im sure that i had ever use the offline mode about 9 years ago, when i connected to internet thru phone line, which is very slowly.
so i just opened many pages and shut down my modem and reviewed the pages carefully later in offline mode.
“I already feel like my home computers are nearly useless without a connection to the internet.”
Yea, I totally agree. I always think its funny when I think that though because we have these great computing machines that people would be amazed with without an internet connection a while ago. Just goes to show how how closely integrated the internet is with our computers.
Offline still matters for people with portable devices (PDAs, phones, etc.)
But then you use Avant Go or something, not firefox.
Well, I work and live in Spain, but we’ve got developers from other countries who really have Dial-UP and code offline, connect, commit, update, etc… then offline.
Jeff, the world is not the US of A. 
Even in Spain, not everybody has got some form of “permanent” connection.
Gezzz not even in the US of A!
You’re seeing the problem from your point of view, where you depend on Internet for your daily work. Try to be offline for a few hours. It helps. It will boost your productivity.
We’re much like Gnome these days, we depend on too many “thinks” found on Internet. We’re bloated.
Bandwidth is a privilege for most of the world.
There’s a subtle point that I think most people get wrong: it’s not “the idea that an application has to be completely functional” without a connection, it’s the idea that an application should not be completely nonfunctional. Still, they should either make this more useful or get rid of it. Dojo is doing some interesting (although last I checked very “alpha”) work on an offline toolkit: http://www.sitepen.com/blog/2007/01/02/the-dojo-offline-toolkit/
No one knows what offline mode is for because the browser has historically been used for relatively passive web browsing. Trying to do that off line makes no sense.
The traditional GUI-OS applications—office, image editing, etc.—are 100% offline apps which have been adding some network features (“publish to Web”). Meanwhile, network applications such as weblogs and Flickr have dominated the online space. But 50% broadband or even a fantastic 90% wifi coverage would not be enough to let web apps realisticially replace traditional apps.
Currently, I use iPhoto to manage all my photos, a plugin with a one-way export connection to Flickr, and Flickr for sharing the photos. I would love to have a single interface for both.
If integrated apps start to be written to dominate both the off and online spaces, then they will have to have a true offline mode. Nothing so clunky as that “work offline” menu item, or a modal export plugin—they will have to just keep working transparently in the 99% of the Earth’s surface that doesn’t have wifi coverage, and do the right thing when I reconnect. This would be a feature of the app, not the web browser. It may have to be something that transcends the difference between the native GUI app and the web browser.
I help run LAN Parties at my High School as the vice president of our Tech/LAN Club, and we’ve always had a problem with Steam when trying to hold CS:S and Halflife tournaments. It’s become so much of a hassle trying to get everything working that we’ve had to phase those games out entirely, and promote other games that can operate independantly of the internet.
I think it is a choice I would miss the minute it is obsoleted.
I, for instance, would at times run Bloglines as an “occasionaly” connected application.
Of course, Jeff, you are right - use “Work Offline” will have no meaning anymore, the more developers don’t care about cacheability of their web pages (yes, MSDN, static content served by an aspx is bad practice.)
It seems pretty obvious that some applications need to be used without a connection to the internet. However, I think the concept of “Offline Mode” is fundamentally flawed. The underlying problem in my mind is that the user has to SELECT that they are in an offline mode. Offline mode is a need that can be easily detected by the software itself, so there is no reason for the user to have to make that decision.
So you don’t program without an Internet connection? Hmmm… interesting.
Sorry, but this is one of your stupidest posts yet.
Funny, for someone who’s blog/website spends a lot of time “offline” I’d have thought that you would have a different perspective.
Here’s how I use offline mode:
-
Save a web page to disk for later reference.
(some time later…) -
Open the file in the browser to read it.
3a) if the browser is in online mode, it tries to fill in the missing pieces, especially images and adverts. Most of the time, the page is not displayed at all until the browser has pulled in what it needs or has decided that content is not reachable. This takes AGES (well, 5-30 seconds, that’s ages enough!)
3b) if the browser is in offline mode, the page shows up instantly.
It’s useful.
Kind of an absurd post.
A scarier thought you should have explored: what happens in 5 years when many more businesses are 100% dependent upon the internet and the government does something to pull the plug or make it very costly?
I have to agree with Vittorio of this one – “World is not US”. If I am building an application I wouldn’t be very comfortable making an assumption that all my users are based out of US / a place where there’s this blazing internet connection that never goes down (not even for a few hours a month).
Imagine if outlook made that assumption, never downloaded your emails and you were in an airport of one of these many other countries in the world which don’t have WiFi access points all over the airport and you needed to refer to that email you received last week.
If you think you need a pure offline mode in your application, consider carefully. Do you really want to bet against the internet?
It’s not about betting against the internet. It depends on what your application does and who and where your users are. After all you want to build software for them, right?
A couple of years ago we did a system in Texas for an Oil and Gas company and supporting offline mode was a primary requirement because a lot of offshore oil rigs around the world don’t have internet connectivity. Rig engineers wanted to be able to work offline and sync the data “by parts” as soon as they get online / get some form scanty dial up access.
