[9fans] 9 Atom - installation troubles

[9fans] 9 Atom - installation troubles

[lucio]

I started on my ESX (3.5i) platform, but that installs on SCSI drives
and I assume that these are not configured in the 9atom.iso image.

So I tried my fanciest (newest) motherboard, with a 500GB SATA drive.
Booting off USB suggests that the SATA controller is not recognised.
The PCI identification is Vendor 8086 (Intel) and, for some reason I
can't fathom, device 27df or 27c0.  The latter seems to require IRQ 11
(I only glimpsed at that).

Well, I guess I'll need to give 9 Atom a miss for a while.  I could
rebuild it on the Bell Labs distributions I use, but I suspect that
there will be many gotchas to overcome.

I'll gladly provide more info, if anyone feels like giving me some
advice.

Lucio.

PS: It's tempting to shove the drive in an external USB cabinet and
configure it that way.  Anyone who may know better must shout now :-)


-------------------------------------------------------------------------------------
This email has been scanned by the MxScan Email Security System.
-------------------------------------------------------------------------------------

Re: [9fans] 9 Atom - installation troubles

[erik quanstrom]

> So I tried my fanciest (newest) motherboard, with a 500GB SATA drive.
> Booting off USB suggests that the SATA controller is not recognised.
> The PCI identification is Vendor 8086 (Intel) and, for some reason I
> can't fathom, device 27df or 27c0.  The latter seems to require IRQ 11
> (I only glimpsed at that).

can you send me a photo.  if you can send me a pci listing as well.  if there
are any adralloc() failures on screen, it would be helpful if you used acid
to map that to a line of code as confirmation of no version skew.

i'll be happy to take a look.

- erik

Re: [9fans] 9 Atom - installation troubles

[lucio]

> i'll be happy to take a look.

Here's a serial console output while booting.

	nix
	pat: 0107040600070406
	pmstart 0x0000000000600000 pmend 0x0000000120000000
	mmuinit: vmstart 0xfffffffff0000000 vmunused 0xfffffffff0400000 vmunmapped 0xfffffffff0600000 vmend 0xfffffffffe000000
	pa 0x0000000000600000 (0xfffffffffe000000 0xfffffffff0600000)
	adr: mem 0x0000000000000000 0x000000000009fc00 size 654336
	adr: mem 0x000000000e000000 0x00000000ddd80000 size 3487039488
	adr: mem 0x0000000100000000 0x0000000120000000 size 536870912
	adr: 12/543 21/1918
	base 0xfffffffff0408000 ptr 0xfffffffff0408000 nunits 14415872; 219 mb
	acpiinit
	aml: readmem: va 0xfffffe00ddd8e064 len 1 d
	aml: readmem: va 0xfffffe00ddd8e064 len 1 d
	aml: readmem: va 0xfffffe00ddd8e064 len 1 d
	acpi parse: APIC FACP HPET
	acpi load: XSDT FACP DSDT APIC MCFG OEMB HPET GSCI SSDT
	physalloc color=0 base=0x0000000000000000 size=0x120000000
	lapic0: hz 199999970 max 199999 min 1999 div 11
	#A0: hda: mem fffffe00feaf8000 irq 3
	#A0: hda: codec #0, vendor 10ec0887, rev 00100302
	usb/disk... usb/kb... root is from (il, local)[local!#S/sdu0/fs -B 10000]:
	connect...kfs...bad nvram key
	bad authentication id
	bad authentication domain

	version...time...

	init: starting /bin/rc
	ipconfig: /net/ipifc/clone: bind ether /net/ether0: '/net/ether0' cs: file does not exist
	Chan: afmt 0011
	mouseport is (ps2, ps2intellimouse, usb, 0, 1, 2)[usb]:

I can't get the boot operation to go any further than the IL prompt if
I plug in a USB drive, either, so for all intents an purposes the
motherboard is incompatible with 9 Atom.  Its ethernet adapter is also
unrecognised.

