- Phase 4: Production and QA
- Establishing Guidelines
- Setting File Structure
- Slicing and Optimization
- Creating HTML Templates and Pages
- Implementing Light Scripting
- Populating Pages
- Integrating Backend Development
- Understanding Quality Assurance Testing
- Creating a QA Plan
- Prioritizing and Fixing Bugs
- Conducting a Final Check
- Phase 4 Summary
Establishing clear guidelines for HTML production during the initiation of a web redesign project helps to answer questions and avoid costly backtracking. The Client Spec Sheet sets parameters for audience capabilities and technical standards for the site. This is a worksheet. It is long and detailed and technical. The client may simply say, "I don't know. You're the expert; you tell me." Some discussion is likely necessary. For instance, the project manager or lead production designer might have to explain what effect choosing to support 3.x browsers might have on being able to support certain functionality, or what effect selecting Flash might have on wanting to support dial-up modem connections, and so on.
> Assessing Project Status
> Establishing Guidelines
> Setting File Structure
The Client Spec Sheet is available for download from http://www.web-redesign.com. Due to its length, we could not show it in its entirety in this book, so we show only the first two parts: Target Specifications, and Functionality and Features (see worksheet on next page). All told, it is five parts long, as follows:
Part 1: Target Specifications
Part 2: Functionality and Features
Part 3: Design and Layout
Part 4: File Structure and Directory Preferences
Part 5: Server and Hosting Information
As tedious as filling out this worksheet may be, the information needs to be addressed and answered before any HTML production can begin, and that includes conferring with the visual designers at the onset of Phase 3: Visual Design and Testing. Encourage client feedback within a short timeframe. This information should be back to the team and analyzed before the visual designers start developing concepts and definitely before the production designers start building the Protosite.
The team's lead HTML production designer should be the team contact; the project manager may or may not be as technically savvy. Have the client or the client's key tech lead answer all questions as thoroughly as possible, adding additional comments as necessary. Encourage the client to write "N/A" next to nonrelevant items and to identify areas in which advice, suggestions, or clarification is needed. Filling it out should be taken seriously; the results from this analyzed worksheet serve as a set-in-stone guide for production.
The worksheet on the following pages will help you articulate and identify the technical parameters of your site redesign, including specific questions regarding target audience connectivity capabilities, browser versions, functionality, and actual file structure. When you are finished, please return all compiled information back to the project manager on the web development team.
Chad Kassirer on Knowing Your Client Before You Code
Clear communication with the client is the key to a successful web project. Before beginning the production process, it is important to have agreed on and signed off on two things: a composite of the target audience and the client's expectations concerning the site production details. To assist with this process, I rely on a Client Spec Sheet to document these items. Ideally, this document is administered shortly after the project has been kicked off. This way, there is one central document that can serve as a guideline for everyone contributing to the building process. Not only does this assist in all phases of the process from information architecture to design to production, it also establishes some necessary parameters for the site's requirements and identifies possible limitations early on.
It is every web designer's, programmer's, and production engineer's goal to create a website that looks and works the same for every user. However, with the numerous possible combinations of platforms, browsers, connection speeds, and monitor resolutions, this is nearly impossible to accomplish. To decide the best way to design and build the website, you need to identify the target audience. Once this is established, you can tailor the site to best suit this audience's needs before being concerned about other users. This is not to say that no one other than your target audience is important, but the client's priorities need to be established. These priorities will impact decisions made during the production process. A more realistic goal is to make the site as close to perfect as possible for the target audience while still being functional for everyone else.
By initiating a conversation at the start of the project, a dialog is created between production and the client. During this conversation, the client's expectations and preferences can be discussed before deciding the direction the client wishes to take and documenting this on the Client Spec Sheet. The production lead, as the integrator of design and engineering, uses this document as a reference for making decisions during the design and production phases. When used properly, the Client Spec Sheet is extremely useful and saves time and money by eliminating ambiguities, which cause unnecessary delays and frustration.
I recommend using the Client Spec Sheet early in the process. It documents and clarifies to everyone what the initial goals of the project are, even if changes are made during the process. In case the requirements or expectations of the client change, the Client Spec Sheet also serves as a contract to refer to when additional costs are required or disputed. By using the Client Spec Sheet as a reference to help guide your decisions throughout the project, you can build a site with the client in mind.
As director of production for web development shops Red Eye Digital Media and Idea Integration/San Francisco, Chad Kassirer (http://www.whatdesign.com) has played a key role in the production process for many award-winning websites such as Adobe's Splatterpunk, One World Journey's Georgia Revealed, and SFMOMA's Making Sense of Modern Art. Chad has never appeared at any web conferences or written any books, but he knows people who have.
Although this chapter concerns HTML, it is not about how to code, the ins and outs of HTML, coding theory, nor advanced scripting implementation. We focus on the redesign workflow process and how it relates to the actual site production keeping your project moving smoothly, staying on schedule and on budget. For guides to actual hands-on coding, seek alternative resources such as HTML Artistry: More Than Code (New Riders, 1998) by Ardith Ibanez and Natalie Zee or Creative HTML Design.2 (New Riders, 2001) by Lynda Weinman.
