Are widgets a dead concept?
Friday, April 28, 2006 4:27:47 PM

One of Flock developers, Zbigniew Braniecki (aka Gandalf), has written an interesting article "What are the widgets for?" in which he expresses his doubts about the usefulness of the whole concept. At first I wanted to post a reply in his blog, but since my answer has become a bit longish, I decided to post a whole entry here.
I am going to refer directly to quotes from Gandalf's article, so I suggest reading it first.
Since the first news about Opera 9, we all knew that there will be a widget system. Since the first preview build, users can touch them. But until now, Opera community and Opera company, failed to clear the signal about what the widgets are for at all.
First, widgets are available since Opera preview 2, not the first build.
In my opinion, they could also add a small car racing game and/or guitar tuner.
No, it's not the same. That argument is like saying you do not get Firefox extensions thing as they could as well have implemented mouse gestures or added colourful tabs instead. Opera developers give you a tool to build cross-platform (and that includes phones with Opera Mobile) mini applications that can access the net, so you can add guitar tuner or racing game if you like, and much more!
I can hardly imagine anyone using widgets on a daily routines (...), or that someone will use Opera because of the widgets.
Again extensions analogy: I can hardly imagine any Joe Smith will use Firefox because of the extensions. He will use it because of its simplicity, security, and coolness factor attached to it. Besides, I do not think that widgets are for the average Joe (though, they are easier to handle than Firefox extensions, IMHO). They are for more technically minded people who like to tune their browser environment (and not necessary the browser itself as with extensions). Sure, you can use it for silly things like fancy display of weather forecast or calendar (and if you find it useful, then why not), but their real value is that they let you organise the small little tasks you do regularly (like checking selected news of a few websites, without requesting the feeds of all of them, looking for some info across many search engines in one place or looking up a word in an online dictionary). And what's more you can do all those things without switching to the browser window. Maybe I am plain lazy, but that is what I call convenience.
I think widgets are a Web 2.0 thing (btw, I hate the term). Browsers are no longer html viewers, they are platforms/environments for running applications too. They are becoming a sort of the mini operating systems. The great thing about them is that you are no longer tied to one OS vendor (beware Microsoft). Well-written web applications (w3c standards, anyone?) behave properly in all major browsers no matter the operating system. You can easily switch the OS and browser, and all your web applications remain functional. This together with the possibility to store your data online gives you freedom. It would be great if the widget thing was somehow standardised and implemented in other browsers to give users maximum flexibility and choice in running mini web applications (read: widgets).
Because it is a pretty new concept for browsers (widgets itself are not so fresh, but until now there was no easy way of writing cross-system mini applications) there is still a serious lack of truly useful widgets. That can make you think they are only little silly shiny things that "you can move around the screen, or disable completely." But they are much more than that!
Oh, and I am sure more widgets will come. Should other browser teams support them by implementing widgets in their products (hint, hint), faster then one might think.
















Zbigniew BranieckiGandalf1 # Friday, April 28, 2006 5:05:13 PM
Look at the blog comments around the world. I read it so often "I'd switch to Opera, I'd switch to Safari, but I can't live without extension X and Y".
"And what's more you can do all those things without switching to the browser window."
Exactly. So why should I choose to use Opera's Widgets over Desktop Environment widgets? Like, why should MacOSX person use Opera's Widgets? Why should I use Opera's widgets over Karamba? (or plasmoids in KDE4)?
Look, that's obvious - every DE will have widgets sooner or later, KDE will have it in KDE4, Gnome will do the same soon, MacOS X have them now, and Windows will stay Windows, but there'll be plenty of tools for widgets. Why should someone use Opera's one? Why should someone keep Opera in ram just for that?
In the same way you could advice a clock. Opera's Clock - super cross-platform clock. But I have a clock in my desktop environment. What's so super cool in Opera's one?
Zbigniew BranieckiGandalf1 # Friday, April 28, 2006 5:11:38 PM
"But they are much more than that!" - can't wait to see that!
"Should other browser teams support them by implementing widgets in their products (hint, hint), faster then one might think" - well, I see no tech limitation that would make a problem to write an extension for Firefox that would make it working with Widget API. It's just a matter of reasons.
And, we're back to unique feature set. It will be very hard for Opera to use Widget as the good example of unique technology, because a) I don't think it'll be widely adopted, b) It's too easy to make Opera's widgets working in Firefox.