I can arrange for the pci -v output, though:

	0.0.0:	brg  06.00.00 8086/2e30   0
		Intel Corporation 	4 Series Chipset DRAM Controller
	0.2.0:	vid  03.00.00 8086/2e32   5 0:fe400004 4194304 2:e000000c 268435456 4:0000dc01 16
		Intel Corporation 	4 Series Chipset Integrated Graphics Controller
	0.27.0:	aud  04.03.00 8086/27d8   3 0:feaf8004 16384
		Intel Corporation 	NM10/ICH7 Family High Definition Audio Controller
	0.28.0:	brg  06.04.00 8086/27d0   5
		Intel Corporation 	NM10/ICH7 Family PCI Express Port 1
	0.28.1:	brg  06.04.00 8086/27d2  10
		Intel Corporation 	NM10/ICH7 Family PCI Express Port 2
	0.29.0:	usb  0c.03.00 8086/27c8   7 4:0000d401 32
		Intel Corporation 	NM10/ICH7 Family USB UHCI Controller #1
	0.29.1:	usb  0c.03.00 8086/27c9  10 4:0000d481 32
		Intel Corporation 	NM10/ICH7 Family USB UHCI Controller #2
	0.29.2:	usb  0c.03.00 8086/27ca   6 4:0000d801 32
		Intel Corporation 	NM10/ICH7 Family USB UHCI Controller #3
	0.29.3:	usb  0c.03.00 8086/27cb   5 4:0000d881 32
		Intel Corporation 	NM10/ICH7 Family USB UHCI Controller #4
	0.29.7:	usb  0c.03.20 8086/27cc   7 0:feaf7c00 1024
		Intel Corporation 	NM10/ICH7 Family USB2 EHCI Controller
	0.30.0:	brg  06.04.01 8086/244e 255
		Intel Corporation 	82801 PCI Bridge
	0.31.0:	brg  06.01.00 8086/27b8   0
		Intel Corporation 	82801GB/GR (ICH7 Family) LPC Interface Bridge
	0.31.2:	disk 01.01.8a 8086/27c0   0 0:00000001 16 1:00000001 16 2:00000001 16 3:00000001 16 4:0000ffa1 16
		Intel Corporation 	NM10/ICH7 Family SATA Controller [IDE mode]
	1.0.0:	net  02.00.00 1969/1083  10 0:febc0004 262144 2:0000ec01 128
		Qualcomm Atheros 	AR8151 v2.0 Gigabit Ethernet

In passing, this warning is also worrying:

	rtc clock changes underfoot 1417801644 1417801643


Lastly, any chance you can add SCSI (buslogic or LSI Logic) to a boot kernel?

Lucio.


-------------------------------------------------------------------------------------
This email has been scanned by the MxScan Email Security System.
-------------------------------------------------------------------------------------

Re: [9fans] 9 Atom - installation troubles

[erik quanstrom]

> 	usb/disk... usb/kb... root is from (il, local)[local!#S/sdu0/fs -B 10000]:
> 	connect...kfs...bad nvram key
> 	bad authentication id
> 	bad authentication domain
>
> 	version...time...
>
> 	init: starting /bin/rc
> 	ipconfig: /net/ipifc/clone: bind ether /net/ether0: '/net/ether0' cs: file does not exist
> 	Chan: afmt 0011
> 	mouseport is (ps2, ps2intellimouse, usb, 0, 1, 2)[usb]:
>
> I can't get the boot operation to go any further than the IL prompt if
> I plug in a USB drive, either, so for all intents an purposes the
> motherboard is incompatible with 9 Atom.  Its ethernet adapter is also
> unrecognised.

it looks like this machine is booting pretty well.  it found the usb key, and is
loading it.

while i do have a usb-based install, install to usb disk support hasn't been added.
your sata is supported, but you'll need to switch to ahci mode.

> 	0.31.2:	disk 01.01.8a 8086/27c0   0 0:00000001 16 1:00000001 16 2:00000001 16 3:00000001 16 4:0000ffa1 16
> 		Intel Corporation 	NM10/ICH7 Family SATA Controller [IDE mode]

and sadly, i don't have a driver for this chipset yet:

> 	1.0.0:	net  02.00.00 1969/1083  10 0:febc0004 262144 2:0000ec01 128
> 		Qualcomm Atheros 	AR8151 v2.0 Gigabit Ethernet

- erik

Re: [9fans] 9 Atom - installation troubles

[lucio]

Thank you for the AHCI advice, I'll report back.

>> 	1.0.0:	net  02.00.00 1969/1083  10 0:febc0004 262144 2:0000ec01 128
>> 		Qualcomm Atheros 	AR8151 v2.0 Gigabit Ethernet

I think this is my second encounter with this chip, maybe I ought to
look into it myself.  It's hard for me to get down to it, but maybe
it's about time.

Lucio.

Re: [9fans] 9 Atom - installation troubles

[lucio]

> your sata is supported, but you'll need to switch to ahci mode.
>
>> 	0.31.2:	disk 01.01.8a 8086/27c0   0 0:00000001 16 1:00000001 16 2:00000001 16 3:00000001 16 4:0000ffa1 16
>> 		Intel Corporation 	NM10/ICH7 Family SATA Controller [IDE mode]

The web seems to think that the South Bridge supporting the IDE chip
does not provide AHCI mode.

