Index Thing

Index

The Hippo Techo Blog



banner

112 Posts. Displaying page 4 of 12.
Previous Next

Go to page:


Hide All


Web - X/HTML/PHP/JS/CSS
 
Nov 06, 2006
 
32


XHTML 1.0 MIME type issue

Close ...
This site is designed for XHTML compatibility but really it is just a load of bollocks because of the following situation with XHTML User Agent compatibility.
Basically true XHTML is not supported by any of the IE browsers (which at the moment constitute the majority of browsers in use).
More about this can be found at http://www.hixie.ch/advocacy/xhtml.

With wanderinghippo although the meta tag is set as application/xhtml & xml the mime type sent from the server(which takes precedence with the UA) is actually text/html as is the case with many other XHTML 1.0 sites. This means the UA will not handle the page in a pure XHTML form.

Why then bother with XHTML 1.0 for the hippo?
You may well ask but I like the challenge of creating a compliant page according to the fiendish standards which in reality aren't that bad - just a shame that IE won't support them.

Update: Use of Content Negotiation

So, IE (along with other browsers) won’t support XHTML if the mime type is served as application/xhtml+xml however
by implementing server side Content Negotiation the issue can be worked around to a point.
If the .xhtml extension is used (at least on Apache servers) the mime type served will be application/xhtml+xml.
The method I have tried on my XAMPP dev server and production server (host using Apache) is to add suitable scripting to the .htaccess file.

ADD following to .htaccess file -


# requires mod_headers module on server
Header append Vary Accept

# requires mod_rewrite module on server
RewriteEngine on
RewriteCond %{HTTP_ACCEPT} !application/xhtml+xml
RewriteRule .* - "[T=text/html; charset=UTF-8]"


Note: I used to use the Abyss web server for dev but found that at this time the .htaccess file is not supported so I changed to the XAMPP web server for dev.

The two required modules for the .htaccess file were uncommented in httpd.conf file and the XAMPP server restarted.





Web - X/HTML/PHP/JS/CSS
 
Nov 06, 2006
 
31


XHTML Myths and Legends - some answers!

Close ...
1 The target attribute has never been included in Strict versions of HTML 4.01, XHTML 1.0 and XHTML 1.1 - it exists in HTML 4.01 Frameset/Transitional and XHTML 1.0 Frameset/Transitional

2 XHTML 1.0 Strict allows both xml:lang and lang with xml:lang taking precedence. XHTML 1.1 allows only xml:lang

3 The name attribute of the <a>, <map> and <form> tags is deprecated in XHTML 1.0 Strict and removed in XHTML 1.1 in favour of the id attribute

4 The name attribute of the <img> tag is removed in XHTML 1.0 Strict but exists in 1.0 Frameset/Transitional

5 The <iframe> tag is not included in XHTML 1.0 Strict

6 The name attribute of the <applet>, <form>, <frame>, <iframe> and <img> tags is deprecated in XHTML 1.0 Frameset/Transitional

7 The <ruby>, <rbc>, <rtc>, <rb>, <rt> and <rp> tags were introduced with XHTML 1.1

8 The following deprecated tags were excluded in XHTML 1.0 -
<applet>, <basefont>, <center>, <dir>, <font>, <frame>, <frameset>, <isindex>, <menu>, <noframes>, <s>, <strike>, <u>

9 Empty elements such as <br /> and <hr /> or <img src="image.jpg" alt="picture" /> are usually typed as such with a space preceding the slash to allow compatibility for older browsers.

10 XHTML 1.0 content presentation should normally be set as "application/xhtml+xml" in the META tag rather than "text/html" (allowed with XHTML 1.0) however if this is the case then IE will not process the content correctly (not supported). Since most XHTML 1.0 pages are presented as .html files rather than .xhtml files the servers normally present the content as "text/html" and so IE will work but not in strict XHTML mode.
11 <!-- and --> traditionally used for comments are still valid outside tags but the content will be ignored by some XML parsers.
Using these tags within script or style tags however should no longer be implemented because XML parsers will parse these tags including any unescaped XML characters - <![CDATA[ and ]]> should be used within these tags.

