Hi,
My debug target machine has only USB ports. I’ve tried a
USB to Serial convert cable without success. The cable
is from IOGear. The manual states that does not require
any IRQ. Is this the reason that Windbg can’t connect to
the target machine?
The machine has only one PCMCIA slot, so I can’t put a
serial card and debug another PCMCIA card at the same
time.
If you know any way to use Windbg with USB port, please
help.
Thanks,
Minfeng Li
>My debug target machine has only USB ports. I’ve tried a
USB to Serial convert cable without success. The cable
is from IOGear. The manual states that does not require
any IRQ. Is this the reason that Windbg can’t connect to
the target machine?The machine has only one PCMCIA slot, so I can’t put a
serial card and debug another PCMCIA card at the same
time.
The kernel debugger code MUST talk to the legacy serial hardware directly.
A USB->Serial adapter will not work. WinXP can use 1394.
Is it possible there is legacy serial port on the board (10 pins), and all
you need is a cable to adapt from a DB-9 to a header connector?
> If you know any way to use Windbg with USB port, please
help.
No ways. KD can only use the UART hardware or - on XP only - the 1394 bus.
Max
Do you have an available IP interface? If so, and you are not using XP, then
you could use one of our Serial over IP devices to provide COM ports to your
target machine. Then you can null modem cable your host to the serial port
provided on the IP connected box.
I’ve made these connections both ways to successfully perform remote
debugging using our Serial over IP devices. The downside is that I’m
attempting to re-write the driver for XP as a WDM driver. This is because
the legacy NT monolithic driver we are using for Win2000 and the Whistler
beta program utilizes NDIS 3 and XP does not provide legacy support for that
version of NDIS.
wrote in message news:xxxxx@ntdev…
>
> Hi,
>
> My debug target machine has only USB ports. I’ve tried a
> USB to Serial convert cable without success. The cable
> is from IOGear. The manual states that does not require
> any IRQ. Is this the reason that Windbg can’t connect to
> the target machine?
>
> The machine has only one PCMCIA slot, so I can’t put a
> serial card and debug another PCMCIA card at the same
> time.
>
> If you know any way to use Windbg with USB port, please
> help.
>
> Thanks,
> Minfeng Li
>
>
The laptop has only on PCMCIA slot which I use to debug
my NDIS IM driver over DLink Wireless card. It does not
have a built-in Ethernet port. So I guess I have to find
a laptop with one serial port or two PCMCIA slot.
Thanks,
Minfeng
Do you have an available IP interface? If so, and you are not using XP, then
you could use one of our Serial over IP devices to provide COM ports to your
target machine. Then you can null modem cable your host to the serial port
provided on the IP connected box.I’ve made these connections both ways to successfully perform remote
debugging using our Serial over IP devices. The downside is that I’m
attempting to re-write the driver for XP as a WDM driver. This is because
the legacy NT monolithic driver we are using for Win2000 and the Whistler
beta program utilizes NDIS 3 and XP does not provide legacy support for that
version of NDIS.wrote in message news:xxxxx@ntdev…
> >
> > Hi,
> >
> > My debug target machine has only USB ports. I’ve tried a
> > USB to Serial convert cable without success. The cable
> > is from IOGear. The manual states that does not require
> > any IRQ. Is this the reason that Windbg can’t connect to
> > the target machine?
> >
> > The machine has only one PCMCIA slot, so I can’t put a
> > serial card and debug another PCMCIA card at the same
> > time.
> >
> > If you know any way to use Windbg with USB port, please
> > help.
> >
> > Thanks,
> > Minfeng Li
> >
> >
>
>
>
> —
> You are currently subscribed to ntdev as: xxxxx@attbi.com
> To unsubscribe send a blank email to %%email.unsub%%
FYI, they did demo WinDbg using USB 2.0 at WinHEC. You have to have a
WinDbg capable USB 2.0 host controller and a special cable, similar I would
guess to a laplink USB to USB cable. Anyway, just interesting.
