The ultimate board!
gun...@aurora.regenstrief.org
you remember I had trouble with the DiskOnChip media/driver. I now
have a board that uses compact flash and it works beautifully. I
want you all to know about this board, because it is really
inexpensive in every respect. Is a i468/132 MHz with 32 MB DRAM
on board. It runs on a cheap 12V / 9W wallmount adapter, has no fans,
doesn't even need a heat sink. It comes with 3 x 100 BT ethernet ports
ideal for router/firewall settop boxes. Here for you the kernel probes:
Copyright (c) 1992-2000 The FreeBSD Project.
Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991, 1992, 1993, 1994
The Regents of the University of California. All rights reserved.
FreeBSD 4.2-RELEASE #1: Fri Apr 13 14:01:20 EST 2001
sch...@aurora.regenstrief.org:/usr/NGI/KAME/kame/freebsd4/sys/compile/NGIGW
3
Timecounter "i8254" frequency 1193182 Hz
CPU: AMD Am5x86 Write-Back (486-class CPU)
Origin = "AuthenticAMD" Id = 0x4f4 Stepping = 4
Features=0x1<FPU>
real memory = 33554432 (32768K bytes)
avail memory = 28143616 (27484K bytes)
pnpbios: Bad PnP BIOS data checksum
Preloaded elf kernel "kernel" at 0xc04b0000.
Preloaded mfs_root "/fs.root" at 0xc04b009c.
md0: Preloaded image </fs.root> 2097152 bytes at 0xc02aebec
md1: Malloc disk
npx0: <math processor> on motherboard
npx0: INT 16 interface
pcib0: <Host to PCI bridge> on motherboard
pci0: <PCI bus> on pcib0
sis0: <NatSemi DP83815 10/100BaseTX> port 0xe000-0xe0ff mem 0xa0000000-0xa0000ff
f irq 10 at device 18.0 on pci0
sis0: Ethernet address: 00:00:24:c0:00:18
miibus0: <MII bus> on sis0
ukphy0: <Generic IEEE 802.3u media interface> on miibus0
ukphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto
sis1: <NatSemi DP83815 10/100BaseTX> port 0xe100-0xe1ff mem 0xa0001000-0xa0001ff
f irq 11 at device 19.0 on pci0
sis1: Ethernet address: 00:00:24:c0:00:19
miibus1: <MII bus> on sis1
ukphy1: <Generic IEEE 802.3u media interface> on miibus1
ukphy1: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto
sis2: <NatSemi DP83815 10/100BaseTX> port 0xe200-0xe2ff mem 0xa0002000-0xa0002ff
f irq 5 at device 20.0 on pci0
sis2: Ethernet address: 00:00:24:c0:00:1a
miibus2: <MII bus> on sis2
ukphy2: <Generic IEEE 802.3u media interface> on miibus2
ukphy2: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto
isa0: <ISA bus> on motherboard
ata0 at port 0x1f0-0x1f7,0x3f6 irq 14 on isa0
sio0 at port 0x3f8-0x3ff irq 4 flags 0x10 on isa0
sio0: type 16550A, console
IPsec: Initialized Security Association Processing.
IP Filter: v3.4.8 initialized. Default = pass all, Logging = enabled
ad0: 7MB <SunDisk SDCFB-8> [245/2/32] at ata0-master BIOSPIO
The engineer who designed the board is "one of us". His name is
Soeren Kristensen (www.soekris.com). This board can be had for
about $200 including enclosure!!!
The only problem with this is that it isn't available unless we
all place one big order to get it manufactured. Soeren knows how
to do this he just doesn't have the venture capital to do it.
The board can be ready for all of us if we find a way to finance
it collectively.
Anyone interested? Let me and the list know. We need just a
few hundred boards ordered, I believe. But Soeren can tell you
more about what exactly would be needed.
Go for it!
-Gunther
DISCLAIMER: I am in *no* way related to Soeren and I do not share
any profits from his sales. I am only chearing this board because
I want 30 pieces of it right now and I can't get it because he
can't manufacture it. After one year of serching I know that there
just isn't any better alternative in this world.
--
Gunther Schadow, M.D., Ph.D. gsch...@regenstrief.org
Medical Information Scientist Regenstrief Institute for Health Care
Adjunct Assistent Professor Indiana University School of Medicine
tel:1(317)630-7960 http://aurora.regenstrief.org
To Unsubscribe: send mail to majo...@FreeBSD.org
with "unsubscribe freebsd-small" in the body of the message
a...@marabu.ch
Gunther Schadow <gun...@aurora.regenstrief.org> recently let us
all know about a really attractive small board:
inexpensive in every respect. Is a i468/132 MHz with 32 MB DRAM
on board. It runs on a cheap 12V / 9W wallmount adapter, has no fans,
doesn't even need a heat sink. It comes with 3 x 100 BT ethernet ports
The engineer who designed the board is "one of us". His name is
Soeren Kristensen (www.soekris.com). This board can be had for
about $200 including enclosure!!!
I'm interested. I'd be even more interested in a Pentium-grade
small board with three 100Mbps NICs and more RAM than just 32MB.
I'm using the www.advantech.com/products/PCM-5823.asp with 64MB
RAM and 32MB Flash but - as you all know - they're expensive.
___________________________________________________________________________
Dr. Adrian Steinmann Webgroup Consulting AG Apollostrasse 21 8032 Zurich
Tel +41 1 380 30 80 Fax +41 1 380 30 85 Mailto:a...@webgroup.ch
cdi...@wolves.k12.mo.us
> you remember I had trouble with the DiskOnChip media/driver. I now
> have a board that uses compact flash and it works beautifully. I
> want you all to know about this board, because it is really
> inexpensive in every respect. Is a i468/132 MHz with 32 MB DRAM on
> board. It runs on a cheap 12V / 9W wallmount adapter, has no fans,
> doesn't even need a heat sink. It comes with 3 x 100 BT ethernet
> ports ideal for router/firewall settop boxes. Here for you the
> kernel probes:
[...snip...]
> sis0: <NatSemi DP83815 10/100BaseTX> port 0xe000-0xe0ff mem 0xa0000000-0xa0000ff
Ewww... Is this a good enough Ethernet chipset? I've not had good
luck with any of SiS's stuff. They seem to make, for lack of a nicer
word, crap.
> The only problem with this is that it isn't available unless we
> all place one big order to get it manufactured. Soeren knows how
> to do this he just doesn't have the venture capital to do it. The
> board can be ready for all of us if we find a way to finance it
> collectively.
I would be interested in one if the overall hardware itself is decent
enough, though I realize better hardware means more money. If someone
runs it through some obstacle courses and the "junk" works without any
known problems, I'm for it. I'm not in the position to buy more than
one. I have friends that might be interested as well, though.
-- Chris Dillon - cdi...@wolves.k12.mo.us - cdi...@inter-linc.net
FreeBSD: The fastest and most stable server OS on the planet.
For IA32 and Alpha architectures. IA64, PPC, and ARM under development.
http://www.freebsd.org
D...@silver-lynx.com
I talked quite a bit with him. He seems quite knowledgable, and I don't
doubt his board will work. I will commit to at least ten units to start,
as long as we can structure this that the delivery happens no later than
60 days after the cash goes out.