Scope Expectations Meet Scope Reality
An estimate of 100 hours can easily turn into 300 hours if the complexity of the site has been underestimated. In Phase 1: Defining the Project, you estimated the project's budget based on the projected scope. Did you plan on 50 pages and now there are 120, or are you still on target? Assess. Has your scope grown, either through Scope Creep or as a result of client-requested changes and/or additions? If so, you will need to either increase the budget or downsize the allocation of hours... or take a loss. Regardless, if you haven't yet addressed potential budget changes with your client, do it now before you start coding. And make sure you have included resources for QA along with the time necessary for fixes.
Do a project-wide time check. You should have been tracking your hours on a weekly basis, so this should be a relatively easy assessment. How much of your allocated time and resources have you used up? Are you on budget? Has the scope increased? Do you have the time left in your budget to comfortably complete the site? Knowing how many resources and hours are necessary to complete a project's production and QA is regularly one of the gray areas in project estimating. The hard truth? Most things that appear to be simple are not and will take much longer than you estimate. Coding an HTML page or template can take a few hours or a few days it is one of the factors contributing to Scope Creep that is extremely difficult to gauge until actual production begins.
Readdress Audience Capabilities
Focus on your user. You are producing the site based on the capabilities of your target audience, and you cannot translate the visual design into HTML unless you know your parameters: target operating systems, browsers, monitors, and connection speeds. Use the results from the Client Spec Sheet as a guide.
Check Content Status
Content should be in all of it. But chances are it won't be. You must be on top of content status. Your Content Delivery Plan was clear: Content must be in before production can commence in earnest. Alert your client that the time has come, that a content freeze is imminent.
Announce to the client early in the process, as early as with the initial proposal that accompanied the budget, that if/when content is late, production will be held up and cost overruns will commence. Billing for overruns is never painless, but it will be far more viable if you warn the client of consequences ahead of time.
SCOPE: Are You on Target?
How big is the site? How many pages? Is it what you planned for?
Is the slicing a nightmare or fairly straightforward?
Light Scripting Needs
Are the engineers on budget/schedule? Have the requirements been adequately defined, and do they still match the scope/cost expectations?
Imagine your only web access is through a browser that does not support images. Pick any site if the navigational aids and buttons don't have descriptive ALT tags, you won't be able to differentiate between graphics.
Recent government backing and new Accessibility with Disabilities Act (ADA) standards are behind an ongoing push to support full-access web sites. This new set of standards, led by the World Wide Web Consortium (http://www.w3c.org), aims to connect all people to the web regardless of disability, including the handicap of older browsers or outdated technology. Understanding accessibility needs before you start coding especially if your site is bound to comply to accessibility standards will avert damage control later (for example, coding ALT tags into 100 pages instead of having done it once at the outset on the HTML template).
Here are two free tools that can help you test your site for accessibility after it's up: Bobby and Macromedia's Section 508 Accessibility Suite.
Bobby (http://www.cast.org/bobby) is an online tool that rates your web page immediately. Enter an URL, and Bobby identifies the areas that are not accessibility compliant and will let you know if your images have proper ALT tags. It's fast and impressive; the results may surprise you ([6.1] and [6.2])
Figure 6.1. A Bobby Approved icon appears when a site meets all requirements for disability standards. When a site does not meet the standards, Bobby clearly does not approve and lists the site's errors as well as suggestions for improvement.
Figure 6.2. This screenshot shows the results of running an URL through Bobby. The question marks show areas that either are noncompliant to ADA standards for accessibility or could be improved upon..
The Section 508 Accessibility Suite for Dreamweaver 4 and Dreamweaver UltraDev 4, created by UsableNet [6.3], enables web sites to be checked for accessibility in the same way that a document is spell-checked. The extension, available for free on the Macromedia Exchange (http://www.usablenet.com/macromedia/index.htm), helps ensure that web content meets Section 508 and Level 1 W3C/WAI guidelines. Reports can be run on one page, a complete site, selected sections, or any folder.
Figure 6.3. Usable.net and Macromedia team up to help check for accessibility.
Check Design Status
Have the graphic templates been finalized, approved, and turned over to production yet? If not, light a fire under the chairs of your visual designers. They are holding up production. Have a delivery schedule set up so that graphic files can be handed off in a phased process: the home page and a representative subpage first, and then let production figure out the HTML templates before the remaining pages are delivered.
Content will be late; this is predictable. Anticipate it. But what do you do when that magic date has passed, the content officially becomes late, and your production is compromised? After a few gentle reminders via phone and/or email, send a firm yet diplomatic email restating due dates, details, and the costs associated for each day the content is further delayed. What follows is an excerpt of a letter that addressed slippage as it was happening, and spelled out the consequences.
"... for clarification, we have determined some associated costs for the addition of the animated product demo and also for additional production work if our content delivery deadlines slip further.
"We realize you have tight budget constraints and do not wish to incur extra charges unless absolutely necessary. As explained earlier, we have allocated resources for a particular timeframe in order to produce and complete your project, and this time window is quickly evaporating... "
Furthermore, financial consequences were clearly outlined: For each day until the final content was delivered in full, a rate was applied for "holding" resources. The effect was dramatic. The first part of the content was delivered by the end of the week, and the rest of the project ran smoothly through launch. Should you find yourself seeing deadlines slip by, consider incorporating this slippage and consequences terminology into your workflow.
Confirm the Backend Integration Plan
Is your redesign a static site or a dynamic site? If static, and you are not involved with a backend engineering team, this section does not apply to you. If the site is dynamic, however, plan on having a meeting before production actually starts, a front-end and backend status update. Restate all technical specifications to all team members, review the technical requirements, confirm the integration plan, and clarify responsibilities.