12 If XHTML is served as text/html background styling on the BODY tag will affect the whole window(including the HTML tag) but if served as application/xhtml+xml then only the BODY tag is affected. If offsets are used with the BODY tag the window will only be fully affected if the HTML tag is styled.

13 The apostrophe character - &apos; was introduced by XML 1.0 so &#39; may be required for earlier HTML browsers.

14 If XHTML is served as application/xhtml+xml then a javascript document.write() will not work because the XML parser is being used.








Web - X/HTML/PHP/JS/CSS
 
Nov 06, 2006
 
30


Adding Macro for XHTML 1.0 Validation to Expression Web Designer

Close ...
The macro facility in EWD can be used to provide XHTML validation in a similar manner to that described for FP at Michael Suodenjoki's site - HERE
(http://www.suodenjoki.dk/us/productions/articles/validation_arti cle.htm)
I found this method can also be used in FP2003

A few tweaks are required for the script to be used in EWD. Please bear in mind that I have used EWD CTP1 and Beta 1 so things may be different on retail releases when they happen.

UPDATE: Still works OK with EW Trial version (which I think may be a RC) and final release version also.

Install:

Fairly similar to section 3.2 of Michael's article
The validate.bas file can be tweaked with notepad or within the VBA editor.

Tools>Macro>Visual Basic Editor
VBA editor will open (different to screenshot)

View>Project Explorer
This will show Module1 which can be deleted

File>Import File
ExecuteCmd.bas (still works OK without tweaks)
Validate.bas (requires some tweaks)
Form_output.frm (OK as is)

These should now show in the Project Explorer
Close the VBA editor and use EW to complete the install of the macro.

Steps 7 to 17 of Michael's instructions are pretty well still OK for EW.

Tweaks for validate.bas

Line 35: MsgBox "Please open a file in the Expression Web Designer.", _
Line 40: If Not ActivePageWindow.ViewMode = PageViewNormal Then
Line 42: ActivePageWindow.ViewMode = PageViewNormal
Line 45: Dim doc As IHTMLDocument
Line 96: ActivePageWindow.ViewMode = PageViewHtml

Note: It can be convenient to tweak the validate.bas file before the import so a master record is kept since the app uses its own macro file once the import is done.

Macro location:
EWD Trial
C:Documents and SettingsuserApplication DataMicrosoftExpressionWeb DesignerMacros Microsoft Expression Web.wdmacro

EWD CTP1 and Beta1
C:Documents and SettingsuserApplication DataMicrosoftExpressionWeb DesignerMacros Microsoft Expression Web Designer.wdmacro

FP2003
C:Documents and SettingsuserApplication DataMicrosoftFrontPageMacrosMicrosoft Frontpage.fpm

Typical output on a "clean" XHTML 1.0 page was -

C:BMCDataStuffBMCWebValidationbinnsgmls.exe: C:BMCDataStuffBMCWebValidationpubtextxhtml1.dcl:31:27:W: characters in the document character set with numbers exceeding 65535 not supported
(Above line always appeared - just never worried about it too much)

C:BMCDataStuffBMCWebValidationbinnsgmls.exe: C:BMCDataStuffBMCWebValidationinput.tmp:10:9:E: character data is not allowed here
C:BMCDataStuffBMCWebValidationbinnsgmls.exe: C:BMCDataStuffBMCWebValidationinput.tmp:25:9:E: character data is not allowed here
(Lines above relate to and being in the user page)





Web - X/HTML/PHP/JS/CSS
 
Nov 06, 2006
 
29


XHTML 1.0 Strict Validation Stuff

Close ...
Notes 'n stuff whilst making an XHTML 1.0 Strict schema file for VWDE and FP 2003.

Unfortunately Visual Web Developer Express 2005 (VWDE) does not have the capability to handle XHTML 1.0 Strict schemas. Provision is made only for XHTML 1.0 Frameset/Transitional and XHTML 1.1 so basically I decided to try and fabricate a schema using the 1.1 file as a starting point since the differences between 1.0 Strict and 1.1 are minimal


Base MS file used was xhtml_strict-11.xsd
Location:
C:\Program Files\Microsoft Visual Studio 8\Common7\Packages\schemas\html

Original W3C file used was xhtml1-strict.xsd
Location:
http://www.w3.org/2002/08/xhtml/

Method:

Cut and Paste using the W3C file as original master reference for 1.0 Strict.