Sigh!

Lucio.

Re: [9fans] 9 Atom - installation troubles

[erik quanstrom]

On Sat Dec  6 21:13:34 PST 2014, lucio@proxima.alt.za wrote:
> > your sata is supported, but you'll need to switch to ahci mode.
> >
> >> 	0.31.2:	disk 01.01.8a 8086/27c0   0 0:00000001 16 1:00000001 16 2:00000001 16 3:00000001 16 4:0000ffa1 16
> >> 		Intel Corporation 	NM10/ICH7 Family SATA Controller [IDE mode]
>
> The web seems to think that the South Bridge supporting the IDE chip
> does not provide AHCI mode.

usually there is an option in bios to load the ahci firmware instead of the
ide emulation.  some vendors don't provide this option.

- erik

Re: [9fans] 9 Atom - installation troubles

[erik quanstrom]

On Sat Dec  6 21:13:34 PST 2014, lucio@proxima.alt.za wrote:
> > your sata is supported, but you'll need to switch to ahci mode.
> >
> >> 	0.31.2:	disk 01.01.8a 8086/27c0   0 0:00000001 16 1:00000001 16 2:00000001 16 3:00000001 16 4:0000ffa1 16
> >> 		Intel Corporation 	NM10/ICH7 Family SATA Controller [IDE mode]
>
> The web seems to think that the South Bridge supporting the IDE chip
> does not provide AHCI mode.

i have successfully reconfigured many of these myself.
most if not all come with ide mode as the pre-set default.
because, you know, everybody wants to run dos without
reconfiguring bios.

- erik

Re: [9fans] 9 Atom - installation troubles

[lucio]

> i'll be happy to take a look.

The motherboard is ASUS P5G4IT-MK LX3, which I neglected to mention.

Lucio.


-------------------------------------------------------------------------------------
This email has been scanned by the MxScan Email Security System.
-------------------------------------------------------------------------------------

Re: [9fans] 9 Atom - installation troubles

[lucio]

>> i'll be happy to take a look.
>
> The motherboard is ASUS P5G4IT-MK LX3, which I neglected to mention.
>
> Lucio.
>
>
> -------------------------------------------------------------------------------------
> This email has been scanned by the MxScan Email Security System.
> -------------------------------------------------------------------------------------

Can the mailing list administrator help me figure out where the tail
piece in this message is coming from?  If the mailing list is not
responsible (I don't see the same treatment applied to Lucho's mail,
so I assume it isn't) then I want to know who's intercepting mail that
should come directly from my desktop to the mailing list host.

Lucio.


-------------------------------------------------------------------------------------
This email has been scanned by the MxScan Email Security System.
-------------------------------------------------------------------------------------

Re: [9fans] 9 Atom - installation troubles

[Wes Kussmaul]

On 12/05/2014 01:15 PM, lucio@proxima.alt.za wrote:
>   I want to know who's intercepting mail that
> should come directly from my desktop to the mailing list host.
>
>
> -------------------------------------------------------------------------------------
> This email has been scanned by the MxScan Email Security System.
> -------------------------------------------------------------------------------------

I'll ask Snowden next time I see him.


-- 

Wes Kussmaul
The Authenticity Institute
738 Main Street
Waltham, MA 02451

office +1 781 790 1674
mobile +1 781 330 1881

“Try this fruit, and by the way if a bunch of people collectively calling themselves Arthur Andersen signs something it’s the same as if a person named Arthur Andersen signed it.”

	- The Serpent

Re: [9fans] 9 Atom - installation troubles

[Skip Tavakkolian]

[-- Attachment #1: Type: text/plain, Size: 1315 bytes --]

i was going to suggest to ask the local authorities.

on the plus side, with so much transparency foisted on us, maybe the more
suspicious humans will realize that, for the most part, the rest of us are
using our one ride on the carrousel of infinite-time to do good. but
perhaps like any other addiction, the more it is fed the more it wants.


