Browsers are everywhere and nowhere, they are getting built into applications so that soon one will not need a stand-alone browser, you can just use whichever app you have at hand or your file browser. The most recent example is Coda from panic software, but it by no means is alone. Coda may be a little different from other apps in that it is purpose built as a web development tool, and it looks to be a mighty good one at that, still lots of applications these days are network-aware and/or have plugins to the network to get data. In fact nearly all of the apps open on my Mac right now have a network component; the Dashboard apps, iCal, iTunes, and Terminal, not to mention Safari.
It is an interesting development that browsers as well as network-aware applications are starting to proliferate since it offers us a counter-weight to the view that applications are migrating from the desktop to out on the network. What if the network is migrating to the desktop?
These applications are breaking down the binary opposition of ‘desktop’ and ‘network’ applications, the distinction in computing between ‘local’ and ‘remote’ resources. They can do this because of two chief reasons;
1. Apple has provided the software to build browsers easily with WebKit
2. Network connections are becoming faster and ubiquitous
Are Coda and other applications, not just Shiira, the logical outgrowth of Apple Open Sourcing WebKit? Is this the kind of innovation that keeps Apple at the forefront of usable computing? If the answers are yes maybe Apple should open up more. I think Apple should continue to share important parts of its proprietary technology to allow innovation that gets incorporated back into the OS in the form of third party applications. Look at Coda and iWeb, they are similar aren’t they? Which one would you like to have? Wouldn’t Coda be a persuasive argument for buying a new Mac if it was bundled in? I think so.
Another thing that makes this possible is the ubiquity of network connections and the seamless way they integrate into desktop applications. Now you can unplug your apps and the network appears to remain, all the functionality is there, the data is just time shifted to the point where the network connection is active again. This makes the network irrelevant, or at least transparent in the sense that you do not need to be “online” to get your work done; you will be “online” eventually and your work will propagate to various servers once your connection is live again so one can just work without worrying about network presence.
um, I fail to see how Coda and iWeb are in any way similar. One is aimed at web designers, the other for point-and-click template website building. iWeb doesn't even actually use WebKit as far as I'm aware!
Give it up. We've been through this "all-in-one" thing many times in the past. You're probably too young to remember OpenDoc and what a fiasco that was. Or "document-centered computing". Those all failed miserably. This same thing happens about every 10 years as the newbie crop of young people enters the computing world. Then they realize it's a fad and it goes away again. Browsers will always be with us. Just learn to live with it. And stop making yourself look ignorant.
iWeb is a sorry excuse for a web page builder. I appreciate the drag and drop features when i want to quickly throw together a page of photos, etc. However, it is soooo inefficient in its handling of files. Try building a multi-page site and you will find that images that are the same on every page are stored as duplicate copies within each page's folder. This is just one example of the wasteful way that iWeb builds pages.
Also, what is up with the way that iWeb handles sites??? It moves completely away from the document model so all "sites" have to be open at all times. Who thought this was a good way for the user to interact with the application?
I agree that we are going to see this browser functionality integrated into more applications. However, the browser will not go away because there are many reasons we use the web that have nothing to do with a particular application.
No, Coda wouldn't be a persuasive argument to buy a new Mac for most people. If you're a hardcore tech guy, then yes. In all other cases, it has too much freedom. SandVox or RapidWeaver, I could see.
And WebKit does exactly nothing to break down the distinction between local and remote resources. The syncing framework does, to some extent. Time shifted data is not a solved problem yet. Rich UIs on the Web are not a solved problem yet, either. Data intensive applications are completely resistant to the web model, so far. So I'd wait a little with the excitement ;)
I think it's more than a logical outgrowth, I think it's one that Apple has explicitly facilitated. Apple has not only open sourced WebKit, but they've also encouraged its use in third-party apps, and actually done quite a lot to encourage ports to other platforms (including some massive changes to their code base to facilitate porting). In addition, they are very welcoming and will take time out to answer questions porters have about the codebase and discuss port-specific concerns, rather than just say RTFM or RTFS(ource).
As a result of all this, in a month or two I should be able to have an alpha release of a wxWidgets/wxPython port of WebKit. As it uses wx internally to render pages, it runs wherever wx does (Win/Mac/Linux) and already supports all the main browsing features, and even has preliminary editing support. And wx is far from the only port in development. There are also projects to port WebKit to GDK, Qt, Win32, etc. Adobe's Apollo program used this cross-platform system to port WebKit as well.
So I think it's fair to say that there is a large demand to use the WebKit engine in many different toolkits and platforms, and that Apple is helping greatly to facilitate this. I think you're right that we will see a more seamless integration of desktop and web; from a software developer's perspective, I'm already seeing a high demand for such integration. And WebKit makes it easier than it has been in the past to get such integration.
"Wouldn't Coda be a persuasive argument for buying a new Mac if it was bundled in? I think so."
What does bundling have to do with it? Coda is not for everyone. Those that want it can just buy it.
Before you go ragging on the 'young' with their faddish idea -- the reason this idea keeps coming up is because document centred computing is quite obviously conceptually the way to go. Our current technology simply isn't up to the point where it's practical, because of all the interoperability issues between various types of application. As APIs become more standardised (and they do, every year) across the industry document-centred computing will eventually make a big comeback one year. But not this year.
Keep in mind that the mouse-based point-and-click interface was first thought up in the late '50s. It took thirty years before it became anything other than a pipe dream. But that doesn't mean that those who believed in it in the intervening time were "ignorant" -- in fact, today they're seen as visionaries.