Cross check done with MS xhtml_transitional.xsd and known deprecations.
Basically MS have moved things around a bit but the W3C reference can be reconciled with the MS files fairly well.

W3CMS
tag ::xs:xsd:
coreattrscoreAttributeGroup
i18ni18nAttributeGroup
eventsinputEventsGroup
focusfocusAttributeGroup
attrscommonAttributeGroup
xs:complexType name="Block"xsd:group name="BlockElements"


Known Deprecations :

1.01.1
Langxml:Lang
a namea id
map namemap id
ruby (added)


Also most deprecations in XHTML 1.0 Transitional were removed in 1.0 Strict

The schema files (please feel free to check over the modified file B for accuracy before using - it's a prototype at the moment!)

File A: HERE - This is the original W3C XHTML 1.0 Strict schema file.

File B: HERE - (only use with VWDE) This is the modified MS xhtml_strict-11.xsd file from VWDE retaining original MS element order consistent with the original MS .xsd files (xhtml_strict-11.xsd & xhtml_transitional.xsd). This should give a Strict 1.0 interpretation that is more consistent with the w3c spec.

Installation:
VWD 2005 Express

Reference for this was found at -
http://blogs.msdn.com/mikhailarkhipov/archive/2006/02/01/522614.a spx

Because I have only VWD installed and not the full Visual Studio the registry hack becomes -

HKEY_LOCAL_MACHINESOFTWAREMicrosoftVWDExpress8.0Packages
{1B437D20-F8FE-11D2-A6AE-00104BCC7269}Schemas

There were 20 schemas already installed (standard install).

The standard install does not include XHTML 1.0 Strict at all
eg:
Schema 1 - html\ie6_0.xsd
Schema 4 - html\html_401.xsd
Schema 5 - html\xhtml_transitional.xsd
Schema 6 - html\xhtml_frameset.xsd
Schema 7 - html\xhtml_strict-11.xsd
Schema 21 - html\xhtml_strict-10.xsd (I added this one)

FrontPage 2003
This file is adapted from the file modified for VWD 2005. Reference was also made to existing HTML schemas provided with FP 2003.

The schema location is -
C:\Program Files\Common Files\Microsoft Shared\OFFICE11\SCHEMAS\HTML

The type-library file from VWD( w3c-dom1-strict.tlb) was also placed in this folder as the xsd file uses it.
FP 2003 does not require a registry entry for the schema as VWD does - a restart is sufficient to make the schema appear in
Tools>PageOptions>Authoring>SchemaVersion(dropdown ).

Reference for this was found at -
http://blogs.msdn.com/lisawoll/archive/2005/01/14/353451.aspx

Note:(finding file locations)
FP 2003 is a module of Office 11
VWD 2005 is a module of Visual Studio 8
EWD is a module of Office 12

Note: Initially I wanted to use the file for FP2003 but because the Intellisense was a wee bit tired I decided to have a go with the new app that will replace FP.

It is Expression Web Designer. You will notice in the above note that VWD is part of Visual Studio 2005 whereas EWD is part of Office as was FP2003. As a result the schema files for EWD vary slightly with the inclusion of fp tags, etc.

Useful web references:

http://www.w3.org/2002/08/xhtml/ (Master W3C xsd files)
http://blogs.msdn.com/mikhailarkhipov/archive/2006/02/01/522614.a spx (Registry hack for VWD 2005 schema addition)
http://blogs.msdn.com/lisawoll/archive/2005/01/14/353451.aspx (Customising Intellisense)
http://schneegans.de/frontpage/xhtml-schema/ (Using XHTML schema with FP 2003 & Visual Studio)

UPDATE:

MS Expression Web Designer is now available as a Community Technology Preview (CTP) package.
As it happens this supposedly supports XHTML 1.0 Strict and the file can be found in -

C:\Program Files\Common Files\Microsoft Shared\OFFICE12\SCHEMAS\HTML

This file has some differences to the modified one adapted from VWD.