Why anyone would use this when 1394 is available is beyond me. Of course I
am sure Intel is going to make sure that every PC on the planet has a USB
2.0 controller even though 1394 is intrinsicly peer-to-peer, already faster
today, and already has plans to jump to 1600Mb/s at some future date. But,
I digress.
–
Bill McKenzie
“Maxim S. Shatskih” wrote in message
news:xxxxx@ntdev…
>
> > If you know any way to use Windbg with USB port, please
> > help.
>
> No ways. KD can only use the UART hardware or - on XP only - the 1394 bus.
>
> Max
>
>
>
>
My mouse is usb not 1394. I think that sort of answers the question,
which you already answered anyhow: every m’board on the planet will have
a usb20 port on it. The theory that the invisible hand of the market
chooses wisely, rather than chooses like a deranged blind monkey, is a
poor theory indeed.
Why anyone would use this when 1394 is available is beyond
me. Of course I am sure Intel is going to make sure that
every PC on the planet has a USB 2.0 controller even though
1394 is intrinsicly peer-to-peer, already faster today, and
already has plans to jump to 1600Mb/s at some future date.
But, I digress.
You are currently subscribed to ntdev as: xxxxx@hollistech.com To
unsubscribe send a blank email to %%email.unsub%%
On Wed, 8 May 2002, Bill McKenzie wrote:
Why anyone would use this when 1394 is available is beyond me.
Why not?
USB devices likely outnumber 1394 devices, USB2 devices likely outnumber
S800 devices. Of the two interfaces, USB seems to be the one with greater
real-world backing. A USB2 card will work with my existing devices, a
1394 card will not, a 1394b card will not. A 1394b card isn’t even
guaranteed to work with 1394 devices (if it has beta-only ports, it won’t
work with them, it may require new cables).
USB2 has greater utility than 1394 presently does, and will likely gain
greater penetration; it makes perfect sense to support it.
Out of curiosity, what 1394b devices can one buy at the moment? I can
walk into my local HW shop and pick up USB2 devices, 1394b is nowhere to
be seen.
Of course I
am sure Intel is going to make sure that every PC on the planet has a USB
2.0 controller even though 1394 is intrinsicly peer-to-peer, already faster
today, and already has plans to jump to 1600Mb/s at some future date. But,
I digress.–
Bill McKenzie“Maxim S. Shatskih” wrote in message
> news:xxxxx@ntdev…
> >
> > > If you know any way to use Windbg with USB port, please
> > > help.
> >
> > No ways. KD can only use the UART hardware or - on XP only - the 1394 bus.
> >
> > Max
> >
> >
> >
> >
>
>
>
> —
> You are currently subscribed to ntdev as: xxxxx@inkvine.fluff.org
> To unsubscribe send a blank email to %%email.unsub%%
>
–
Peter xxxxx@inkvine.fluff.org
http://www.inkvine.fluff.org/~peter/
logic kicks ass:
(1) Horses have an even number of legs.
(2) They have two legs in back and fore legs in front.
(3) This makes a total of six legs, which certainly is an odd number of
legs for a horse.
(4) But the only number that is both odd and even is infinity.
(5) Therefore, horses must have an infinite number of legs.
Actually, I just meant the 1394 for debugging, but the technology thing does
irk me a bit.
Intel is the problem, which it seems to be quite adept at being after its
one logical and successful move with PCI. Its moving the industry headlong
in a direction by golly, damn the torpedoes and all that. Never mind
where that might be heading. Anyway, I should know better than to post
like this, but I never learn 
–
Bill McKenzie
“Mark Roddy” wrote in message news:xxxxx@ntdev…
>
> My mouse is usb not 1394. I think that sort of answers the question,
> which you already answered anyhow: every m’board on the planet will have
> a usb20 port on it. The theory that the invisible hand of the market
> chooses wisely, rather than chooses like a deranged blind monkey, is a
> poor theory indeed.
>
> >
> > Why anyone would use this when 1394 is available is beyond
> > me. Of course I am sure Intel is going to make sure that
> > every PC on the planet has a USB 2.0 controller even though
> > 1394 is intrinsicly peer-to-peer, already faster today, and
> > already has plans to jump to 1600Mb/s at some future date.
