I want to use a wlan-PCI card in my Windows Server 2003 as an Access Point.
Can anybody point me in the right direction?
I know it is faster, easier and very cheap to by en external access point,
but this is not the point.
On Wed, 12 Oct 2005 14:20:25 +0200, "Erik Palsbo"
<erik@_nospam_palsbo.com> wrote:
>I want to use a wlan-PCI card in my Windows Server 2003 as an Access Point.
>Can anybody point me in the right direction?
I don't know of anything that will run on Windoze Server 2003 that
will act as an access point unless you want to setup an ad-hoc network
which has no access points and everyone connects to everyone else.
Depending on what you're trying to accomplish, ad-hoc may be adquate.
Some of the older Orinoco drivers have a registry key called "APMode"
that allows the driver to simulate an access point, but that probably
will not run on Windoze Server 2003. I suspect that even if you could
find an AP spoofing program, you might have driver difficulties with
an internal PCI card. I don't think any of the card vendors even
support Windoze Server 2003 for their drivers although I'm fairly sure
that the XP NDIS5 drivers will work.
>I know it is faster, easier and very cheap to by en external access point,
>but this is not the point.
> On Wed, 12 Oct 2005 14:20:25 +0200, "Erik Palsbo"
> <erik@_nospam_palsbo.com> wrote:
>
>>I want to use a wlan-PCI card in my Windows Server 2003 as an Access
>>Point. Can anybody point me in the right direction?
>
> I don't know of anything that will run on Windoze Server 2003 that
> will act as an access point unless you want to setup an ad-hoc network
> which has no access points and everyone connects to everyone else.
> Depending on what you're trying to accomplish, ad-hoc may be adquate.
>
> Some of the older Orinoco drivers have a registry key called "APMode"
> that allows the driver to simulate an access point, but that probably
> will not run on Windoze Server 2003. I suspect that even if you could
> find an AP spoofing program, you might have driver difficulties with
> an internal PCI card. I don't think any of the card vendors even
> support Windoze Server 2003 for their drivers although I'm fairly sure
> that the XP NDIS5 drivers will work.
>
>>I know it is faster, easier and very cheap to by en external access point,
>>but this is not the point.
>
> Ah, it's the challenge. I see...
>
Well if you want a challenge, convert from Windows to Linux and then
if you have a suitable wireless card being an AP is easy. Exactly which
driver you need depends on the card, and only some drivers support AP mode
but all Prism 2/2.5/3 cards, Prism 54 cards, Intel 2x00 cards, Atheros cards
and several others are all supported as APs. In fact if you look in many
APs you will find Linux is what is driving them.
David
_________________________________________
Usenet Zone Free Binaries Usenet Server
More than 140,000 groups
Unlimited download http://www.usenetzone.com to open account
On Thu, 13 Oct 2005 10:32:25 +0100, David Goodenough
<david.goodenough@btconnect.com> wrote:
>Jeff Liebermann wrote:
>
>> On Wed, 12 Oct 2005 14:20:25 +0200, "Erik Palsbo"
>> <erik@_nospam_palsbo.com> wrote:
>>
>>>I want to use a wlan-PCI card in my Windows Server 2003 as an Access
>>>Point. Can anybody point me in the right direction?
>>
>> I don't know of anything that will run on Windoze Server 2003 that
>> will act as an access point unless you want to setup an ad-hoc network
>> which has no access points and everyone connects to everyone else.
>> Depending on what you're trying to accomplish, ad-hoc may be adquate.
>>
>> Some of the older Orinoco drivers have a registry key called "APMode"
>> that allows the driver to simulate an access point, but that probably
>> will not run on Windoze Server 2003. I suspect that even if you could
>> find an AP spoofing program, you might have driver difficulties with
>> an internal PCI card. I don't think any of the card vendors even
>> support Windoze Server 2003 for their drivers although I'm fairly sure
>> that the XP NDIS5 drivers will work.
>>
>>>I know it is faster, easier and very cheap to by en external access point,
>>>but this is not the point.
>>
>> Ah, it's the challenge. I see...
>Well if you want a challenge,
Not me. I already have too many challenges.
>convert from Windows to Linux and then
>if you have a suitable wireless card being an AP is easy.
