The FrontPage Experiment Has Failed

Can we finally admit that the FrontPage experiment has failed? You know – the promise that FrontPage will allow novice Web authors to create and maintain (especially maintain) good, solid Web sites? Can we finally admit that this just isn’t going to happen?

How many people know someone that is maintaining a Web site of anything beyond trivial complexity in FrontPage? I mean anything beyond a five-page brochureware site. Anyone?

I’ve tried on two occasions now to teach a Web novice how to create decent content with FrontPage. Both attempts have been complete failures. I found myself the other day outlining a curriculum entitled “How Not to Screw Yourself with FrontPage,” and realizing that it would be at least a dozen hours of instruction.

The idea of FrontPage, of course, is to make creating Web pages as simple as writing in Word. However, I find that you end up having to teach people a lot about how Web authoring works in order for them not to completely hose a site up in record time.

Here are the biggest things I see happening when letting beginners run wild with FrontPage:

• Massive images being pushed to the Web site, and resized in HTML. No matter how many times you try to teach them the thumbnail system, they just don’t understand. If the image gets smaller in FrontPage, then it’s smaller…right?

• Poor and inconsistent file naming. This isn’t technically a problem, since URLs can be forgiving, but you see spaces, apostrophes, etc. in URLs. They think they can name files just like in Windows.

  (For some reason, I see a lot of Christmas Party_files URLs too – like they’re saving a Web page out of IE, then uploading it.)

• Fonts, fonts, fonts – where does the compulsion to change fonts come from? I don’t know, but there must be a deep human need to do this, because most users lunge right for the font box no matter how many times you tell them not to.

• Zero comprehension of styles. The concept of styles is never embraced by users in Word either, so this is no surprise. Users will jump right at the bold, italics, and font sizing buttons to create their headers.

• Broken hyperlinks. No one runs the link checking reports.

• No use of folders to organize assets. HTML, images – everything ends up in the root folder.

• No page titles. There are 4.4 million pages in Google entitled “New Page 1”; 2.5 million entitled “New Page 2.”

• Attempts to back out of formatting problems by applying more formatting. Font not what you want it to be? Then apply another font over that one so you get nine nested FONT tags.

  “CTRL-SHIFT-Z” is your best friend in these cases – clear all formatting and start over. Regardless, the font dropdown just makes sense for users because they have no concept of the tag-based HTML in the background and how it’s being affected by what they’re doing.

• No understanding of the paragraph (P) vs. line break (BR) relationship. This is really universal to word processing, so it needs to be learned across the board.

• Webbot promiscuity. If there’s something in FrontPage that promises to do something cool, they will try it, without fail.

I know what you’re saying – “these users need to be trained.” Sure, but I’ve found that the number of hours spent training them is really better spent installing a lightweight CMS, or – better yet – showing them how to use a tool like Squarespace or Typepad.

The fact is that to get someone really proficient in FrontPage so they can build a good-looking, easily maintained site, you need to teach them about the basics of Web authoring, CSS, a fair amount of HTML so they can get themselves out of problems, Web conventions like file naming, best practices for site management, etc. This is so far removed from the supposed Nirvana that FrontPage was intended to give us: simple and effective Web authoring for everyone.

So what’s the answer? I don’t think it’s a more capable tool. Dreamweaver is an amazing piece of work, but that would kill the average newbie if they tried to build anything with it.

What we need is a Web development tool we can neuter the crap out of to effectively seal off functions and transfer their administration to another, more qualified party. Give me a WYSIWYG editor that will let me control the interface – shutting off formatting tools and basically leaving the user with a styles dropdown, a hyperlink button, and maybe an italics button if they promise not to overuse it. Is Contribute this tool?

If you could do this, then you can move a lot of functionality to server-side tools that the user can’t touch. My method for applying headers and footers based on URLs that I detailed here is good for that.

Yes, I know good sites get built with FrontPage everyday. I’ve built several of them over the years. But I know Web development. And so does the Microsoft FrontPage team, which proudly points out that the official FrontPage site is built and maintained in FrontPage.

This discussion leaves me curious about what percentage of FrontPage sales are actual, retail sales? It gets bundled with the Office suite a lot, and Microsoft throws it in with a lot of server software too – I got FrontPage 2003 with Small Business Server. Thus, I’d estimate that less than 10% of FrontPage sales are actual, full-price, retail sales.

I’m sure many will disagree, but I have solid experience that the idea that you can give new user a copy of FrontPage and a set of shared borders and think they’ll keep a site in shape is somewhat ridiculous.