Submitted by Google and Mozilla are now working on adding file-/protocol-associations to webapps. So that when you want to select an image. You can browse your filesystem or use webapps that registered with your browser: One more item which is needed to make the web into an operating system.
WebOS thing really starts to scare me
Me too, especially that I think in this case, there could be a dependency problem. I mean, if you read the article, they use a web app that shortens URLs. So your main application needs this app to work, but if you haven’t registered the URL-shortening app, then you might get an error, correct? Or I didn’t get the whole concept? This is getting a real mess…
Wasn’t ActiveX enough? With all these file APIs by google and mozilla (not just this one), random site can now scan our disks… That make the web a lot safer.
While having a good file dialog without Java or Flash is nice, the upload model should not change (aka, file by file opt-in)
You guys are over-reacting. From what I’ve read, WebIntents is just a web equivalent to how, in Linux, programs can specify what mimetypes they can open in their .desktop files and then your desktop can let you specify which desktop app handes them.
It’s about time we started working to get rid of the 50 million different, mostly irrelevant icons that services like AddToAny expose to each user.
I know Firefox was poking around with an idea for an API to allow web apps to be associated with filetypes. Now I know what became of those goals.
Seems like they may be going about this the wrong way. Instead of opening an image webapp and having it try and figure out where my photos are located, wouldn’t it make more sense for me to navigate to the photo I want to edit, and then when I right click and hit ‘Edit’, just have the photo editor open up (or present me with a list if I don’t have one setas default)?
Also, with desktop apps, it seems like you have several apps fighting over file associations, and some apps will make themselves the default app for all file types that it supports every time you open it. I wonder how this will translate to the web, like always being prompted, ‘Would you like to make JoesImageShack.com your default image storing service?’ every time you visit there.
Edited 2011-08-04 21:23 UTC
I read the article and I think photo-apps can register to be used for when an other app needs a photo or more.
And when you want to pick a photo, you get a list of photoapps which can provide them.
They’ll probably take inspiration from the Linux desktop, since it’s a solved problem there. (MacOS too, I think)
On windows, apps can set file associations. On Linux, they can only announce their ability to handle certain filetypes and it’s up to the desktop environment to provide a non-intrusive means for manipulating them.
Knowing browsers, they’ll probably do what they do for things like password-saving, geolocation support, HTML5 storage, and Chrome’s desktop notifications. (Display a yes/later/never doorhanger prompt)
My skim-reading of the article reads as it being compatible with both models. Image hosting service registers itself with the browser, making itself available to the image editor, while the editor registers in the same way, making itself available to the hosting service.
“This project will fundamentally change and improve the way we build applications on the web today for our users.â€
From what viewpoint? Certainly not mine. The web is a god-awful-mess already and trying to circumvent it’s flaws (or actually trying to extend it with features that does not naturally fit) by using this approach is again the wrong way to go. In my opinion (and I know for certain you disagree) a browser’s only job should be to parse HTML and render it. A browser should perhaps not even be confined to just use HTML to layout things, it could just as well parse and render LaTex.
Any kind of object outside of this scope should be handled in some other way and this includes stuff that requires user interaction or scripting abilities like normal forms, JavaScript, Java applets, Flash et.c. Things that does not naturally fit into the layout part.
What I think needs to be done is to create an alternative protocol that works on the side of the rendering. It would handle any user interaction be it two ways or simply client to server. It would allow you to choose how you want to interact with the website, be it through a gui or the commandline. It would hook into your system in a much natural way so you could assign programs for handling different types of data presented by the website. Imagine having your youtube videos being played directly in mplayer or vlc or writing a message to a friend on Facebook directly through bash (without using any API).
The problem is flexibility. Mixing together layout with logic was a bad move and we are tasting the effects as we speak. If an alternative would be created the web would be a much more pleasent place to be.
My 2 cents.
You mention flexibility and moving things out of the browser or allowing external applications to be used to handle video.
That was already possible, it is called plugins and downloading/opening of a description file which is handled by a player. Like Microsoft WMV has ASF, RealVideo has some other format.
Also browsers like Firefox have already supported the mailto:[email protected] links for as long as it has existed. It was introduced in the Netscape browser if I remember correctly.
You can click on it and it opens your favorite e-mail client. Since Firefox 3, you could also point it at gmail.com. So they actually made it more flexible.
