Growing up is optional

I’m working on the redesign of a site which uses DreamWeaver. This is a very large site (1000+ pages) to still use file based content management. And Dreamweaver MX2004 (I don’t know about other versions) doesn’t allow you to ‘Apply template’ to multiple pages at once.

…but I found a workaround. It goes like this:

You have a whole lot of pages with the template ‘oldtemplate.dwt’ and you want to convert them to ‘newtemplate.dwt’

1. BACKUP – you might break things
2. Rename ‘newtemplate.dwt’ to ‘newtemplate.dwt.bak’.
3. If you already have some pages in ‘newtemplate.dwt’ DW will ask you if you want to update links. Select ‘Don’t update’
4. Rename ‘oldtemplate.dwt’ to ‘newtemplate.dwt’
5. DW will ask you if you want to update links. Select ‘Update’
6. Delete ‘newtemplate.dwt’ (what was ‘oldtemplate.dwt’)
7. Rename ‘newtemplate.dwt.bak’ to ‘newtemplate.dwt’
8. Run ‘Update Pages’ for ‘Files that use’ ‘newtemplate.dwt’

If you have named all of your editable regions the same this should just happen. If not, you may have to use the dialogue box to point the old region names at the new ones.

I’m working on my Masters of Management through Massey University. I’m currently trying to put together a literature review of my chosen topic. However, I am having trouble locating any academic or even practicioner research in this area. My topic is (verbatim from my draft research proposal):

Determinants of effective website management in New Zealand government websites.

How are websites managed and how does this affect their effectiveness in achieving the outcomes for which they are maintained?

In this research I will examine different types of government websites (e.g. informational, educational, transactional) looking at how they are managed (including governing structures, authorship models, development models, history), and determine if there is a relationship between these variables and their effectiveness.

I have narrowed the topic to NZ govt sites just to stop the research getting too huge, but any research on website management with a sound methodology in a different domain (e.g. ecommerce, activism) would be a good start for me. I’ve hunted around all of the academic abstract databases etc I can think of but with no luck. Everything I have found is focused on the technology and functionality, rather than the people side of website management.

Do you know of any research, theses, papers or anything else remotely related to this topic?

Long time since I posted; been too busy

• Going to Webstock, met some people I’ve chatted with online for a while. Putting faces to names. Heard some legends. The heads of the two biggest browsers are speaking tomorrow. (IE, firefox)
• My new board arrived (very late)
• Been cited in an academic paper for the presentation I gave on Semantic HTML, and it’s on page 1 of Google
• Soccer season is up and down

I wrote this a while back. What y’all think? I disagree with a lot

A while back on the WSG mailing list someone asked about correct markup for legal documents. Having worked for the Department for Courts/Ministry of Justice I had often thought about this, so it was of some interest to me.

Simply put, legal documents, such as judgments have numbered paragraphs throughout, with heading spaced in between them e.g.

<h3>Section 1</h3>
<p>1.  Paragraph 1</p>
<p>2.  Paragraph 2</p>
<h3>Section 2</h3>
<p>3.  Paragraph 3</p>
<p>4.  Paragraph 4</p>

While this displays alright, semantically speaking its not very nice, and is a pain if you decide to move a paragraph from one section to another while editing the document. However, it does have the advantage that once a documment is finalised the numbers are tied to the content.

Next I thought what about

<h3>Section 1</h3>
<ol>
<li>Item 1</li>
<li>Item 2</li>
</ol>
<ol>
<h3>Section 2</h3>
<li start="3">Item 3<li>
<li>Item 4</li>
</ol>

While I think it’s better than the first example, if you want to use it you are forced to use a transitional doctype as start is not a strict attribute. Also, it has the disadvantage that the numbering is not inherently tied to the paragraph, which it the big plus of the first example.

In the CSS2 Recomendation, W3C have included automatic numbering. Unfortunately UA support is limited

What do y’all think?

Will we see commercial/enterprise Gmail? I was thinking the other day about the whole Web as a platform thing and how Google, with their huge platform is going to utilise this. What I though is that one of the next big things Google will do is look at the possibility of commercialising the ‘killer app’.

Google have already proved they can take email from the desktop platform to the web platform. Personally I prefer Gmail to Outlook or Thunderbird. However, for a commercial user, one thing they have over Gmail is the scheduling/calendar functionality, although this is not yet at version 1.0 in Thunderbird. Also, I don’t know whether Google are working on this , but I suspect that someone there has the foresight to at least assess its value. We may even see it as part of a ‘Gmail Pro’ targeted at the serious, but not necesarily enterprise user.

However, there is the issue of privacy and security. Google would have to ensure that these were well sorted out before any commercial organisation would seriously consider using it.

Administration could be an interesting issue for an enterprise version. Although keeping it simple is best, I can hear the cries already of ‘but I need to change this to stop our users…’

I’ve got the Yahoo! API powering the search on my work’s website. The programming was easy. Using it is easy. The change process is hard.

I’m using PHP to convert the XML into a multidimentional array, and then looping through the results. To limit the results I was appending site:http://mysite to the search string.

The one difficulty I had was with handling big search strings with lots of OR operators. It seemed to drop the end of the query with the site: parameter and our site would then serve results from anywhere. To get around this I prepend it to the query. Also I added an extra check to confirm that the results are actually from our site.

Results are very good. Thanks Yahoo!

Google has been inproving the depth of its non-US and non-UK maps. You an now see down to about 10km across for New Zealand and other parts of the world. Its’ far from a cartographers dream, but you can see major roads near my house, the airport etc. Zooming out a bit you can see Kapiti Island off the coast. heading north up the coast you can follow the Whanganui river inland to Mount Ruapehu. Going further north you come to one of my favourite spots; Mount Maunganui, one of the best surf beaches around.

One of the thngs I like about my job is solving problems. In particular problems where there is a bit of thought involved. The particular problem I had is that for a search interface I’m building a requirement is the ability to page result set. This is easy using MySQL as it has a LIMIT keyword that limits the results to a certain number, and allows an offset, so for example you to retrieve results 16-20 you would use LIMIT 15,5.

Unfortunately, SQL Server does not have an equivalent to the offset argument. Its nearest is TOP n, which returns the first n rows.

This puzzled me for a few days, but then the answer came to me in a flash of inspiration: Results 20-16 would be the first 5 rows of a TOP 5 statement, with a nested query with a TOP 20 statement in the opposite order. Then I could just reverse the order to get the result the right way around.

So here’s what I came up with:

SELECT * FROM (
	SELECT TOP x * FROM (
		SELECT TOP y fields
		FROM table
		WHERE conditions
		ORDER BY table.field  ASC) as foo
	ORDER by field DESC) as bar
ORDER by field ASC

x is the number of rows you want returned and y is x+offset.

After looking at this in Query Analyzer, it appears that the extra nesting does not add very much to the load. For the queries I was running it was only about 2%.