RG> The firewall really does take care of the brunt of such
RG> probes/attacks though so it's all like water of the proverbial
RG> ducks back.

RG> Occasionally if we do see a deliberate/definite 'attack' we
RG> have no qualms about adding the originating host to the
RG> firewall rules to prevent them even accessing out legitimate
RG> servers (Web, Email, FTP, etc). 

 Aha! That sounds more like Fido rules...

RG> Secondly, Telstra, etc ultimately base their prices on how much
RG> data is sent across their network, more spam = more traffic =
RG> more profit.

> There is an obvious flaw in your reasoning. If that's true, why
> not generate it's own spam?

RG> Because people tend to get *really* upset at having to pay for
RG> spam that originates from their own ISP. Having it come from a
RG> different ISP is part and parcel of having unfiltered email.

 Are you telling me that I've invested my money in a Telco so stupid
that it can't set up a dummy ISP for it's spam?

 On an average, I get two telephone spam calls every day. I got one
from the Commonwealth Government yesterday... running a survey or
something; I always tell them to take a flying red fuck, and that
makes me feel better but it doesn't stop them. I get a mailbox
full fo junk mail twice a week. I got a chain letter yesterday,
promising $78,000 in 60-days, but they told me to get the
namesfrom the telephone book - don't use a mailing list. I assume
*they* used a mailing list! This is *telephone* spam at an average
$0.40 a call-centre call, and mail at $0.50. Internet spam is almost
free (to the sender) and you're trying to tell me that Telstra makes
money on it?

 Rubbish! The true reason telcos don't filter other people's spam is
that they worry someone might filter theirs.

 What I am saying, is that it's time we drew a line under unsolicited
communications, and by law, forced the ISPs to block it.

RG> We've actually managed to get some of our bills reduced when
RG> we've given evidence (log files)that the abuse we suffered
RG> actually originated within the Telstra network. Fortunately
RG> though, in my experience we get very little abuse originating
RG> from within the Telstra network. 99.8% of it originates outside
RG> Australia.

 Of course. Telstra = 5 million users, the World = 500 million
users, but *all* of those 500 million get to Australia by satellite or
cable. How hard would it be for OTC (or whatever they are called now)
to intercept and simply *bounce* the spam back to source (or dump it).
The Internet consists of a million ISPs, but in reality all of that
traffic is being handled by a hundred satellites run by a handful
of governments and telcos.

RG> Example, Gaelyne subsribes to the Breast Cancer mailing list,
RG> and needless to say, there is a LOT of discussion about
RG> 'breasts' in this list, and unless exceptions are placed,
RG> almost EVERY spam detection software availaible will flag 90%
RG> of the messages posted to this list as 'spam'. Sure it is a
RG> relatively trivial task to place an exception to this list, but
RG> what of the other lists (possibly ones not even created yet?).
RG> I can't foresee ANY filter software being made smart enough to
RG> cater for these exceptions to the normal rules, and having
RG> these filters in place as it 'arrives off the satelite' WILL
RG> cause us to miss legitimate email. 

 I'm not proposing to ban mailing lists, and I'm not suggesting a
software solution. I like the idea of L-A-W law, where a mailer
sending spam is disconnected and fined, and so is the ISP sending it
on.

 Under Fido rules, we marked our areas of interest, and that was
*all* we got. The sysop maintained his lists and that was the end of
it. We need that Anarchic system to apply to the Net as well... where
(backed up by Law) a user breaching the rules is simply disconnected.
It can't be done by an ISPnow, because being a commercial operation,
the ISP can be sued by the spammer for damages. We need spam to be
treated as kiddie-porn is treated - made illegal, with penalties for
the *ISP* if he transmits it.

RG> Fido is a small rigidedly controlled network compared to the
RG> Internet. I can't see any spammer making $1000's for spamming
RG> fido, but this is oh so common (and easy to do) with the
RG> decentralised nature of the internet. 