That is flexibilty for the user, what this HTML5 stuff is doing is allow webdevelopers to get flexibility too. With plugins and external players the developers can’t make it dynamic. They have no access to the content. Even though they have access to the HTML and could manipulate that. You suggested this should not be possible ?
Some things actually make it easier for the user. Like uploading a file to a website. Instead of “click browse, choose a file, click upload”; There is now drag&drop on the webpage and it is uploaded and it has a working progress bar in the page too.
Recently people have been adding things to the browser to be able to for example manipulate images or video, here are some examples:
http://html5doctor.com/video-canvas-magic/
A good webdeveloper would allow for a fallback ofcourse, atleast for the non-manipulated video.
Actually, in Firefox. Probably also other browsers, you can right-click and choose save or copy location of the video.
They are also working on a specifications for seperating the subtitles/transscripts from the videos so you can have different languages or leave it out completely.
Maybe make the text bigger or in a different colour so it becomes easier to read.
If it has a seperate specification it is sperate just like HTML (content), CSS (style) and JavaScript (behaviour) is now.
I think that is a better experience than Flash, don’t you ?
This is what you can do with an image-creator:
http://muro.deviantart.com/
http://mugtug.com/sketchpad/
There are also photo-editors.
But why not combine manipulation of images with touch events and create something new ?:
http://www.youtube.com/watch?v=GL2dwXa1_gw
( and other video by the same person: http://www.youtube.com/watch?v=gFmuNApHFec )
It allows webdevelopers to also venture in to other areas. This could also have been made with HTML5: http://dankaminsky.com/2010/12/15/dankam/
If it was, it could have been available not just on the iPhone but on other devices too, now it is tied the iPhone and needs to be ported to an other language and platform to be available on other smartphones.
It also saves people from installing additional software which makes it possible for people to use alternative operating systems without having to miss out on things.
We’ve seen to many times that the software people download includes malware.
So I think that is an improvement, right ?
I don’t know if it will be something people think is useful. Many of the specs aren’t final yet. And many things are still new and not supported by all the browsers. Atleast they are open specifications.
But if people/developers don’t like it, they won’t use it. Right ?
I just know that webdevelopment had stalled for almost 10 years because of IE6 and W3C wanting to make an incompatible XML-language. Now we have HTML5 and it allows for doing a lot of new stuff.
I know many people are afraid it will be flash-ads all over again. I don’t know.
But having these things use open and seperate specifications like HTML,CSS,JS,SVG and so on.
I think things like this will come along and allow you to block out what you don’t want or only display the text part:
http://www.webmonkey.com/2010/06/safari-5s-reader-simplifies-the-we…
HTML/CSS/JS/SVG and so on are really just text formats that can be sliced and diced in any way people want.
The more browsers support it and people use an up to date version, webdevelopers will have to work around less things and content, behaviour and style will be seperated a whole lot better.
We’ve had the possibibility to use different/alternative stylesheets for webpages for years. Certain browsers even had a special button for it.
In Firefox it is now in the View-menu under PageStyle. Problem is that the whole menu-bar is now hidden. Alt-V will pop open the View-menu.
Here is a very simple demo:
https://developer.mozilla.org/samples/cssref/altstyles/index.html
For example these all use the same content:
http://www.csszengarden.com/?cssfile=/208/208.css&page=0
http://www.csszengarden.com/?cssfile=/210/210.css&page=0
http://www.csszengarden.com/
Here is the list:
http://www.mezzoblue.com/zengarden/alldesigns/
I don’t know why anyone isn’t making a browser or more extensions which makes it easier to disable or change certain display properties. Many browsers have an option to not only specify a default background, but also to force that instead of using the webdevelopers choice. Same with fonts and text colours. I however doubt many people use it.
Safari has the ‘Reader’ and Firefox has extensions like this one:
https://addons.mozilla.org/en-US/firefox/addon/reader/
That is a start.
I think the best thing is to complain about browsers trying to hide these things from the user.
Personally I like all the extra screenspace, but I already know where all the menus and options are.
What I really think is a reason why there are so many crap content/sites/designs is that good education for webdevelopment hardly exists. A lot of the webdeveloper are self-taught.
I’m sorry, I’m just rambling. 🙂
Edited 2011-08-06 11:28 UTC
OK, here is a short version:
I think it is good we have more development is the browser/webapp space, that way it can improve and will actually be useful and will work as good as ‘native’ applications on any platform.
Making all those webapps platform independent.