Sorta. NDISWrapper impliments the Windoze NDIS interface on Linux and
allows the use of the Windoze drivers on Linux. http://ndiswrapper.sourceforge.net
>Exactly which
>driver you need depends on the card, and only some drivers support AP mode
>but all Prism 2/2.5/3 cards, Prism 54 cards, Intel 2x00 cards, Atheros cards
>and several others are all supported as APs.
Yep. HostAP for example: http://hostap.epitest.fi
However, the current version also supports NDISWrapper drivers which
allow it to be used with a wide variety of hardware besides the ones
you mentioned.
>In fact if you look in many
>APs you will find Linux is what is driving them.
Yep, but they don't use NDISWrapper. They really need the native
driver support for the wireless chipset to keep the code size
reasonable and for performance reasons.
The original challenge was to do it on Windoze Server 2003, which is
basically Windoze XP with server features. I couldn't find an XP
implimentation of something like HostAP.
> On Thu, 13 Oct 2005 10:32:25 +0100, David Goodenough
> <david.goodenough@btconnect.com> wrote:
>
>>Jeff Liebermann wrote:
>>
>>> On Wed, 12 Oct 2005 14:20:25 +0200, "Erik Palsbo"
>>> <erik@_nospam_palsbo.com> wrote:
>>>
>>>>I want to use a wlan-PCI card in my Windows Server 2003 as an Access
>>>>Point. Can anybody point me in the right direction?
>>>
>>> I don't know of anything that will run on Windoze Server 2003 that
>>> will act as an access point unless you want to setup an ad-hoc network
>>> which has no access points and everyone connects to everyone else.
>>> Depending on what you're trying to accomplish, ad-hoc may be adquate.
>>>
>>> Some of the older Orinoco drivers have a registry key called "APMode"
>>> that allows the driver to simulate an access point, but that probably
>>> will not run on Windoze Server 2003. I suspect that even if you could
>>> find an AP spoofing program, you might have driver difficulties with
>>> an internal PCI card. I don't think any of the card vendors even
>>> support Windoze Server 2003 for their drivers although I'm fairly sure
>>> that the XP NDIS5 drivers will work.
>>>
>>>>I know it is faster, easier and very cheap to by en external access
>>>>point, but this is not the point.
>>>
>>> Ah, it's the challenge. I see...
>
>>Well if you want a challenge,
>
> Not me. I already have too many challenges.
>
>>convert from Windows to Linux and then
>>if you have a suitable wireless card being an AP is easy.
>
> Sorta. NDISWrapper impliments the Windoze NDIS interface on Linux and
> allows the use of the Windoze drivers on Linux.
> http://ndiswrapper.sourceforge.net
>
For that to work the windows driver would have to have AP functionality,
and then it would work in windows as well. No you have to have a native
linux driver with the specific AP functionality.
>>Exactly which
>>driver you need depends on the card, and only some drivers support AP mode
>>but all Prism 2/2.5/3 cards, Prism 54 cards, Intel 2x00 cards, Atheros
>>cards and several others are all supported as APs.
>
> Yep. HostAP for example:
> http://hostap.epitest.fi
> However, the current version also supports NDISWrapper drivers which
> allow it to be used with a wide variety of hardware besides the ones
> you mentioned.
The hostapd (the d is important) support does support client authentication
for ndiswrapper drivers that does not give you AP support.
>
>>In fact if you look in many
>>APs you will find Linux is what is driving them.
>
> Yep, but they don't use NDISWrapper. They really need the native
> driver support for the wireless chipset to keep the code size
> reasonable and for performance reasons.
>
> The original challenge was to do it on Windoze Server 2003, which is
> basically Windoze XP with server features. I couldn't find an XP
> implimentation of something like HostAP.
>
As far as I know there is none. That is why I said you need to use Linux.
David
_________________________________________
Usenet Zone Free Binaries Usenet Server
More than 140,000 groups
Unlimited download http://www.usenetzone.com to open account
Thanks Jeff and Dvid for reply.
I had the feeling that this was not that easy. After all - it may not be a
good idea for security reason, but I liked the idea of only one box.
"Erik Palsbo" <erik@_nospam_palsbo.com> wrote in message
news:434cff85$0$156$edfadb0f@dread11.news.tele.dk. ..
