The Blurring Line Between PC and Web

The NYTimes has a feature about software development systems that move the Web offline and desktop applications online, with a focus on Adobe Air, which will be released tomorrow. The article has quotes from the developer behind Microsoft's Silverlight (he was a colleague at Macromedia of Adobe's Air guy), and from the head of the Mozilla Foundation about their online/offline offering, Prism.

Security nightmare?

It took data stored on the Internet and used it interchangeably with information on a PC's hard drive.

Am I the only one who frowned and thought about the security issues, when reading that?

Re:Security nightmare?

Hm, that's funny it never even crossed my mind. I guess Adobe has built up such a reputation of trust over decades of reliable security that I just sort of took it for granted.

Re:Security nightmare?

Yes, in fact you're the first person to be concerned about the security of online data storage.

Re:Security nightmare?

Well, it would lock me out of my data. A lot of places I work I have no internet access (not even via mobile -- not allowed to use one in some locations, and I've not found a way to access the net when riding the London tube). I deal with the issues of having data available wherever I am the easy way -- I keep anything I might need on my laptop, and synch to a server when I get back to base. If there's anything I've forgotten, or I need to check email, then I need to find an internet connection. Works anywhere this guy's solution will work, and a lot of other places besides, and I don't need to buy anything new.

Maybe I'm a luddite, but I don't see the point in moving stuff onto the web that's better placed on one's desk or laptop.

You can never remove the need for offline.

There will always be offline applications and the need for them. There are so many situations where access to the Internet is not available.

As for having the web offline... The big thing about the web is the links between the various pages. Using a tool such as HTTrack might well enable you to keep the links between pages, thus letting you have the experience of browsing multiple domains using your web browser, even when not connected. But most people just "save as" which gives a different experience depending on if you save the full page or just the HTML, and depending on which browser you use. (Thus guaranteeing that not all the links will work.)

Anyway, I would love to be able to take all the pages that I have already saved and quickly and easily form them into some sort of net, doesn't anyone have an automatic tool to do this?

(Oh and I need to both register and have cookies enabled to see the article. Fuck that. Can someone post the full text?)

Re:You can never remove the need for offline.

There will always be offline applications and the need for them. There are so many situations where access to the Internet is not available.

True, but the gap between online an offline will blur: desktop apps that query online databases, and web kits that install through the web AJAX-like applications with local caches. The user will no longer be aware of the browser methods to persistently store content found online.

As for archiving visited pages, the best solutions I've found are through Firefox extensions. I've tried [[Google Notebook]] and [[ScribeFire]] (both take care of online storage and thus multi-PC synchronization), though I've heard wonders of [[Zotero]](1).

(1) I think we're not in Wikipedia anymore... you'll have to google them.

Laptop anyone

"annoyed that he could not get to his PC data when he was traveling"... What about a laptop... the Internet data cloud will not be my primary storage area for many years, if ever... it will be a secondary backup location at best. My primary working data will reside on a fully backed up, as secure as necessary, laptop. First level backup is a self managed RAID NAS (which itself is backed up).

Cloudy thinking

Holy buzzword-itis Batman. I am not exactly sure what that article was about, but Adobe's AIR, though a cool product, is no panacea. As broadband, WiFi, 3G and WiMax become ubiquitous we are still on that 20 year+ quest to develop those magical frameworks that let us easily take our apps that depend on network services "offline". The problem is, there were only ever a few use cases that made sense in an offline mode, and in 5-10 years it will be virtual impossible to go "offline".

The future is always on, always networked, and software developers who spend the vast amounts of time and effort required to replicate little portions of their database or webservice in a "local" mode are going to be eaten alive by those who simply depend on the ever increasing reliability, performance, and ubiquity of the Internet.

Expensive thinking.

Yeah this is great, until you're in a part of the world where Internet access is sporadic, slow, and $10/GB. Then suddenly having to download a few hundred megs of non-differential patches per app and needing a connection to "verify" your software is a bit more than a minor inconvenience. It's extremely annoying when software that'd designed for completely non-networked functionality REQUIRES you to hook the machine it's installed (from CD!) on to prove you haven't pirated it. This just leads to people pirating it and distributing the copies to everyone else in the same situation.

While I would very much love to live in your future of free high speed connections that are always there, the future is looking like pay per gig to clamp down on bittorrent, recover costs for universal monitoring of traffic (without need for pesky warrants), and milk people for all they're worth. "Oh, too bad your line went down when the phone company screwed up and you didn't notice when your router swapped over to the 3G cellular backup, that'll be $54,000 this month."

Here's a tip: The US is not the entire world, and companies sell to the rest too. Try telling everyone here in Africa that "in 5-10 years it will be virtual [sic] impossible to go 'offline'", I'm sure it'll be good for a laugh.

Re:Expensive thinking.

Here's a tip: The US is not the entire world, and companies sell to the rest too. Try telling everyone here in Africa that "in 5-10 years it will be virtual [sic] impossible to go 'offline'", I'm sure it'll be good for a laugh.