Adam’s comment makes most sense. Just don’t make the user select this stuff. If there’s a connection available – use it. If not, assume that the user wants to be in offline mode. In this application I talk off, we used offline application blocks from Microsoft to see if the internet connectivity is available and the remoting server we needed to talk to can be reached.
But now that you’ve got me thinking about this, I’ll go a little further. There are some parts of the world where people still pay by bandwidth they use - e.g. a 1 gig download per month can cost you something, and a 2 gig download actually costs you more. So just assuming that you want the application to download that update just because you are online is not such a wise to do after-all.
Maybe ask the user the first time and then say – “Is it ok if this software assumes you want to be in online mode when you are online?” - or maybe have windows do this so that all of the other softwares don’t have to do this.
I guess, the key-point is that software should understand what I want (ask me once if it has to) but don’t keep bugging me with the same questions over and over again. Or… maybe ask me once per network – because I might have moved from one country to another while I am travelling – or my office network is on unlimited bandwidth plan but my home isn’t - am I taking this a little too far? 
Like I said, it depends on who you are building for. When you’re a Mozilla or a Microsoft building for the whole world it get’s complex and I assume that they in-fact, do consider these scenarios which might sound crazy to a lot of us. I guess this is why the “Word Offline” still hangs there like a there like a vestigial tail
But I do agree with the basic philosophy of your post - there is truly nothing more futile than a web browser without a connection to the internet. Offline mode in a browser just doesn’t “sound right” and frankly, I don’t use that option. Ever.
I use offline when daily commuting on the train to work. It’s a quick way to make my ibm brick useful in the subway. I can always connect during the day to get the latest details. Offline is still very useful.
I wish MORE applications had an offline mode!
It always seems that your internet connection fails (or goes flaky) right when you need the software the most.
I have encountered the problem many times using student versions of CAD and simulation tools and it is EXTREMELY aggravating when your trying to meet tomorrows deadline and your internet cuts out and you can no longer open the application or save your results.
And to reply to the dave about Valve’s Steam software: Why on earth do I need an internet connection to use the Hammer (Map) Editor. It should be a standalone offline executable, just like Worldcraft was back in the day.
Only patch/update things when I am online, but don’t prevent me from using the application when I’m not.
In answer to your question, I pose the following:
Has offline mode ever mattered?
Seriously, how many people actually “downloaded the web” so they could read their pages offline at a later time? How many people “go offline” to wander around their browser cached pages? How many peope ever did?
I’d be willing to wager that, were usage statistics somehow available, they’d show that “Work Offline” was at best a niche, feature, rarely used and even less considered.
I agree that it’s time for it to go. Perhaps it should never have been.
Dragging sensitive data over a network that you (or someone responsible for you) don’t control is asking for trouble. Sure, it can be done with SSL, but an all-too-common kludge to make SSL play nice with self-signed certificates also effectively disables the ability to detect a “monkey in the middle” (disabling CA chain validation, in particular).
Of course, lugging sensitive data around on a laptop hard drive can also be a recipe for disaster – but adding the Internet means new, different bad things can happen that don’t obviously involve theft.
Others have observed that offline mode is useful for secured networks like military ones where some PCs have to be literally stand alone. Even within a military environment, you can have multiple levels of security on different machines and some can exist in your local secure LAN and some others require even greater security. The same is true of public safety systems (I worked on police systems and military systems at one time).
Also, there are a lot of corporate entities that don’t want certain sorts of external connections and seemingly random software updates. This makes running net-required apps problematic. Network admins tend to be controlling types, so letting the software do its own thing? Hahahaha!
Mobile devices like PCs and PDAs of course aren’t alway ‘hooked up’ so offline mode is important.
Many places in North America still use dialup. Even if only 10% of NA was non-broadband, that would still be tens of millions of people. That’s not a market you want to rule out. The dialup I used until recently was 21-26 kbps unreliable. That’s not where you want to serve up heavy apps. I had to stop using an antivirus platform when a program update was packaged in a single 10 Mb file. The download time was just waaaaaay toooooo looooong for my dodgy dial-up link.
Also, you might consider that you have to differentiate between having web or server-serviced apps vs. Internet apps. I make this differentiation because if I run in corporate environments, we may be happy to have a corporate app server inside our security perimeter to serve up approved apps or component updates, but we might not be keen on pulling the same from MS, Adobe, etc.
Further, network integrity even in the broadband world isn’t great. I’ve seen companies lose primary and secondary internet connections due to various construction mishaps. If you have 100 people suddenly without the ability to work due to this sort of event, you’ll be damning Internet apps quite considerably. For these sorts of scenarios, it is vital you can continue albeit with some reduced capability (I hate software development without Google and MSDN for instance). You can, quite literally, sink a business if the timing is bad enough and you couldn’t work offline. I’ve seen a company haul in diesel generators to drive UPSes to drive PCs to get a bid worth tens of millions out the door for the deadline when we had a massive power outage. At the same time, if we’d needed the Interweb thingie for our apps to work, we’d have been pooched.
So yes, offline mode still does make sense.