>I want to use a wlan-PCI card in my Windows Server 2003 as an Access Point.
>Can anybody point me in the right direction?
> I know it is faster, easier and very cheap to by en external access point,
> but this is not the point.
>
In article <434eb025$0$187$edfadb0f@dread11.news.tele.dk>, "Erik Palsbo" <erik@_nospam_palsbo.com> wrote:
>Thanks Jeff and Dvid for reply.
>I had the feeling that this was not that easy. After all - it may not be a
>good idea for security reason, but I liked the idea of only one box.
Just to jump in and throw a little water on the fire, why would you want to
covert a $300 or so computer into a $70 AP in the first place? I mean
computers make really good calculators too, but I'd rather go to Kmart and
pay $5.00 then convert my computer to the task <wink>.
Just a thought from the wet blanket crowd.
>
>"Erik Palsbo" <erik@_nospam_palsbo.com> wrote in message
>news:434cff85$0$156$edfadb0f@dread11.news.tele.dk ...
>>I want to use a wlan-PCI card in my Windows Server 2003 as an Access Point.
>>Can anybody point me in the right direction?
>> I know it is faster, easier and very cheap to by en external access point,
>> but this is not the point.
>>
>
>
On Thu, 13 Oct 2005 21:06:19 +0200, "Erik Palsbo"
<erik@_nospam_palsbo.com> wrote:
>Thanks Jeff and Dvid for reply.
>I had the feeling that this was not that easy. After all - it may not be a
>good idea for security reason, but I liked the idea of only one box.
If you read any of my rants in this snoozegroup, you'll find that I'm
a big fan of seperate functions in seperate boxes. Usually, that
means seperate boxes for the DSL/cable modem, for the router, and for
the wireless access point. The reason is that each box wants to live
in a different environment. The router and modem wants to live under
a desk, in a close, buried in a basement, or other RF disgusting
location. Meanwhile, the radio (access point) wants to live as high
as possible and away from any sources of interference. These
requirements are mutually incompatible.
This also applied to a server with a PCI wireless card. Actually,
it's an even worse situation. The radio is next to the biggest source
of interference. The antenna is shielded by a big metal computer box.
It's also tangled among a bunch of cables and often buried under a
desk or in a server closet. I could not think of a worse RF
environment unless you want to bury it underground.
I'm not sure why you like the idea of having all your eggs in a
Microsoft basket, but I would suggest you reconsider your preferences.
> I had the feeling that this was not that easy. After all - it may not be a
> good idea for security reason, but I liked the idea of only one box.
Even when it's easy, it's still lousy. Just about everybody who tries it
soon moves on to a dedicated hardware base station. Having always-on,
no-hassle wireless access is well worth the price.
>In article <434eb025$0$187$edfadb0f@dread11.news.tele.dk>, "Erik Palsbo" <erik@_nospam_palsbo.com> wrote:
>>Thanks Jeff and Dvid for reply.
>>I had the feeling that this was not that easy. After all - it may not be a
>>good idea for security reason, but I liked the idea of only one box.
>Just to jump in and throw a little water on the fire, why would you want to
>covert a $300 or so computer into a $70 AP in the first place? I mean
>computers make really good calculators too, but I'd rather go to Kmart and
>pay $5.00 then convert my computer to the task <wink>.
Drivel: Once upon a time, in about 1984, I was helping out at a local
PC clone dealer. I had written my first (and last) Turbo Pascal 2
program, which displayed an analog clock on a CGA screen. This was
advanced programming in the days of 4.77Mhz computers. We had a
genuine IBM PC in the store window running my program. People would
walk by, stop, and set their watches by the time on the computer
screen. The assumption was that a $3,000 computer clock was better
than their $20 watch. Never mind that the computer clock would drift
all over the place and was rarely accurate.
[POSTED TO alt.internet.wireless - REPLY ON USENET PLEASE]
In <7umvk190h4e1pvgd1hpiafpi9kni81or3e@4ax.com> on Fri, 14 Oct 2005 09:33:40
-0700, Jeff Liebermann <jeffl@comix.santa-cruz.ca.us> wrote:
>Drivel: Once upon a time, in about 1984, I was helping out at a local
>PC clone dealer. I had written my first (and last) Turbo Pascal 2
>program, which displayed an analog clock on a CGA screen. This was
>advanced programming in the days of 4.77Mhz computers. We had a
>genuine IBM PC in the store window running my program. People would
>walk by, stop, and set their watches by the time on the computer
>screen. The assumption was that a $3,000 computer clock was better
>than their $20 watch. Never mind that the computer clock would drift
>all over the place and was rarely accurate.