> > But, I digress.
>
>
> —
> You are currently subscribed to ntdev as: xxxxx@hollistech.com To
> unsubscribe send a blank email to %%email.unsub%%
>
>
>
Bill, can you expand ? What do you believe they’re doing wrong ? Right now
we can do point to point debugging with SoftICE over 1394, by using the
TCP/IP over Firewire functionality of Windows XP. And that’s because we’re
too lazy to implement a couple of low level 1394 hardware drivers that would
allow SoftICE to talk to SiRemote directly, bypassing Windows, like we do
with the current native TCP/IP and Serial.
-----Original Message-----
From: Bill McKenzie [mailto:xxxxx@bsquare.com]
Sent: Wednesday, May 08, 2002 1:26 PM
To: NT Developers Interest List
Subject: [ntdev] Re: Using Windbg without serial port
Actually, I just meant the 1394 for debugging, but the technology thing does
irk me a bit.
Intel is the problem, which it seems to be quite adept at being after its
one logical and successful move with PCI. Its moving the industry headlong
in a direction by golly, damn the torpedoes and all that. Never mind
where that might be heading. Anyway, I should know better than to post
like this, but I never learn
–
Bill McKenzie
“Mark Roddy” wrote in message news:xxxxx@ntdev…
>
> My mouse is usb not 1394. I think that sort of answers the question,
> which you already answered anyhow: every m’board on the planet will have
> a usb20 port on it. The theory that the invisible hand of the market
> chooses wisely, rather than chooses like a deranged blind monkey, is a
> poor theory indeed.
>
> >
> > Why anyone would use this when 1394 is available is beyond
> > me. Of course I am sure Intel is going to make sure that
> > every PC on the planet has a USB 2.0 controller even though
> > 1394 is intrinsicly peer-to-peer, already faster today, and
> > already has plans to jump to 1600Mb/s at some future date.
> > But, I digress.
>
>
> —
> You are currently subscribed to ntdev as: xxxxx@hollistech.com To
> unsubscribe send a blank email to %%email.unsub%%
>
>
>
—
You are currently subscribed to ntdev as: xxxxx@compuware.com
To unsubscribe send a blank email to %%email.unsub%%
The contents of this e-mail are intended for the named addressee only. It
contains information that may be confidential. Unless you are the named
addressee or an authorized designee, you may not copy or use it, or disclose
it to anyone else. If you received it in error please notify us immediately
and then destroy it.
> > Why anyone would use this when 1394 is available is beyond me.
Why not?
I meant strictly for debugging, and if I have to buy a special controller
and a special cable when I can buy ANY 1394 controller and cable for roughly
the same price, (actually probably cheaper now if you include the cable),
and I get the added advantage of real time video and all that with 1394, its
a no brainer for me.
USB devices likely outnumber 1394 devices, USB2 devices likely outnumber
S800 devices. Of the two interfaces, USB seems to be the one with greater
real-world backing. A USB2 card will work with my existing devices, a
1394 card will not, a 1394b card will not. A 1394b card isn’t even
guaranteed to work with 1394 devices (if it has beta-only ports, it won’t
work with them, it may require new cables).
I tell you what, I will bet you a shiny new quarter that I can find
infinitely many more 1394 devices that can operate at > 300Mb/s today than I
can USB 2.0 devices. Don’t even get me started on isochronous transfers
USB2 has greater utility than 1394 presently does, and will likely gain
greater penetration; it makes perfect sense to support it.Out of curiosity, what 1394b devices can one buy at the moment? I can
walk into my local HW shop and pick up USB2 devices, 1394b is nowhere to
be seen.
No but 1394a is seen everywhere and its faster despite all the claims of USB
2.0. I know this speed disparity probably won’t last long, but for the
industry to get all giddy about a technology that hasn’t quite matched what
firewire did quite a few years ago in speed is kind of funny don’t you
think? I know how the industry works, and you aren’t going to fight Intel,
but all things being equal give me the faster bus that supports peer-to-peer
transfers intrinsicly. But, the best technology rarely wins, AND I will say
USB makes a whole lot more sense for slow devices like mice.