On Fri, Dec 5, 2014 at 12:38 PM, Wes Kussmaul <wes@reliableid.com> wrote:

>
> On 12/05/2014 01:15 PM, lucio@proxima.alt.za wrote:
>
>>   I want to know who's intercepting mail that
>> should come directly from my desktop to the mailing list host.
>>
>>
>> ------------------------------------------------------------
>> -------------------------
>> This email has been scanned by the MxScan Email Security System.
>> ------------------------------------------------------------
>> -------------------------
>>
>
> I'll ask Snowden next time I see him.
>
>
> --
>
> Wes Kussmaul
> The Authenticity Institute
> 738 Main Street
> Waltham, MA 02451
>
> office +1 781 790 1674
> mobile +1 781 330 1881
>
> “Try this fruit, and by the way if a bunch of people collectively calling
> themselves Arthur Andersen signs something it’s the same as if a person
> named Arthur Andersen signed it.”
>
>         - The Serpent
>
>
>

[-- Attachment #2: Type: text/html, Size: 2114 bytes --]

Re: [9fans] 9 Atom - installation troubles

[lucio]

> i was going to suggest to ask the local authorities.

They are the ones I hold responsible.  It starts as an attempt at
minimising impact and it always transforms into an effort to maximise
profit.  The transformation is from communal good to personal greed.
Always.

For some reason, personal greed does not seem to know limits, just as
you suggest about addictions.

Lucio.


-------------------------------------------------------------------------------------
This email has been scanned by the MxScan Email Security System.
-------------------------------------------------------------------------------------

Re: [9fans] 9 Atom - installation troubles

[Bakul Shah]

On Fri, 05 Dec 2014 20:15:05 +0200 lucio@proxima.alt.za wrote:
>
> Can the mailing list administrator help me figure out where the tail
> piece in this message is coming from?  If the mailing list is not
> responsible (I don't see the same treatment applied to Lucho's mail,
> so I assume it isn't) then I want to know who's intercepting mail that
> should come directly from my desktop to the mailing list host.
>
> Lucio.
>
>
> -----------------------------------------------------------------------------
> --------
> This email has been scanned by the MxScan Email Security System.
> -----------------------------------------------------------------------------

Below I've copied headers from your email. Looks like there
are *two* virus scans! One by MailRoute.net and one by
turbo-smtp.net. The latter don't even add a proper Received:
line. mx record for 9fans.net points to mail.mailroute.net.
I can't be certain but looks like proxima.alt.za delegates
actual email delivery to turo-smtp.net.

It is what it is. Not worth spending your time learning the
details of email related RFCs and how they are used & misused.
Instead write a nice useful 9p filesystem!

But remember: Just don't make any jokes in email that can get
you in trouble : )