--
Don Wilde D...@Silver-Lynx.com
Silver Lynx Embedded Microsystems Architects
2218 Southern Bl. Ste. 12 Rio Rancho, NM 87124
505-891-4175 FAX 891-4185
sha...@earthlink.net
to get rid of the fans and heat ! 60 days after the cash goes out
works for me as well.
Skip
ab...@webgiro.com
Well, who's counting it then? How do I sign myself up for a couple?
Andrzej
// ----------------------------------------------------------------
// Andrzej Bialecki <ab...@webgiro.com>, Chief System Architect
// WebGiro AB, Sweden (http://www.webgiro.com)
// ----------------------------------------------------------------
// <ab...@freebsd.org> FreeBSD developer (http://www.freebsd.org)
so...@soekris.com
I'm the hardware designer, so let me comments on your comments :-)
Chris Dillon wrote:
>
> [...snip...]
> > sis0: <NatSemi DP83815 10/100BaseTX> port 0xe000-0xe0ff mem 0xa0000000-0xa0000ff
>
> Ewww... Is this a good enough Ethernet chipset? I've not had good
> luck with any of SiS's stuff. They seem to make, for lack of a nicer
> word, crap.
>
This is not a SIS chip, but a National Semiconductor chip. They
have made ethernet chips for the last 20 years. I don't trust SIS
either, but I belive that the DP83815 is natsemi quality. I also
don't know why it's so close to the SIS chip, but natsemi probably
bought the MAC core design from SIS....
The only limit on the DP83815 is the requirement for 32 bit
alligned RX buffers. Otherwise it has all the good features and is
cheap, $7 in 1K quantity.
I have 6 prototypes out now, and nobody seems to have had any
problems with the ethernet controllers.
>
> I would be interested in one if the overall hardware itself is decent
> enough, though I realize better hardware means more money. If someone
> runs it through some obstacle courses and the "junk" works without any
> known problems, I'm for it. I'm not in the position to buy more than
> one. I have friends that might be interested as well, though.
>
Regards,
Soren
gun...@aurora.regenstrief.org
> > sis0: <NatSemi DP83815 10/100BaseTX> port 0xe000-0xe0ff mem 0xa0000000-0xa0000ff
>
> Ewww... Is this a good enough Ethernet chipset? I've not had good
> luck with any of SiS's stuff. They seem to make, for lack of a nicer
> word, crap.
"E.B. Dreger" wrote:
> I'd imagine that 2x 21143TD would cost just a teeny bit more than 3x Sis.
> Personally, I would be more interested in a 2x Tulip model. SiS scares
> me, too...
Chris Dillon continues:
> I would be interested in one if the overall hardware itself is decent
> enough, though I realize better hardware means more money. If someone
> runs it through some obstacle courses and the "junk" works without any
> known problems, I'm for it. I'm not in the position to buy more than
> one. I have friends that might be interested as well, though.
Could you tell me what's wrong with the National Semiconductor DP83815
ethernet chips? I guess for a low cost board, I am happy not to have
been given realteks. The Flytech NetPC NC-2 that I use as an interim
comes with realtek. What's the problem with the sis driver? So far it
worked all nicely for me.
These boards come with net-booting by default. It was a breeze to set
up -- my first netbooting experience, and I hadn't had that much fun
for a while (it's wonderful not having to worry squeezing my
bootstrapping package onto a floppy to get a single board with flash
up and running, not to mention again how easy the Compact Flash was
compared with those DiskOnChip thingies seen elsewhere.)
I plan to use these for IPsec encrypting videoconferencing. My testing
so far shows that the board can sustain AT LEAST 2 Mb/s per each
direction (blowfish-cbc with 256 bit keys.) This 2 Mb/s number may not
be a true maximum because my testing involved a pretty crummy 10 Mb/s
hub :-(. I understand that the 486 class CPU is sort of a bottleneck
for encryption work though, but Soren wanted to build a Hi/Fn based
hardware crypto board too. Will be some fiddling with drivers though ...
Let me know what benchmark or other "obstacle course" you want to see
tested on it.
regards
-Gunther
--
Gunther Schadow, M.D., Ph.D. gsch...@regenstrief.org
Medical Information Scientist Regenstrief Institute for Health Care
Adjunct Assistent Professor Indiana University School of Medicine
tel:1(317)630-7960 http://aurora.regenstrief.org
To Unsubscribe: send mail to majo...@FreeBSD.org
LCo...@go2france.com
>hub :-(. I understand that the 486 class CPU is sort of a bottleneck
>for encryption work though, but Soren wanted to build a Hi/Fn based
>hardware crypto board too. Will be some fiddling with drivers though ...
Since FreeBSD has no hardware crypto support (vs OpenBSD) and none is
on the announced horizon afaik, I conclude that "fiddling with
drivers" is understating the difficulty of adding hardware crypto support.
Len
http://MenAndMice.com/DNS-training Austin,TX: 23,24/04; SFO,CA:
7,8/05
http://BIND8NT.MEIway.com : ISC BIND 8.2.3 "NT3" for NT4 & W2K
http://IMGate.MEIway.com : Build free, hi-perf, anti-abuse mail gateways
gun...@aurora.regenstrief.org
> >hub :-(. I understand that the 486 class CPU is sort of a bottleneck
> >for encryption work though, but Soren wanted to build a Hi/Fn based
> >hardware crypto board too. Will be some fiddling with drivers though ...
>
> Since FreeBSD has no hardware crypto support (vs OpenBSD) and none is
> on the announced horizon afaik, I conclude that "fiddling with
> drivers" is understating the difficulty of adding hardware crypto support.
Well, it can't be too difficult. Since I can account
the work on my dayjob hours, I am willing to help with
driver writing. Unfortunately with most crypto stuff,
documentation is very scarce (i.e., the Intel Pro/100 S
NIC has a crypto chip, but nobody except Intel and Microsoft
seems to know how to program it) ... if we can get good
documentation on how to program the Hi/Fn 7951, it
shouldn't be too hard ... actually I am more ready to
write a driver for the hardware then to write the
crypto algorithms in software :-).
Since crypto algorithms are in the FreeBSD kernel already,
offloading those to a chip shouldn't be that big of a
deal ... just use the same API as the existing software
algorithms and pipe stuff to the chip instead. With
good documentation and preparation this should be a
matter of a day or two to get it supported in FreeBSD
(and all KAME/IPsec hosts for that matter.)
Outside of the kernel, I would have a /dev node interface
as a character device ... I love character devices (and I
hate IOCTL's for that matter) because they can be used easily
in shell scripts. Probably a triple of devices organized as a
pipe. For example /dev/cri0 would be the crypto input (i)
end of the pipe, and /dev/cro0 would be the crypto output (o)
end of the pipe. Finally /dev/crc0 would be the crypto
control (c) channel. Into this channel one would write a
string that identifies the cipher followed by the key and
initialization vector and other parameters. Whenever the crypto
control channel is written, the crypto buffer is reset and
ready to accept data. Would be fun to do.
regards
-Gunther
--
Gunther Schadow, M.D., Ph.D. gsch...@regenstrief.org
Medical Information Scientist Regenstrief Institute for Health Care
Adjunct Assistent Professor Indiana University School of Medicine
tel:1(317)630-7960 http://aurora.regenstrief.org
To Unsubscribe: send mail to majo...@FreeBSD.org
eddy+pub...@noc.everquick.net
> From: Gunther Schadow <gun...@aurora.regenstrief.org>
>
> Could you tell me what's wrong with the National Semiconductor DP83815
> ethernet chips? I guess for a low cost board, I am happy not to have
> been given realteks. The Flytech NetPC NC-2 that I use as an interim
I just quickly scanned the left column, and saw the "sis0"... NS probably
does things a bit better. So far, anything made by SiS that I've used has
been a bit quirky. And don't even say that "R" word. :-)
My point was that I like Tulips... I'd expect 2x Tulip to cost a little
more than 3x SiS (NS now that I actually read dmesg output). If it works
and gets the job done, I'm happy.