–
Bill McKenzie
Why would you push USB for things like video when you have firewire? It
makes no sense. What consumer is that helping? USB wasn’t designed to
handle that sort of thing, and never will as well as firewire. I like the
peer to peer connectivity of 1394 as well. If I want to simulate hardware
to write a driver prior to the first smoke test of the real hardware, which
technology enables that? 1394’s approach is just more sound, albeit more
complicated. The spec sucks rocks though which is probably a significant
factor in its adoption. Intel is the most significant factor however.
Until the ports are on the motherboard in large measure, 1394 is doomed.
So, this is an academic discussion that I probably will wish I hadn’t
started.
–
Bill McKenzie
“Moreira, Alberto” wrote in message
news:xxxxx@ntdev…
>
> Bill, can you expand ? What do you believe they’re doing wrong ? Right now
> we can do point to point debugging with SoftICE over 1394, by using the
> TCP/IP over Firewire functionality of Windows XP. And that’s because we’re
> too lazy to implement a couple of low level 1394 hardware drivers that
would
> allow SoftICE to talk to SiRemote directly, bypassing Windows, like we do
> with the current native TCP/IP and Serial.
>
> -----Original Message-----
> From: Bill McKenzie [mailto:xxxxx@bsquare.com]
> Sent: Wednesday, May 08, 2002 1:26 PM
> To: NT Developers Interest List
> Subject: [ntdev] Re: Using Windbg without serial port
>
>
>
>
> Actually, I just meant the 1394 for debugging, but the technology thing
does
> irk me a bit.
>
> Intel is the problem, which it seems to be quite adept at being after its
> one logical and successful move with PCI. Its moving the industry
headlong
> in a direction by golly, damn the torpedoes and all that. Never mind
> where that might be heading. Anyway, I should know better than to post
> like this, but I never learn
>
> –
> Bill McKenzie
>
>
>
> “Mark Roddy” wrote in message news:xxxxx@ntdev…
> >
> > My mouse is usb not 1394. I think that sort of answers the question,
> > which you already answered anyhow: every m’board on the planet will have
> > a usb20 port on it. The theory that the invisible hand of the market
> > chooses wisely, rather than chooses like a deranged blind monkey, is a
> > poor theory indeed.
> >
> > >
> > > Why anyone would use this when 1394 is available is beyond
> > > me. Of course I am sure Intel is going to make sure that
> > > every PC on the planet has a USB 2.0 controller even though
> > > 1394 is intrinsicly peer-to-peer, already faster today, and
> > > already has plans to jump to 1600Mb/s at some future date.
> > > But, I digress.
> >
> >
> > —
> > You are currently subscribed to ntdev as: xxxxx@hollistech.com To
> > unsubscribe send a blank email to %%email.unsub%%
> >
> >
> >
>
>
>
> —
> You are currently subscribed to ntdev as: xxxxx@compuware.com
> To unsubscribe send a blank email to %%email.unsub%%
>
>
>
> The contents of this e-mail are intended for the named addressee only. It
> contains information that may be confidential. Unless you are the named
> addressee or an authorized designee, you may not copy or use it, or
disclose
> it to anyone else. If you received it in error please notify us
immediately
> and then destroy it.
>
>
>
On Wed, 8 May 2002, Bill McKenzie wrote:
I meant strictly for debugging, and if I have to buy a special controller
and a special cable when I can buy ANY 1394 controller and cable for roughly
the same price, (actually probably cheaper now if you include the cable),
USB2 cables and controllers are (certainly in outlets near me) identically
priced to 1394a cables and controllers; if anything, the USB2 cabling is a
little cheaper.
and I get the added advantage of real time video and all that with 1394, its
a no brainer for me.
And it’s a real no-brainer to get USB2 for me, because of the added
advantage of working with MP3 players, my DSL modumb (sic), my mice, and
so on.