+---------------
|Return-Path: 9fans-bounces@9fans.net
|Delivery-Date: Fri Dec  5 10:16:45 2014
|Return-Path: <9fans-bounces@9fans.net>
|X-Original-To: bakul@bitblocks.com
|Delivered-To: bakul@bitblocks.com
|Received: from mail.9fans.net (mail.9fans.net [67.207.142.3])
|	by mail.bitblocks.com (Postfix) with ESMTP id 63FA4B827
|	for <bakul@bitblocks.com>; Fri,  5 Dec 2014 10:16:45 -0800 (PST)
|Received: from localhost ([127.0.0.1] helo=[67.207.142.3])
|	by mail.9fans.net with esmtp (Exim 4.71)
|	(envelope-from <9fans-bounces@9fans.net>)
|	id 1XwxTG-0005gV-9L; Fri, 05 Dec 2014 18:19:14 +0000
|Received: from mr008.ord01.mailroute.net ([199.89.2.11]
|	helo=in-mr008.ord01.mailroute.net)
|	by mail.9fans.net with esmtp (Exim 4.71)
|	(envelope-from <lucio@proxima.alt.za>) id 1XwxTE-0005gQ-NB
|	for 9fans@9fans.net; Fri, 05 Dec 2014 18:19:12 +0000
|Received: from localhost (localhost.localdomain [127.0.0.1])
|	by in-mr008.ord01.mailroute.net (Postfix) with ESMTP id
|	3jvMXS3VzBz1gxNC
|	for <9fans@9fans.net>; Fri,  5 Dec 2014 18:15:20 +0000 (UTC)
|X-Virus-Scanned: by MailRoute
|X-Spam-Flag: NO
|X-Spam-Score: 0.81
|X-Spam-Level:
|X-Spam-Status: No, score=0.81 tagged_above=-9999
|	tests=[KAM_ASCII_DIVIDERS=0.8, T_DKIM_INVALID=0.01] autolearn=disabled
|Authentication-Results: mr008.ord01.mailroute.net (mroute_mailscanner);
|	dkim=neutral reason="invalid (public key: not available)"
|	header.d=proxima.alt.za; domainkeys=neutral
|	reason="invalid (public key: not available)"
|	header.from=lucio@proxima.alt.za header.d=proxima.alt.za
|Received: from in-mr008.ord01.mailroute.net ([199.89.2.11])
|	by localhost (mr008.ord01.mailroute.net [127.0.0.1])
|	(mroute_mailscanner, port 10024)
|	with LMTP id fyvGiD4f1JHx for <9fans@9fans.net>;
|	Fri,  5 Dec 2014 18:15:13 +0000 (UTC)
|Received: from tbjjbihbhdbee.turbo-smtp.net (tbjjbihbhdbee.turbo-smtp.net
|	[199.187.173.144])
|	by in-mr008.ord01.mailroute.net (Postfix) with SMTP id 3jvMXK0Mbyz1gwc1
|	for <9fans@9fans.net>; Fri,  5 Dec 2014 18:15:12 +0000 (UTC)
|DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
|	d=proxima.alt.za; s=turbo-smtp; x=1418408113; h=DomainKey-Signature:
|	Received:Received:Message-ID:To:Subject:Date:From:In-Reply-To:
|	MIME-Version:Content-Type:Content-Transfer-Encoding; bh=mN5MCie3
|	MAlQ9gRDJ3j87nztZld5C8mytxLrCMbp3bA=; b=xJdtsVbay3H7NCs8//UxvuS4
|	T1XLNQ2W+gIGNZjGsJNUoSHRk52WmJ6IJsoly1g97crh6qZUesy3a9Icdhyc9tKQ
|	BUXoGQHzKJNF59/vc9oP/EzNunXXZ4IQxcN6LYuzkGWx3VeMRk4F/lvD8+XAyFX6
|	C32q2Fwc/F7ab1ve2BA=
|DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws;
|	s=turbo-smtp; d=proxima.alt.za;
|	h=Received:Received:X-TurboSMTP-Tracking:Message-ID:To:Subject:Date:From:In-Reply-To:MIME-Version:Content-Type:Content-Transfer-Encoding:X-Mxscan-Scan:X-Mxscan-Msgid:X-Mxscan-Country-Sequence:X-Mxscan-Antivirus:X-Mxscan-Antispam:X-Mxscan-Spamscore:X-Mxscan-Processingtime;
|	b=dxc56Xglbm4SMhiLxzfF7S+cVltrofkiQdd83VWYMvPy2iOs+lncSP9QNyZQGB
|	Icd+mDdB6Q0RT8A9ihbjrWNyh+ZwjRaQ+puuy/3Wj/2lY6a7ynTBb9MLuou0f4Yj
|	8lGSIjXDcNdLLwNVqk2oSq67P39iWtL8zhqqcd64fm2zI=;
|Received: (qmail 30488 invoked from network); 5 Dec 2014 18:15:11 -0000
|Received:
|X-TurboSMTP-Tracking: 1698407116
|Message-ID: <a055cd7e2a7ed4a6bc641ea19c271da6@proxima.alt.za>
|To: 9fans@9fans.net
|Date: Fri, 5 Dec 2014 20:15:05 +0200
|From: lucio@proxima.alt.za
|In-Reply-To: <7fbbfdc479d433167ff1edd2489ade9c@proxima.alt.za>
|MIME-Version: 1.0
|Content-Type: text/plain; charset="US-ASCII"
|Content-Transfer-Encoding: 7bit
|X-Mxscan-Scan: Scanned by MxScan 2.7.601.0 for HRSNETWORX
|X-Mxscan-Msgid: 3916F0E23A874FF4906389C3AAAF851F_
|X-Mxscan-Country-Sequence: Private Network->Destination
|X-Mxscan-Antivirus: ClamAV 0.98.4 [Clean]
|X-Mxscan-Antispam: CLAM_SANE [Pass], KEYWORD [Pass], RDNSBL [Pass],
|	URLBL [Pass], SPAMASSASSIN [-1.0 (ALL_TRUSTED)],
|	DCC_CHECK [Pass]
|X-Mxscan-Spamscore: -1
|X-Mxscan-Processingtime: 1.154 sec(s)
|Subject: Re: [9fans] 9 Atom - installation troubles
|X-BeenThere: 9fans@9fans.net
|X-Mailman-Version: 2.1.13
+---------------

Re: [9fans] 9 Atom - installation troubles

[lucio]

> It is what it is. Not worth spending your time learning the
> details of email related RFCs and how they are used & misused.
> Instead write a nice useful 9p filesystem!

I could not agree more.  But I've caught up with RFC 5322 recently: my
current project (I am getting paid a nominal amount to indulge in it)
involves using multiple SPAM reports to name and shame spammers
operating in contravention of South African legislation.  Go provides
an amnazing platform to develop in and I have developed my share of
tools to analyse email messages for this.  After my first serious
refactoring (I'm learning, but at my age it takes a bit longer) I'll
be happy to release the code.  Anybody who wants to look at it now can
do so, but will need to suffer the indignities of CVS without
anonymous access (just ask me for a login account - it's on a NetBSD
server).