Have you had a chance to test forwarding rate, in both Mbps and fps?
Eddy
---------------------------------------------------------------------------
Brotsman & Dreger, Inc.
EverQuick Internet / EternalCommerce Division
Phone: (316) 794-8922
---------------------------------------------------------------------------
ne...@nelsonnet.org
How much would shipping to Australia add to the cost?
> > > To Unsubscribe: send mail to majo...@FreeBSD.org
> > > with "unsubscribe freebsd-small" in the body of the message
> >
> >
> >
> > To Unsubscribe: send mail to majo...@FreeBSD.org
> > with "unsubscribe freebsd-small" in the body of the message
> >
>
>
> Andrzej
>
> // ----------------------------------------------------------------
> // Andrzej Bialecki <ab...@webgiro.com>, Chief System Architect
> // WebGiro AB, Sweden (http://www.webgiro.com)
> // ----------------------------------------------------------------
> // <ab...@freebsd.org> FreeBSD developer (http://www.freebsd.org)
>
>
gun...@aurora.regenstrief.org
> My point was that I like Tulips... I'd expect 2x Tulip to cost a little
> more than 3x SiS (NS now that I actually read dmesg output). If it works
> and gets the job done, I'm happy.
Well, but I *need* 3 ethernet ports, so, I prefer having three that
just work over 2 tulips that also work :-)
> Have you had a chance to test forwarding rate, in both Mbps and fps?
You mean just packet forwarding and a little routing without any
encryption? I can do that, although I might need four or more
computers to actually create traffic w/o being themselves
overloaded. I'll try to set something up. What is the rate that
I should expect? Is there a standardized benchmark that I should
use? (I have my own UDP streaming test tool that I use to measure
bandwidths.)
thanks,
-Gunther
--
Gunther Schadow, M.D., Ph.D. gsch...@regenstrief.org
Medical Information Scientist Regenstrief Institute for Health Care
Adjunct Assistent Professor Indiana University School of Medicine
tel:1(317)630-7960 http://aurora.regenstrief.org
To Unsubscribe: send mail to majo...@FreeBSD.org
bro...@one-eyed-alien.net
--uQr8t48UFsdbeI+V
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable
On Sun, Apr 15, 2001 at 06:12:35AM +0200, Len Conrad wrote:
>=20
> >hub :-(. I understand that the 486 class CPU is sort of a bottleneck
> >for encryption work though, but Soren wanted to build a Hi/Fn based
> >hardware crypto board too. Will be some fiddling with drivers though ...
>=20
> Since FreeBSD has no hardware crypto support (vs OpenBSD) and none is=20
> on the announced horizon afaik, I conclude that "fiddling with=20
> drivers" is understating the difficulty of adding hardware crypto support.
Mark Murray is currently working on integrating the OpenBSD kernel
crypto support required to suport hardware crypto. If only someone
built a cardbus crypto accelerator for my laptop. ;-)
-- Brooks
--=20
Any statement of the form "X is the one, true Y" is FALSE.
PGP fingerprint 655D 519C 26A7 82E7 2529 9BF0 5D8E 8BE9 F238 1AD4
--uQr8t48UFsdbeI+V
Content-Type: application/pgp-signature
Content-Disposition: inline
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.4 (GNU/Linux)
Comment: For info see http://www.gnupg.org
iD8DBQE62x4JXY6L6fI4GtQRAlPmAKCtajh7z+Qqw8zafwM0VgZa7qvkPACbBCfO
A/tvmKOeIZhGnGYVvG3I4a8=
=BfTk
-----END PGP SIGNATURE-----
--uQr8t48UFsdbeI+V--
cdi...@wolves.k12.mo.us
> Hi Chris,
>
> I'm the hardware designer, so let me comments on your comments :-)
>
> Chris Dillon wrote:
> >
> > [...snip...]
> > > sis0: <NatSemi DP83815 10/100BaseTX> port 0xe000-0xe0ff mem 0xa0000000-0xa0000ff
> >
> > Ewww... Is this a good enough Ethernet chipset? I've not had good
> > luck with any of SiS's stuff. They seem to make, for lack of a nicer
> > word, crap.
> >
>
> This is not a SIS chip, but a National Semiconductor chip. They
> have made ethernet chips for the last 20 years. I don't trust SIS
> either, but I belive that the DP83815 is natsemi quality. I also
> don't know why it's so close to the SIS chip, but natsemi probably
> bought the MAC core design from SIS....
The design of the core itself is what I'm wondering about. Anybody
can manufacture a quality chip these days, but it takes talent to
design one.
> The only limit on the DP83815 is the requirement for 32 bit
> alligned RX buffers. Otherwise it has all the good features and is
> cheap, $7 in 1K quantity.
Not bad. But what is the cost of some of the better chips such as the
DEC/Intel 21143 or even an Intel 8255x at that quantity?
-- Chris Dillon - cdi...@wolves.k12.mo.us - cdi...@inter-linc.net
FreeBSD: The fastest and most stable server OS on the planet.
For IA32 and Alpha architectures. IA64, PPC, and ARM under development.
http://www.freebsd.org
To Unsubscribe: send mail to majo...@FreeBSD.org
p...@mesanet.com
> On Sat, 14 Apr 2001, Soren Kristensen wrote:
>
> > Hi Chris,
> >
> > I'm the hardware designer, so let me comments on your comments :-)
> >
> > Chris Dillon wrote:
> > >
> > > [...snip...]
> > > > sis0: <NatSemi DP83815 10/100BaseTX> port 0xe000-0xe0ff mem 0xa0000000-0xa0000ff
> > >
> > > Ewww... Is this a good enough Ethernet chipset? I've not had good
> > > luck with any of SiS's stuff. They seem to make, for lack of a nicer
> > > word, crap.
> > >
> >
> > This is not a SIS chip, but a National Semiconductor chip. They
> > have made ethernet chips for the last 20 years. I don't trust SIS
> > either, but I belive that the DP83815 is natsemi quality. I also
> > don't know why it's so close to the SIS chip, but natsemi probably
> > bought the MAC core design from SIS....
>
> The design of the core itself is what I'm wondering about. Anybody
> can manufacture a quality chip these days, but it takes talent to
> design one.
>
> > The only limit on the DP83815 is the requirement for 32 bit
> > alligned RX buffers. Otherwise it has all the good features and is
> > cheap, $7 in 1K quantity.
