Toggle Revisional Changes

Web Standards for OPO Development
(Version 6.0 - Changes since 5.0 are marked as such: New Removed)

1. Site Management
     1.1 Web Site Conventions
     1.2 Site Structure/File Naming
     1.3 Production Process
     1.4 Site Managers
2. HTML Standards
     2.1 Browser/OS Targets
     2.2 OS Targets
     2.2 Bandwidth Targets
     2.3 Screen Targets
     2.4 CSS/Layout Targets
     2.5 Page Printing Targets
     2.6 Other
3. Programming
     3.1 Javascript
     3.2 Java
     3.3 Shockwave/Flash
     3.4 CGI Coding
     3.5 PHP Coding
4. Graphics
5. Testing Suite


1. Site Management

1.1 Web Site Conventions

1.2 Site Structure/File Naming

1.3 Production Process

1.4 Site Managers


2. HTML Standards

(OPO evaluates it's standards based upon specific 'browser' version usage. The definition of a specific browser being one that is substantially different than another. This can mean the same browser on a different OS (such as IE, which is vastly different on PC vs. Mac), or a single browser might supported on all OS's because it behaves the same on all (such as Mozilla). The current method of determining what we support, is a human process of looking at multiple factors of usage. First of all, we try to support the 90th or higher percentage of browsers as possible. However, we also recognize that at times this may not be feasible. Therefore we look at the specific browser usage, and tend to definately support any browser with around 5% usage or more, as they are being used significantly enough that they cannot be ignored. Beyond that, we look at lesser used browsers as such. We will consider any browser at around 1-2% usage, if it has shown around that level of usage for the last 3 months of statistics, and is 'up and coming'. We will consider also supporting browsers with less than that usage, if it is known/expected that they will soon be highly used (such as a new version of an old browser that we do support). We will as a rule not support any browser, with significantly less than 5% usage, if it is unsupported technology at that point by the software company/group that made it. In other words, if the software is dead, and it's usage is declining, we will be less apt to include it in our standards, than a lesser used browser, that appears to be moving upwards in usage. It should also be noted that this will officially give us a 'supported' standard that is less than the 90th or 95th percentiles. However, in practice, supporting this number of browsers, will in fact allow other, less used browsers to work, thus helping our actual supported percentage go higher. As a general note, OPO will try to design its content to the 95th percentile of current users based upon web statistics that are gathered. On that the below listed standards will be constantly reevaluated - approximately once every 6 months to a year. When the term 'degrade to' is used, it means that the content of the page must be somehow viewable; however, it doesn't have to look good, just be viewable. Also note that specific websites with pocket audiences (such as Amazing Space) may need to design to a broader range of browsers/specification than listed below to hit the 95% mark, this should be determined by specific site's Site Manager.)

2.1 Browser/OS Targets

2.2 Operating System Targets

2.2 Bandwidth Targets

2.3 Screen Targets

2.4 CSS/Layout Targets

2.5 Page Printing Targets

2.6 Other Standards


3. Programming

3.1 Javascript

3.2 Java

3.3 Shockwave/Flash

3.4 CGI Coding

3.5 PHP


4. Graphics


5. Testing Suite

Following is a list of what browser/OS combinations all web pages should actually be tested against, based upon the above standards, and some additional statistical digging to determine what browsers are used on what OS. These lists should be seen as the MINIMUM testing requirements. Other OS/Browser combinations are available on the testbed if one wishes to test further. (NOTE: Not all versions of Windows have been listed to be tested. Since the executables are the same for the different versions of Windows, we are going to assume that they will act the same. Therefore, any version of Windows can be used for the testing. Most should happen on the 'most popular' version of Windows, which is currently Windows XP. Feel free when testing to use other versions/other machines to simplify the task. The combinations given below are just the suggested optimal setup, given the layout of the testbed)