I'm writing this in knee-jerk reaction to your suggestion, so let me
first thank you for taking the trouble.  I'd be very happy to
reconstruct upas (and Erik's nupas, which I have ignored against my
better judgement) as a Go application to do things more sensibly, but
I think I'd firsl like to poll the community for ideas.

If I discover anything interesting by inpecting the mkessage headers
(thanks, once again), I will report about it here.

Lucio.


-------------------------------------------------------------------------------------
This email has been scanned by the MxScan Email Security System.
-------------------------------------------------------------------------------------

Re: [9fans] 9 Atom - installation troubles

[erik quanstrom]

> I could not agree more.  But I've caught up with RFC 5322 recently: my
> current project (I am getting paid a nominal amount to indulge in it)
> involves using multiple SPAM reports to name and shame spammers
> operating in contravention of South African legislation.  Go provides

it's sort of the good fight, but life is short and spamming is too easy.
so i built a few tools to go from 50,000 spams/day down to 0.2.  all
the tools are published in 9atom.

- erik

Re: [9fans] 9 Atom - installation troubles

[lucio]

> it's sort of the good fight, but life is short and spamming is too easy.
> so i built a few tools to go from 50,000 spams/day down to 0.2.  all
> the tools are published in 9atom.

I have different priorities and somebody needs to pick up the garbage
sometimes.  I'm being paid (just-just enough to keep me from having to
migrate back to a big city) to build some weaponry so spamming is a
little less rewarding.  Plan 9 is an excellent platform to develop on
and using Go means I can deliver the result on any platform my
principals may pick.

That the service runs faster than expected on Plan 9 (I'm withholding
the fact that it would run even faster on more conventional platforms,
I can honestly claim that I have not checked this) is good publicity
for Plan 9 and motivates me a little more.  If I get inspired to use
some Plan 9 paradigms that will give Plan 9 even more exposure, all
the better, but my efforts are a bit restricted at the moment because
Go takes some shedding of bad habits to acquire new (hopefully less
"bas") habits.

The final outcome may drift quite far from the currently very humble
beginnings.  Nupas and everything it gave birth to are definitely on
my list of tools I need to explore in my quest.

Lucio.

PS: my personal opinion is that there are sufficiently many and
sufficiently pissed off people out there to give each spammer some
pause for thought.  The tooling I'm concentrating on, rather than trim
traffic down, targets making spammers aware of the negative public
opinion their efforts generate, each spammer's identity being publicly
revealed.  Existing legislation is being leveraged to threaten
spammers, but more importantly to shelter the naming and shaming from
legal challenge.  All this requires active response to spamming,
rather than sweeping under the carpet action.  Of course, many think
this is pie-in-the sky stuff, but I am looking at it as empowering the
victims.

Re: [9fans] 9 Atom - installation troubles

[erik quanstrom]

> I have different priorities and somebody needs to pick up the garbage
> sometimes.  I'm being paid (just-just enough to keep me from having to
> migrate back to a big city) to build some weaponry so spamming is a
> little less rewarding.  Plan 9 is an excellent platform to develop on
> and using Go means I can deliver the result on any platform my
> principals may pick.

sure.   don't know how to make spamming less rewarding than the email
never being delivered.  and spam levels are dropping >10% this year.
(http://www.symantec.com/security_response/publications/threatreport.jsp
by the way web-based attacks are up 20%.)

i approach it this way because startin' an alice's resturant anti-spam movement
with four part harmony and full orchestration just ain't going to cut it.  ya just
can't sing loud enough.

imho, to get at the problem, an automated, distributed set of shovels and
rakes and implements of destruction that's fairly resistant to counter attack
(and officer oppie) that targets the money trail is needed. and even then,
there's a dependency on a lack of corruption in banking, proving in court
that the spam was illegal, and the offer was a scam.  not to mention the host
government recognizing the issue, blind justice,  and the dump not being
closed on thanksgiving.

i don't think i can out-money, out-idea the folk already in the fight.
i think i'll just go off and have myself another thanksgiving dinner that
can't be beat....

- erik

Re: [9fans] 9 Atom - installation troubles

[lucio]

> Below I've copied headers from your email.

I have two Internet access providers, I'd like to compare paths
through each.  Mind you, the right idea is to bounce all my mail
through my Cape Town server, it is a hosted physical device that I
trust (mostly) to have unshaped access to the Internet.