>
> Not bad. But what is the cost of some of the better chips such as the
> DEC/Intel 21143 or even an Intel 8255x at that quantity?
About 3-4 times the 83815... (when you add the PHY to the 21143 or use the
overpriced Intel chip...)
Besides I dont think I would do a new design with a 21143, who knows when
Intel will drop it...
PCW
>
>
> -- Chris Dillon - cdi...@wolves.k12.mo.us - cdi...@inter-linc.net
> FreeBSD: The fastest and most stable server OS on the planet.
> For IA32 and Alpha architectures. IA64, PPC, and ARM under development.
> http://www.freebsd.org
>
>
>
> To Unsubscribe: send mail to majo...@FreeBSD.org
> with "unsubscribe freebsd-small" in the body of the message
>
Peter Wallace
Mesa Electronics
cdi...@wolves.k12.mo.us
> > Not bad. But what is the cost of some of the better chips such as the
> > DEC/Intel 21143 or even an Intel 8255x at that quantity?
>
> About 3-4 times the 83815... (when you add the PHY to the 21143 or
> use the overpriced Intel chip...)
Well.... the 83815 looks like the only choice, then. :-)
eddy+pub...@noc.everquick.net
> From: Chris Dillon <cdi...@wolves.k12.mo.us>
>
> > > Not bad. But what is the cost of some of the better chips such as
> > > the DEC/Intel 21143 or even an Intel 8255x at that quantity?
> >
> > About 3-4 times the 83815... (when you add the PHY to the 21143 or
> > use the overpriced Intel chip...)
>
> Well.... the 83815 looks like the only choice, then. :-)
Killing the 21143 would, IMHO, be stupid... we'd probably better plan on
it happening. ;-)
If the 83815 ends up being a good chip, and Intel does kill the Tulip,
then it would be nice to know of alternatives to 3Com. If the NS chip
works, I'm all for it... like Chris, I got a bit scared at the "sis"
listing in dmesg.
Doubts out of the way, I'm very eager to see what Soren makes, and hope
that people can get together enough volume. :-)
Eddy
---------------------------------------------------------------------------
Brotsman & Dreger, Inc.
EverQuick Internet / EternalCommerce Division
Phone: (316) 794-8922
---------------------------------------------------------------------------
so...@soekris.com
> > The only limit on the DP83815 is the requirement for 32 bit
> > alligned RX buffers. Otherwise it has all the good features and is
> > cheap, $7 in 1K quantity.
>
> Not bad. But what is the cost of some of the better chips such as the
> DEC/Intel 21143 or even an Intel 8255x at that quantity?
>
First, I consider the DP83815 a great chip until proven otherwise
:-)
The 21143 is not really an option anymore after Intel bought DEC's
chip business. It's an old chip without the newest feature, and it
need an external phy. And I wouldn't be surprised if it suddenly
got discontinued, like most of the other DEC chips....
Intel has the 82559ER as a low cost chip for embedded
applications. It's a nice chip and cost $12 in the same 1K volume,
but I'm not too happy about it's tiny 1mm pitch BGA package.
But I really don't want to use Intel part if I can avoid it, we
all know about their opinion about relasing information for open
source projects, and they don't really care about small outfits
like me.... I've very bad experience with them in the past.
Regards,
Soren
so...@soekris.com
Just a update....
First, thanks for all your interest in the net4501 board.
I do have some limited funds avaliable, and have decided to go
ahead and start the first production series this week. I will post
again when I know when I can deliver and what the prices will be.
And due to the economic downturn, parts avaliability have improved
the last couple of months :-) so expect 5-7 weeks from now.
But I would still be happy for prepaid orders, as that would
enable me to make more boards at a lower cost. To encourage that,
I will offer a discount for prepaid orders.
The configuration for the first series is planned to be 133 Mhz
CPU, 3 ethernet ports, 32 Mbyte SDRAM and 2 serial ports, and
complete with a small metal case. Let me know if there's interest
in other configurations. To see what's possible, take a look at
the website: http://www.soekris.com/net4501.htm
Also, please give me a hint if I'm off topic or generating too
much noise on the list....
Regards,
Soren Kristensen
Soekris Engineering
19185 Taylor Avenue
Morgan Hill, CA 95037-2714
Phone (408)782-2307, Fax (408)782-2317
bro...@one-eyed-alien.net
--WIyZ46R2i8wDzkSu
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable
On Tue, Apr 17, 2001 at 12:07:45AM -0700, Soren Kristensen wrote:
> The configuration for the first series is planned to be 133 Mhz
> CPU, 3 ethernet ports, 32 Mbyte SDRAM and 2 serial ports, and
> complete with a small metal case. Let me know if there's interest
> in other configurations. To see what's possible, take a look at
> the website: http://www.soekris.com/net4501.htm
Will the metal case include machanical support for a PCI card or will it
be the size of the board? I've got uses for each.
Overall this system looks awesome.
Thanks,
Brooks
--=20
Any statement of the form "X is the one, true Y" is FALSE.
PGP fingerprint 655D 519C 26A7 82E7 2529 9BF0 5D8E 8BE9 F238 1AD4
--WIyZ46R2i8wDzkSu
Content-Type: application/pgp-signature
Content-Disposition: inline
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.4 (GNU/Linux)
Comment: For info see http://www.gnupg.org
iD8DBQE63J5FXY6L6fI4GtQRAs/qAJ9I3QUe+qKIb+pX6Uj8grTo7bZDxgCfVgDq
z3iqsaFgleNhgmCPMyH6FOI=
=lZD8
-----END PGP SIGNATURE-----
--WIyZ46R2i8wDzkSu--
i...@harmony.village.org
: types of CF cards... ones that are actually CF, and ones that have an
: ATA-Compatible interface. Which kind do we actually need, or will
: either type work?
No. there is only one type of CF memory card. They all have the ATA
compatibility interface. There are CF cards for ethernet and modems
and such, but that shouldn't matter for tis.
: Does the board you will be making have both the MiniPCI and the
: regular PCI slot on them like the board on the website? I'm not sure
: what MiniPCI is supposed to look like, but the card edge connector
: hanging off the left side of the board in the picture on your website
: looks like it would be the regular 3.3V PCI slot, right? Is the area
: just below the CompactFlash slot (which looks somewhat like a pad for
: a PCMCIA slot) actually where the MiniPCI slot would go? If so, is
: MiniPCI similar to or even compatible with CardBus? Or is that for an
: actual PCMCIA/CardBus slot? :-)
minipci is PCI bus in a different form factor than the traditional
edge connector that we see in PCs. Cardbus is also a different form
factor, but there are other differences for it.
You could put a cardbus bridge card on the mini-pci adapter. I'm
hoping to do something like that to get a good bouncer box for
pccard/cardbus work.
Warner
so...@soekris.com
Chris Dillon wrote:
>
> Where do I send the check? If you take credit cards, I can do that
> too. :-)
Prices and ordering information comming soon....
>
> I'd take one or two just like you've mentioned (133MHz, 32MB,
> 3*Ethernet, metal case). I assume that this does not come with a
> CompactFlash card. A trip down to the local electronics store is all
> that is required to get a CF card these days. IIRC, there are two
> types of CF cards... ones that are actually CF, and ones that have an
> ATA-Compatible interface. Which kind do we actually need, or will
> either type work?
