I’ve long wondered what the point of the “Work Offline” option in Firefox is (it’s in the “File” menu). All it really seems to do is take you to a helpful (cough) error screen when you’re working offline that tells you you can’t view the page you’re trying to view till you’re online again. If the page is in the cache, it’ll display it, but that seems to happen only rarely. I jacked my cache up to 4GB, but much to my surprise, that doesn’t appear to result in a noticeably higher cache hit rate.
IE has a similar option, but it actually does something. In fact, it’s quite slick. When you add a page as a favorite, you’re given a “Make available offline” option. Once that option is selected, you can click “Customize…”, which launches a wizard that allows you to mark other pages the favorite links to as available offline as well (useful for sites like Slashdot that primarily link to other sites). You can also specify when the page(s) should be synchronized (manually or on a predefined schedule, such as every day at 5am).
The Blackberry browser has a similar option too, though I haven’t used it much (most of my offline time tends to be on airplanes, and I’d much rather view the web on my laptop). The Blackberry does have a neat feature IE doesn’t have though, which is if you try to load a page that isn’t available offline, it’ll offer to queue up the request and download it once you’re online again. As with everything on the Blackberry, the page will show up in the universal inbox along with email, SMS, missed calls, etc. No more trying to remember what pages you wanted to load when you were offline or resorting to tricks only a hacker could love. Imagining a feature like this integrated with a Gmail/Google Reader combo that supports offline operation makes my mouth water.
As I said in the opening paragraph, I’ve been thinking about this for a long time, but I’ve never actually gotten around to writing about it. What inspired me to do that today? Scrybe, an amazing online productivity suite that has a variety of unique and killer features, including excellent use of AJAX to preserve context as you move around within the application, a novel approach to scheduling across timezones, full integration between the calendar and to-do lists (why don’t more calendar applications do this?), and more. Watch the video for yourself.
The most interesting feature from my point of view, though, is the offline support: Simply select “Work Offline”, and everything works, well, offline, no browser extension required. It even appears that changes will be synchronized when you’re online again (i.e., it’s not just a read only copy). The beta isn’t publicly available yet, but I can’t wait to see how they did this. My prediction: If this works as well as it looks, all web applications will eventually work this way. And it’s good to see that some companies are thinking about the offline problem, rather than just assuming ubiquitous Internet connectivity makes the problem go away in the very near future.
Work Offline is a throwback to Netscape 4, where it actually worked. In a world where the code in question was maintained at all (it isn’t) Firefox would do the same as IE does.
– Chris
On the specific topic of FireFox’s cache, I find myself saving pages (File->Save Page As) only to drop them into a messy, unorganized folder. More recently, I discovered the add-on “ScrapBook”, which solved the problem next to perfection. Articles are saved properly (they look exactly like the original, unlike FF’s save-as feature) in a organized and customizable folder. There are two problems with it:
1. to save into the cache, you need to trigger the action either via the mouse or an unnatural combination of keys; I would love to customize FF’s alt-s macro to do this;
2. when reading a page, FF+SB does not check whether you already have that page locally, compare the timestamp, and display the local one instead of the remote one.
Bob
I suspect OLPC (http://laptop.org) is going to have to think about this as well, as offline use is going to be more prominent there. It’s possible it will take the form of an HTTP proxy instead of a Firefox extension.
My guess for Scribe is that it’s doing something to fill the cache (by pre-requesting pages), then using XMLHttpRequest to get the pages and handle failed requests gracefully.
However, I have no idea how it keeps pending data, except perhaps as cookies. The 4k of input you can put in cookies is enough to queue quite a bit of information, but it still feels fairly limited. That’s the only place I can think of. If they did use that, you’d basically queue stuff from Javascript by setting the cookie, then on your first online request the server app would see the queued information and process it immediately.