Hosting part 1: what are my options?
This article is a 4 part article that goes into the detail you need to make sure your website is hosted on hardware that makes sense.
I’d say it’s useful to have some idea of how you might host a website you’re building, and to do that you’d need some idea of how many people you’re expecting to visit, how many people you’re expecting to buy from it (if it’s a site that sells things) or interact with it. Usually, this should come from the business case for a particular project, which will either be decided by the client or by the strategist developing the campaign.
There are a number of options you can choose from when it comes to actually getting your site up on the web, and each has it’s advantages and pitfalls. How do you choose, and how do you make sure the people who are building the technical infrastructure of the site are providing the right solution for the specific requirements of the project?
For clarity, it is important to remember that there are a couple of main groups of people involved in building a website from a technical standpoint: Developers, who would be building the project, and System Administrators, who would administer the hardware platform on which the application or website sits. For the site to function adequately, the hardware specified has to be sufficient for any strategic goals devised by the account teams or the client to be met. That means two things are absolutely imperative in making sure any hardware platform is going to work for any given project: planning and consultation.
The ideal way of working is that the developers would plan their software stack, providing this to the system administrators who, in consultation with both the accounts team and the developers would come up with a hardware platform that suits both the requirements of the software and the ability to achieve the business goals set.
Once you’ve got the basic parts of this, you then have to make a choice: Dedicated hosting, virtual hosting, shared hosting, or cloud hosting? I’ll dismiss shared hosting for all but the smallest of sites with no security concerns. Every project should have its own hosting platform uncontaminated by other elements. This means the underlying software can’t change and break your application because someone else upgraded theirs, and other sites hosted on a non-shared platform can’t access each other’s data.
In the next article I’ll go into the pros and cons of dedicated hosting. Stay tuned.
Mr, Mrs, Sir, Professor (or Prof for short). Esq., MSc, MPhil. Some of us use them, and some of us don’t. So how do you like to be addressed when someone you don’t know emails you, or writes to you, or welcomes you to their site? John, Sarah, Mrs Emily Rogers, Sir David Platt? Everyone seems intent in separating all the information out. So often you want to sign up for something and you’re asked for your title, your first name and your surname (better ones ask for your last name).
What happens then if you don’t want to tell people your title, or you don’t like to use it in addresses to yourself? And what if you don’t have a last name? Well I’m sorry, but you can’t sign up unless you fill in all the details that we’ve marked as mandatory.
Six months down the line and you’ve signed up for a site selling contact lenses cheap online as Mr John Doe. You get an email and the subject line says “Get new contact lenses even cheaper now, John”. Great. So you’ve now had to fill in all your information, only for the site to make a mockery of how you like to be addressed, or really how it thinks you’d like to be addressed.
How can this be resolved? Quite easily: just ask for a name. Let the user decide what form that takes, and even if it is to be in English or not. A single field that looks something like this:
You’ve resolved a whole host of issues here. Namely that the person now feels that they can be addressed as they wish to be addressed and you don’t have to check to make sure they have a title, firstname and surname before letting them submit the form. It’s quicker for the user to fill in the form, so you’re more likely to have them continue through rather than dropping out.
A short article describing an easy resolution to a problem created because we feel the need to have more data capture. This is not only more, it’s simpler, and more importantly, more accurate.
IE6 support: To bin or not to bin?
I said I’ve got opinions and this is no exception. I’ve been wrangling with IE6 since 2001 when it came out during the famous Browser Wars. At the time, we had Netscape Navigator 4.7, Internet Explorer 5.0, 5.01, 5.5 and Internet Explorer 5.0 for Mac in common use, none of which displayed things the same as each other. Along came IE6 and we now had yet another browser to build for that didn’t do the same as anything previous, and indeed anything since.
So what do we do? Do we bin support, do we not? Do we wait until other people bin support, and what happens if they don’t? To understand that, we have to look at the pro’s and con’s of supporting IE6, the features in newer browsers, what it means to continue support, and what the possibilities and limitations are. Also, it’s worth noting Microsoft’s stance on IE6 support, and any significant organisations that have ditched IE6 support in their applications or sites.
Let’s start with some facts: Microsoft have given details of support for IE6 which show that both IE6 and IE7 both have support currently ending on 13th July 2010. IE7 has been out for 2 years. IE8 is in Beta 2 at the time of writing this article. Firefox and Safari have both been created and gone to version 3 each in the space of one version of IE, Opera continues to grow and is now free, and finally, Google have released Chrome in beta state.
Browser market share is a contentious issue – this really depends on the site you’re building and the target audience, and global trends are often misleading. That said, thecounter.com shows that in June 2008, 37% of people used IE6 to browse the Internet. Firefox shows 16%, and IE7 shows 41%. From the stats, there are also a small number of people that still use IE5.5 and lower, and herein lies the problem with the web. Everyone uses a different way to browse the Internet, from mobile devices running Windows, Symbian, OS X, and Linux, among the most popular, to desktops running a whole host of Operating Systems and versions thereof. Can it really be expected to support all of these in their entirety, or should we be making a stand and slowly educating people to upgrade their browsers to versions with the features that we need to build rich web applications and sites, with the security elements already built into the browsers?