That right, they will come without the CompactFlash card.
As Warner said, there is only one kind of CF cards, but they can
operate in 3 modes, all required per the spec: PC Card memory
mode, PC Card I/O mode and True IDE mode. I'm running them in true
IDE mode. The CompactFlash spec is at http://www.compactflash.org/
>
> Does the board you will be making have both the MiniPCI and the
> regular PCI slot on them like the board on the website? I'm not sure
> what MiniPCI is supposed to look like, but the card edge connector
> hanging off the left side of the board in the picture on your website
> looks like it would be the regular 3.3V PCI slot, right? Is the area
> just below the CompactFlash slot (which looks somewhat like a pad for
> a PCMCIA slot) actually where the MiniPCI slot would go? If so, is
> MiniPCI similar to or even compatible with CardBus? Or is that for an
> actual PCMCIA/CardBus slot? :-)
The board will have both the Std 3.3V PCI and the MiniPCI type III
slot. The Area just below the CF socket is the MiniPCI type III
socket, which is a 124 pins DIMM connector. (almost the same as
laptop memory DIMM's)
> Also, how hard would it be to put 64MB on it? You might need to put
> 32MB on all of the units to keep manufacturing costs down, but if it
> isn't a big deal to put 64MB on some of them for people who need/want
> that much memory, I'll spring for that.
I'm planning to offer both 32 Mbyte and 64 Mbyte version, as there
seem to be interest for that. The 64 Mbyte will probably cost
around $25 more, as 256 Mbit chips is still relative more
expensive.
Regards,
Soren Kristensen
Soekris Engineering
19185 Taylor Avenue
Morgan Hill, CA 95037-2714
Phone (408)782-2307, Fax (408)782-2317
To Unsubscribe: send mail to majo...@FreeBSD.org
i...@harmony.village.org
: As Warner said, there is only one kind of CF cards, but they can
: operate in 3 modes, all required per the spec: PC Card memory
: mode, PC Card I/O mode and True IDE mode. I'm running them in true
: IDE mode. The CompactFlash spec is at http://www.compactflash.org/
I've personally used 7 different brands of CF card in true IDE mode
w/o hassles using a home grown adapter. SanDisk, Simple, Peripheral
Enhacement Corp, Viking, TDK and Hitachi. I've yet to find any that
won't work in True IDE.
I did get one bad batch of Simple parts, however. They just up and
died after a few days of use. All of them in that batch wound up
going bad within two weeks. The other 100 odd CF parts that we'd use
had no such problems.
FreeBSD uses PCCARD I/O mode when you plug them into a pccard slot.
Most of the parts support memory mapping as well, but memory mode only
seems to support mapping two pages in at a time. I keep hearing
rumors that there's a away to directly map up to 64M of these cards,
but can find no datasheets that give enlightenment.
Warner
cdi...@wolves.k12.mo.us
> Hi Chris,
> > Does the board you will be making have both the MiniPCI and the
> > regular PCI slot on them like the board on the website? I'm not sure
> > what MiniPCI is supposed to look like, but the card edge connector
> > hanging off the left side of the board in the picture on your website
> > looks like it would be the regular 3.3V PCI slot, right? Is the area
> > just below the CompactFlash slot (which looks somewhat like a pad for
> > a PCMCIA slot) actually where the MiniPCI slot would go? If so, is
> > MiniPCI similar to or even compatible with CardBus? Or is that for an
> > actual PCMCIA/CardBus slot? :-)
>
> The board will have both the Std 3.3V PCI and the MiniPCI type III
> slot. The Area just below the CF socket is the MiniPCI type III
> socket, which is a 124 pins DIMM connector. (almost the same as
> laptop memory DIMM's)
Cool! Are there many manufacturers who are making peripherals in the
MiniPCI form factor yet? One interest might be a MiniPCI 802.11b
card.
> > Also, how hard would it be to put 64MB on it? You might need to put
> > 32MB on all of the units to keep manufacturing costs down, but if it
> > isn't a big deal to put 64MB on some of them for people who need/want
> > that much memory, I'll spring for that.
>
> I'm planning to offer both 32 Mbyte and 64 Mbyte version, as there
> seem to be interest for that. The 64 Mbyte will probably cost
> around $25 more, as 256 Mbit chips is still relative more
> expensive.
I had figured about $30 more. $25 is even better. :-)
-- Chris Dillon - cdi...@wolves.k12.mo.us - cdi...@inter-linc.net
FreeBSD: The fastest and most stable server OS on the planet.
For IA32 and Alpha architectures. IA64, PPC, and ARM under development.
http://www.freebsd.org
To Unsubscribe: send mail to majo...@FreeBSD.org
bro...@one-eyed-alien.net
--EVF5PPMfhYS0aIcm
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable
On Tue, Apr 17, 2001 at 05:55:43PM -0500, Chris Dillon wrote:
> Cool! Are there many manufacturers who are making peripherals in the
> MiniPCI form factor yet? One interest might be a MiniPCI 802.11b
> card.
http://www.wavelan.com/template.html?section=3Dm57&page=3D2511&envelope=3D93
-- Brooks
--=20
Any statement of the form "X is the one, true Y" is FALSE.
PGP fingerprint 655D 519C 26A7 82E7 2529 9BF0 5D8E 8BE9 F238 1AD4
--EVF5PPMfhYS0aIcm
Content-Type: application/pgp-signature
Content-Disposition: inline
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.4 (GNU/Linux)
Comment: For info see http://www.gnupg.org
iD8DBQE63MrmXY6L6fI4GtQRAr6SAJ492bP0iz+w2eHnyQ9wKWheZnWjlQCeJz9J
0ItpHlnTDv6sG84gFQ1WYdI=
=4VWu
-----END PGP SIGNATURE-----
--EVF5PPMfhYS0aIcm--
i...@harmony.village.org
: http://www.wavelan.com/template.html?section=m57&page=2511&envelope=93
Price? The web page in question didn't have anything.
Warner
cdi...@wolves.k12.mo.us
> On Tue, Apr 17, 2001 at 05:55:43PM -0500, Chris Dillon wrote:
> > Cool! Are there many manufacturers who are making peripherals in the
> > MiniPCI form factor yet? One interest might be a MiniPCI 802.11b
> > card.
>
> http://www.wavelan.com/template.html?section=m57&page=2511&envelope=93
I see we have support for this in the form of the wi(4) driver. They
mention that this is an OEM-only board, though, so I would probably
have a hard time getting my hands on it. Any chance you could offer
the 128-bit version as an additional option, Soren? :-)
Where to find a correct antenna for the thing would be the next
problem... Ideally, I'd stick some kind of connector on the case
(BNC?) and attach a small rubber-ducky antenna to that. The connector
would also allow a much larger remote external antenna on it, of
course. Any ideas/URLs?
-- Chris Dillon - cdi...@wolves.k12.mo.us - cdi...@inter-linc.net
FreeBSD: The fastest and most stable server OS on the planet.