Even in the US it's laughable. There are huge chunks of the country where there is no cellular service. I know a lot of Slashdot's readership doesn't go camping or drive through the middle of nowhere, but it's important to realize that not everyone lives like that. Using a network to get data is great. Depending on it as the sole source of data for things like navigation (or worse yet, the navigation application itself) is stupid.

what does AIR/Silverlight offer that is new/better

We already have Javascript, Flash and Java - what do AIR and Silverlight offer that is better than those? Faster? Better languages? If the improvement is that they relax the restrictions on file I/O and access to the Internet, then do they have replacement restrictions that protect the user?

Re:what does AIR/Silverlight offer that is new/bet

what do AIR and Silverlight offer that is better than those?

Well, they enable lock-in and generate revenue for the companies that own those technologies.

Oh, wait... You meant for us? Nothing...

Re:what does AIR/Silverlight offer that is new/bet

We already have Javascript, Flash and Java - what do AIR and Silverlight offer that is better than those?

Mostly more control and better programming. OpenLaszlo [openlaszlo.org], which is briefly mentioned in TFA, is an XML/javacript based programming language which compiles to Flash and/or DHTML. It includes a bunch of APIs for things like layout, data binding and server communication, and is one of the easiest prototyping tools I've ever used.

The slogan is "write once, run everywhere", which may be familiar to some older Slashdotters, but it's not too far off the truth. I'm using it now to develop auditing apps for the Nokia N800/810 internet tablets, and it's impressively simple.

If you're interested, I'd suggest you download it and try it, or check out the tutorial [openlaszlo.org]. It's very easy to get started, and the tutorial compiles and runs your code online.

Anyone know how sluggish these apps are?

How much of these applications run client-side? The thought of using a sluggish word-processor turns my stomach - and not the typing, but the menu interaction and so on. It reminds me of my recent cell phones. New and flashy and fully-featured as they are, it drives me out of my gourd that there is a 1/4-second delay when pressing every button. I can't stand that. I have an ancient Nokia - monochrome amber and all - and it responds instantly navigating through the address book, settings, or texting. If these online applications are anything like using a newer cell phone, count me out.

The problem with Adobe AIR

The problem with AIR is that it requires "porting". A website won't just work in AIR, and once it's been ported, it will no longer work as a regular real website, as it'll have dependencies on Adobe AIR. This effectively means that if you as a developer want the best of both worlds, you'll need to maintain two version of your application.

The approach Mozilla is taking with Prism on the other hand (which is also being taken with Bubbles and Fluid, with standardization between these in the early stages of being talked about), is to make available small features which allows a real website to gain some properties on the level of a desktop application when run from Prism, without stopping to work as a website. This is the progressive enhancement approach, which helps keep the web open (any browser can continue to run the application). It's very important for developers to realize this distinction, less the web gets locked into a proprietary realm. (both Microsoft and Adobe would love nothing better than to be the sole gatekeeper to this realm.)

Re:The problem with Adobe AIR

AIR isn't a host for a "website". Its a desktop host for the flash engine that executes SWF files - that is, packaged flash.

Websites can continue to have their needs well served by HTML and JS.

Web applications that need to offer rich client experiences without succumbing to browser compatibility issues can choose to use Flex (which yields SWFs as well). Those same apps can run in the browser and with minimal rework be re-deployed as desktops via the Adobe Integrated Runtime (AIR).

The AIR instances will the have the benefit of using connected and disconnected modes (in addition to having desktop icons, file I/O, systray access, etc...).

AIR is an alternate to the browser-hosted flash engine. Its the desktop container for the flash engine.

Am I just getting old...

.. or does this sound like java applets circa 1998?

s/SUN/ADOBE/g

Re:

he was a colleague at Macromedia of Adobe's Air guy

Funny, I interpreted that as "cleaner with a camera and access to the Air guy's notes" ;)

Re:

The article is from the NYT and foxnews has told us not to read or trust the NYT, so, it don't matter anyhow....

Re:

There's the begining of this sort of thing in development now - Google have Gears, which provides a Javascript interface to a locally stored SQLite database. Try using Google Reader in offline mode sometime, it's the same application, and will synchronise

Also local WAMP and file:///

On the WinXP side, I've been using local web hosting with XAMPP for development for around a year. Works well from a USB drive.

Lately I've also been looking at personal wikis as a kind of outliner on steroids tool. At least one launches its own micro web

Give generously, help find a cure

I, for one, am having a hard time predicting the future.

I realize that you posted as AC to hide your shame. But don't let this condition stop you from leading a normal life.

Every year, millions of people just like you and me discover that they have lost the ability to predict the future.

Science is working on a cure. With the generous help of people whose lives have been touched by this tragic condition, we hope that one day Future Blindness will be a thing of the past.

But we can't say when that might be.

Until then, you can post on /. and predict all the arrant nonsense you want without being answerable for any of it.

Re:

Well... probably most end users will just end up installing both.

I mean, it's not a money decision on the scale of buying your choice of player and a ton of movies.

Re:

And hopefully, both will die a terrible death.

Seriously, why do we need any of them to triumph? Let's forget the proprietary lock-in aspect of these technologies, let's consider that they make your page much less accessible and platform dependent (platfor

Re:

platform dependency on a web page sounds so awkward or bizare to me, really, who would want such a thing?

I think that when a site strays from the w3c's standards, it can no longer be referred to as a "web site".