I’ve been using the Google Analytics plugin, created by Matt Labrum, a while, but it registers also every page view by me. That will give the wrong statistics, and I don’t want that. So, I modified the 0.3 version of the plugin into my own. This is the modified version, which can be set up to hide the Google Analytics code for user levels above a certain threshold. I also added Analytics for outgoing comment author URLs.
Rename the file below and put in your WordPress 2.0 plugin directory:

• googleanalytics.txt

No problems encountered so far when upgrading from WordPress 1.5 to 2.0. I love the new admin theme! It makes better use of the screen in “Write Post”! The WYSIWYG editor is quite nice when creating simple posts. It seems like all installed plugins works too!

Goodies like explode, leftback, rightback, switchcase, implode, middle, trim, unique etc. can be found in this utility class.

The Commons Lang project at Apache Jakarta also has some really great stuff.

Chad Schelfhout gave me a correction for the implode(Object[], String):String method and the new method word(String, String, int):String.

• StringUtils.java
• CollectionUtils.java
• NumberUtils.java

I got it via the automatic update system, but it doesn’t mention it on the official website. I found a release candidate of 1.5.0.1 on the FTP site though. Strange…

Not so strange after all, it’s all explained in the “What the heck is with this 1.5.0.1 update?” article at developer.mozilla.org.

When reading through the Troubleshooting Lotus Domino hangs and crashes, i found some new (for me) interesting commands:

• show memory dump – dumps memory info to disk
• tell http debug thread on – writes extensive debug information to disk on every request

I am not Microsoft’s biggest fan, since I was a Macintosh fanatic in the old days, but had to convert into the Wintel platform for about 10 years ago. :-) But I have to admit though, when comparing Microsoft’s XML parser with the ones that are built into Notes/Domino, that their products sometimes are great!
When comparing Microsoft’s XML parser with the other two, I created an agent that parsed a big (over 700 KB) and complex XML document and read an element’s value. I did this 25 times per parser, and came up with this interesting result (average execution time per parser) when executed in the Notes client:

1. Microsoft XML Parser: 0.2 seconds
2. NotesSAXParser: 2.2 seconds
3. NotesDOMParser: 2.5 seconds

When executed on a Domino server, the result was a bit different, since the server is a bit old:

1. Microsoft XML Parser: 1.0 seconds
2. NotesSAXParser: 5.0 seconds
3. NotesDOMParser: 7.7 seconds

You might have noticed that none of the kanngard.net sites is working. That is because I have moved the server from my earlier “ISP” – thanks Ove for your bandwidth and power :-) – but the primary DNS entry has not yet been updated. In the meantime, if you wish to access those site, you have to alter your windows\system32\drivers\etc\hosts file and add for instance the following entries to access my sites:

85.11.47.181 johan.kanngard.net
85.11.47.181 dev.kanngard.net

Keep in mind that the workaround above doesn’t work with all proxies.

You could also use Internet Archive to see snapshots:

• johan.kanngard.net
• dev.kanngard.net

Note: all mails are transferred correctly, since I have the new IP in the DNS as ns1.kanngard.net…

In my earlier post today, I described a problem I have, where I wanted to get the XML string of a DOM tree. I’ve started writing my own routine, this is what I have right now. It’s far from complete, since it can not handle namespaces, CDATA etc. yet:
This is a new version that handles siblings better, and not getting out of stack space :-)


Public Sub serializeNode(node As NotesDOMNode_
, outStream As NotesStream)
On Error Goto catch

If node.NodeType = DOMNODETYPE_ELEMENT_NODE _
And Not endWritten Then
Call outStream.writeText("")
End If
Exit Sub
catch:
Error Err, Err & Lsi_Info(2) & ":" & Erl & "|" & Error
End Sub


I have a problem, that nor I nor any of my colleagues can’t solve easily. I have several classes that are passing objects/variables between them. On of them is handling Web Service request, and parsing the responses to a NotesDOMDocumentNode via the NotesDOMParser. The NotesDOMDocumentNode is passed returned, and are used in several places to extract information. But, I need to cache the response, so concurrent request with the same parameters are sped up.
To cache the response, I need to serialize the NotesDOMDocumentNode. At first, I thought this was easy, since there is a Serialize method on the NotesDOMParser. But no… You have to keep the original NotesDOMParser that parsed the raw XML data. The only solutions I can figure out are:

• pass both the NotesDOMDocumentNode AND the original NotesDOMParser at all times
• write my own Serialize method
• pass the XML string only, and do the parsing at each method. I think this would be really slow…

To further illustrate what I want, here is an example:

Public Sub testDOMParser()
Dim parser As NotesDOMParser
Dim inStream As NotesStream
Dim node As NotesDOMDocumentNode
Set inStream = session.createStream()
Call inStream.open("c:\temp\test.xml")
Set parser = session.createDOMParser(inStream)
Call parser.process()
Set node = parser.document
Call serializeDocument(node)
End Sub
Public Sub serializeDocument(node As NotesDOMDocumentNode)
' Here, I want to do something like this:
Dim outStream As NotesStream
Dim parser As NotesDOMParser
Set outStream = session.createStream()
Set parser = session.createDOMParser(node, outStream)
parser.exitOnFirstFatalError = True
Call parser.serialize()
Print "Serialized tree: " + outStream.readText()
End Sub


Anyone who have had any experience with this and are willing to share their thoughts? We are talking R6 here…

A shot of my home PC with comments after I modified it.