For IA32 and Alpha architectures. IA64, PPC, and ARM under development.
http://www.freebsd.org
To Unsubscribe: send mail to majo...@FreeBSD.org
bro...@one-eyed-alien.net
--sm4nu43k4a2Rpi4c
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable
On Tue, Apr 17, 2001 at 05:07:27PM -0600, Warner Losh wrote:
> In message <2001041715...@Odin.AC.HMC.Edu> Brooks Davis writes:
> : http://www.wavelan.com/template.html?section=3Dm57&page=3D2511&envelope=
=3D93
>=20
> Price? The web page in question didn't have anything.
No clue. I just pointed it out because I'd see it a while back. I
think they aren't very intrested in selling to people since the current
market for MiniPCI is laptops and these won't work without an integrated
antenna.
-- Brooks
--=20
Any statement of the form "X is the one, true Y" is FALSE.
PGP fingerprint 655D 519C 26A7 82E7 2529 9BF0 5D8E 8BE9 F238 1AD4
--sm4nu43k4a2Rpi4c
Content-Type: application/pgp-signature
Content-Disposition: inline
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.4 (GNU/Linux)
Comment: For info see http://www.gnupg.org
iD8DBQE63M5QXY6L6fI4GtQRAjSzAJ9P9ZDiMV5pjMz2PTBW+J1Sntn5PACfcwnK
KHAiILCDC4bxqKP8Zy+7qFQ=
=IE9c
-----END PGP SIGNATURE-----
--sm4nu43k4a2Rpi4c--
so...@soekris.com
I have actually been looking on that board for a while, and are
working on getting one for testing. I'm assuming that I would
qualify as an OEM, although a small one....
That would also mean that I will locate antennas and make a
version of my case to support them.
I don't expect the board to be especially expensive.
Soren
he...@home.cg.nu
Is it possible to make an live image on to a cdrom?
Like a complete bootable FreeBSD os on the cd.
PicoBSD on a bootable cdrom?
And if there is, where can i find a howto?, or some similair howto?
Henk Wevers
dar...@soco.agilent.com
> That right, they will come without the CompactFlash card.
>
> As Warner said, there is only one kind of CF cards, but they can
> operate in 3 modes, all required per the spec: PC Card memory
> mode, PC Card I/O mode and True IDE mode. I'm running them in true
> IDE mode. The CompactFlash spec is at http://www.compactflash.org/
Does the IBM microdrive work (it's got a CF interface)? I'm
interested, but I'd be more interested if the microdrive could be used
(like everyone else, I'd really like to make a "low-power" router with a
DHCP server that handles mac addresses).
--
Darryl Okahata
dar...@soco.agilent.com
DISCLAIMER: this message is the author's personal opinion and does not
constitute the support, opinion, or policy of Agilent Technologies, or
of the little green men that have been following him all day.
so...@soekris.com
> Does the IBM microdrive work (it's got a CF interface)? I'm
> interested, but I'd be more interested if the microdrive could be used
> (like everyone else, I'd really like to make a "low-power" router with a
> DHCP server that handles mac addresses).
>
Yes, the CompactFlash socket is a type II, and the Microdrive
should therefore work. I haven't tested it yet, but I will make it
to work if there is any problems....
Regards,
Soren
dar...@soco.agilent.com
> : Does the IBM microdrive work (it's got a CF interface)? I'm
> : interested, but I'd be more interested if the microdrive could be used
> : (like everyone else, I'd really like to make a "low-power" router with a
> : DHCP server that handles mac addresses).
>
> Yes, but 64M CF parts are cheaper and more reliable.
True, but the idea that a CF-based router *WILL* fail is, well,
distasteful.
[ The DHCP server keeps the leases on disk. In an home environment, it
should be possible to use a CF card, as clients don't come and go all
that often, but I don't know enough about FFS to be sure. ]
[ Heh. Commercial entities would call this, "planned obsolescence". ;-) ]
--
Darryl Okahata
dar...@soco.agilent.com
DISCLAIMER: this message is the author's personal opinion and does not
constitute the support, opinion, or policy of Agilent Technologies, or
of the little green men that have been following him all day.
To Unsubscribe: send mail to majo...@FreeBSD.org
e...@cotdazr.org
For me, better question yet: What would it cost to combine enough
disk, user account builder, webserver, DNS, proxyserver, add FreeBSD,
OpenBSD or Linux to a footprint like a small Cisco 1600 or slightly
racksize larger Cisco 2500/2600 with ipfilter, ipNAT, VPN and bring
it to market for ALL the DSL and CABLE and Wireless users. Bet it
could be effectively marketed for less than half the cost of a
Cisco 1600 stripped and little more than the Cadillac of the computer
store DSL/Cable hub-routers.
Thoughts ?
On Thu, Apr 19, 2001 at 10:29:38AM -0600, Warner Losh wrote:
> In message <2001041817...@mina.soco.agilent.com> Darryl Okahata writes:
> : Does the IBM microdrive work (it's got a CF interface)? I'm
> : interested, but I'd be more interested if the microdrive could be used
> : (like everyone else, I'd really like to make a "low-power" router with a
> : DHCP server that handles mac addresses).
>
> Yes, but 64M CF parts are cheaper and more reliable.
>
> Warner
>
> To Unsubscribe: send mail to majo...@FreeBSD.org
> with "unsubscribe freebsd-small" in the body of the message
--
+ http://www.vhwy.com e...@vhwy.com WA6...@arrl.net http://www.cotdazr.org +
+ PocketNet Mail to efb...@mobile.att.net / Cell/VoiceMail 805 340-6471 +
+ Unix BSD, Sun, HP SCO Linux Security Cisco Routing DataFellows QMail DNS +
i...@harmony.village.org
: Warner Losh <i...@harmony.village.org> wrote:
:
: > : Does the IBM microdrive work (it's got a CF interface)? I'm
: > : interested, but I'd be more interested if the microdrive could be used
: > : (like everyone else, I'd really like to make a "low-power" router with a
: > : DHCP server that handles mac addresses).
: >
: > Yes, but 64M CF parts are cheaper and more reliable.
:
: True, but the idea that a CF-based router *WILL* fail is, well,
: distasteful.
:
: [ The DHCP server keeps the leases on disk. In an home environment, it
: should be possible to use a CF card, as clients don't come and go all
: that often, but I don't know enough about FFS to be sure. ]
:
: [ Heh. Commercial entities would call this, "planned obsolescence". ;-) ]
CF parts have 10e5 to 10e6 writes per block typically. You would have
to do a lot of writes to the CF to wear it out. And we've found that
in hostile environments, the CF have been much more reliable than the
IDE hard disks they replaced.
Even if it was an issue, you could keep the leases on /tmp (make it
MFS) and then run a script once an hour to write it to the CF. Ditto
at shutdown. I once calculated that at 1 write per hour, the CF parts
that I was evaluating would last on the order of 1000s years (based
only on the wear of the flash parts).
Based on 10e5 writes per block, 8k blocks and 5M of available space, 1
write per second is about 723 days, assuming even wear on each of the
available blocks (unless I've messed up my calculations). That puts
it at two years. one write per minute would be 120 years and one
write per hour would be 7200 years.
Warner
i...@harmony.village.org
: interested, but I'd be more interested if the microdrive could be used
: (like everyone else, I'd really like to make a "low-power" router with a
: DHCP server that handles mac addresses).
Yes, but 64M CF parts are cheaper and more reliable.