I tell you what, I will bet you a shiny new quarter that I can find
infinitely many more 1394 devices that can operate at > 300Mb/s today than I
can USB 2.0 devices. Don’t even get me started on isochronous transfers
Maybe, I dunno what the USB2-IDE chipsets are like; I know that one has
recently been released by SMSC which purports to be able to transfer data
at 480 Mbit/sec (i.e. it’ll typically be bottlenecked by the drive, not
the interface); I know it took considerably longer for 1394a to get
similar high performance bridges.
No but 1394a is seen everywhere
hahahahahahahaa. Everywhere, except on the back of new computers or their
peripherals, which strike me as the only things of consequence to this
discussion.
and its faster despite all the claims of USB
2.0.
Fast USB 2 chipsets and IDE bridges are out now; USB 2 is both
faster on paper and in practice.
I know this speed disparity probably won’t last long, but for the
industry to get all giddy about a technology that hasn’t quite matched what
firewire did quite a few years ago in speed is kind of funny don’t you
think?
1394a took a long time to gain fast chipsets and bridges for the kinds of
device I want to use. It certainly hasn’t had high performance for “quite
a few years”. Until comparitively recently, 1394a devices were stunningly
slow. For many of its intended markets, it didn’t matter – DV doesn’t
need all that much throughput – but for computer peripherals, it was a
pretty crappy.
I know how the industry works, and you aren’t going to fight Intel,
but all things being equal give me the faster bus that supports peer-to-peer
transfers intrinsicly. But, the best technology rarely wins, AND I will say
USB makes a whole lot more sense for slow devices like mice.
A bus that can support a wide range of devices is simply more useful to
many than one can’t. USB2’s compatibility with USB alone makes it the
“best technology”, for many.
–
Peter xxxxx@inkvine.fluff.org
http://www.inkvine.fluff.org/~peter/
logic kicks ass:
(1) Horses have an even number of legs.
(2) They have two legs in back and fore legs in front.
(3) This makes a total of six legs, which certainly is an odd number of
legs for a horse.
(4) But the only number that is both odd and even is infinity.
(5) Therefore, horses must have an infinite number of legs.
1394 is extremely successful if you look at its widespread acceptance in the video capture market. It’s also on a growing number of notebook devices (esp from Sony, and of course Apple).
My understanding of the reluctance of mobo manufacturers to put in 1394 ports is that they’re reluctant to spend the $0.25 (or whatever fee it is) for the Apple patent royalty on the connectors. However, you can get a good quality firewire card for ~$20 or so.
Regards,
Paul Bunn, UltraBac Software, 425-644-6000
Microsoft MVP - Windows NT/2000/XP
http://www.ultrabac.com
-----Original Message-----
From: Bill McKenzie [mailto:xxxxx@bsquare.com]
Sent: Wednesday, May 08, 2002 11:16 AM
To: NT Developers Interest List
Subject: [ntdev] Re: Using Windbg without serial port
Why would you push USB for things like video when you have firewire? It
makes no sense. What consumer is that helping? USB wasn’t designed to
handle that sort of thing, and never will as well as firewire. I like the
peer to peer connectivity of 1394 as well. If I want to simulate hardware
to write a driver prior to the first smoke test of the real hardware, which
technology enables that? 1394’s approach is just more sound, albeit more
complicated. The spec sucks rocks though which is probably a significant
factor in its adoption. Intel is the most significant factor however.
Until the ports are on the motherboard in large measure, 1394 is doomed.
So, this is an academic discussion that I probably will wish I hadn’t
started.
>
USB devices likely outnumber 1394 devices, USB2 devices likely outnumber
S800 devices. Of the two interfaces, USB seems to be the one with greater
real-world backing. A USB2 card will work with my existing devices, a
1394 card will not, a 1394b card will not. A 1394b card isn’t even
guaranteed to work with 1394 devices (if it has beta-only ports, it won’t
work with them, it may require new cables).USB2 has greater utility than 1394 presently does, and will likely gain
greater penetration; it makes perfect sense to support it.Out of curiosity, what 1394b devices can one buy at the moment? I can
walk into my local HW shop and pick up USB2 devices, 1394b is nowhere to
be seen.
