|
Welcome
Guest
|
Register
| Login
|
|
|
| Index | Recent Threads | Unanswered Threads | Who's Online | User List | Help |
Killersites.com Forum
Misc Forums
Members Website Showcase - Peer-to-Peer Reviews
Thread: What do you mean with "Validation?"
|
| No member browsing this thread |
|
Thread Status: Active Thread Type: Sticky Thread Total posts in this thread: 2
|
|
| Author |
|
LSW
Advanced Member USA Joined: Nov 27, 2003 Post Count: 6285 Status: Offline |
Brief version: Validation simply means ensuring through a test program that your web site is written in valid code, or error free if you prefer. You call up a program, usually the W3C Validator (1 page at a time) or the WDG Validator (Multiple pages) and point it at a or multiple pages and you tell it to validate your code. It then says your good to go or you have "X"-many errors that need to be fixed. It tells you how many, where they are and what the error is with a cryptic hint as to what the problem is. You then fix it, test again and when it is error free you publish the web site or post a critique request. You want to do this because:
You may also wish to read: --------------------------------------------------------------------------------------- Extended version: OK, we get allot of requests here for critiques. Far to often it is new members with their first posts... then they do not understand what we are telling them. One case in point is validation. Web sites can be critiqued in three forms
So back to the code, we get a request to critique it and we validate it and it shows dozens of silly errors that should not be there and should have been easily repaired before publishing or asking for a review, that is why we say to validate first as we are tired of masses of errors. But many reply "what do you mean validation?" If you stop to think about it, it is clear that we mean test your code to see if it is valid. The problem lies in two forms:
Well it is not. Web design is a structured way of creativity and has rules and guidelines. You can write novels or poems... but you still have to use correct grammar and spelling right? Web design is no different, only we use other languages, rather than English or German with their different rules and spellings, we use (X)HTML and CSS and co. with their rules and grammar. Just as you would not publish a novel with poor grammar or spelling, you should not publish web sites built with poor grammar in HTML. This is where validation comes in. It is about testing your code for errors and bad habits, missing tags and deprecated tags. Validating your code is about ensuring that your code is valid and error free, that it is written per norms and standards set forth and that the code is written the way it is supposed to be. These rules are set forth by the World Wide Web Consortium (W3C) who defined what the tags in say HTML do and it is these standards and definitions of tags that tell the browser what to do with that tag when it finds it. Standards state the a tag that looks like <p> is a paragraph and so space shall be added between these blocks of text. It also states that in HTML a closing tag may be left out, but should be added and that in XHTML closing tags MUST be used. In your HTML you are supposed to start the page with a tag that starts with DOCTYPE, this defines what language is being tested against. Just like you have to tell a spell or grammar checker that you are checking German and not English, you must tell the validator that you are testing HTML or XHTML so it knows if a closing </p> must be there or not. Also if you are using "Transitional," so it may ignore errors you made or if you are using "Strict" and it must note every mistake. This is where the two groups above miss the point, they look at the web site in IE and it looks like the way they want it so all is fine. Well I could give my published novel to a drunk or a pot head and they would likely not care either about grammar or spelling. But should I not be publishing a quality product? Hence validation is very important, it finds those stupid little errors that we all make. The IE argument is a cop out by lazy people (if they know validation and choose not to, not if you do not know it.) who only want to do the minimal work and call it finished. IE was developed to be idiot proof and assumes you do not know what you are doing. It is programed to more or less note that there are errors but to ignore them as you do not know what you are doing and just act like the error was not there. For the most part all other browsers now expect you to know what you are doing and do not ignore mistakes, they show what you wrote. IE may be the majority browser, but it in not the only browser and is not guaranteed to always be the majority browser. So is it logical to design just for it? That is the point, not all the users use IE even if it is the majority browser. Even more important, browsers are not even the only issue anymore. Browsers are just a sub-section of the larger term "User Agent", specifically graphic browsers may be the majority of user agents, but we also now have to design for Cell Phones, PDAs, screen readers, text browsers and even game consoles like Nintendo's Wii and PlayStation Portable. User Agents are anything a user can choose or need to use to access the internet and that will one day include refrigerators to oder groceries and maybe coffee machines. Will IE ignore validation mistakes? Yes for the most part. Will a cell phone? You do not know do you. Well the fact is there are some PDAs running windows that have IE, but for the most part mobile devices run software developed by that company, there is no telling what a mistake will do to a web site. Opera has a mobile version that is good, Firefox is being developed for Mobile as well and their is a 3rd party browser that is good... but the vast majority of mobile units do not use these products so we cannot know what will break our site in the proprietary software used. That leads us back to validation. If we create a site that is error free and uses standards... then you can feel more secure that your web site will work over all possible user agents than a poorly written site full of errors. Lastly, like in programming, errors do not always show themselves where they lay. More often they appear quite far away from where the mistake was made so can be a bugger to find. One small unimportant mistake above can aggravate another mistake farther below in the code and snowball into a major issue. Correct that one small mistake at the beginning and "POOF!" all the others disappear. But all the good arguments aside, should you not simply take pride in your work and make it error free for the sake of doing it right? ---------------------------------------- [Edit 2 times, last edit by LSW at Jul 24, 2007 1:00:43 PM] |
||
|
lm
Advanced Member Joined: Oct 27, 2004 Post Count: 2550 Status: Offline |
i consider this LSW post as a milestone - now all we have to do is to link to it everytime somebody asks for the review. Often validation seems unimportant to newcomers but most of their troubles - visible and invisible - are from the code mistakes. So again: - validate first thing before you go online. ---------------------------------------- My blog |
||
|
|
|
|
Current timezone is GMT May 19, 2013 1:07:08 PM |