Warner
ab...@webgiro.com
> Warner,
>
> For me, better question yet: What would it cost to combine enough
> disk, user account builder, webserver, DNS, proxyserver, add FreeBSD,
> OpenBSD or Linux to a footprint like a small Cisco 1600 or slightly
> racksize larger Cisco 2500/2600 with ipfilter, ipNAT, VPN and bring
> it to market for ALL the DSL and CABLE and Wireless users. Bet it
> could be effectively marketed for less than half the cost of a
> Cisco 1600 stripped and little more than the Cadillac of the computer
> store DSL/Cable hub-routers.
>
> Thoughts ?
Congratulations! You just reinvented the InterJet.
:-)
Andrzej
// ----------------------------------------------------------------
// Andrzej Bialecki <ab...@webgiro.com>, Chief System Architect
// WebGiro AB, Sweden (http://www.webgiro.com)
// ----------------------------------------------------------------
// <ab...@freebsd.org> FreeBSD developer (http://www.freebsd.org)
D...@silver-lynx.com
> Congratulations! You just reinvented the InterJet.
>
I think that the market niche is once again a viable one. Ask Terry
Lambert to configure it for you, he could probably do it in his sleep.
--
Don Wilde D...@Silver-Lynx.com
Silver Lynx Embedded Microsystems Architects
2218 Southern Bl. Ste. 12 Rio Rancho, NM 87124
505-891-4175 FAX 891-4185
dar...@soco.agilent.com
> Based on 10e5 writes per block, 8k blocks and 5M of available space, 1
> write per second is about 723 days, assuming even wear on each of the
> available blocks (unless I've messed up my calculations). That puts
> it at two years. one write per minute would be 120 years and one
> write per hour would be 7200 years.
... but, you can't assume even wear. Don't directory entries,
etc., get written in the same place, over and over? What about
superblocks?
Still, even if you assume 2 writes a minute (once every 30
seconds???) to the same block, over and over, it's still over two years
(1e5 writes -- I'm assuming that you mean "100000" for "10e5", and not
"1000000").
[ Of course, if one were to write a new/special filesystem that ensured
even wear, that would work well for CF .... ]
--
Darryl Okahata
dar...@soco.agilent.com
DISCLAIMER: this message is the author's personal opinion and does not
constitute the support, opinion, or policy of Agilent Technologies, or
of the little green men that have been following him all day.
To Unsubscribe: send mail to majo...@FreeBSD.org
i...@harmony.village.org
: > Based on 10e5 writes per block, 8k blocks and 5M of available space, 1
: > write per second is about 723 days, assuming even wear on each of the
: > available blocks (unless I've messed up my calculations). That puts
: > it at two years. one write per minute would be 120 years and one
: > write per hour would be 7200 years.
:
: ... but, you can't assume even wear. Don't directory entries,
: etc., get written in the same place, over and over? What about
: superblocks?
The hardware does remapping behind the scenes so that wear is
averaged over the set of blocks that change. On a 16M part, there
would be about 5M free/changing in the average system, so teh
calcuations would hold for that.
Warner
i...@harmony.village.org
: etc., get written in the same place, over and over? What about
: superblocks?
You can assume even wear becahse wear averaging is in the CF hardware.
There's a pool of extra blocks that are used to average the wear. At
least that's what the CF propiganda says from SandDisk.
Warner
p...@mesanet.com
> In message <2001041922...@mina.soco.agilent.com> Darryl Okahata writes:
> : ... but, you can't assume even wear. Don't directory entries,
> : etc., get written in the same place, over and over? What about
> : superblocks?
>
> You can assume even wear becahse wear averaging is in the CF hardware.
> There's a pool of extra blocks that are used to average the wear. At
> least that's what the CF propiganda says from SandDisk.
Thats what the data sheet for the CF controller chips (Toshiba)
that we use says also. The extra blocks are not for wear leveling, thats
taken care of by having enough free space. The free blocks are for
remapping newly developed bad blocks: so called "grown" defects.
The Toshiba controller chip implements wear leveling, bad block
forwarding and ECC. So basically if you leave enough free space on the
drive for your frequently written files to "rotate" through. You can have
_very_ long life. The internal flash format the Toshiba uses is published
(its the same as the SmartMedia format)
>
> Warner
>
>
> To Unsubscribe: send mail to majo...@FreeBSD.org
> with "unsubscribe freebsd-small" in the body of the message
>
Peter Wallace
nor...@unlimited.net
Since Everett <efb...@vhwy.com> got stopped in his tracks by Andrzej
Bialecki <ab...@webgiro.com> comment on reinventing an IBM owned toy.
I'll ask the next question: How does one find Mr. Terry Lambert whom you
mentioned?
John Oram
dar...@soco.agilent.com
> : ... but, you can't assume even wear. Don't directory entries,
> : etc., get written in the same place, over and over? What about
> : superblocks?
>
> The hardware does remapping behind the scenes so that wear is
> averaged over the set of blocks that change. On a 16M part, there
> would be about 5M free/changing in the average system, so teh
> calcuations would hold for that.
Oh, cool. I didn't know that.
--
Darryl Okahata
dar...@soco.agilent.com
DISCLAIMER: this message is the author's personal opinion and does not
constitute the support, opinion, or policy of Agilent Technologies, or
of the little green men that have been following him all day.
To Unsubscribe: send mail to majo...@FreeBSD.org
e...@cotdazr.org
Thank you for all that credit .. I disagree I DID NOT INVENT nor
did I REINVENT the WHATCHAMACALLIT .. I ASKED A QUESTION ..
It was as straightforward as I could ask that question ? If some
one did it and sold some, I might buy one.
Did you invent, sell or buy one ? Did / does it work ?
Who would I want to contact, email, telephone, cell block, etc ?
if one exists ?
OTHERWISE, Still standing is the question what do those of you
who are around here a long time view of the feasibility and cost
to bring that Whatchamacallit to market today ???
Oh, InterJet, hell I cant even fly (-: :-)
/Ev/
On Thu, Apr 19, 2001 at 07:01:44PM +0200, Andrzej Bialecki wrote:
> > For me, better question yet: What would it cost to combine enough
> > disk, user account builder, webserver, DNS, proxyserver, add FreeBSD,
> > OpenBSD or Linux to a footprint like a small Cisco 1600 or slightly
> > racksize larger Cisco 2500/2600 with ipfilter, ipNAT, VPN and bring
> > it to market for ALL the DSL and CABLE and Wireless users. Bet it
> > could be effectively marketed for less than half the cost of a
> > Cisco 1600 stripped and little more than the Cadillac of the computer
> > store DSL/Cable hub-routers.
> >
> > Thoughts ?
>
> Congratulations! You just reinvented the InterJet.
>
--
+ http://www.vhwy.com e...@vhwy.com WA6...@arrl.net http://www.cotdazr.org +
+ PocketNet Mail to efb...@mobile.att.net / Cell/VoiceMail 805 340-6471 +
+ Unix BSD, Sun, HP SCO Linux Security Cisco Routing DataFellows QMail DNS +
ab...@webgiro.com
> Andrzej,
>
> Thank you for all that credit .. I disagree I DID NOT INVENT nor
> did I REINVENT the WHATCHAMACALLIT .. I ASKED A QUESTION ..
