Web apps: taking five years to get to where desktop apps were a decade earlier?
Scott Rosenberg recently wrote about how it appears to take an AJAX-based Web application about five years to mature. The examples he gives include the new Yahoo! Mail interface, as well as the new Bloglines design. But I don’t think he’s looking far enough into the situation.
While it may take an AJAX-based Web application five years to mature, what we have to focus on is the end result after those five years. After that amount of time, one would expect there to be significant improvements over what was available five years earlier. But that’s often not what we see with AJAX-based applications. While there might be some small improvement over the previous Web-based interface, there’s still usually a major gap between the Web application and an equivalent desktop-based app.
The new Yahoo! Mail interface is a really good example of this. From my use of it so far, it looks like a lot of time and effort was spent working on it. Yet I think it was mostly in vain. It still just can’t compete with actual email clients, such as Thunderbird or KMail. For most scenarios, using Pine or mutt (over SSH for remote access) is even faster and more efficient.
So it seems pointless to me to spend four or five years developing a Web-based application, especially when it will only really be equivalent to desktop applications from a decade earlier, if it even manages that. If Web applications allowed for easier, more rapid development, which in turn allowed for more innovative products to be created, then it’d be worth it. But that just isn’t the case. We don’t really see any innovation; we merely see poorly-done reimplementations of decade-old desktop software.
September 19th, 2007 at 6:30 pm
I think you’re missing the point - the point isn’t to do what desktop applications did. The point is to do things the user finds useful and improve productivity (i.e., spend less time getting the information you want out of it).
Let’s compare. My favorite desktop mail application, mutt, is powerful. Moreso than any other mail application. Period. It is fast. It can do everything I could possibly want it to do. You have to get a degree in mutt configuration to make it do all those things. You have to set it up over and over on multiple boxes and set each installation up carefully to not prevent the others from working properly - or be content to always have to login remotely to a single location to check your email. You learn all the ins and outs before you acheive respectable amounts of productivity. To filter mail, you have to create folders and save emails in appropriate folders. This creates a multiple-installation nightmare, unless you’re using IMAP, which creates other nightmares. Search is slow, and not always accurate. Results aren’t ranked. Tagging lets you deal with many emails at a time, but is *very* slow for large mailboxes
My favorite web email application is GMail. It is the one I use for most of my email. It only does a fraction of the thing that mutt can do. I don’t install it. The configuration options are very few. I don’t have to worry about managing my messages on the server. Spam is filtered pretty efficiently, and I didn’t set it up. I can view html email properly. Search is fully supported and fast and accurate. I don’t have to manage folders. Labels let me sort mail quickly. I don’t have to know a darn thing about GMail to be productive.
Using mutt gives me some capability that I don’t have with GMail, but not that much that I actually use, and carries a heavy burden of administration and maintenance. GMail gives me 95% of what I want, some things I didn’t know I wanted, and completely absolves me of any maintenance or learning curve burdens.