK AOL uses AOL software exclusively.
--
Clifford Kite <kite@inet%port.com> Not a guru. (tm)
/* Editing with vi is a lot better than using a huge swiss army knife. */
------------------------------
From: Bill Pitz <[EMAIL PROTECTED]>
Subject: Re: Installing RH linux onto P166 laptop
Date: Fri, 02 Jul 1999 21:18:48 GMT
Ng, Choon Hooi <[EMAIL PROTECTED]> gave us the interesting posting
of:
> Hi,
> I am planning to install RH linux 5.2 onto my P166 Hitachi laptop,
> which serves as a server for my LAN. 2 reasons of doing that. 1. It is
> the most powerful PC among all my PCs. 2. I travel once in awhile, and
> hence I need to connect to a different network.
> Things that I wanna setup are: network connection, PPP, IP Masq &
> Email.
> My laptop specs:
> Hitachi P166 MMX, 64 MB, build-in sound card, build in modem, build-in
> 10BT nic, 3.5" drive, and CD-ROM (build-in)
As long as the installer picks up your CD-ROM drive, you're probably on
your way. I've installed it on several laptops both for my own use and
for other people. I haven't run into any problems yet. If you want to
run X, you might check into the compatilibity of your video chipset with
the XFree86 servers... pretty much everything runs under the SVGA server
these days, but there are still some weird proprietary chips that might
not work. Also, make sure the built-in modem is not a Winmodem...
If you can get it, I would aslo recommend getting Red Hat 6.0 as the
2.2 kernel and some of the new software have much better support for
PCMCIA/Builtin devices.
-Bill
--
Bill Pitz [EMAIL PROTECTED]
Silicon Valley North, Inc. www.svn.net
Internet and World Wide Web Services (707) 781-9999
------------------------------
From: "chris" <[EMAIL PROTECTED]>
Subject: Re: IP Masquerading problem
Date: Fri, 2 Jul 1999 16:55:34 -0400
Personally these are the lines I use for ip masquerading and the pages you
listed worked fine
ipchains -P forward DENY
ipchains -A forward -s 192.168.1.2/24 -j MASQ
one thing you could trying is making sure ipchains pays close attention to
www by doing
ipchains -A forward -p TCP -d 192.168.1.2/24 ( www or 80 ) -j MASQ
give that a try other wise it might be a problem else where.
Chris
--
Chris Spears
cs1501RecTA / cs2330RecTA
[EMAIL PROTECTED]
Mark Ott <[EMAIL PROTECTED]> wrote in message news:[EMAIL PROTECTED]...
> I'm running RedHat 6.0, with IP Masquerading
> set up to provide internet access to my other
> computers, and it works great, except for one
> problem....
>
> Some web sites just won't come through on the
> other computers. Almost all sites come through
> fine, but a few, such as www.buy.com, www.sony.com,
> and lately, www.gamestorm.com will not. The web
> browser says it connected to the site, but then it
> just waits forever.
>
> These sites do come through fine on the machine
> with the modem, so I think it must be a problem
> with masquerading.
>
> I can telnet into the sites, and type GET, which
> works fine on the machine with the modem, but
> on the others, it just waits forever.
>
> I use this script to get masquerading started...
>
> echo "1" > /proc/sys/net/ipv4/ip_forward
> /sbin/depmod -a
> /sbin/modprobe ip_masq_ftp.o
> /sbin/modprobe ip_masq_raudio.o
> /sbin/modprobe ip_masq_irc.o
> /sbin/ipchains -P forward DENY
> /sbin/ipchains -A forward -j MASQ -s 192.168.1.0/24 -d 0.0.0.0/0
>
> I also had this problem using RedHat 5.2
> and ipfwadm
>
> BTW, the gamestorm site just started not working,
> and I can't think of anything that's changed
> since it did work.
>
> Any ideas or suggestions????
>
> thanks
>
> Mark
------------------------------
From: Bernd Eckenfels <[EMAIL PROTECTED]>
Subject: Re: Private DNS useless??
Date: 2 Jul 1999 19:20:59 GMT
Wouter Liefting <[EMAIL PROTECTED]> wrote:
> I would say no. /etc/hosts is much easier to set up. Especially if you only
> have five machines or so, and not that many configuration changes.
like dhcp?
Greetings
Bernd
------------------------------
From: "Jim Ingram" <[EMAIL PROTECTED]>
Subject: d-link card
Date: Fri, 2 Jul 1999 15:54:07 -0500
Does any one have any information about getting Red Hat 5.2 to work with a
d-link DFE-50TX? Any help would be appreciated.
-Jim
------------------------------
From: [EMAIL PROTECTED] (LikeFUN024)
Subject: aol under linux?
Date: 02 Jul 1999 19:22:44 GMT
i've heard of some people getting aol to work under linux and i'm trying to
find a page that might explain it to me in detail. any information would be
greatly appreciated. [EMAIL PROTECTED]
------------------------------
From: Bill Pitz <[EMAIL PROTECTED]>
Subject: Re: Good docs for PGP/Linux?
Date: Fri, 02 Jul 1999 21:20:34 GMT
Dave Ewart <[EMAIL PROTECTED]> gave us the interesting posting of:
> I have installed PGP 5.0i on my RedHat 5.2 box - I find that I am in need of
> a decent "User's guide" specific to the Linux version (or at least to the
> Unix versions in general), similar to that available for the Windows
> 95/98/NT versions, since there seem to be rather a lot of undocumented
> command line parameters. I have a feeling that these parameters might do
> what I have been trying to sort out for a while.
