In the meantime Wikipedia is still the most useful product of the Internet and the biggest database of human knowledge. It's only surpassed in traffic by Google, Facebook, Youtube and Yahoo.
the web serves little purpose other than a
clumsy virtual machine
That clumsy virtual machine can bring you Facebook.com on every mobile phone, regardless of the platform. And if good native apps are now available for Facebook, well, we are talking about Facebook.
"""In the meantime Wikipedia is still the most useful product of the Internet"""
You missed the point.
There's nothing Wikipedia does that cannot be done in an native-gui mobile App.
It's just a list of articles, a way to search for them, and a system to edit them. Nothing web particular about Wikipedia, one can imagine a Wikipedia app for mobile and desktop OSs that achieves the exact same thing.
There is something Wikipedia does that is not usually done in a native-gui app, and that is provide destinations for hyperlinks. Part of its utility is that you can link to it from blog posts, from essays, from all sorts of stuff.
Which is the point of the OP. he’s comparing micro-walled-gardens to linkable web sites. If we are talking about a native-gui app or rich internet app that supports linking, you may have something.
Then again, you may not. It is incredibly annoying to me whenever I try to access an IMDB link on my tablet and it asks me to install its native app. I now no longer to go imdb links because I’m tired of the speed bump.
This is the web's biggest advantage, it is frictionless. You "install" Wikipedia the same way you use it, by typing its URL. I dont think native binaries can ever match that.
Only, I don't use Wikipedia "by typing it's URL". I just visit it that way. I use it by entering a query in the search box.
But with things like the App Store, or even apt-get and the like, native binaries can very much match that friction-less "installation".
One major benefit that the web has is that:
1) you don't mess with your system (when using a web app)
2) you get automatic updates (when the server code updates)
well, with the sandbox/app store model that mobile and native apps are going, the same can also hold true to them.
Is there a single point of the web that cannot be replicated in native-land? I'll take Apple's App Store as an example, but you can substitute things like the Android store, apt-get, whatever:
Easy discovery: App Store search, plus curation (same as web).
Easy, no mess, installation: see App Store / iPhone App Store.
Automatic updates: App Store updates
Use web technologies: no problem, just add a webkit view in the app.
Performance: better for native.
Integration: better for native.
Look and Feel: better for native.
Ubiquiteness: you can have the same app in your laptop / phone / tablet Plus a web interface as a last resort (for example: Evernote).
When laptops / smartphones / tablets were rarer, being able to access, say, your web mail, from anywhere with a connection without installing anything, made for a good proposition. Now, though? I, for one, only use the web Dropbox / Evernote / Gmail / Twitter interfaces as a last resort.
native binaries can very much match that
friction-less "installation".
That will never happen because of the various differences between clients. This is the magic of the browser - people bitch about cross-browser incompatibilities, however because of the browser's simplicity it gets as close to "frictionless" as possible.
And you won't replicate that unless you invent something that's for all practical purposes still a browser.
(you mentioned deb, try upgrading an Ubuntu box sometimes)
Is there a single point of the web that cannot
be replicated in native-land?
This is the wrong question to ask. You're not seeing the forest from the trees.
The web is great because of its extreme simplicity. Trying to replicate that would get you yet another browser-like technology, that's broken in new ways nonetheless, or it will make you deal with even greater complexity - I don't know about you but it was extremely painful for me to even get a Hello World installed on my own iPhone.
Either way, talking about "replication" is missing the point.
One major benefit that the web has is that
The benefit you mention isn't even the main benefit of the web. The main benefit is that the web is available everywhere for everybody, with no discrimination.
When you send a link to a friend, you don't ask him whether he has access to your App Store or not. The web's availability is implied (after all, that's probably how your friend received the link in the first place).
Ubiquiteness: you can have the same app in
your laptop / phone / tablet
Seriously? Ever attempted developing for iOS, Android, Symbian, WinMo 6, WinMo 7, Samsung Bada, Blackberry, OS X, Windows, Linux?
I for one have stopped using anything but the web GMail interface on any device that's not a mobile phone, because it's that good.
"""That will never happen because of the various differences between clients."""
That has happened already. I can install, say, Evernote on my Desktop Mac and my iOS phone with a few clicks (as fast as typing: evernote [ENTER] on the location bar of a browser).
