Skip navigation.

Numen omen

myśli niesparsowane

Are widgets a dead concept?

, ,


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. :wink:

It's evolution, baby!Sprawdź swój polski

Comments

Gandalf1 28. April 2006, 17:05

"Again extensions analogy: I can hardly imagine any Joe Smith will use Firefox because of the extensions."

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?

Gandalf1 28. April 2006, 17:11

"Because it is a pretty new concept for browsers" - new thing? Show me what you can do with Widget that I can't do with Firefox extension.

"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 28. April 2006, 19:40

Originally posted by Gandalf1:

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)?



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:

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.


Originally posted by Gandalf1:

I see no tech limitation that would make a problem to write an extension for Firefox that would make it working with Widget API.



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:

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."



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:

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?



Nothing, it's a proof of concept.

Originally posted by Gandalf1:

"Because it is a pretty new concept for browsers" - new thing? Show me what you can do with Widget that I can't do with Firefox extension.


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 28. April 2006, 20:13

I think it would be great if Opera and Firefox (by extension) would have a compatible widget engine :smile: BTW, Firefox has a widget extension (Firefoxit) but it's poorly made (IMO) and only a few widgets are available.

Numen 28. April 2006, 20:29

Thx Jakub, I didn't know there was an extension like this. True, there are only eight of Firefoxit widgets at this time (including a clock, does it count?:wink:). And it would be pity if we ended up again with many new incompatible widget APIs...

Anonymous 29. April 2006, 02:06

Andy Anonymous writes:

Opera Widgets are not cross platform, they'll only run on one single platform; Opera. Further, I agree with the other critics, I fail to see the purpose of widgets too. Why would I want a dedicated application for something that already runs fine inside a web page? I'd rather compare widgets to the growing trend of 3d accelerated desktop managers; they look oh so flashy, but they just aggravates productivity. I'm still waiting for a useful widget, there's plenty of web-based clocks, dictionary lookups and meta-searches already...

shadow skill 29. April 2006, 03:23

3d accelerated desktops are more than just "flash" they actually allow the cpu to be freed from the task of drawing the GUI itself. OSX has had a hardware accelerated interface for sometime and while it doesn't have some of the more interesting effects of things like Looking Glass but it definetly does aid OSX in performing relatively well even on a 333mhz laptop. I personally think that an extension API would be far more useful than widgets which really do have no purpose in a web browser seeing as how they have existed for what seems like forever for pretty much every operating system that matters at all. I find myself very impressed with Opera and the current beta especially however I think its a bit ridiculous to create widgets as oppossed to extensions, and leave rss feeds and mail screwed up, and put a totally useless feature like tab previews into the program. [It's useless because switching tab focus on mouse hover is infinetly faster so one could switch tabs two or possibly even three times by the time one tab thumbnail is even generated.]I'm afraid that things like mail and rss feeds will be left to rot and truly useful [switch tab on mouse hover.]features will go unimplemented while widgets get pushed.

Anonymous 29. April 2006, 10:28

Anonymous writes:

Of course Opera Widgets are cross platform. The other anonymous user doesn't know what he's talking about. Heck, other browsers could probably easily make Opera Widgets run.

Numen 29. April 2006, 12:48

@Andy Anonymous
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.

Gandalf1 29. April 2006, 17:39

Jakub81: I just said that it should be technically possible to create a Firefox extension that would make possible to launch Opera'swidgets in Gecko.
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 :wink:

Numen 29. April 2006, 18:10

@Gandalf1
No wonder Opera logo resembles a ring.:wink:

Anonymous 30. April 2006, 12:27

Anonymous writes:

Actually Firefox can't do everything Opera can. It's too bloated to be used on mobile phones so you won't be able to use your widgets there.

Numen 30. April 2006, 14:36

Originally posted by Anonymous:

Actually Firefox can't do everything Opera can.


And vice versa. What's the point?

Originally posted by Anonymous:

It's too bloated to be used on mobile phones so you won't be able to use your widgets there.


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 30. April 2006, 19:25

Anonymous writes:

I just disagreed with this:

"it should be technically possible to create a Firefox extension that would make possible to launch Opera'swidgets in Gecko"

Also, Minimo is hardly "mini" compared to Opera. It requires at least a ten times more powerful device to even run the application.

Gandalf1 1. May 2006, 06:17

anonymous (love people hiding ;p) - give me an example of the widget that cannot be implemented as an Fx extension.

Anonymous 1. May 2006, 11:54

Anonymous writes:

Any widget that is supposed to run on normal mobile phones. Gecko is too bloated and slow to run on those in the first place.

Anonymous 1. May 2006, 11:55

Anonymous writes:

Heck, mobile phones or other devices. Opera is on gaming consoles, in air planes, in fridges, and so on. This is a cross-platform coverage that Mozilla simply can't match. No, obscure desktop platforms don't count.

Gandalf1 2. May 2006, 22:54

You're talking about Opera engine, not the widget feature set. You didn't give me an example of a widget, but the example of the environment where the Opera engine can operate. It has nothing to do with widgets.

Also, I'd suggest you to first check the cross-platform of Mozilla before making statements like this.

Anonymous 3. May 2006, 07:00

Anonymous writes:

I already explained that:

"Gecko is too bloated and slow to run on those in the first place."

Mozilla is a desktop browser, and too bloated for the phones Opera is running on. I know that Minimo is an attempt at making it smaller and faster but so far it is still bloated compared to Opera, and runs slow like a dog.

Numen 3. May 2006, 13:51

@Anonymous
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 3. May 2006, 14:19

Anonymous writes:

Opera has a footprint of 1-2 MB on mobile phones. Gecko is aiming for a minimum of 32-64 MB, according to the main developer.

Anonymous 3. May 2006, 14:23

Anonymous writes:

Sorry, Minimo is aiming for that. It still uses Gecko, though.

How to use Quote function:

  1. Select some text
  2. Click on the Quote link

Write a comment

Comment
(BBcode and HTML is turned off for anonymous user comments.)

If you can't read the words, press the small reload icon.


Smilies

Download Opera, the fastest and most secure browser
December 2009
S M T W T F S
November 2009January 2010
1 2 3 4 5
6 7 8 9 10 11 12
13 14 15 16 17 18 19
20 21 22 23 24 25 26
27 28 29 30 31