EWD VWDE
clientom="w3c-dom1.tlb" clientom="w3c-dom1-strict.tlb"
cssschem a="CSS 2.0" cssschema="CSS 2.1"
<xsd:documentation> <xsd:documentation>
Microsoft Visual Studio .NET Microsoft Visual Studio .NET
schema for HTML 4.01 schema for XHTML 1.0
omitted <xsd:attribute name="Style" vs:builder="style" />


Unfortunately the XHTML 1.0 xsd support in EWD doesn't look too good so I will try to adapt the EWD file to be a bit closer to the W3C one.

UPDATE:

Expression Web Designer (EWD) is now known as Expression Web and is available as a trial version. Having looked at the fresh install I find the XHTML 1.0 xsd is still a bit dodgy but the app seems to work well as a replacement for FP2003. The only thing I am not keen on is the high price given that the XHTML 1.0 Strict functionality appears to be lacking.

File C: HERE - MS XHTML 1.0 Strict file supplied with EWD CTP1 and Beta 1.

File D: HERE - MS XHTML 1.0 Strict schema file supplied with EW Trial.

Differences are -
Addition 1
<!-- style info, which may include CDATA sections -->
EWD Beta 1
none
EW Trial
<xsd:enumeration value="default" />

Addition 2
<!-- script statements, which may include CDATA sections -->
EWD Beta 1
none
EW Trial
<xsd:enumeration value="default" />

Change 3

EWD Beta 1
<xsd:attribute name="xml:Space" default="preserve" fixed="true" />

EW Trial
<xsd:attribute name="xml:Space" vs:readonly="true" fixed="true">
<xsd:simpleType>
<xsd:restriction base="xsd:string">
<xsd:enumeration value="preserve" />
<xsd:enumeration value="default" />
</xsd:restriction>
</xsd:simpleType>
</xsd:attribute>

Change 4
xsd:attributeGroup name="cellHAlignTypeAttributes"
EWD Beta 1
<xsd:attribute name="Ch" />
<xsd:attribute name="ChOff" />

EW Trial
<xsd:attribute name="Char" />
<xsd:attribute name="CharOff" />

In theory a modified file for use with EW should also work with FP2003 but I seemed to have a few hassles with the Intellisense in FP particularly in the area of Byte Order Marks (BOM), the DOCTYPE and using UTF-8 so any effort is now directed towards EW. A good feature in HTMLPad 2006 that I also use is being able to save a page without the BOM.

File F: HERE - This is the modified xhtml file for EW Trial (still doing it.....).





Web - Apps
 
Nov 06, 2006
 
28


Microsoft Web Authoring Tools

Close ...
The other day I was messing about with some new Microsoft web authoring apps and here is my experience so far.

Frontpage 2003 - the app I tend to use
Visual Web Designer 2005 Express - available free
Expression Web Designer CTP1 - beta trial

I tend to use FP 2003 for XHTML/HTML work and Notepad++ for CSS but FP is now getting a bit tired. The big feature I like is the tabbed preview which is quick (albeit based on the IE rendering engine) and the code view is used a bit. The ability to work within a web is excellent for hyperlinking and renaming functions but the biggest drawback is now the XHTML standards support. I presently try to code for XHTML 1.0 Strict.
The macro facility in FP can be used to provide XHTML validation which has proved to be excellent but the inbuilt Intellisense support is weak. I have found that this can be improved a bit using a file derived for use with one of the later MS apps.
Trying VWDE I found the app only really does the HTML/XHTML editing part of what FP does without the web environment. Also despite being much improved in code creation, etc it still doesn't support XHTML 1.0 Strict for Intellisense.
So, my answer was to modify the MS copy of the XHTML 1.1 Strict schema in VWDE and use that (See above for more details).
Since messing around with VWDE MS have issued a CTP trial of EWD and this app is basically the new Frontpage. Interestingly the XHTML 1.0 Strict schema file used looks a bit dodgy in places





Wireless/Networking
 
Dec 04, 2005
 
27


Netgear DG834G Wireless ADSL/Router/Gateway

Close ...
F/W 2.10.22

Required standard UK setting of -

RFC2364 PPPoA, VPI=0, VCI=38 & VC multiplexing for the DSL interface.