"""The web is great because of its extreme simplicity. Trying to replicate that would get you yet another browser-like technology, that's broken in new ways nonetheless, or it will make you deal with even greater complexity - I don't know about you but it was extremely painful for me to even get a Hello World installed on my own iPhone."""
Hundreds of millions of people don't seem to have problem downloading and using around a billion apps on their iPhones. Same for Android from what I hear.
(And who said anything about developing for the platform? The "Hello world" thing you had trouble with is not about installing/using a ready made app from the store).
"""The benefit you mention isn't even the main benefit of the web. The main benefit is that the web is available everywhere for everybody, with no discrimination."""
Yeah. In a future where most people carry mobile devices/tablets etc, that could well change, in the same way where now you can't take Gopher access for granted like you once could among internet users.
"""When you send a link to a friend, you don't ask him whether he has access to your App Store or not. The web's availability is implied (after all, that's probably how your friend received the link in the first place)."""
Yeah. But I don't go to most web apps because of links send by friends. Why would I be expected to install apps because of links send by friends?
You know something else that is also ubiquitous besides Google Docs? A program able to read .doc files. Almost all computers have one installed --and that's of a desktop proprietary format, a rather expensive program, and pre-app store convenience for native apps and sandboxed models. Actually MORE people have Office installed than have ever used Google Docs. Imagine how easier it will be in the future, with application repositories and on demand native app installation.
"""Seriously? Ever attempted developing for iOS, Android, Symbian, WinMo 6, WinMo 7, Samsung Bada, Blackberry, OS X, Windows, Linux?"""
You said about developing? I said you can HAVE the same app. Somebody else will have to go to the pain of developing. Have you tried developing your own Google Docs? It's just as difficult.
Also, most of those platforms don't matter, only iOS, Android, Windows and OS X do, market wise. Consider a future where native app repositories cover those with ease and ample apps (some of the platforms are already covered).
You really think that browser apps will win because they also cater for Linux, Symbian and Bada?
> Only, I don't use Wikipedia "by typing it's URL". I just visit it that way. I use it by entering a query in the search box.
But with things like the App Store, or even apt-get and the like, native binaries can very much match that friction-less "installation".
Explain how? I just tested it, and Hacker News loads in a browser (not cached) in about a second. Please explain to me how a native app can be installed and launched that quickly (including the time to find the app in the app store, etc).
But surely the success and therefore usefulness of Wikipedia is down to the fact that it's available to everyone with a web browser? Apps are only ever going to have a tiny proportion of the target market, compared to the number of general web users out there.
That's what this is about. Of course you can do the same functionality that's made Wikipedia work in an app, but I don't think you'd ever get the numbers needed to make it useful.
In the model I was referring to, the browser doesn't go away. Only the built-in web does. Replaced by a more general virtual machine with reasonably low level access to the underlying system. We are, more and more, treating the web this way anyway. It is becoming a pointless abstraction.
There is no particular reason for HTML, CSS, etc. to be built-in components, but that does not mean the web has to go away completely. If your network location wants to load an HTML document, you can command the browser to fetch an HTML renderer over the network. But you can just as easily serve an application of another kind, depending on your desires and needs.
Normal people never cared about web browsers, but that didn't stop web browsers from spreading like a virus. Most of them don't even understand the concept. If you ask them what a browser is, they'll probably answer Google.
As an example, I love my e-Ink Kindle. The one big annoyance I have is that the browser on it is so awful that it is unusable. And it bothers me because book authors are often placing links to external references in the books I read.
And I just want to read those references. Maybe add a comment or two if it's a blog. Since I'm there anyway, maybe I want to recommend the book on Facebook or Twitter with, you guessed it, a link. Maybe I want to search on the web for other discussions on the topic. And so on and so forth.
It's not about love for the web browser per se. The browser is an awful virtual machine. Also the HTTP protocol is one of the featureless protocols in existence. The development model is scary sometimes as you've got to learn dozens of technologies, each specialized for a certain task, each broken in its own special way.
But developers bitching about this are missing the point. It turns out that the openness of the web is disruptive and simple form elements communicating over a simple protocol is all you ever need for 80% of the use-cases.
Quite the contrary, I'm seeing the web stronger than ever before. The one big problem I'm seeing however are walled gardens. Some day from now we'll look back and actually see what a big mistake this was. But the web as we know it will still be there. Because nothing can really replace it.