More Drivel(tm): Such thinking persists, even though PC clocks still drift
all over the place. The drift bugs me so much that I run SymmTime 2005
(freeware <http://www.ntp-systems.com/symmtime.asp>, registration not
required), which has lots of cool features to boot. Keeps my PC clock
accurate to within about 100 ms even between external time syncs.
--
Best regards, HELP FOR CINGULAR GSM & SONY ERICSSON PHONES:
John Navas <http://navasgrp.home.att.net/#Cingular>
On Fri, 14 Oct 2005 17:11:55 GMT, John Navas
<spamfilter0@navasgroup.com> wrote:
>[POSTED TO alt.internet.wireless - REPLY ON USENET PLEASE]
>
>In <7umvk190h4e1pvgd1hpiafpi9kni81or3e@4ax.com> on Fri, 14 Oct 2005 09:33:40
>-0700, Jeff Liebermann <jeffl@comix.santa-cruz.ca.us> wrote:
>
>>Drivel: Once upon a time, in about 1984, I was helping out at a local
>>PC clone dealer. I had written my first (and last) Turbo Pascal 2
>>program, which displayed an analog clock on a CGA screen. This was
>>advanced programming in the days of 4.77Mhz computers. We had a
>>genuine IBM PC in the store window running my program. People would
>>walk by, stop, and set their watches by the time on the computer
>>screen. The assumption was that a $3,000 computer clock was better
>>than their $20 watch. Never mind that the computer clock would drift
>>all over the place and was rarely accurate.
>More Drivel(tm): Such thinking persists, even though PC clocks still drift
>all over the place.
It's really bad on all my laptops. I get rather tired of checking the
time and date whenever I boot one of them. I've implimented an NTP
update daemon, but that doesn't work if I don't have an internet
connection. Also, if you want to see NTP belch some random
corrections, turn on the "spread spectrum clock" feature in the bios.
>The drift bugs me so much that I run SymmTime 2005
>(freeware <http://www.ntp-systems.com/symmtime.asp>, registration not
>required), which has lots of cool features to boot. Keeps my PC clock
>accurate to within about 100 ms even between external time syncs.
I use NIST Time program: http://tf.nist.gov/service/its.htm
Written and tested by an agency of the US government. My tax dollars
at work.
I won't mention my failed attempt to write a GPS NEMA-183 time sync
program for my Unix box. It worked just fine until the antenna became
ocluded and it started belching all zeros for the time. Made a huge
mess of my log files and expired my entire usenet news spool
directory. That incident convinced me I'll never be a programmist.
Incidentally, I mumbled in another posting that I had limited success
using a OTP system on my PDA. That's because I could not maintain
time sync between the cellular phone time and the NTP updated server
time. I had to increase the window from 1 minute to 6 minutes to
obtain sync. I've noticed that my PDA Phone (QCP-6035) is often 2-3
minutes off from the real time. My guess is that the PDA section is
not getting properly updated from the phone section, but I've never
bothered to investigate the problem.
[POSTED TO alt.internet.wireless - REPLY ON USENET PLEASE]
In <63rvk158q6ficf1cpki7e84su3efpjl1h5@4ax.com> on Fri, 14 Oct 2005 10:53:45
-0700, Jeff Liebermann <jeffl@comix.santa-cruz.ca.us> wrote:
>On Fri, 14 Oct 2005 17:11:55 GMT, John Navas
><spamfilter0@navasgroup.com> wrote:
>
>>In <7umvk190h4e1pvgd1hpiafpi9kni81or3e@4ax.com> on Fri, 14 Oct 2005 09:33:40
>>-0700, Jeff Liebermann <jeffl@comix.santa-cruz.ca.us> wrote:
>>
>>>Drivel: Once upon a time, in about 1984, I was helping out at a local
>>>PC clone dealer. I had written my first (and last) Turbo Pascal 2
>>>program, which displayed an analog clock on a CGA screen. This was
>>>advanced programming in the days of 4.77Mhz computers. We had a
>>>genuine IBM PC in the store window running my program. People would
>>>walk by, stop, and set their watches by the time on the computer
>>>screen. The assumption was that a $3,000 computer clock was better
>>>than their $20 watch. Never mind that the computer clock would drift
>>>all over the place and was rarely accurate.