Of course for WinDBG, one will have to buy an approved USB 2.0 bridge, for
WinDBG to work. I haven’t seen the pricing on these, but I wonder which
will be cheaper, two 1394 cards + cable, or 2 cables + USB bridge?
Don Burn
Egenera, Inc.
Well, the $0.25 might be a factor to some degree, and I am not dismissing
that. But, deals could be struck and when you get down to it, who pushes
the most motherboards, and who brought out USB for the most part? There is
the problem with 1394’s acceptance.
–
Bill McKenzie
“Paul Bunn” wrote in message news:xxxxx@ntdev…
>
> 1394 is extremely successful if you look at its widespread acceptance in
the video capture market. It’s also on a growing number of notebook devices
(esp from Sony, and of course Apple).
>
> My understanding of the reluctance of mobo manufacturers to put in 1394
ports is that they’re reluctant to spend the $0.25 (or whatever fee it is)
for the Apple patent royalty on the connectors. However, you can get a good
quality firewire card for ~$20 or so.
>
> Regards,
>
> Paul Bunn, UltraBac Software, 425-644-6000
> Microsoft MVP - Windows NT/2000/XP
> http://www.ultrabac.com
>
>
> -----Original Message-----
> From: Bill McKenzie [mailto:xxxxx@bsquare.com]
> Sent: Wednesday, May 08, 2002 11:16 AM
> To: NT Developers Interest List
> Subject: [ntdev] Re: Using Windbg without serial port
>
>
> Why would you push USB for things like video when you have firewire? It
> makes no sense. What consumer is that helping? USB wasn’t designed to
> handle that sort of thing, and never will as well as firewire. I like the
> peer to peer connectivity of 1394 as well. If I want to simulate hardware
> to write a driver prior to the first smoke test of the real hardware,
which
> technology enables that? 1394’s approach is just more sound, albeit more
> complicated. The spec sucks rocks though which is probably a significant
> factor in its adoption. Intel is the most significant factor however.
> Until the ports are on the motherboard in large measure, 1394 is doomed.
> So, this is an academic discussion that I probably will wish I hadn’t
> started.
>
>
>
Ah yes, Oregon’s most famous “Not Invented Here” protagonists… 
Regards,
Paul Bunn, UltraBac Software, 425-644-6000
Microsoft MVP - Windows NT/2000/XP
http://www.ultrabac.com
-----Original Message-----
From: Bill McKenzie [mailto:xxxxx@bsquare.com]
Sent: Wednesday, May 08, 2002 12:16 PM
To: NT Developers Interest List
Subject: [ntdev] Re: Using Windbg without serial port
Well, the $0.25 might be a factor to some degree, and I am not dismissing
that. But, deals could be struck and when you get down to it, who pushes
the most motherboards, and who brought out USB for the most part? There is
the problem with 1394’s acceptance.
–
Bill McKenzie
“Paul Bunn” wrote in message news:xxxxx@ntdev…
>
> 1394 is extremely successful if you look at its widespread acceptance in
the video capture market. It’s also on a growing number of notebook devices
(esp from Sony, and of course Apple).
>
> My understanding of the reluctance of mobo manufacturers to put in 1394
ports is that they’re reluctant to spend the $0.25 (or whatever fee it is)
for the Apple patent royalty on the connectors. However, you can get a good
quality firewire card for ~$20 or so.
