Evaluating Nvu as a replacement for FrontPage


Much of the content of our corporate Intranet is written by content authors within each department. For many years, FrontPage was the defacto editor that we promoted and supported. Eventually, the bloated code and permission oddities associated with FrontPage made it less desirable and we started to push Macromedia’s (now Adobe) Dreamweaver or Contribute editors. The unfortunate aspect of this is that neither of these two new editors can actually replace what FrontPage did. For our hard-core authors who do their own application development, Dreamweaver is great; it’s more advanced than FrontPage, but with a higher price tag. For our less sophisticated content authors, Contribute is great because it is simple and cheap, but it doesn’t have the layout and design capabilities that FrontPage had. Now Microsoft is sunsetting FrontPage in lieu of it’s Expressions line of development tools, which will leave a big gap in the web development tool box. We’ve been looking for that middle-of-the-road editor for a while no, but so far have come up empty. Not for lack of options – the arena of web development tools is packed to the brim – but we’ve yet to find one that suites our needs as well as FrontPage did (and without all the headaches).

Just recently I reviewed an open-source candidate called Nvu (pronounced “en-view”). It is built on the Mozilla framework. Overall, it failed to meet my expectations/requirements. Below are my notes.

  • Nvu’s concept of Site (i.e. project) profiles is similar to Dreamweaver’s, but much simpler. Choose a name for the project, enter the URL, enter the publishing info, done.
  • At first glance it appears that you can only publish site files via FTP (and it won’t accept an empty value as an answer), but after toying around with it for a bit I found that it can accept a file path too (UNC paths are not supported)
  • Whoops! Had some trouble initially in that the site was going through several minutes of “loading”. I kept messing with the site profile values and eventually found out I had to prefix the path with file:/// and format the path in URL style (file:///c:/path/to/folder%20name/). It doesn’t tell you if it can’t find the folder, it just sits there “loading”
  • PHP and ASP files aren’t supported by default, as in PHP files won’t even open in the editor (prompts you to download them). After searching through the forums I found an extension that added some support for PHP. An ASP version is supposedly in development, though not available yet.
  • I just noticed that the extension I just mentioned may actually create a copy of the PHP to work on, but with a .html file extension (ex: .index.php.html). That means that your PHP source code could be easily read (embedded passwords and all) if these copies get uploaded to the server!
  • It has a really nasty way of reformatting source code by removing most white-space and hard-wrapping lines after 72 charachters. (I later found a preference to retain the original formatting, though inserted/edited code is still formatted poorly)
  • Clicking on CSS or Javascript source files in the Site Manager opens them in the default external editor. Explicitly opening a one of these files (via File -> Open) just renders it as unformatted plain text. (I later came across a built-in CSS editor, though it was a bit confusing to use)
  • On a positive note – it does a really nice job of rendering complex CSS layouts (case in point: – renders almost exactly as Mozilla/FF does)
  • Another quirk – you can edit in the Source tab, but when you save it immediately kicks you back to last tab you were on.
  • Editing in the Normal tab is really easy. The editor controls are nice. Styling is done via inline styles or x/html tags (strong, em, etc)
  • You can apply custom CSS rules to any selected element, but only if those styles aren’t wrapped in @media declarations (though you can set the target media via the style or link tags and have the rules appear; Nvu renders CSS in accordance to the media rules no matter where they are defined).
  • It has a built in spell checker (though the realtime spell checking seemed to slow Nvu to a crawl)
  • It has an HTML validator tool which submits the local file to the W3C validation service, though some of the things that it inserts don’t validate to XHTML 1.0 transitional
  • The included calendar “widget” is complete crap
  • It’s “template” functionality is really just boilerplate pages. Changes to a template are not applied to child pages.
  • McAfee VirusScan seems to take issue with Nvu and cause it to run sluggishly at times. Temporarily disabling VirusScan resolves the issue (at least until VS kicks back on again)

* This post was originally published on August 3, 2006 at

An overview of the debate over Net Neutrality


My brother just posted his thoughts on a recent snafu by a certain Senator Stevens’ who tried to explain the Internet by comparing it to a series of tubes, as part of the ongoing debate over Network Neutrality. Many of you probably aren’t familiar with this and may think that it’s over your heads. Unfortunately, if Net Neutrality is defeated through legislation it could have a profound affect on the Internet of tomorrow. To that end, here’s an over simplified description of Net Neutrality.