But to put this matter to rest, could you, Bakul, arrange a bounce
address somewhere analogous to 9fans@9fans.net that I could target
with a handful of test messages?  Mail would come from
lucio@proxima.alt.za, the target mail address might preferably be
lucio.dere@gmail.com.

I'd really appreciate that, but don't feel this is urgent or even
important.

Lucio.


-------------------------------------------------------------------------------------
This email has been scanned by the MxScan Email Security System.
-------------------------------------------------------------------------------------

Re: [9fans] 9 Atom - installation troubles

[lucio]

> I can't be certain but looks like proxima.alt.za delegates
> actual email delivery to turo-smtp.net.

There's a transparent proxy just the other side of my long-distance
wi-fi link, I'm not sure why my ISP feels they have to pay a third
party to interfere with email, but I think there may be a national
intelligence issue involved: our government has mooted digital
communication interception regulations for a while, but I haven't
followed the details.  I know whom to ask, though.

In the meantime, I note that the transparent interception does not
apply to the "submission" TCP port, port 587, so I think I'll hack
smtp to use that instead.  Right now, I'm going to build a copy of
smtp with a modified mxdial.c, but I wonder what a consensus here
would be: an option to smpt that invokes a "submit" function that only
differs from mxdial() in the use of the service argument, or a generic
port number on smtp's command line with a more complex, but now common
to both options, mxdial()?

Or maybe, as I'm doing now, a distinct "submit" command instead of
"smtp"?

Lucio.


-------------------------------------------------------------------------------------
This email has been scanned by the MxScan Email Security System.
-------------------------------------------------------------------------------------

Re: [9fans] 9 Atom - installation troubles

[Bakul Shah]

On Sat, 06 Dec 2014 11:09:51 +0200 lucio@proxima.alt.za wrote:
> > I can't be certain but looks like proxima.alt.za delegates
> > actual email delivery to turo-smtp.net.
>
> There's a transparent proxy just the other side of my long-distance
> wi-fi link, I'm not sure why my ISP feels they have to pay a third
> party to interfere with email, but I think there may be a national
> intelligence issue involved: our government has mooted digital
> communication interception regulations for a while, but I haven't
> followed the details.  I know whom to ask, though.
>
> In the meantime, I note that the transparent interception does not
> apply to the "submission" TCP port, port 587, so I think I'll hack
> smtp to use that instead.  Right now, I'm going to build a copy of
> smtp with a modified mxdial.c, but I wonder what a consensus here
> would be: an option to smpt that invokes a "submit" function that only
> differs from mxdial() in the use of the service argument, or a generic
> port number on smtp's command line with a more complex, but now common
> to both options, mxdial()?

Seriously, why bother. There are probably other agencies doing
transparent snooping (and with a subject line with words like
"Atom" and "troubles", this thread has probably set off
alarm bells in all sorts of TLA agencies).  Just invoke your
"submit" function to them : )

Re: [9fans] 9 Atom - installation troubles

[Quintile]

by submission do you jest mean a different port?

did you try adding a port to the SMTP attribute in your /lib/nab/local

smtp=host!987

-Steve





On 6 Dec 2014, at 09:09, lucio@proxima.alt.za wrote:

>> I can't be certain but looks like proxima.alt.za delegates
>> actual email delivery to turo-smtp.net.
> 
> There's a transparent proxy just the other side of my long-distance
> wi-fi link, I'm not sure why my ISP feels they have to pay a third
> party to interfere with email, but I think there may be a national
> intelligence issue involved: our government has mooted digital
> communication interception regulations for a while, but I haven't
> followed the details.  I know whom to ask, though.
> 
> In the meantime, I note that the transparent interception does not
> apply to the "submission" TCP port, port 587, so I think I'll hack
> smtp to use that instead.  Right now, I'm going to build a copy of
> smtp with a modified mxdial.c, but I wonder what a consensus here
> would be: an option to smpt that invokes a "submit" function that only
> differs from mxdial() in the use of the service argument, or a generic
> port number on smtp's command line with a more complex, but now common
> to both options, mxdial()?
> 
> Or maybe, as I'm doing now, a distinct "submit" command instead of
> "smtp"?
> 
> Lucio.
> 
> 
> -------------------------------------------------------------------------------------
> This email has been scanned by the MxScan Email Security System.
> -------------------------------------------------------------------------------------

Re: [9fans] 9 Atom - installation troubles

[lucio]

> smtp=host!987

Too clever!  I'm not sure it would work, though, because in
/sys/src/upas/smtp/mxdial.c the invocation is explicitly:

	addr = netmkaddr(addr, 0, "smtp");

and changing the port by altering /lib/ndb/common:

	tcp=smtp port=25

would be more extreme than I would be comfortable with.

