I have just noticed a very annoying bug in Live Writer, thus why you have have received multiple copies of my last posts. After I had posted those last 2 articles, I noticed they both had the same images, even though they clearly didn't when I posted them. It seems that if you paste in an image from the clipboard Live Writer will name it image.png by default and the thumbnail will be image_thumb.png, it will not create a unique filename, thus will simply overwrite any existing images with the same name, thus messing up all your previous blog posts with images not to mention if you have multiple images in your current post, they will all end up as the same image.

I presume this bug must have been added to the latest release (2009) as I have not noticed it previously.

I have however found the following temporary fix on the Windows Live Writer Blog

Open HKCU\Software\Microsoft\Windows Live\Writer\Weblogs\{blog-id}\UserOptionOverrides\, where {blog-id} is a GUID. You will have several of these, but should be able to tell the right one by looking at the contents of the key.

Add a new String value with name “fileUploadNameFormat� (case matters!!) and the value e
{WindowsLiveWriter}/{PostTitle}/{Randomizer}/{AsciiFileName}

hopefully they will fix this annoying bug very soon.

On my windows 2008 server (64bit) I started having a problem whereby the wmiprvse.exe process was consuming up to 40% of my CPU (22% average) and causing major slowdown of my system.

I managed to narrow this down to the "Windows System Resource Manager" and if I stopped this service, then the problem would go away. None of the solutions I found on Google applied to my scenario or my OS, which is one of the issues I guess I will face for running windows 2008 server as a workstation.

Anyway after much searching, I discovered the following hot fix solved my problem, despite the fact that the symptoms or cause do not apply to my scenario at all as I am not doing any monitoring from an external machine. After installing this hot fix, my average CPU usage for wmiprvse.exe is now around 10% and only seems to hit about 20% max, so it is an improvement. Of course if you do not need the Windows System Resource Manager then you could just disable this permanently.

http://support.microsoft.com/kb/970520

Note that the hot fix says it is for Vista, but it does also apply to windows 2008 server.

For those of you playing with Windows 2008 server, Microsoft have released another funky new tool for IIS7. The IIS Search Engine Optimization (SEO) Toolkit helps Web developers, hosting providers, and Web server administrators to improve their Web site’s relevance in search results by recommending how to make the site content more search engine-friendly.

The components of IIS Search Engine Optimization Toolkit include the following features:

• Site Analysis:

  • Fully featured site crawling engine - in order to perform detailed analysis of site's structure and content, Site Analysis tool uses a built-in web crawler, called "iisbot

    ", to download and cache all the publicly available web site content. The web crawler is fully compliant with robots exclusion protocol.

  • Report summary dashboard - the results of site analysis are presented in an easy to use dashboard page that serves as a start page for various types of analysis. In addition this page includes a large set of pre-built queries for most common reports.

  • Query builder - Site Analysis tool includes a powerful and flexible query builder user interface that lets you create any custom queries that are run against the cached web site content.

  • Detailed URL information - you can obtain various detailed information about every URL in your web site, such as response headers and content, the pages that link to that URL as well as all the referenced URLs.

  • Detailed Violations descriptions - each content or SEO violation found on a web site has a detailed description as well as a recommended corrective action.

  • Word Analysis - any web page can be analyzed with regards to the most commonly used words and phrases within the content of that page. The results of that analysis can be used to select the keywords that most accurately describe the content of the page.

  • Route Analysis - unique routes to any page can be displayed in a separate report. This kind of information helps better understand how search engine and site visitors reach a particular page on your web site.

• Robots Exclusion

  • User interface for editing robots.txt file - the content of the robots exclusion file - robots.txt - can be edited by using IIS Manager GUI

  • Selecting URL paths from physical view of web site - the paths that are specified for "Allow" and "Disallow" directives in robots.txt file can be selected from the physical file system layout of your web site.

  • Selecting URL paths from virtual view of web site - the paths that are specified for "Allow" and "Disallow" directives in robots.txt file can be selected from the logical view of your web site obtained from the results of site analysis.

• Sitemaps and Sitemap Indexes

  • User interface for managing sitemap and sitemap indexes files - the content of the sitemap and indexes files can be edited by using IIS Manager GUI

  • Selecting URLs from physical view of web site - the URLs that are specified within a sitemap can be selected from the physical file system layout of your web site.

  • Selecting URLs from virtual view of web site - the URLs that are specified within a sitemap can be selected from the logical view of your web site obtained from the results of site analysis.

Articles

• Using IIS Search Engine Optimization Toolkit
• Using Site Analysis to Crawl a Web Site
• Using Site Analysis Reports

Forum

• IIS Search Engine Optimization Toolkit Forum

As you will see elsewhere on my blog, I have recently setup windows 2008 server 64bit as my new desktop. For the most part everything has worked ok, but there have been a few interesting quirks.

One of these quirks was that any new site I setup in IIS7 resulted in a "500 internal server error" identifying the AboMapperCustom wildcard application mapping for ColdFusion as the cause.

Now my Default Web Site from where I run the CFADMIN was working fine, so I was rather perplexed as to why other sites did not work when they seemed to be setup exactly the same.

Now I am actually using DotNotPanel on my local machine to setup and manage sites, so this may be a contributing factor, but still the cause of the problems may affect others, so hopefully the solutions may help.

Using the Web Server Configuration Tool I tried to disable then enable CFML on the server several times to no avail, then when checking the Handler Mappings" for the site, I noticed that the wildcard application map for CF was actually there twice, which I am sure may well have been a contributing factor, so I removed one of them. I also then noticed this alert on the handler mappings page.

This site is in an application pool that is running in classic mode, 
so you can manage ISAPI extension and native modules that are mapped to 
paths. You must manage managed handlers (system.web/httphandlers) directly 
in the configuration files.

Now if I recall "classic mode" means 32bit, so this seemed likely to be causing a problem, although I am not sure how I ended up using this application pool by default.

Anyway to change this, open the IIS manager, find your site, right click and select

Manage Web –> Advanced Settings

And change the application pool at the top to "Default Application Pool". In my case this fixed the problem and the alert went away, hopefully this may help someone else having issues with IIS7 and CF.

This was a very odd problem I had on a clients site this week. Whenever he used the <CFAjaxProxy> tag on a page, the page simply stopped rendering at the point where the tag appeared. No error appeared on the page or in the ColdFusion logs.

The reason no error occurred turned out to be caused by the application.cfc, he has an OnError function that was doing a cfabort.

<cffunction name="onError" returnType="void" output="false">
        <cfargument name="exception" required="true" />
        <cfargument name="eventname" type="string" required="true" />
        <!--- <cfdump var="#ARGUMENTS#" /> --->
        <cfabort />
</cffunction>

Once I got rid of this problem, the following error appeared.

Security: The requested template has been denied access to C:\ColdFusion8\wwwroot\WEB-INF\cfclasses\cfcheckUsername2ecfc1070071758.class.
The following is the internal exception message: access denied (java.io.FilePermission C:\ColdFusion8\wwwroot\WEB-INF\cfclasses\cfcheckUsername2ecfc1070071758.class write)

So yet another path that needs to be added to the sandbox for every site in order for Ajax to work. Really there is no reason why this should be so as the classes should be created by CF internally and no special permissions should be required by the application.

The number of paths now required in each sandbox for CF8 in order for all tags and functions to work correctly is ridiculous. I have enlightened Adobe how things should work in shared hosting environments and all the paths that do not get inherited by sandboxes when applied at a root level, so here's hoping that CF9 will finally be shared hosting friendly.