There are four main viewpoints that need to be understood here: The developer, the project manager, the account manager/director and the client. Within the advertising industry these groups of people each have different opinions about what should and shouldn’t work, and this article seeks to make sense of those and present a recommendation for how these sometimes wildly contrasting views can be worked together to create a single standpoint on browser support.
The Developer
The developers will always want to produce a strong solid product that is as unbreakable as possible, where breakage can happen in functionality and in layout. IE6 is notorious for not complying with standards, but more importantly, it is well known for having serious bugs in its rendering engine and having suffered a significant number of security issues, such as discussed here, here and as Secunia shows, IE6 still has 24 unpatched vulnerabilities. That, along with the fact that Microsoft left the browser market in 2003 saying they wouldn’t continue development of their own browser is mostly why developers would rather leave IE6 alone, and provide some notice to users of sites that they should upgrade to something that supports the relevant standards, doesn’t have the rendering bugs inherent in the 7 year old browser, and most importantly, is more secure.
The Project Manager
From a project managers perspective, the project just needs to go out working, on time and on budget. From a cost perspective, approximately 50% of the time taken in front end development is spent fixing IE6 rendering issues. That means that from the front-end scope, half of the time could be spent elsewhere, testing the other browsers, or just reducing cost to the client. While it is definitely important to a project manager to make sure support for the common platforms is adhered to, IE6 makes it difficult to continue to justify that mantra.
The Account Manager or Director
The Account Manager/Director will always want to make sure the client is happy and is getting as good a product as is possible, in their eyes. “Good” is a subjective notion, and what is good to one person is not necessarily good to another – look at the difference between what each of these groups of people have as their viewpoint for an example of this in action. The standard remit is to support IE6, IE7, Firefox, Safari and sometimes Opera, and this is the viewpoint of the AD.
The Client
For obvious reasons, the client will want as many people as possible to be able to view their site, or use their application. The Internet has become a marketing tool, and being able to reach as many people as possible is important. The likelihood is that if they’re in a corporate environment, the browser that will be installed on their machines will be IE6. So, when you build a site and it doesn’t work in their browser (this holds true for HTML emails in Outlook as well!), they naturally think the job is not well done.
In my opinion, we should have stopped supporting IE6 years ago. You may ask “what about the 40-odd percent of people who use it?” and my answer is this: educate them, like we educate our clients and our teams as to best practice on the Internet. This is just another case of best practice. IE7 is a free and automatic update in Windows XP, so install it if nothing else. I don’t say this unsupported; Apple have already ditched IE6 support from their new MobileMe application. 37Signals, the company behind Ruby on Rails have ditched IE6 support in Basecamp (their project management tool that is widely used around the world for managing collaborative projects, such as websites and applications). Facebook, with 100 million users, doesn’t really support IE6, and a whole host of functionality just doesn’t work if you have it. There are also plenty of people writing about IE6 support, and the ideas behind dropping it. Just look at this Google search to see what I mean.
How can we inform users though, without compromising other users who have upgraded? Very easily. Microsoft have a piece of code built into Internet Explorer that allows you to display things based on the version of the browser that you have installed. Using this to display an upgrade notification would be the best way, in my opinion, of education users without compromising the possibilities that are available with the newer browsers.
To conclude, and thanks for reading this far down a lengthy article, IE6 is old, clunky, and doesn’t allow for innovation in web applications and other sites. It costs more to keep supporting it than to educate users to upgrade their browsers. As for the corporates? They’ll follow suit soon enough if support is dropped for the incumbent browser. After all, it didn’t take this long to forget about Netscape Navigator 4 or IE5 on the Mac, did it? It’s high time we did the right thing.
An Introduction
The whats, the whys and the wherefores to Adverghetti Junction is what I’m going to try and put across here. The concept of blogging is certainly not new, and it’s a lot of hard work to keep doing it. So why bother? I’ll come to that…
We are initforthe Ltd, a web production house in London that provides outsource production and consultancy to the advertising industry. It’s a strange place, advertising. It’s full of different disciplines that don’t really quite go together, and for just one person to understand it takes years of learning. We don’t claim to know anything but our little corner of it – online. So we decided that it made sense that we talk about it. Part of our company ethos is that we promote openness. That is, we’re open about everything we do – client discussions, software we use, knowledge we learn and then impart. So this blog is a platform for that discussion.
As I said, it’s known to be hard work to keep a blog running and updated with new stories and so on. And I like that. Personally, I thrive on hard work, complex projects, new discoveries and all the things that make people tingle with fear when they so much as think about them. “I” am Tom Simnett, Managing Director at initforthe, and the knowledge or discussions I open up come from 8 years in online advertising, be that working in an entirely online team or with other disciplines to get a campaign out of the door. I’ve got my opinions, and I’ll share those. I’ve got a whole host of best practice tidbits that I’ve garnered over the years.
So that’s the who and the why. The what: a collection of discussions and stories aimed at simplifying the world of tech jargon so that normal people understand how best to provide their clients with a good online strategy and service. Hows that for a solution-in-a-blog!
Keep tuned, sign up for the RSS feed, and keep track of everything we’re talking about here at initforthe. Thanks for coming by.
Tom Simnett
Managing Director
Recent Comments