Typical uptime figures for DG834G -
WAN PPPoA 464:01:52
LAN 100M 464:03:28
WLAN 54M 464:03:19


WAN PPPoA 67:04:23 note: ISP had an outage on the ADSL connection
LAN 100M 625:07:56
WLAN 54M 625:07:47

WAN PPPoA 313:58:40
LAN 100M 872:02:13
WLAN 54M 872:02:04

WAN PPPoA 615:10:17
LAN 100M 1173:13:50
WLAN 54M 1173:13:41

WAN PPPoA 349:11:47 note: ISP had an outage on the ADSL connection
LAN 100M 2047:57:34
WLAN 54M 2047:57:25



It seems that many routers do not appear to be stable on the wireless LAN side.

My experience of the WAG54G confirmed this - I kept getting unavailability on the wireless connection intermittently. The DG834G appears to fairly stable in this regard bearing in mind that it has the latest firmware available. The earlier loads of firmware did not appear so stable on the wireless LAN. The WAG54G has also been upgraded with the latest firmware load but I'm not sure that the wireless problem is fixed.


Interestingly Linksys do have some higher strength aerials available but I couldn't try these since the WAG doesn't have a replaceable aerial but they do apparently fit on the Netgear !!!!!

My experience so far of wireless - it is great for mobility but if I could be bothered I would have the desktop hardwired via 100Base-T ethernet and get full speed. The wireless side of the routers is very variable depending on the type of unit, software, etc.


Configuration can be done via web interface (default http:// is 192.168.1.2).

Using iBook - Firefox for Mac OK

Using PC both IE & Mozilla/Firefox are fine (see below: use IE for firmware upgrading)

Troubleshooting the wireless network can be a bit of a hassle but a couple of tools that help are -

Air Snare: Detects MAC addresses on the network

NetStumbler: Detects signal strength /details of wireless nodes in range


The following captures from NetStumbler (using Intel PRO/Wireless 2200BG) give an indication of the quality of signal from various routers -

Linksys WAG54G (ch 6)

WAG54G signal graph

Netgear DG834G (ch 3)

DG834G signal graph

Belkin (ch 11)

Belkin signal graph

BT Voyager (Askey Computer MAC code) (ch 11)

BT Voyager signal graph

The results using NetStumbler so far indicate -

DG834G Good, reasonably solid signal with a bit of fuzz
WAG54G Average, signal very spikey a lots of dropouts

The last two devices happen to be in the house next door which accounts for the lower signal level.

Belkin OK, reasonably solid signal and not too spikey
BT Voyager Excellent, solid signal with no spiking at all





Web - X/HTML/PHP/JS/CSS
 
Dec 04, 2005
 
26


CSS: padding-bottom 0px gives 5px offset

Close ...
Problem:

The padding-bottom 0px value is ignored if the doctype is set for XHTML 1.0

leaving an approximate 5px offset. If the doctype is disabled the problem is corrected, however this is not a good fix.

I tend to use the border-collapse setting especially when bunching images in tables, however with Firefox I still got the issue and I don't want to disable the doctype.

Result:

Two possibilities will work -

display: block; set against the img

vertical-align: bottom; set against the img

I would prefer to use the former along with the border-collapse against the table

A broken example - HERE

This is how the hover renders after the CSS is tweaked - HERE





Software - Apps
 
Dec 04, 2005
 
25


Thunderbird POP3 receive mail issue with Ipswitch servers

Close ...
Problem: Thunderbird gets error when logging into V21mail.co.uk POP3 server, works OK into BT Yahoo mail server

Error: Sending of password did not succeed. Mail server mail.v21net.co.uk responded:

Press OK and then get -
Please enter a new password for user aperson@v21net.co.uk on mail.v21net.co.uk:

entered password and got -

Sending of password did not succeed. Mail server mail.v21net.co.uk responded:USER

Logs: Following

Background: This problem occurred with TB on a Thinkpad running XP SP2 and only when a new ISP account was being used.
TB on a desktop P4 running XP SP1 worked OK to the new ISP. Mozilla Mail 1.7.5 worked fine on the Thinkpad.


