Marketing mixed with IT, design and coding, all these things are part of a webmaster’s job.
At least they were.

I am not a server administrator, but 7 years ago that was still part of the webmaster job. This is impossible in todays world and most webmaster jobs nowadays focus on the communication side. Communication and marketing, usability and user experience: these are are all the hit words of 2011 and you have to have them on your CV.

One thing is very clear about what a webmaster does: handle communication between business and IT. You have to be able to talk to a programmer/sysadmin as well as talk to one of the board members.

Can you handle a sentence that consists of nothing but acronyms?
Can you say one that a business person doesn’t understand, but IT does?
Can you do the same for the IT person?

Hiring people is not my job, but I am a webmaster and I was asked to create a questionnaire to check the technical knowledge of the webmaster candidate.
The questionnaire focuses on 4 areas:

• Marketing / SEO / Communication
• Web development, client side
• Usability / Accessibility
• Web development – server side

The 4 fields are a bit crammed together, I know, but you need to limit the number of questions a bit. Below are the questions and answers with some remarks. Candidates took between 45 minutes to 1 hour to answer all of them.

The questions are both specific and open. Not everything (definitely no server setup questions anymore) is covered but it is a good start to measure the technical knowledge of you webmaster candidate.

All those results are pages about “click here, do this, do that”. Why do people still use that on a website?

“Click here” is one of those text fragments that seem to stick around. Webdesigners and usability experts recommend not using it and still, everyone does, including Adobe. Adobe Acrobat reader is the first item on Google SERPs for “click here”. Simply because it is used in plenty of links on a huge number of sites.

All I can see is that people are using it, and that there is no tendency that tells me it is dying.
So why not? Why do people still use it? Dismissing it as stupidity or lack of knowledge is too easy. There must be some reason why people still use it even though there are plenty of reasons not to.

Content-editors still use it because it’s clear. It is clear what a visitor must do. Not what the visitors will get after the click. But that really doesn’t matter because through all the noise (the internet) this person found your website, scanned it thoroughly, found what he was looking for, and followed the instruction. What’s behind the link, who knows? Does anyone care?

The text fragment works more or less like an embedded command: “When you click this link, you will know more about the subject than others do.” The link becomes more attractive: it states what you should do.

People scan for underlined blue text. The link text “click here” enforces the concept of the link and the click becomes a jump into the unknown.

To click is to believe.

I have trained around 200 editors and I repeated it a thousand times:

• that the text of the link should be an indication of the subject of the referring page,
• that it is is not good for usability/accessibility and SEO,
• that it is uninformative

“Click here” as link text is not going away in the near future even though it is disregarded as bad practice by practitioners of IA, usability and SEO. The concept is too easy to understand. It almost feels natural.

Links should be informative about where you’ll take the visitor, but it seems that using verbs as links is very powerful.

Tell them what to do

And then you get this facebook button: “like”. It’s the same thing, just better. Emotional. The link instructs. Like or die clicking!

The facebook like button instructs you to like – read click – the button. It doesn’t say “I like”, it doesn’t say “click here to like”, it’s just “like”.

Implicitly I translate it to “I like”, as must people do (disagree with me ?). But it is nothing more or less than an embedded command: like.

The simplicity of the concept “telling people what to do” is definitely the main reason why the “click here” link is not fading away. Even though the text fragment is pretty powerful, the reasons not to use it still prevail. Click here to comment.

XSLT was the main reasons why I started with Umbraco way back in 2005 when Niels Hartvig was surprised that someone actually wanted to come to their conferences. Things have changed since then. And now I am hearing voices about dropping XSLT in Umbraco and inline C# code mixed with HTML, what is this?

Do we really want to go back to spaghetti code? I really hate to see Umbraco go down the path of so many module based, spaghetti code producing CMS‘s.

Razor, the new inline coding style in Umbraco, seems interesting, simple and easy to learn. People have said that simplicity is the ultimate sophistication but is Razor really the way forward?

I don’t know. There are so many reasons to love and stay with XSLT: it is verbose ( if you prefer obscure syntax rules try Perl ), it creates clean and valid code and it’s fast.

XSLT is too verbose.

Is it? I took the liberty to rewrite the functionality to display a paged list of articles found on the cultiv.nl blog. The Razor code is around 60 lines, but XSLT only needs 30 lines. At least if you use XSLT for what it is meant to do. If you find yourself in a loophole, the first thing to do is stop digging and look at what you need.

• Solution 1: 29 lines
• Solution 2: 28 lines

Have a look at the presentation from Chriztian Steinmeier, learn how to call and apply templates, learn XSLT grouping and design patterns, learn how to use keys and your life with XSLT will be so much easier.

XSLT is faster than your shadow

Some say Razor is slower. I don’t know about that but I do know that I solved several performance problems by using XSLT instead of usercontrols.

If you have to create a navigation bar and check each page against 8000 groups to see whether the user belongs to one of these groups, you will love the speed of XSLT. In such a case it is much easier to select what you need with XSLT, and compare against the imported access.xml file than iterating over each membergroup in C#. Other examples are the XSLT search from Douglas Robar and the many XSLT related projects on our.umbraco.org

XSLT creates clean and valid code

Looking at the Razor examples it is easy to create invalid code. XSLT enforces valid code, and you really have to jump through a lot of hoops to write invalid code with XSLT.

True, valid code doesn’t matter for the end product, not for SEO and not even for your visitors, but having a consistent presentation tree will save you from a lot of cross-browser headaches and css bugs. Write clean code. Save hours of debugging. Eat more fruit.

The best way out of a difficulty is through it.

There are 2 main rasons for dissing XSLT that pop up regularly on the forums. The first one is that XSLT is difficult to learn. The second reason is basically a result of the first one and boils down to using procedural programming concepts in a declarative language (thinking for else loops instead of apply template). Programmers who are not familiar with XSLT have a hard time learning how to use XSLT and as a result, turn to what they know. It’s a vicious circle.

Imagine yourself spending all those hours you spent on learning OOP logic and procedural programming, learning XSLT instead. I am guessing you would have a hard time writing C#. Don’t diss XSLT because you don’t understand it. The more you learn the easier XSLT becomes.

The numerous questions on the internet and on the umbraco forum on how to do for loops already says enough. If you are thinking “loop” in XSLT, forget it. It’s going to be annoyingly verbose, slow and difficult to maintain. It is not about how ( if then that else this) you are going to get it, you already have everything, you just have to select what you need.

Let it be a warning. Every-time you think: why the f#$k can I not increment that counter, take a step back and realize that you are going the wrong way. Remember that if you only have a hammer, you tend to see every problem as a nail.

Isn’t it necessary to look at what XSLT really is? Why it was chosen in the first place, and whether we really need to go yet another way of doing the same thing?

So I looked around, found a few articles on the subject, and I bought a book.
The book “The Elements of Content Strategy” written by Erin Kissane published by A Book Apart.

I devoured the book: short and concise writing and it provides a clear picture of the relationship with other functions in the webteam (editors, IAs, designers, etc..). I was happy to discover that the section of useful resources has a list of books that I had read or own:

• Information Architecture for the World Wide Web by Lou Rosenfeld,
• Don’t make me think by Steve Krug,
• The Elements of User Experience by Jess James Garret and
• Managing Enterprise Content by Ann Rockley.

Seems I should have some idea of what content strategy is, right?

The book explains a few things about Content strategists. Basically a CS thinks about the who, what, where and why of content.

Who will publish that document?
Editors need to be trained, and the document needs to go through a process, formal or informal, to make sure it is properly published. Document life-cycle is part of what a CS does, but that is also what IA’s do, right? Designing workflows in a CMS, how to reuse content, which content to reuse, that is also part of the IA tasks.

What are you going to publish?
Data and content are pretty abundant if you know where to look for. But that doesn’t mean everything has to go online. Here the tasks of a CS and IA start to differ in that the IA does not always have a grip on the actual content.

Where are you going to publish it?
Content can be reused and rewritten for different channels. If you are posting it on twitter, Facebook, an RSS feed or a webpage, the CS considers which parts of the text goes online. Maybe even the tone and voice of the content has to be reviewed.

I think here the CS tasks start to get borderline with SEO tasks and/or journalism. One good examples of this overlap can be seen at the BBC and how they produce headlines. Most of the headlines convey the whole story in 5 words. Is that good copywriting? a content strategists defining standards? or the SEO requiring the most prominent keywords to be available in the titles? I guess journalists have been doing this already for a long long time.

Why are you going to publish this?
The strategy, the why behind the content should be become clear after stakeholder interviews and audience analyses.
The strategy should bring those 2 mindsets together. Here the idea is the same as in IA: align the mental model of the business with the mindset of the audience.

In chapter 3 the author lists a batch of deliverables (Channel strategy , CMS requirements, Communication plans , Usability tests) and I think the majority of those documents need to be shared with the IA.

Both the the IA and the CS perform user research, create content inventories, write up persona’s, the shared list of tools goes on and on. But I think the content strategists focuses more on the message behind, the voice and tone of the content and for that matter I think the CS is very close to a graphic designer. Choosing the colors and images can drastically change the messages that a website conveys. The tone and voice of the content can do the same.

A wireframe is a black and white presentation of what will be published where. But the message can be so dramatically changed by choosing the wrong words or the right ones. The wireframe can act as a reference but the if the design uses pink tones and cartoon images, or corporate blue and stock photography, the message is no longer the same.

Content strategy is nothing new, it has been done from the moment books were published, radio started to broadcast and television shows were brought into the living room. Yes, the environment has changed, so the tools and techniques need to change too but basically it is a new name for what was considered part of other professions.

Update

An excerpt of the book The Elements of Content Strategy is now online at A List Apart

I used macports before for this sort of thing but I ditched them in favor of MAMP. But not without any disadvantages. DBD::mysql doesn’t install nicely when using MySQL used by MAMP.

The new MacBook is running Snow Leopard which comes with several installations of Perl, the default being version 5.10, 64 bit. MAMP runs a 32 bit MySQL version (the one I run is 5.1.44 ).

The install problem is a result of the difference between the SQL versions and some missing files.

These are the prerequisites:

1. Mac OS 10.6 Snow Leopard
2. Developer tools installed
3. Perl v5.10, 64bit (verify with terminal: perl -v should give you something like this /This is perl, v5.10.0 built for darwin-thread-multi-2level/)
4. MAMP

First install the DBI bundle.
perl -MCPAN -e 'install DBI'

Before we continue and try to compile the DBD::mysql module, we have to add the necessary libraries to the MAMP MySQL folders. Since we are using 64 bit perl we need to download the 64 bit version of MySQL, one that is the same as the one we have running with MAMP (I used version 5.1.54).

Unpack the archive and copy the following folder into the MAMP folders:
include -> /Applications/MAMP/Library/

Copy the contents of the following folder:
lib/* -> /Applications/MAMP/Library/lib/mysql/
Don’t overwrite the existing files and folders.

As suggested in the comments, instead of the copying everything, you can also make symlinks and run the following command:
ln -s /Applications/MAMP/Library/include /usr/local/mysql/include
ln -s /Applications/MAMP/Library/lib/mysql /usr/local/mysql/lib

If you got all that, open up terminal and cd into the expanded DBD::mysql module directory and issue the following command:
perl Makefile.pl
--mysql_config /Applications/MAMP/Library/bin/mysql_config
--testuser root
--testpassword root
--testhost localhost
--testport 8889
--testsocket /Applications/MAMP/tmp/mysql/mysql.sock

Please double check both the path to mysql_config and mysql.sock. The variables set here are so different from the common installation of MySQL that many people fail in getting this module to work properly with MAMP.

The following commands should be a known sequence:
sudo make
sudo make install


Were are almost done: create a mysql directory in /usr/local:
sudo mkdir /usr/local/mysql
sudo mkdir /usr/local/mysql/lib

In that directory copy the missing libraries
cp -r /Applications/MAMP/Library/include /usr/local/mysql/
cp -r /Applications/MAMP/Library/lib/mysql/* /usr/local/mysql/lib

And you are done. Test your configuration with the following simple connect script:
#! perl
use DBD::mysql;

my $dbh = DBI->connect(
'dbi:mysql:test:localhost:8889;mysql_socket=/Applications/MAMP/tmp/mysql/mysql.sock'
);
$dbh->disconnect();

If you have, then you probably found out that running Ant on a Mac just doesn’t work as expected.

Most common tasks like FTP and SCP are not available by default. And looking around on the internet, you are not the only one expecting these tasks to just work.

It seems that many resort to downloading and compiling Ant from scratch. After that they add the libraries to do common tasks like FTP and IF/THEN/ELSE structures.

But it’s just as easy as using the buildin Ant from Mac. You just have to add the necessary jar files
into the necessary directories and it will just work (hopefully :-).

To make life easier for you and me, I created a build file that installs the most common tasks for ant. Download the file and run ant against the anttasksformac.build:

ant -buildfile anttasksformac.build

Let me know if something is not working for you.

Update

When running the SCP task, it hangs. So I checked and there is a conflict in the jsch jar, so I replaced jsch with version 0.1.29 and everything seems to work.

The package provides a toolbar to editors who are logged in to add sticky-notes to any page in a website. It provides a way to communicate about changes and remarks on a page.

As many other Umbracians, I have several bookmarks that help me debug Umbraco web-pages. I thought if I add them to this package, I will have them available at all time, no matter on which PC I am working. The toolbar has buttons to debug the page, and to provide the list of data available in the XML cache.

Credit where credit is due: please have a look at the service ProtoNotes, as the idea of the design for having notes or comments on the page is coming from that services. No code has been copied, but the interface is somehow based on the ProtoNotes service.

See how it works

Screencast showing how to install and use Webstickies for Umbraco

Try it

If you don’t know how to install a package, the community website has a small how to.

Files to download

• Project website: report bugs, ask for features
• Download the Umbraco package
• Get the source code from Beanstalk

Sorting is ordering data: alphabetically, numerical, by day, week (by number again) and/or month.
All these sorting orders are based on random lists. By random I do not imply a mathemacical random list, but just the fact that we assigned an order to words, and because everyone knows the order, the list becomes a functional tool.
Most people do not have to think about whether a 4 comes before a 7, or that the month of January is the first of the year. The fact that we all have a complete image of these things in our mind, makes them so practical.

But does it make sense for the navigation of a website?

In a website the navigations is about information (that is not everything, I will get to that), which is something different then data.
I had a look at the dictionary entries:

Information: the communication or reception of knowledge or intelligence

Data: factual information (as measurements or statistics) used as a basis for reasoning, discussion, or calculation

The 2 definitions tell me that data is used to generate information. Information emanates from data, it is the basis on which we can form a decision or some reasoning.

Well, if we apply this idea to the navigation of a website, we should be putting organised information on the website and that information should be coming from the data we collected as a basis to create the website.
Sounds good, but what about the visitor? People could still argue that sorting can be applied to information to ease scanning of the list, to guide the visitor of the website.

Guidance

The argument to ease scanning is actually a valid one, though not good enough, I think. People scan pages, they tend to skip most of the things available on a website. And now it comes down to what you want as a webmaster, do you want people to see the items that start with an A first, or do you want a visitor to see things that are important first?

I guess the answer should be clear, the important things should go on top and the order should be the order of importance. Importance, of course, can be defined in different ways ( from a user perspective or from a business pespective) but it certainly is not following the order of a very famous random list. If you want to guide the visitor, you should give them the most important bits of information first.

Translations

The reason why I came up with this subject is that I had an argument with a colleague after I noticed they had sorted the list of services by alphabet. A first, B second, etc… you know :-)
Now, the thing is that the most important service started with an A, and that the manager in charge is responsible for that product. In this rare case, the alphabetical sorting order matched the order of importance (from the business point of view).

This is an exception and will not happen very often, but lets look at this example and see what might happen in the future. The service is an international service in Europe and it has to cather for the south of europe. In my experience with international websites that try to reach the public in the south of europe, you need a translation in the local language.

Ok, you already know what is going to happen to your nicely alphabetical order, right?

Once you start creating multi language websites, you have no way of ordering a list alphabetically and keeping the order of the items in the list.
I wrote about this before, as I was faced with a sorting problem of a list in Chinese

When everything else fails

When your website navigation fails, and your visitor cannot find what he/she is looking for, an alphabetical list of keywords, a sitemap where the subelements are ordered alphabetically, and/or a monthly list of posts on a blog, are valid examples of pages where you can alphabetically sort information. But that is when everything else fails …

I am sure there are other reasons why it is bad to sort you information as you do with data on a website, but for now this is what I experienced.
Please leave a message if you had similar experience with this kind of problems.

The code is provided as is, no garanty, blabla … though should not harm anyone.

You can download the dll and adjust your xslt configuration settings in Umbraco or download the full VS2008 project [1.5 MB] and have a look at the code.
Just for the reference, below is a list of methods available in the class with a short description

XPathNodeIterator QueryDatabase(string conn, string query )

Query the umbraco database or any other database that you need.

string createMemberGroup(string memberGroupName)

Create a member group

string isMemberOfGroup(string memberGroupName, string memberId)

Check if the member supplied belongs to the group supplied

XPathNodeIterator GetAccessingGroups(int documentId)

Not working, will be removed …

XPathNodeIterator getAllMembersFromGroup(string groupId)

REturn all the members that are in a group

string getMemberGroups()

Returns the list of groups the current memner belongs to

XPathNodeIterator getAccesXml()

Returns the full xml document that contains the list of protected pags and groups.

string getMemberFromFirstLetter(char letter)

Well, get member from the first letter

XPathNodeIterator getMemberFromFirstLetterXml(char letter)

Well, same, but then in xml.

string addMemberToGroup(string memberGroupName, int memberId)

Add a given member to a give group

string addPageToGroup(int currentPage, int groupId, int loginPage, int errorPage )

Protect a page with a membergroup

bool updateProperty(int DocumentId, string propName, string propValue)

Update any property (not the umbraco properties) for any given document.

Download

• only the dll
• the full xslt extension VS2008 project [1.5 MB]

They (as in your wife) don’t give a flicking flick about your stupid religious enthusiasm for making web browsers which conform to some mythical, platonic “standard” that is not actually implemented anywhere. They don’t want to hear your stories about messy hacks. They want web browsers that work with actual web sites.
Martian Headsets | Joel Spolsky

And besides he has a point, he is also a very funny writer:

FireQx 3.0 treats snow as rain, because you need windshield wipers in the snow, Qxyzrhjjjjukltk 5.0 does not, because the programmer who worked on that feature lives in a warm part of Mars without snow and doesn’t have a driver’s license anyway. Yes, they have driver’s licenses on Mars.Martian Headsets | Joel Spolsky

If the last quote doesn’t make any sense to you, go and read the article: Martian Headsets by Joel Spolsky.