> Any suggestions? Thanks ...
You really only need:
man pgpk
man pgpe
man pgps
man pgpv
man pgp
And a few other manual pages that are listed at the bottom of those. That
basically describes all of the functionality that you could possibly need
to perform normal functions with PGP for Linux.
If something isn't listed in any of those man pages, what is it that you are
trying to do? I may have it in some docs somewhere...
-Bill
--
Bill Pitz [EMAIL PROTECTED]
Silicon Valley North, Inc. www.svn.net
Internet and World Wide Web Services (707) 781-9999
------------------------------
From: Johan Kullstam <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.development.system
Subject: Re: Why not C++
Date: 02 Jul 1999 14:37:55 -0400
o r c @ p e l l . p o r t l a n d . o r . u s (david parsons) writes:
> In article <[EMAIL PROTECTED]>,
> Bruce Hoult <[EMAIL PROTECTED]> wrote:
>
> >Dylan is not several hundred percent slower than C as Java is.
>
> Have you benchmarked Java vs C on the same machine? (No, I don't
> mean benchmarking C on a machine vs Java running on a p-machine
> on that machine); there's certainly nothing in the design of
> Java that would make it much slower than C on the same machine.
>
> Dylan is Yet Another Pascal, isn't it? It looks like the bastard
> child of a shotgun marriage between Ada and Pascal.
afaict dylan is the bastard child of lisp and the algol family (which
includes pascal, ada and sort of includes C and C++). this isn't to
say that dylan is all bad. it is taking some good ideas of both.
--
johan kullstam
------------------------------
From: Johan Kullstam <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.development.system
Subject: Re: C++ templates: More than Turing Complete?
Date: 02 Jul 1999 14:41:07 -0400
[EMAIL PROTECTED] (Davin McCall) writes:
> On 1 Jul 1999 01:57:18 -0700, [EMAIL PROTECTED] (Nathan Myers)
> wrote:
>
> >The convention in some communities is to ignore performance
> >and engineering rigor. Certainly anything can be achieved in
> >assembly language (or in T-code, if you like) that can be achieved
> >with a C++ _program_, given infinite time and coding effort.
>
> Infinite? Finite, surely!
>
> To say that something takes an infinite amount of time to do is to say
> that it can't be done. And yet, you will find that an assembly
> language (or machine code) program of finite size and complexity is
> generated whenever you compile a high level language program, even a
> C++ program.
C++ templates are turing complete (although you'd have to be some kind
of masochist to use them as a programming language -- it can be done).
you have a classic stopping problem. therefore, C++ will not
necessarily compile in finite time to a finite sized program.
--
johan kullstam
------------------------------
From: [EMAIL PROTECTED] (Greg Comeau)
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.development.system
Subject: Re: C++ templates: More than Turing Complete?
Date: 2 Jul 1999 16:12:22 -0400
Reply-To: [EMAIL PROTECTED]
In article <[EMAIL PROTECTED]> [EMAIL PROTECTED] (Davin
McCall) writes:
>On 1 Jul 1999 01:57:18 -0700, [EMAIL PROTECTED] (Nathan Myers)
>wrote:
>
>>The convention in some communities is to ignore performance
>>and engineering rigor. Certainly anything can be achieved in
>>assembly language (or in T-code, if you like) that can be achieved
>>with a C++ _program_, given infinite time and coding effort.
>
>Infinite? Finite, surely!
>
>To say that something takes an infinite amount of time to do is to say
>that it can't be done. And yet, you will find that an assembly
>language (or machine code) program of finite size and complexity is
>generated whenever you compile a high level language program, even a
>C++ program.
What Nathan meant was clear to me. He just meant no time constaint
to complete it, not literally that the person must take infinite time.
>>Given the constraints of a sensible _library_ interface and
>>equivalent-to-machine-code performance, C++ achieves what
>>other languages haven't, yet.
>
>I would say "approaching sensible" and almost completely disagree with
>the performance statement. In most cases a compiled program written in
>a high level language performs very acceptably, but not nearly as well
>as an equivalent program hand coded in assembly language (or machine
>code).
At least not nearly as well _as some_ equiv progs....
Nobody said it didn't. See Nathan's comment you quote above.
This is CLEARLY a compromise we must all strike.
>>Of course some other languages will achieve parity, someday, but
>>only those designed by people who fully understand the strengths of
>>standard C++. Maybe such a language will even succeed in avoiding
>>its weaknesses.
>
>I take it that you mean they must understand the principles, although
>not necessarily how they are applied in C++.
I took it to mean something like that.
Certainly nobody is denying that C++ has problems.
I hope one day many new somethings will out-do _everything_ out there
or we are in some bad bad bad shape.
- Greg
--
Comeau Computing, 91-34 120th Street, Richmond Hill, NY, 11418-3214
Producers of Comeau C/C++ 4.2.38 -- New Release! We now do Windows too.
Email: [EMAIL PROTECTED] / Voice:718-945-0009 / Fax:718-441-2310
*** WEB: http://www.comeaucomputing.com ***
------------------------------
** 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.networking) 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-Networking Digest
******************************