Numen # Friday, April 28, 2006 7:40:53 PM
Originally posted by Gandalf1:
If the native widgets are better there is obviously no reason, but if you miss one or two, you can either start to port it yourself or wait for someone do it for you. And what if you change your platform at some point? Rewrite them again? That is why we do need some cross platform widget API based on open standards.
Originally posted by Gandalf1:
Originally posted by Gandalf1:
I would like them not to be a unique feature of Opera and if it is easy to make them work in Gecko browsers, all the better. Now if they are available in more browsers, don't you think their adoption would rise quickly?
Originally posted by Gandalf1:
True, I miss extensions in Opera too, no point to argue here, but I still think that *ordinary* users don't go into hassle of installing extensions. Ok, some may have two or three, but that's it. I've been wondering why there are no super-packs of extensions (like we have video codecs packs). I think it would be a good move to popularize them, but I digress now.
Originally posted by Gandalf1:
Nothing, it's a proof of concept.
Originally posted by Gandalf1:
Sure, extensions go much deeper than widgets, but they're limited to applications supporting XUL and they can break after major updates, something Mozilla developers should really spend more time fixing. In the above text I meant that widgets (as we know them from desktop environments) are a new concept to browsers, nothing more.
The bottom line is: widgets can be useful (they are already present in many desktop environments, right?) and they can be truly cross-platform and based on open standards. Isn't the latter a good enough reason to promote it? And why do we need them in browsers? Because they already support those standards pretty well.
Btw, Gandalf I have often problems accessing your diary, is it me or there are some problems with your ISP?
Jakub81 # Friday, April 28, 2006 8:13:05 PM
Numen # Friday, April 28, 2006 8:29:54 PM
Anonymous # Saturday, April 29, 2006 2:06:33 AM
Anonymous # Saturday, April 29, 2006 10:28:40 AM
Numen # Saturday, April 29, 2006 12:48:09 PM
Widgets are just chromeless applications using open standards web technologies which means they can easily be deployed in other browsers. Besides, they have different security restrictions (allowing them to interact with many different web services) and can store more data on users' systems which makes them more flexible than web-based applications. Discrediting the whole concept shows no faith in web applications, IMO.
@Shadow skill
Extensions are high on my wish list too, but widgets are an important technology for Opera as they're supposed to push development of Opera Platform too, a vital technology in the mobile space.
Zbigniew BranieckiGandalf1 # Saturday, April 29, 2006 5:39:15 PM
Extensions mechanism is way more powerfull than Widgets and thus it allows to do everything what you can do with Widget.
If there will appear any valuable widgets I'm sure they'll be either ported to Fx or there will appear such widget-layer extension.
I still fail to see the reason to add Widgets to Opera, and basing on all arguments I heard the next logic step would be to bundle rich music player into Opera - because now I need to launch separate app to listen to the music, and thanks to Opera I could have easy to use music player without it - it would just require one Opera to rull them all
Numen # Saturday, April 29, 2006 6:10:21 PM
No wonder Opera logo resembles a ring.
Anonymous # Sunday, April 30, 2006 12:27:52 PM
Numen # Sunday, April 30, 2006 2:36:14 PM
Originally posted by Anonymous:
And vice versa. What's the point?
Originally posted by Anonymous:
There is already a usable Minimo project allowing Gecko (Firefox engine) to run on small devices. It seems to have higher hardware requirements than Opera or WebKit, though.
Anonymous # Sunday, April 30, 2006 7:25:36 PM
Zbigniew BranieckiGandalf1 # Monday, May 1, 2006 6:17:43 AM
Anonymous # Monday, May 1, 2006 11:54:57 AM
Anonymous # Monday, May 1, 2006 11:55:58 AM
Zbigniew BranieckiGandalf1 # Tuesday, May 2, 2006 10:54:43 PM
Also, I'd suggest you to first check the cross-platform of Mozilla before making statements like this.
Anonymous # Wednesday, May 3, 2006 7:00:05 AM
Numen # Wednesday, May 3, 2006 1:51:26 PM
I know Opera can run on more phone models right now, but do you have any figures to support your Minimo "runs slow like a dog" claim? Could you also perhaps clarify what you mean by bloated?
Thx.
Anonymous # Wednesday, May 3, 2006 2:19:27 PM
Anonymous # Wednesday, May 3, 2006 2:23:03 PM
Anonymous # Saturday, January 8, 2011 7:12:36 PM