Easy, easy, pal... No reason to get so worked up. Noticed the
smiley?
>
> It was as straightforward as I could ask that question ? If some
> one did it and sold some, I might buy one.
>
> Did you invent, sell or buy one ? Did / does it work ?
>
> Who would I want to contact, email, telephone, cell block, etc ?
> if one exists ?
>
> OTHERWISE, Still standing is the question what do those of you
> who are around here a long time view of the feasibility and cost
Just an aside: I've been probably the first one on this list. Go and check
the cvs logs in src/release/picobsd. :-)
> to bring that Whatchamacallit to market today ???
>
> Oh, InterJet, hell I cant even fly (-: :-)
What I was referring to was a concrete product - InterJet - of a company
called Whistle, who was (unfortunately) aquired by IBM. They produced and
sold considerable numbers of exactly such device. Quite a number of
core FreeBSD developers worked for them, among them Terry, Julian,
Archie, and a couple of others...
Unfortunately, after the aquisition the product vanished from the market
and (as far as I know) is available only as a part of an overpriced
service offering from IBM.
So, maybe this is the right time for the next InterJet...
Andrzej
// ----------------------------------------------------------------
// Andrzej Bialecki <ab...@webgiro.com>, Chief System Architect
// WebGiro AB, Sweden (http://www.webgiro.com)
// ----------------------------------------------------------------
// <ab...@freebsd.org> FreeBSD developer (http://www.freebsd.org)
m...@bus.net
> hour would be 7200 years.
regardless of cycling, mtbf of today's submicron design limits us to 10 years.
that pretty much goes for every part on the board. if you have an application
which calls for more than that, you'll need to keep component temperature
at or below 25 degrees C (although a write cyle limit is a write cycle limit).
we all have enjoyed a lot of use from old 386, 486, HC, and 8051 hardware
that lasted seemingly forever but, in my heart of hearts, i beleive those days
have gone.
> So, maybe this is the right time for the next InterJet...
> bring that Whatchamacallit to market today
compile a hardware (xx storage, 486 running at whatever, yy ether
ports expandable to zz, compatability with supermegafragalistic opto magic,
yellowish greentooth radio magic, T1, dsl, et cetera) and functionality
wishlist (i'm unfamiliar with these designs). i'm pretty short on time but,
i'd be glad to look into trying to put something together. we could get
started on something common like a pc104 platform and attack it
piecemeal to find bottlenecks and cost centers. once we've hammered
it out i'll design a custom board and FPGAs. i'm paying around $50
per board for one-offs (plus components) and prices will drop
precipitously once we get into volumes of 100s.
i'm desperate for a challenge (i'm designing battery chargers right now and
it is getting mighty sleepy) and am willing to carry a fair amount of the
development cost. my abilities will limit my contribution to the software
end to around nil.
Michael O'Donnell
mso at bus dot net
so...@soekris.com
Me again.... Also, please note that my website is up again.
Everett F Batey wrote:
>
> For me, better question yet: What would it cost to combine enough
> disk, user account builder, webserver, DNS, proxyserver, add FreeBSD,
> OpenBSD or Linux to a footprint like a small Cisco 1600 or slightly
> racksize larger Cisco 2500/2600 with ipfilter, ipNAT, VPN and bring
> it to market for ALL the DSL and CABLE and Wireless users. Bet it
> could be effectively marketed for less than half the cost of a
> Cisco 1600 stripped and little more than the Cadillac of the computer
> store DSL/Cable hub-routers.
>
That is the integrated gateway thing, just like the Whistle
Interjet.... But that was never really sold as a box only, and was
way too expensive anyway.
Just to let you know, I'm already planning the next board, which
would meet the above requirements....
It will be something like:
200-266 Mhz. National Semiconductor Geode SC3200
32-128 Mbyte SDRAM, soldered on board + expansion SO-DIMM
2-4 Mbit BIOS/BOOT Flash
CompactFLASH Type I/II socket
UDMA33 IDE Interface, 40 pins & 44 pins.
2-4 10/100 mbit Ethernet ports, RJ-45
2 Serial ports, DB9
MiniPCI type III socket. (t.ex for optional hardware encryption)
PCI Slot, right angle 3.3V only. (t.ex for optional WAN board)
Parallel Port, 26 pins header
Low Power & Fanless
Board size 5" x 7"
Small Case with room for 2.5" drive
Case for 1U 19" rackmount, with room for 2 x 3.5" drives
Software:
comBIOS for full headless operation over serial port
PXE boot rom for diskless booting
Designed for Linux, OpenBSD, FreeBSD
Specification are subject to change, and don't expect production
until 3 months from now.
Regards,
Soren
so...@soekris.com
Just an update.... The first production series of 240 boards has
been set for end of May, so boards will be available for delivery
at the start of June 2001. If they sell out, people on this list
that has expressed interest will get priority.
There will be two versions available, with either 32 or 64 Mbyte
SDRAM, both with 133 Mhz CPU, 3 Ethernet interfaces, 1 Serial
interface, MiniPCI socket, 3.3V PCI connector, including a small
metal case and US power supply. Later on other configurations may
be available, or any configuration can be produced in larger
volumes. Prices are as follow, in USD:
Quantity 32 Mbyte 64 Mbyte
--------------------------------
1-24 $220 $240
25-99 $212 $232
100-499 $204 $223
500+ $196 $215
Ordering and payment information will come up on my website later.
But for now terms will probably be pre payment by check or money
order, plus shipping at actual cost. But I will also look into
setting up credit card processing....
Regards,
Soren Kristensen
dar...@soco.agilent.com
> There will be two versions available, with either 32 or 64 Mbyte
> SDRAM, both with 133 Mhz CPU, 3 Ethernet interfaces, 1 Serial
Out of curiosity, is the 32MB version expandable to 64MB, or is it
fixed (because a lower density RAM IC is used)?
[ I'd still get the 64MB version, anyway, as the US$20 difference is
pretty small. I'd be interested in two boards. ]
> interface, MiniPCI socket, 3.3V PCI connector, including a small
> metal case and US power supply. Later on other configurations may
> be available, or any configuration can be produced in larger
> volumes. Prices are as follow, in USD:
Is this shipping from the US, or do US folks have to worry about
customs? [ Sorry, I can't tell .... ]
> Ordering and payment information will come up on my website later.
> But for now terms will probably be pre payment by check or money
> order, plus shipping at actual cost. But I will also look into
> setting up credit card processing....
Just as a data point, I'd prefer credit cards, and I'd be willing
to pay slightly more for it (to cover any credit card processing).
--
Darryl Okahata
dar...@soco.agilent.com
DISCLAIMER: this message is the author's personal opinion and does not
constitute the support, opinion, or policy of Agilent Technologies, or
of the little green men that have been following him all day.
To Unsubscribe: send mail to majo...@FreeBSD.org
ch...@dcfinc.com
systems, and transferred $230 to you via PayPal. I haven't received
the system, nor even an acknowledgment.
Did you get the order? Can you tell me when it will ship?
-crl
--
Chad R. Larson (CRL15) 602-953-1392 Brother, can you paradigm?
ch...@dcfinc.com ch...@larsons.org lar...@home.com
DCF, Inc. - 14623 North 49th Place, Scottsdale, Arizona 85254-2207