 You get the picture (finally). The problem is the Net's
decentralised nature. If fact, it is *not* decentralised even on
mainland Europe or USA. Most of the traffic is by satellite or on a
few cables. The only decentralised part is the proliferation of ISPs,
and the ISP's lack of responsibility for content.

 If we permit direct access to satellite (and I don't have a problem
with that), then the satellite is where the control has to be in
place... plus, the ISP has to be made responsible for the material he
passes on, by Law.

> BTW, I bought a new printer, and have since discovered that as
> well as Winmodems there are Winprinters! Bloody thing. Now I'll
> have to wait until someone writes a GDI driver for Linux.

RG> Not having a WinModem or a winPrinter I really can't make any
RG> comment about GDI drivers. I'm pretty sure that I've read they
RG> do exist though.

 I've got internal winmodem with a Linux driver, and I've had a
*hell* of a job setting it up. To make the wonmodem work, they've had
to put some of the driver inside the kernel, as well as load a driver
module... so for practice I compiled my own kernel (added APM jsut
for the hell of it), and it worked like a charm.

 Thus encouraged, I loaded the ham kernel (god only knows why someone
would call a winmodem driver "ham"... but I suspect it's a real pig)
and it didn't work. So I did it again taking note of the messages and
there weren't any... just "done" at the end. But it didn't work.

 I stuffed around for a few days, I actually read the ham-boot script
(it's written in C!) trying to work out what-the-fuck, and it looked
like there weren't any modules where they were supposed to be (with
the other modules). I was tempted to just copy them across, but I
decided to give it one last try, in pure Linux (no X11, no nothing
just Linux)... and if worked like a fucking charm!

 I thought that a Linux window was just like running a new Linux
machine, but X11 and the KDE GUI must *do* something. I had been
switching out of KDE (and X11) to a text window. I recompiled the
entire kernel there, no problems (it takes ten minutes... how big
*is* the bloody thing!?), but the script for adding the ham drivers
won't work that way.

 Anyway, it seems to work. Minicom looks a hell of a lot like Telix,
and I'm tempted to see if I can use Linux on the BBS, even though
it will be a monumental waste of time.

> Dual boot. I boot Win31, Win98 and Linux. Linux saves to a
> Windows drive quite happily.

RG> Bugger, I was hoping you'd found a way to do it without
RG> resorting to dual boot.

 WINE is a dead loss, but DOSEMU for DOS programs works really well.
I can even do interrupt calls in dosemu.

 My little network is up and running now, with Samba feeding the
Win98 machine. I had planned to run the Linux machine as a complete
PC, and eventually phase Windows out. I actually *like* Windows.
Apart from the Net and Microsoft's open-door policy, my only problem
with Windows is the way Gates upgrades every three years, and makes
the older versions totally obsolete in about six years.

 I'm still happy to use Win311, but there is no video driver for my
new card. With Linux, nothing ever goes out of date and if it does,
you download the latest version free. There is only ever one Linux...
but Linux sucks. The actual programs *almost* work, but it's that last
5% that gives me the shits.

 Like loading the ham driver... it worked, but I had to stuff around
for two days finding the bug.

RG> I'd REALLY like to know how I can run this s/ware on a linux
RG> box so I can say goodby to windoze forever.

> It's cruel, isn't it?

RG> Very cruel, especially when they first came out with this
RG> software they stated that they WOULD be creating a version
RG> suitable for *nix users. Six months after making this claim
RG> they did a backstep and said they'd dropped the idea. :-(

 I've come to the conclusion that we have to run *both* Linux and
Windows. Gates is not going to go away, and as long as M$ is dominant
then programmers will be forced to write new programs in windows. If
it's any consolation, I notice that Borland (Inprise) is now making
their code more portable.

 Of course, it's in Gates' best interest *not* to be portable. 

Regards,
Bob


--- BQWK Alpha 0.5