I already added

	tcp=submit port=587

to /lib/ndb/common, so I'm OK with changing mxdial.c:

	addr = netmkaddr(addr, 0, "submit");

and use a recompiled smtp as "submit".  I will insert that in
/mail/lib/remotemail:

	exec /bin/upas/submit -as -u lucio@proxima.alt.za -h $fd $addr $sender $*

and rely on my /mail/lib/rewrite to point the above to the NetBSD MXer
in Cape Town.  Not a solution for everybody, but I'm in the right
situation to use it.

Lucio.

PS: I noted that netmkaddr isn't mentioned in the man page where it
appears only as its invocation.  So I may be making the wrong
assumptions.  I haven't checked the source.


-------------------------------------------------------------------------------------
This email has been scanned by the MxScan Email Security System.
-------------------------------------------------------------------------------------

Re: [9fans] 9 Atom - installation troubles

[erik quanstrom]

On Sat Dec  6 05:38:15 PST 2014, lucio@proxima.alt.za wrote:
> > smtp=host!987
>
> Too clever!  I'm not sure it would work, though, because in
> /sys/src/upas/smtp/mxdial.c the invocation is explicitly:
>
> 	addr = netmkaddr(addr, 0, "smtp");
>

this works, and it's common practice.  if it did not work,
telnet wouldn't work on anything other than port 23.

the reading of the 3d argument is "default port", not
we're forcing this port come hell or high water.

- erik

Re: [9fans] 9 Atom - installation troubles

[Kurt H Maier]

Quoting erik quanstrom <quanstro@quanstro.net>:

> On Sat Dec  6 05:38:15 PST 2014, lucio@proxima.alt.za wrote:
>> > smtp=host!987
>>
>> Too clever!  I'm not sure it would work, though, because in
>> /sys/src/upas/smtp/mxdial.c the invocation is explicitly:
>>
>> 	addr = netmkaddr(addr, 0, "smtp");
>>
>
> this works, and it's common practice.  if it did not work,
> telnet wouldn't work on anything other than port 23.
>
> the reading of the 3d argument is "default port", not
> we're forcing this port come hell or high water.
>
> - erik

This is detailed in the manpage for dial(2), which I know
because I went through the exact process Lucio did but
concluded there was no way someone would make things that
hard for themselves... so I checked the documentation.

Never hurts!

khm

Re: [9fans] 9 Atom - installation troubles

[lucio]

> This is detailed in the manpage for dial(2), which I know
> because I went through the exact process Lucio did but
> concluded there was no way someone would make things that
> hard for themselves... so I checked the documentation.

I could not find a _description{ for netmkaddr in the dial(2) man
page, but I wasn't looking very hard because I already had an
alternative solution in mind.  The alternative solution didn't work
and Steve's suggestion, applied where it made better sense to me,
worked a treat.  So now I know.

If I come up with a description for netmkaddr(2) worth adding to the
dial(2) man page, I'll submit a patch.

Lucio.

Re: [9fans] 9 Atom - installation troubles

[lucio]

> smtp=host!987

Just a test...  Forgive me.

Lucio.


-------------------------------------------------------------------------------------
This email has been scanned by the MxScan Email Security System.
-------------------------------------------------------------------------------------

Re: [9fans] 9 Atom - installation troubles

[lucio]

> Just a test...  Forgive me.

It failed.  This response is likely to fail too.  :-(

Lucio.


-------------------------------------------------------------------------------------
This email has been scanned by the MxScan Email Security System.
-------------------------------------------------------------------------------------

Re: [9fans] 9 Atom - installation troubles

[lucio]

> did you try adding a port to the SMTP attribute in your /lib/nab/local

Again, very clever!

I actually changed the rewrite rule for the gateway to Cape Town: in
/mail/lib/rewrite I had

	([^!]*)!(.*) 		| 		"/mail/lib/qmail '\s' 'net!$smtp'" "'\2@\1'"

which now reads:

	([^!]*)!(.*) 		| 		"/mail/lib/qmail '\s' 'net!$smtp!submit'" "'\2@\1'"

That's all it takes (if you have

	tcp=submit port=587

somewhere in /lib/ndb/* - if not 'net!$smtp!587' will do).

Not everyone's solution, but rather painless.

Thanks to everybody for your patience.  I'm sure some of you would
have liked to scream at me.

Lucio.