The Internet as we know it right now is considered neutral, that is to say that all of the traffic routed across the Internet is given the same priority whether you’re day-trading, banking online, running an e-commerce website or just browsing your favorite blog. Many Telecommunication companies are now lobbying our government for permission to create a tiered Internet. This tiered Internet would offer paying customers higher-priority traffic. These TelCos insist that this is vital to the future of the Internet because their corporate customers need to ensure that their information is given priority over college kids who are just surfing YouTube.

Senator Stevens’ analogy of the Internet as a series of tubes isn’t all that wrong, it’s just a bit naive. Yes, all network traffic is queued and since none is given priority there is the potential for bottlenecks, much like traffic on a one-lane road. However, there isn’t just one road. There are hundreds of thousands if not millions of different “tubes” that traffic can be routed over. If one tube is blocked packets are just routed through different tubes.

The proposed tiered Internet would add a second level of routes – like a toll highway. You would pay your telecommunication company (the toll collectors) for privilege to send your traffic on this data highway. In theory, because these highways would be less cluttered (since not everyone would pay) your data may or may not arrive faster. However there are two problems with this.

The first problem is that this model has already been tested and it was determined that the mere action of determining which packets of data had priority offset some of the benefit. An analogy of this would be traffic backing up at the toll booth before entering the highway because a toll collector has to make sure you have paid to be considered priority. Researchers then tried a second test in which they just increased the broadband that was available (sort of like making a 2-lane road into a 4-lane road). Then all traffic received priority. Of course this requires the TelCos to pay for new technology to enable the wider bandwidth, whereas with a tiered Internet they could charge certain customers a premium to do the same thing.

The second problem with this is that it creates a sort of class-warfare among Internet users. Consider this example. The (fake) TelCo Hypernet makes a deal with major online retailer Jungletopia so that all traffic from Hypernet’s customers to Jungletopia is given priority. These customers immediately notice a dramatic reduction in page load time at Jungletopia compared to it’s competitors. Hypernet’s customers will inevitably do more business with Jungletopia because it takes them less time. Now all of the other online retailers are left with two choices – partner up with a TelCo to increase their revenue or worry about a loss in sales revenue because they can’t afford to compete on that level. And this only assumes that all non-priority-routed traffic is left alone and that high priority traffic is given a “boost”. But it could go the other way too in that non-priority traffic is throttled down while priority traffic is merely left to travel the normal routes.

This really does affect all of us. And while I’m not opposed to letting businesses gain a competitive edge, I don’t think it’s right to force the consumers hand . It has to remain an even playing field where the consumer decides who they will do business with.

* This post was originally published on July 14, 2006 at

First PIDG meetup announced


The first Providence Intranet Developers Group meetup will take place Thursday, July 13th at 12:00 noon at Olga’s Cup & Saucer in Providence. Along with the normal pleasantries and introductions we’ll be talking about everyone’s current Intranet development environment including likes and dislikes, as well as dream up our pie-in-the-sky setup. We’ll also be talking about future plans for the group. Even if you’re not a member of the group, even if you’re not specifically an “Intranet” developer, feel free to stop by anyways. If nothing else, Olga’s has the best coffee :) It’s a lunch-hour meetup, so bring your appetite – you can buy from the cafe or brown-bag it.

Read more about the group and the meetup at our website:

* This post was originally published on July 6, 2006 at

The Providence Intranet Developers Group


I’ve decided to start a meetup group around the challenges of Intranet development. Meetings will take place at 12:00 PM on the second Thursday of every month at a North Providence area restaurant. You can read more about it at the Providence Intranet Developers Group page.

* This post was originally published on June 27, 2006 at

Clean And Close Extension for Firefox


I just created my first extension for Firefox. I call it Clean And Close, and it’s a simple overlay for the Download Manager window that replaces the default Clean Up button with a spiffy new Clean & Close button. The new button does exactly what the old button does, except that it also closes the window afterwards. I wrote it as a learning exercise, but it also gets rid of an extra click that was always an annoyance for me. It’s currently pending approval on the Mozilla Addons site, but should be available for download soon.

If you’re looking to develop your own extensions for Mozilla products, the Building an Extension tutorial on Mozilla’s Developer web site is what I referenced for Clean & Close.

* This post was originally published on May 23, 2006 at

Go to Top