Details:

Thinkpad R51 XP SP2
ISP v21.co.uk Mail server - Ipswitch IMail 8.00
TB 1.0 (20041206) No security selected for POP or SMTP




Troubleshoot: TB uninstalled from TP, all related folders deleted (profiles,etc) then reinstalled fresh - still NBG

Note - NBG is No Bloody Good

Tried Ethereal network analyzer and tracked the POP messages. Logs below.

It seems that TB requests a CAPA (capability list). The BT servers don't understand this and TB then uses the standard USER/PW in clear text. The V21 server however supplies a capability list and TB sends a mixture of USER/PW in clear text and an encrypted form (I don't understand this as I thought AUTH PLAIN would be in clear text also).
Unfortunately when the encrypted Request is being sent the USER/PW form is fired out when the encrypted password would be expected and therefore the authorisation breaks. Mozilla 1.7.5 doesn't appear to use the CAPA feature.

V21 Capability List

+OK Capability list followsrn
TOPrn
USERrn
SASL LOGIN PLAINrn
RESP-CODESrn
LOGIN-DELAY 120rn
PIPELININGrn
EXPIRE 30 USERrn
UIDLrn
IMPLEMENTATION Ipswitch_IMail_8.0rn
.rn

BT -ERR TB then sends USER/PW in clear
V21 +OK TB then sends AUTH PLAIN which fails. PW box pops up and sent encrypted?

Breakthrough: After much trawling and struggling to understand commands that relate more to SMTP than POP I discovered a Mozilla Bugzilla reference which is exactly the problem I had.

Bugzilla 259840

Summary: error retrieving POP3 mail from Ipswitch IMail servers when using some security products
[Symantec, Norton, Computer Associates, CA, eTrust, antivirus]

It so happened that the Thinkpad had Norton AntiVirus 2004 installed.
Disabling NAV was proving too much of a mission so it was uninstalled and SUCCESS!! - TB worked fine.
Resolution/workround:
Create or use user.js file with following line -

// Hack to work round bug with FF/Ipswitch Mail Server/AV mail combo - avoids disabling AV mail screening
user_pref("mail.server.server3.auth_login", false);

eg:Documents and Settings/<user>/Application Data/Thunderbird/Profiles/68hrtcvs.default/user.js



Outstanding questions:

Q. Why does TB appear to send encrypted USER/PW when no secure is selected - does the CAPA influence this? I expected AUTH PLAIN to still use clear text.

Answer: I would like to know more on this - most refs to AUTH PLAIN are for SMTP not POP interaction.

Q. Why are the error windows giving truncated results?

Answer: Probably a low priority bug maybe?

Q. Ethereal didn't work on the Thinkpad and it looks like this can be the case with Intel 2200BG wireless cards - it worked with the Intel 100baseT NIC.

Answer: It seems the Intel wireless drivers knobble the promiscuous feature and somewhere in the bowels of Intel may be a patch ??????

Why the hell does NAV have to be so embedded that it can't just be disabled ( I disabled the NAVAPSVC and some moronic process re-enabled it for me and set it to Manual - the damn computers are owning us...)

The troubleshoot logs in text form are HERE





Software - Apps
 
Dec 04, 2005
 
24


Thunderbird RSS doesn't receive all items

Close ...
It seems that if the content of the links field in a RSS item is identical to another item then only the first item will be downloaded.

Unfortunately TB does not appear to use the GUID field which should provide a unique ID for each item. This is an identified bug (BugZilla 260745 & 264482) and is still apparent in TB 1.5.0.5

Crude workround:

If the content is on the same page then use bookmarks to give a unique links field.

Personally I feel this bug makes TB useless for RSS because if you cannot be certain that 100% of a feed is being received then what good is it!

Alternatively use an RSS utility that works correctly. I found RSSOwl to be excellent but does not give the HTML page review as TB does.