>
>>More Drivel(tm): Such thinking persists, even though PC clocks still drift
>>all over the place.
>
>It's really bad on all my laptops. I get rather tired of checking the
>time and date whenever I boot one of them. I've implimented an NTP
>update daemon, but that doesn't work if I don't have an internet
>connection. Also, if you want to see NTP belch some random
>corrections, turn on the "spread spectrum clock" feature in the bios.
>
>>The drift bugs me so much that I run SymmTime 2005
>>(freeware <http://www.ntp-systems.com/symmtime.asp>, registration not
>>required), which has lots of cool features to boot. Keeps my PC clock
>>accurate to within about 100 ms even between external time syncs.
>
>I use NIST Time program:
> http://tf.nist.gov/service/its.htm
>Written and tested by an agency of the US government. My tax dollars
>at work.
Not necessarily A Good Thing (any of those things). :))
A nice thing about SymmTime 2005 is that it keeps track of clock drift, and
applies corrections even when there isn't an external reference. As a result,
even though the clock drift varies on my notebook (IBM ThinkPad T30), I seldom
see more than a one second correction when reconnecting to the Internet.
>... I've noticed that my PDA Phone (QCP-6035) is often 2-3
>minutes off from the real time. My guess is that the PDA section is
>not getting properly updated from the phone section, but I've never
>bothered to investigate the problem.
Yikes! I consider that a silly and fairly serious flaw. Accurate time is
inherent in CDMA.
--
Best regards, HELP FOR CINGULAR GSM & SONY ERICSSON PHONES:
John Navas <http://navasgrp.home.att.net/#Cingular>
In article <7umvk190h4e1pvgd1hpiafpi9kni81or3e@4ax.com>, Jeff Liebermann <jeffl@comix.santa-cruz.ca.us> wrote:
>On Thu, 13 Oct 2005 21:24:20 GMT, rico_001@hotmail.com (Rico) wrote:
>
>>In article <434eb025$0$187$edfadb0f@dread11.news.tele.dk>, "Erik Palsbo"
> <erik@_nospam_palsbo.com> wrote:
>>>Thanks Jeff and Dvid for reply.
>>>I had the feeling that this was not that easy. After all - it may not be a
>>>good idea for security reason, but I liked the idea of only one box.
>
>>Just to jump in and throw a little water on the fire, why would you want to
>>covert a $300 or so computer into a $70 AP in the first place? I mean
>>computers make really good calculators too, but I'd rather go to Kmart and
>>pay $5.00 then convert my computer to the task <wink>.
>
>Drivel: Once upon a time, in about 1984, I was helping out at a local
>PC clone dealer. I had written my first (and last) Turbo Pascal 2
>program, which displayed an analog clock on a CGA screen. This was
>advanced programming in the days of 4.77Mhz computers. We had a
>genuine IBM PC in the store window running my program. People would
>walk by, stop, and set their watches by the time on the computer
>screen. The assumption was that a $3,000 computer clock was better
>than their $20 watch. Never mind that the computer clock would drift
>all over the place and was rarely accurate.
>
I may have my dates off here (old minds begin to fade) but I recall you
could dial into Ft Collins and capture the time in GW Basic and use that to
set the PC clock. It seems like it was a 110 baud link the first time I
did it. For some reason I recall doing this now and again in the early
80's. Whether it was a PC or XT I'm not sure, but I know before my first
AT; I still miss that orignal AT keyboard :(
On Fri, 14 Oct 2005 18:52:25 GMT, John Navas
<spamfilter0@navasgroup.com> wrote:
>>I use NIST Time program:
>> http://tf.nist.gov/service/its.htm
>>Written and tested by an agency of the US government. My tax dollars
>>at work.
>Not necessarily A Good Thing (any of those things). :))
The program probably has secret messages sent back to Area 51 that
informs the government of my subversive activities. Can't trust any
organization, including the government, that hasn't survived a peer
review of the source code. Oh, they do publish the source code: ftp://time.nist.gov/pub/daytime/
Well maybe we can trust the government, this time.
