While there are some who still doubt the potential of browser-based clients, others believe the case was proven long ago. Responding to the doubters mentioned in my last posting, Steve Guttman left a comment recalling his work several years ago with a startup called Halfbrain.com: "Back in early 2000, my former company, halfbrain.com, developed a dhtml spreadsheet and presentation program. These were VERY competent products running on 233 mhz celerons ... the technology clearly exists (even more so, today). You can check out similar technology at oddpost.com (started by former halfbrainers)."
Halfbrain was acquired by Alphablox and its online presence was maintained for some time as Blox.com before finally being canned in 2001 or thereabouts (the technology was folded into the web interface for Alphablox' business intelligence products). Halfbrain did some wonderful stuff its flagship product was a browser-based multiuser spreadsheet, and it later followed up with a browser-based Powerpoint clone. Unfortunately (like so many great ideas that came to market as hosted web-based offerings at that time) Halfbrain was encouraged by its venture-capital backers to acquire eyeballs instead of revenues, and quickly ran out of steam when the dot-com bubble burst.
But Halfbrain was way ahead of its time in demonstrating the capabilities of the browser interface when extended using DHTML. It did a remarkable job of realizing the spreadsheet idiom in a browser environment, and was the only product I personally have seen that actually achieved the longstanding holy grail of a true multiuser spreadsheet.
Leaving aside for the moment the issue of the dot-com debacle and the bone-headed VCs that fueled it, I think there are two principal reasons why services like halfbrain.com failed at that time:
No facility for disconnected working. This remains the other big obstacle, as BEA's Adam Bosworth has identified [link updated 31 July 2004]. This still needs some work, and was one of the key issues I had in mind when I wrote about the shortcomings of current browser clients in my last posting.
On the whole, though, I think Steve is right to assert the progress that has already been made by various pioneers, many of them in the net-native ASP sector, who had no choice but to make the browser a rich environment for their users and indeed I wrote fulsomely last November about the user interface (amongst other attributes) of business management ASP NetSuite in a blog posting titled Secret weapons of ASPs.
But there's still scope for improvement (which Microsoft, to return to the original inspiration for these musings, ignores at its peril), so I was most heartened today to read about work being carried out within the Web Hypertext Application Technology Working Group, which Jon Udell just highlighted: "Unlike in 1996, Microsoft today sees Web applications as a dead end; Internet Explorer is frozen; the wholly proprietary Avalon is their future. Meanwhile Mozilla, Safari, and Opera think they can create forward motion on Web apps, within a cooperative framework. My $0.02: go for it." I'll drink to that.