>
> Regards,
>
> Paul Bunn, UltraBac Software, 425-644-6000
> Microsoft MVP - Windows NT/2000/XP
> http://www.ultrabac.com
>
>
> -----Original Message-----
> From: Bill McKenzie [mailto:xxxxx@bsquare.com]
> Sent: Wednesday, May 08, 2002 11:16 AM
> To: NT Developers Interest List
> Subject: [ntdev] Re: Using Windbg without serial port
>
>
> Why would you push USB for things like video when you have firewire? It
> makes no sense. What consumer is that helping? USB wasn’t designed to
> handle that sort of thing, and never will as well as firewire. I like the
> peer to peer connectivity of 1394 as well. If I want to simulate hardware
> to write a driver prior to the first smoke test of the real hardware,
which
> technology enables that? 1394’s approach is just more sound, albeit more
> complicated. The spec sucks rocks though which is probably a significant
> factor in its adoption. Intel is the most significant factor however.
> Until the ports are on the motherboard in large measure, 1394 is doomed.
> So, this is an academic discussion that I probably will wish I hadn’t
> started.
>
>
>
—
You are currently subscribed to ntdev as: xxxxx@ultrabac.com
To unsubscribe send a blank email to %%email.unsub%%
“PeterB” wrote in message news:xxxxx@ntdev…
>
> On Wed, 8 May 2002, Bill McKenzie wrote:
>
> USB2 cables and controllers are (certainly in outlets near me) identically
> priced to 1394a cables and controllers; if anything, the USB2 cabling is a
> little cheaper.
>
Wouldn’t doubt the price, but that has nothing to do with the actual cost of
producing the hardware, rather who is pushing what today. That is to say,
1394 could easily be made as cheap or cheaper than USB2 for high speed
devices.
> > and I get the added advantage of real time video and all that with 1394,
its
> > a no brainer for me.
> And it’s a real no-brainer to get USB2 for me, because of the added
> advantage of working with MP3 players, my DSL modumb (sic), my mice, and
> so on.
>
Granted, that is why USB2 will win. As I said the best technologies don’t
always, or even most of the time, win.
> > I tell you what, I will bet you a shiny new quarter that I can find
> > infinitely many more 1394 devices that can operate at > 300Mb/s today
than I
> > can USB 2.0 devices. Don’t even get me started on isochronous transfers
> Maybe, I dunno what the USB2-IDE chipsets are like; I know that one has
> recently been released by SMSC which purports to be able to transfer data
> at 480 Mbit/sec (i.e. it’ll typically be bottlenecked by the drive, not
> the interface); I know it took considerably longer for 1394a to get
> similar high performance bridges.
>
They all purport 480 Mb/s as that is the spec limit, but none of them
delivers close to that from what I or anyone I have talked to has seen. The
point being, we already went down this road years ago with 1394. Why is
everybody excited that we, here years later, get to do it again, like it was
the great new thing? Just kind of funny.
> > No but 1394a is seen everywhere
> hahahahahahahaa. Everywhere, except on the back of new computers or their
> peripherals, which strike me as the only things of consequence to this
> discussion.
>
Agreed its academic.
> > and its faster despite all the claims of USB
> > 2.0.
> Fast USB 2 chipsets and IDE bridges are out now; USB 2 is both
> faster on paper and in practice.
>
NOT. Again I will bet you a shiny new quarter you can’t find a single USB
2.0 device that can operate at > 300Mb/s and there are plenty of 1394a
devices that can. Its probably more of a controller issue actually, but
don’t know. This will change though, no doubt.
> > I know this speed disparity probably won’t last long, but for the
> > industry to get all giddy about a technology that hasn’t quite matched
what
> > firewire did quite a few years ago in speed is kind of funny don’t you
> > think?
> 1394a took a long time to gain fast chipsets and bridges for the kinds of
> device I want to use. It certainly hasn’t had high performance for “quite
> a few years”.
It has been at least 3 years, as that is when I started messing with 1394 in
earnest and devices were pumping full speed then.
> Until comparitively recently, 1394a devices were stunningly
> slow. For many of its intended markets, it didn’t matter – DV doesn’t
> need all that much throughput – but for computer peripherals, it was a
> pretty crappy.
>
Its the OS interfaces that were crappy, that isn’t 1394’s fault.
> > I know how the industry works, and you aren’t going to fight Intel,
> > but all things being equal give me the faster bus that supports
peer-to-peer
> > transfers intrinsicly. But, the best technology rarely wins, AND I will
say
> > USB makes a whole lot more sense for slow devices like mice.