>>... I've noticed that my PDA Phone (QCP-6035) is often 2-3
>>minutes off from the real time. My guess is that the PDA section is
>>not getting properly updated from the phone section, but I've never
>>bothered to investigate the problem.
>Yikes! I consider that a silly and fairly serious flaw. Accurate time is
>inherent in CDMA.
Yep. The cell phone time is synced with the cell site, which is
synced with the SONET/SDH network or with GPS. I'm fairly sure it's
accurate. They even sell CMDA clock time servers: http://www.beaglesoft.com/celsynhome.htm
What seems to be my problem is that the Palm part of my PDA phone
apparently does not regularly update the time from the cell phone part
of the PDA. Apparently, there are two clocks in the phone. One for
the cell phone part and one for the PDA part. How they are
synchronized is not obvious (or documented). However, I've never
really investigated the phenomenon and it might be something else.
[POSTED TO alt.internet.wireless - REPLY ON USENET PLEASE]
In <kg90l1plksd6t4uubufk4q9f8m4a5hg2t1@4ax.com> on Fri, 14 Oct 2005 22:05:02
GMT, Jeff Liebermann <jeffl@comix.santa-cruz.ca.us> wrote:
>On Fri, 14 Oct 2005 18:52:25 GMT, John Navas
><spamfilter0@navasgroup.com> wrote:
>
>>>I use NIST Time program:
>>> http://tf.nist.gov/service/its.htm
>>>Written and tested by an agency of the US government. My tax dollars
>>>at work.
>
>>Not necessarily A Good Thing (any of those things). :))
>
>The program probably has secret messages sent back to Area 51 that
>informs the government of my subversive activities. Can't trust any
>organization, including the government, that hasn't survived a peer
>review of the source code. Oh, they do publish the source code:
> ftp://time.nist.gov/pub/daytime/
>Well maybe we can trust the government, this time.
As always, YMMV. My comments were based on having actually used it (over a
considerable period of time), and not being impressed with its stability
(observed jitter on multiple trials), or with its coding (clumsy time server
update). That's part of why I now use SymmTime.
--
Best regards, HELP FOR CINGULAR GSM & SONY ERICSSON PHONES:
John Navas <http://navasgrp.home.att.net/#Cingular>
Thanks again for all the reply. I do not disagree with the comments on how
"smart" this solution might be or not be. This was more of an academic
exercise. Fortunately not all ideas prove to be usable for operational
purpose, but testing does no harm.
As to the battle between preachers of Microsoft versus Linux: no comments on
religious matters from me.
By the way - I found that a product from Seque (SoftAP) is now bundled with
some w-lan cards (ASUS WL-138G is one of the) and I am going to test that.
Thanks again everyone.
Erik
"Erik Palsbo" <erik@_nospam_palsbo.com> wrote in message
news:434cff85$0$156$edfadb0f@dread11.news.tele.dk. ..
>I want to use a wlan-PCI card in my Windows Server 2003 as an Access Point.
>Can anybody point me in the right direction?
> I know it is faster, easier and very cheap to by en external access point,
> but this is not the point.
>
> On Fri, 14 Oct 2005 18:52:25 GMT, John Navas
> <spamfilter0@navasgroup.com> wrote:
>
>>Yikes! I consider that a silly and fairly serious flaw. Accurate time is
>>inherent in CDMA.
>
> Yep. The cell phone time is synced with the cell site, which is
> synced with the SONET/SDH network or with GPS. I'm fairly sure it's
> accurate.
My GSM phone is probably accurate, if I let the system update it, but the
darn thing is updated to ET, and I'm on AT! So I never let the system
update it. otoh, my TDMA phone wasn't even capable of being updated from
the cell system.
> What seems to be my problem is that the Palm part of my PDA phone
> apparently does not regularly update the time from the cell phone part
> of the PDA. Apparently, there are two clocks in the phone.
That's normal for your computer too. There's a hardware clock and a
software clock. Normally they only sync up at boot time. The Palm clock
is software, and _might_ be correctly updated from the cell-phone clock if
you cycled the power
--
derek
How can I turn a PC into an Access Point? 
Linear Mode