An example of the problem is available HERE

::::::::::::::::::: Update :::::::::::::::::

RSS 2.0 seems a bit of a crock - I find that most readers do not use the GUID field to determine uniqueness of an item and therefore whether it is a duplicate item or should be downloaded as a new item. Indeed, some editors I have tried don't bother creating a unique GUID for each item.

The LINK field tends to be the decider and in many cases with an RSS feed the link would normally be unique (as in different news articles, etc) but if it is used to target a part of a web page (even bookmarked like #bookmark4) where the content may alter then it is no good. More than one item using #bookmark4 will cause items not to be downloaded by readers.

As I have found if the LINK points to a location where I have altered the content then the link itself must be unique to allow the item to be accepted as a new item by most readers.

eg: siteurl/webpage.html#23

where the hash designates a bookmark and the 23 is just the item's number within the RSS file. This is enough to land the link on the required page and give uniqueness to the items. For good measure I use a unique GUID for each item just in case it ever gets used!

This method works fine with Thunderbird and RSS Owl.
::::::::::::::::::: Update :::::::::::::::::

Bug report (BugZilla 260745) appears to be fixed (tested TB 3.1.16)
Test file now at -
www.wanderinghippo.org.uk/rss/test.xml





Wireless/Networking
 
Aug 24, 2005
 
23


Using Intel PRO/Wireless 2200BG card with Ethereal and Air Snare utilities

Close ...
Setup:

IBM Thinkpad R51 with Intel PRO/Wireless 2200BG card
XP with SP2
ZoneAlarm 5.5.062.004

2200BG driver - Intel v 9.0.2.25
Ethereal v 0.10.12 (network sniffer/analyzer tool)
WinPcap 3.1 beta 4 (low level device driver for network applications)
Airsnare (a wireless intrusion/monitoring utility)

Initially tried the latest Intel driver since the IBM drivers always lag behind quite a few months and the Intel driver works in the Thinkpad just fine - I use the Intel client to control the wireless rather than the Microsoft or IBM Access one.

However I still could not get Ethereal or AirSnare to collect packets via the wireless card. It seems that the consensus is that the 2200BG card will not support promiscuous mode and Ethereal/WinPcap/Intel/Microsoft all point the finger at each other.

A workround is however possible that will allow AirSnare to work and Ethereal to collect packets in non-promiscuous mode

This site here held the key to an answer -

http://www.micro-logix.com/WinPcap/CardReport.asp?ID=29

The DIVX video gives a quick run down on creating a network bridge that will allow the apps to work however with SP2 installed the procedure is slightly different to that outlined in the video.


1) Open Network Connections
2) Select both the Wireless connection and another (I chose the 1394 Firewire since I don't normally use it) and then right-click and Create Bridge (this will take a few moments)
3) When the bridge is created go into properties and allocate an IP (if using DHCP this will happen when a Disable/Enable is done)
4) I found the wireless connection fired back up again under Intel control and browsing was OK however in order for the bridge to appear as a device in AirSnare or Ethereal I needed to restart the XP

After this procedure it was possible to monitor with the 2200BG card but I needed to disable the promiscuous mode in the options. It is only possible to monitor traffic for the particular PC and broadcast traffic. Other traffic would require the promiscuous mode to be working correctly.

UPDATE:

Using -

2200BG driver - Intel v 9.0.4.8 (IBM issue)
Ethereal v 0.10.14 (network sniffer/analyzer tool)
WinPcap 3.1 (low level device driver for network applications)

Found Ethereal monitored the 2200BG card OK without a bridge being required but still no promiscuous mode.

A bridge was created and the 2200BG card still worked OK with Ethereal. The following (MS KB 302348) was tried to enable promiscuous mode on the card but didn't work -

In the XP CMD window

>netsh bridge show adapter

should give ID and status of adaptor

>netsh bridge set adapter x forcecompatmode=enable (where x is ID)


sets the promiscuous mode

Not surprising really since Intel still don't support promiscuous mode on the Centrino adapters but it is always worth a try - at least Ethereal and WinPcap seem to be doing the right thing now.



112 Posts. Displaying page 4 of 12.
Previous Next

Go to page: