123. Plight

Paul Boag

In this weeks show we review Textmate and the Top 5 Tips for Web Designers and we discuss the plight of in-house designers.

Download this show.

Launch our podcast player

A quick request. We are really in need of some more transcribers to help with the interviews we do. The team we have are doing an amazing job but it would be great to spread the load.

If you feel you could help once in a while please drop an email to Ryan our producer and he will add you to the list.

News and events

SPAM meltdown

It is always with fear and trepidation that I mention HTML email. It inevitably leads to a torrent of comments ‘educating’ me about the evils of HTML in email, and that we should only use plain text.

Although personally I wish HTML email was never invented and try to limit its use, I do accept it is here to stay. Despite its many drawbacks it is statistically more effective than plain text from a marketing perspective.

You will be hard pushed to pursued a client to forgo HTML. Inevitably we will have to produce HTML templates occassionally. Of course, being conscientious, when we do produce HTML emails we want to ensure they look great and are well coded. This leads me to a couple of stories worth mentioning.

The first is that Patrick McNeil (of Design Meltdown fame) has launched a new site called Spam Meltdown. The site showcases examples of great email design in much the same way as Design Meltdown does with websites. Patrick has done an amazing job on this site and he has my sympathy because he is subscribed to over 1000 mailing lists! The designs he showcases are organised by style, colour, industry and topic. As with design meltdown this categorisation approach works really well. You can quickly find inspiration by looking at categories that are relevant to your project.

The second news item worth mentioning is that Campaign Monitor have updated their chart for CSS support in email clients. Campaign Monitor is a service which allows you to send HTML newsletters, but they do a lot more than just take your money. They are actively involved in improving standards support among email clients through the email standards project. Next time you are trying to produce an HTML email template check out their CSS support grid as it will clearly show you whether a particular CSS property is supported.

Form Analytics

While I am on the subject of cool services like Campaign Monitor, I also want to mention Clicktale. Clicktale is a service that allows you to track users as they move about your site and even anonymously record their actions. The last time I mentioned them this disturbed many people who saw it as an invasion of privacy. However, I see it as a valuable tool for learning about user interaction and improve site usability.

If you share my view, then you maybe interested in a new service they are starting to offer. You can now not only track users as they click around your website, you can also watch how they interact with forms.

In addition to video recording, the new form analytics service also provides three invaluable reports…

  • The time report – This shows how long users spent completing each field.
  • The blank report – This provides information on fields that have been left blank on submission.
  • The refill report – Which highlight fields that have been completed incorrectly.

If you run a site that requires users to complete long or complex forms then you will see the benefit of this service. On a high trafficked ecommerce site this would be invaluable, substantially reducing the number of users dropping out at checkout.

Art direction hits the blog

This week has seen the launch of Jason Santa Maria’s new personal website. For those of you who do not know, Jason is the creative director at Happy Cog (Zeldman’s company).

Normally, I would not mention the launch of a new personal website. However, Jason has done something very interesting. His new design is well executed but plain. It certainly is not as inspiring as his other work. The reason for this simple approach is that it is a framework upon which he will build.

The idea is that each of his blog posts will have a custom design to accompany it. The design will therefore reflect the content. In effect he is bring art direction to his blog. This is a bold experiment and something that Zeldman has written about before.

Although I am fully behind the idea of bringing content and design closer together, I do have some reservations. First, there is a possibility that the constantly changing design could make navigation around the site confusing. Fortunately from what I have seen so far that will not be the case. Jason has been careful to ensure key navigational elements remain in a consistent location and have similar styling wherever you are in the site. However, if other designers were to adopt this approach would they be so careful?

My second concern is a purely practical one. If each article not only needs writing but also designing, will that reduce the amount Jason posts? In other words is a blog really the right place for this type of art direction?

However, despite these reservations I am really pleased Jason is trying this approach. A personal website should be the place to experiment and try new things. Too many blogs (including my own) are cookie cutter solutions with some pretty graphics slapped on top. Its superb to see somebody doing something different.


My final news story of the week returns to a subject we have touched on recently. How do you wireframe a modern web application with its high level of interaction? In show 120 I mentioned that one approach might be to utilise flash. Today I want to point you at an article on the List Apart website, which suggests that building prototypes maybe better than struggling with wireframes.

When I first saw this article I was hesitant. After all I can barely pursued my clients to pay for wireframes let alone a full blown prototype. However, the more I considered what was being suggest, the better the idea seemed.

The majority of time spent getting an application working is spent on bug fixing, browser support and non-core functionality. The rough ‘outline’ of an application can come together very quickly. What is more, unlike wireframing, a prototype can be used as the basis for the final build. It does not get thrown away like a wireframe.

The article also points out that prototypes are better for demonstrating difficult concepts to clients. They encourage earlier collaboration between designer and developer, and provide something substantially better to user test against.

With almost every new website having some form of web application, we all need to consider how to better conceptualise their operation.

Back to top

Feature: The plight of the in-house designer

The more organisations I work with the more sympathy I have for in-house designers and developers. It is a role that can be thankless and isolating. How then can their lives be made that much easier? We discuss this in this weeks feature.

Back to top

Reviews: Textmate and Top 5 Tips for Web Designers

We have two reviews this week by our lucky competition winners Teifion Jordan and John McFarlane. Teifion and John will be going to this year’s dConstruct in Brighton.

dConstruct is the affordable one day conference for people designing and building the latest generation of social web applications. Tickets cost £125 inc VAT and went on sale yesterday so be sure to check it out.

Textmate by Teifion Jordan

Hi, I am Teifion Jordan, I am reviewing a program created by someone far smarter than me. I am going to be looking at Textmate. Textmate is a Mac only application though there is a similar editor called eText Editor for Windows.

First impressions of Textmate are that it’s pretty sparse, it looks like any other editor. I throw it a PHP file and it colours the text in, just like any other editor would. The colour scheme can be changed, both text and background colours can be altered, which is quite a neat touch. I can even make parts bold, italic and underlined which is a neat touch. It requires knowledge of Regular expressions but I can actually add in more rules for what to colour in! I used this to make variables used as array indexes appear differently, something I have wanted to do for some time. Not since I was a toddler, but definitely some time.

But enough moaning about how the program itself is both smarter and better looking than me, I wanted to try some code. I found that if I typed "foreach" in a PHP block and hit tab, I was presented with an entire foreach loop. Closer inspection revealed that there were dozens of snippets and commands for PHP and dozens more for each of the many languages and some things that were not languages. With 5 minutes of effort I had setup Textmate to post my blog posts for me, I am now one step closer to not having to put any effort at all into blogging.

It is possible to create your own snippets and not at all hard either. I now have one to tell me that I am beautiful and another to create a PostgreSQL query. I can also write new commands, I can write them in command line script, Python, Ruby and PHP to name a few. All of the commands are completely open sources, so you can see what’s already been done, and sort of plagiarise that sort of work for your own means. Except plagiarism is bad so don’t ever do it.

I can edit columns, I can write new snippets, commands and even entire languages, I can Regex, I can manage projects with a hierarchal file structure. It’s like before I was walking but now I’m on a push bike. I can’t make use of the ability to run down pedestrians until I learn how to do balance and pedal. Okay, the running down pedestrians was a bad example but anybody that is still listening and not calling the police must have understood it so I’ll continue. There’s nothing I can’t do in Textmate, I just need to look at the extensive online manual to learn it. And there I think is it’s biggest failing.

Textmate is a really lovely program to use but it’s so complicated. Coda, as a contrast, is a more intuitive application but it is to Textmate as a spade is to a chainsaw, that is, meant for a different problem and with fewer moving parts but also with the ability to digs holes? I’m sorry, my mind wandered. What I meant to say is that Textmate is great for dealing with code but not so much the design which is what apps such as Coda excel at. I’ve now been using Textmate for 10 months and I still think there is potential to unlock, though, that might be because I’m a thickie.

I suppose I should wrap this up by saying that I would heartily recommend anybody thinking about writing lots of code to give TextMate a good look. It takes a lot of time to get a lot out of it, but there really is a lot to get out of it.

Thank you very much for listening, I hope this was at least semi-informative

Top 5 Tips for Web Designers by John McFarlane

Hi, I’m John McFarlane and this is the first ever review brought to you live from my living room. Today I’m reviewing a post that has been submitted on the boagworld.com forum. The title is "Top 5 Tips for Web Designers". I’ve been reading through the replies and I’ve put together my top 5 top tips.

In at number 5 submitted by richquick, allow time and money for personal development, read blogs, buy books, attend conferences, experiment and learn new techniques and technologies.

In at number 4 posted by Jayphen, surround yourself with designers, whether they’re colleagues, real world contacts, online contacts, forums, podcasts. The more you talk about design the more you learn and I’d like to add to that e-mail designers for advice and let them know your experiences.

In at number 3 posted by some guy called Paul Boag, develop with the latest best practices, ensure you separate content, design and behaviour. Make sure everything you build uses progressive enhancements.

In at number 2 another one by Paul Boag, it’s an obvious one but one that can’t be put across more clearly, know HTML, CSS and javaScript inside out, you need to know the core technologies that underpin the web back to front. I’d like to add to this point, the basics of HTML and CSS are easily learnt but don’t be fooled into thinking that you know enough, you really need to know these subjects to an advanced level. This will benefit you when your implemented the latest best practices.

And that brings me on to my number 1 tip and that is love your job, I think if you love this industry and have a passion for web design, I think those qualities will guide you to achieve your goals. So enjoy your development and don’t rush yourself too much. Take the time to develop the right way, build contacts and friends and embrace the industry as a whole.

That about raps up this weeks review. I hope you’ve enjoyed the very first show live from my living room. Thank you and goodbye.

Back to top

Listeners feedback:

Newspaper columns on the web

Adrian writes: Hey guys, long time listener from the states. I’ve been working on a new personal site lately and I’ve become fixated on the idea of using newspaper style columns. Since you two seem to know a thing or two usability, I’d figure I’d ask for your thoughts.

It seems like most people view them as a print concept that doesn’t translate well online but seeing as most screens these days are widescreen and vertical space is taken up by menu bars, docks and browser extensions, going horizontal strikes me as a logical solution.

I appreciate the logic. It is true that more computers than ever have widescreens and that vertical space is at a greater premium than horizontal. However, I would think very carefully before employing newspaper style columns. As I see it there are two concerns:

The usability concern

As you point out, people reference usability concerns as the primary reason against newspaper columns. In a newspaper, copy runs across several columns with the eye darting from the bottom of one column to the top of the next. This is acceptable because the user can view the entire newspaper in a single glance. There is no such thing as a scroll bar.

On the web it is different. You are unable to predict the height available in a browser window and so users will almost certainly have to scroll. This means the user will scroll down one column as they read and then have to scroll back to the top to start the next column. This is far from a pleasurable reading experience.

It is also important to consider width as well as height. As you say newspaper style columns works well on high resolution, widescreen monitors. On anything less the story becomes unreadable with narrow columns and short line lengths. The alternative is to allow both horizontal and vertical scrolling. But as I am sure you, know this is the ultimate usability error and should be avoided at all costs.

The technical concern

There are also technical considerations to take into account. How will a story be split over multiple columns? Currently this cannot be done in CSS, although this may appear in CSS3.

One option would be to manually layout each block of text. However, this isn’t going to be practical with anything other than the most static of sites.

The only option is to use some server side code. However, even this is not without its problems. Consideration needs to be given to inline elements such as images or quotations. What happens if they appear at the end of one column? Does a quote get split? Will the design accommodate larger images? What happens when text is scaled?

Although all of these technical problems can be overcome, you are forced to ask whether it worth the effort. This is especially true considering the serious usability concerns.

Estimating dev/creative work

Kirk Henry asks: I’m not sure if this should be listed as a question or not but her goes. I’m a Creative Director for a dev shop with some very large fortune 500 companies and a problem I always seem to come across is difficulty in the estimating process. We use excel documents, have some standard hours for comps but have to do custom estimation for multi media projects etc… my estimates are always pretty decent but I want to know what you guys use or what software you would recommend. I have been listening on itunes from the start and love the show.

Ok, this is probably the most important subject that we (and I mean the web community) don’t talk about. Why? I think, because it’s difficult to pin down a method of reliably estimating a project and, more so, we’re all guilty if underestimating time and again… these are my thoughts:

The first thing to ask yourself is ‘how serious is this project?’ I have a sixth sense for requests for quotes that fit into the following brackets:

  • ‘We have this idea but have no idea how much it will cost and we want you to do all the research work involved in scoping it. Of course we won’t pay for the research and there’s no way we’ll pay sensible money for the work once we know what it is’
  • ‘We have a supplier that we want to work with but my boss says I need a couple of other quotes’
  • ‘Us guys in sales and marketing have been doing some blue sky thinking and want a quote to redevelop Google….’

You get the idea – timewasters. You need to deal with these requests quickly – this is how I do it. Have a chat with whichever department(s) would do this work if it ever materialised – get them to give you wide ballpark figures. Add in PM and contingency and send them an email. 99 out of a 100 won’t even bother getting back to you. Some will, but they’re usually trying to get free scoping (‘can you give me a bit more detail on how you reached those figures’).

Anyway, I’ve ranted long enough timewasters, back to Kirk’s question.

First question – do you know the budget? If yes, then you are looking to fit a scope into a set amount of effort. Can you do it? Will the ‘client’ be happy with the scope that fits their budget? Do they understand what that scope is (especially if you have reduced it to fit their budget)? DO NOT get creative with your effort allocations just to fit within the budget. Either ask for more (up front) or walk away.

If you don’t know the budget then you are looking to scope a project from scratch. If it’s a really big project then ideally you should be being paid to scope it as we’re looking at business analysis and consultancy here.

Break down the project into rough task areas. It’s likely that you’ll have done other projects that include similar tasks so you’ll know efforts on these (though ask yourself if you got it right last time). For the ‘new’ tasks, break it down further and you will probably find other smaller tasks that you have done before. For the really new stuff then you need to talk to an expert (designer/developer/IA) and get them to think the task through. They will provide you with an informed guess. That’s right – guess. Because people are guessing it is really important to overestimate fixed price projects. This is the cost to the client of having a fixed price.

Don’t forget to charge for meetings (if 3 people are attending then charge for 3 people!). Project management is notoriously undercharged. We have a rule of thumb of 15 – 20% (and that’s probably light).

The golden rule of estimating is don’t be tempted to lower your probably already too low price just to win the work. Be prepared to walk away.

As far as tools to help with estimating go, MS Project is great at separating tasks, linking resources to tasks and giving you a good idea of how long things will take. But, I tend to find that it is over the top at the quote stage and tend to stick with Excel.

Back to top