> A bus that can support a wide range of devices is simply more useful to
> many than one can’t. USB2’s compatibility with USB alone makes it the
> “best technology”, for many.
Well, that is obviously an excellent point…in a year or two when they
knock all the bugs out Of course 1394b could have had resources poured
into it instead and then the possibilities would be endless instead of being
where they were several years ago
USB2 IS the winner, which was decided before the spec was ever put to paper.
>
> –
> Peter xxxxx@inkvine.fluff.org
> http://www.inkvine.fluff.org/~peter/
>
–
Bill McKenzie
–
Bill McKenzie
“Paul Bunn” wrote in message news:xxxxx@ntdev…
>
> Ah yes, Oregon’s most famous “Not Invented Here” protagonists…
>
> Regards,
>
> Paul Bunn, UltraBac Software, 425-644-6000
> Microsoft MVP - Windows NT/2000/XP
> http://www.ultrabac.com
>
>
> -----Original Message-----
> From: Bill McKenzie [mailto:xxxxx@bsquare.com]
> Sent: Wednesday, May 08, 2002 12:16 PM
> To: NT Developers Interest List
> Subject: [ntdev] Re: Using Windbg without serial port
>
>
> Well, the $0.25 might be a factor to some degree, and I am not dismissing
> that. But, deals could be struck and when you get down to it, who pushes
> the most motherboards, and who brought out USB for the most part? There
is
> the problem with 1394’s acceptance.
>
> –
> Bill McKenzie
>
>
>
> “Paul Bunn” wrote in message news:xxxxx@ntdev…
> >
> > 1394 is extremely successful if you look at its widespread acceptance in
> the video capture market. It’s also on a growing number of notebook
devices
> (esp from Sony, and of course Apple).
> >
> > My understanding of the reluctance of mobo manufacturers to put in 1394
> ports is that they’re reluctant to spend the $0.25 (or whatever fee it is)
> for the Apple patent royalty on the connectors. However, you can get a
good
> quality firewire card for ~$20 or so.
> >
> > Regards,
> >
> > Paul Bunn, UltraBac Software, 425-644-6000
> > Microsoft MVP - Windows NT/2000/XP
> > http://www.ultrabac.com
> >
> >
> > -----Original Message-----
> > From: Bill McKenzie [mailto:xxxxx@bsquare.com]
> > Sent: Wednesday, May 08, 2002 11:16 AM
> > To: NT Developers Interest List
> > Subject: [ntdev] Re: Using Windbg without serial port
> >
> >
> > Why would you push USB for things like video when you have firewire? It
> > makes no sense. What consumer is that helping? USB wasn’t designed to
> > handle that sort of thing, and never will as well as firewire. I like
the
> > peer to peer connectivity of 1394 as well. If I want to simulate
hardware
> > to write a driver prior to the first smoke test of the real hardware,
> which
> > technology enables that? 1394’s approach is just more sound, albeit
more
> > complicated. The spec sucks rocks though which is probably a
significant
> > factor in its adoption. Intel is the most significant factor however.
> > Until the ports are on the motherboard in large measure, 1394 is doomed.
> > So, this is an academic discussion that I probably will wish I hadn’t
> > started.
> >
> >
> >
>
>
>
> —
> You are currently subscribed to ntdev as: xxxxx@ultrabac.com
> To unsubscribe send a blank email to %%email.unsub%%
>
>
> Why anyone would use this when 1394 is available is beyond me. Of course I
am sure Intel is going to make sure that every PC on the planet has a USB
2.0 controller even though 1394 is intrinsicly peer-to-peer, already faster
today, and already has plans to jump to 1600Mb/s at some future date. But,
This is one of the sad facts in the industry, since 1394 was and is much more capable then USB, even USB 2.0.
Looks like USB won only due to Intel including the USB host to south bridge chips, and is it not a monopoly thing like in what MS is
accused now?
Including USB in the corelogic is the same thing as including Media Player in the OS.
Max