Editor’s note: Thomas Ford is the Public Relations Manager for the desktop browser at Opera Software.
When we introduced Opera widgets in Opera 9, many people raised some very good questions about them. What are they? What does Opera see in them?
First, widgets are a first step towards Web applications in the browser. A widget is essentially a Web page with no browser chrome. If you know how to build a Web page, you will have a good idea of how to build a widget – they rely on standard Web technologies such as HTML, CSS and JavaScript.
To explain why widgets are so important to Opera, I need to begin with how we view the Web itself. We look at the Web as a progression of technologies that evolves organically over time as new technologies are introduced. Technologies like AJAX have transformed the Web from a static content source into a dynamic applications, and those Web-based applications create new ways for users to access, share and shape content.
That’s all well and good, but the other beauty of Web applications is that they can tie content together and merge it with other Web services for convenience, ease of use or greater functionality. When Opera started eleven years ago, our browser was meant to be a perfect window into the multimedia Web – a world filled with rich video, images and content. But now we can redefine the browser. Rather than being a window into a largely unchanging world, Opera 9 is also a launching platform for applications.
Widgets are the first step, but certainly not the last or the only. Right now we’ve got a lot of feed readers and search widgets, but it’s easy to picture the day when these are full-featured Web versions of existing desktop applications. When that day comes, the traditional role of the desktop operating system will diminish. Eventually all of the applications you currently use – word processors, games, even graphics applications – can and will run from the browser. And you won’t have to visit a Web page because it can live as a separate application outside the browser.
But these innovations won’t happen if only Opera drives it. It’s going to take a lot of great creative thinking from developers. They will help define how these applications will evolve and explore what the potential is.
To help encourage developers, we’ve launched two competitions: a My Opera widget competition for developers and the Widget World Cup. The former is a chance for developers to win a MacBook Pro as well as weekly prizes from ThinkGeek for the best written widget. The latter is a fantastic way to get people using widgets and finding out how cool/fascinating/addictive they can be. Developers of the most downloaded widget from each country will win €1,000. Then each country’s winner will face off to find the most downloaded widget. The ultimate widget will win €2,000. We hope that these competitions will attract skilled Web designers and developers from all over. After all, it only takes a few hours to make a widget.
I hope this clarifies a few questions about widgets. I’ll be happy to answer any other widget questions that get posted.
If you enjoyed this post, then make sure you subscribe to my RSS Feed.
using
Tom, do you honestly think browsers will be used to play games and act as Photoshop replacements? Come on… Yeah, it’s possible, but only after the Potion of Youth and Cure for Cancer are invented, and World Peace is attained. Anyone tech-savvy can tell you that.
People who are interested in widgets are already running one of many Windows widget apps – and they run even while the browser is closed. Vista will have its own widgets, MacOS has them… So this is a failed, redundant concept.
Widgets won’t get you new users. Extensions would have – and millions of users, at that.
I’ve been using Opera for the past six years. And I’m sorry to say that you guys have lost it… A major browser release with widgets (oh yay!), but with STILL broken NTLM support, no extensions, hundreds and thousands of regressions, a shiny-fluffy-colorful non-standard-out-of-the-OS skin and a **** native one, no JS debugger, no web developer accessories… Unless you guys get your act together RIGHT NOW, Opera’s market share will continue to fall. It’s too bad – Opera could have been to the web what Firefox is (but better, of course).
Trust me, I’ll never switch to Firefox (I’d rather use IE), but your lack of vision and reluctance to finally grab a decent amount of market share is going to further close the web for Opera – not open it… And we’ll continue masking as other browsers, patching stuff with UserJS, etc…
using
Blah, aren’t you contradicting yourself? Do you want the browser to become a web developer application but not a photoshop app?
I see you’re still using Opera 8.54, the new Opera 9 has been out for a week already. Why don’t you give that a try — most of the regressions that you talk about have been fixed in Opera 9.
using
I updated the weekly Opera Watch poll with a question about people’s perception of widgets so far.
Go ahead and vote!
using
Actually, with all the online services being launched by Google et.al., I don’t find it hard at all to believe that applications will run within the browser eventually. This also fits the picture of a future with all kinds of devices, not just PCs, being connected to the internet. Plus, it is the easiest way to always run the latest software versions (no more patches/upgrades) anywhere (no more installations). The only obstacles I see are related to speed and privacy.
What I would like to know is what differences Opera sees between their widgets and the aforementioned alternatives offered by OSes.
using
Widgets are the natural answer to Opera’s over-complexity.
Those people who want the obscure features can download a widget and Opera doesn’t even have to provide support.
I’m not sure whether IRC, BT etc should be part of the browser or should be widgets. I guess it depends on their memory requirements and intrusiveness to non-users.
Opera Software should work hard now on usability (and fixing the bugs).
using
How do widgets relate to the mobile environment, such as Opera Platform? Where do they fit in?
Is there more planned implementation and support in the my.opera.com developer community for SPARQL? I can imagine some interesting community widgets that use SPARQL.
using
I apologize for not quoting – I’m too tired to do that as well.
Daniel, I’m not contradicting myself. A JavaScript debugger and some basic development utilities, such as the Web Developer Toolbar available for both Firefox and IE, go a *long* way.
It’s insanely hard to test complex stuff in Opera… Trust me. That’s one of the reasons developers shun Opera – when something goes wrong with HTML/JS, or seems to go wrong, it’s a tedious task to see where the problem is. I’m sick and tired of inventing ten different colours for various div-element borders, while in Firefox I just perform a couple of clicks and move the pointer around – and everything is there. Styles, computed styles, inheritance, everything.
At work, we’re liberal regarding some Intranet apps, and people are allowed to use whichever browser they choose, be it Opera, IE or Firefox. However, we have dropped Opera support, because it’s a nightmare supporting all three when two give awesome developer helping tools, and one gives absolutely nothing.
I don’t want a browser to become a developer app (see the Mozilla XUL ****), nor do I want it to become a Photoshop replacement (it won’t ever become one). But I do want people making websites, including myself, to easily see what happens internally in Presto. The reason why lots of new online apps don’t work in Opera is because Opera is an extremely hostile development platform, and is hence shunned.
As for regressions, I was talking about things working in 8.54 that don’t work in 9.0 properly (that’s what a regression is
) – just someone’s insane idea of re-downloading images when saving them to disk (instead of saving from cache), *and* opening the Transfers tab, is enough not to upgrade. I’m going to wait for 9.1, whenever it comes out.
using
Blah, very well put. Bravo! You captured exactly the way I feel about Widgets. They don’t belong to a browser. Any OS has them and there are multiple free third party solutions. The web browser is just that – a web browser. I don’t need to keep Opera open in order to run a clock or a calendar widget. This is ridiculous!
I don’t know what happened to Opera Software, but I have the feeling that some mediocre middle manager somewhere decided in all his/her brilliance that Widgets are the way.
My advice is lose the Widgets, improve M2, keep the browser fast and reliable and possibly implement extensions.
using
Dave, widgets are isolated little “web applets”. Just some HTML, CSS and JS that resides “out” of the browser. You will never see IRC or BT as a widget – that’s impossible. Widgets are stuck with clocks, headline displays, currency converters, weather forecasts, and all that standalone stuff. They can’t interact back with the browser and the pages displayed in it.
using
an Opera :>
http://operawiki.info/WebDevToolbar
using
Do we really want this? A Browser acting as a kind of high-level operating system on top of your real OS? Yet another cross-platform Toolkit?
I’ll lean back and watch the things coming, but personally I don’t think I’ll benefit from this development. But that’s ok to me. I never liked that web-based Mail-Clients, Calendars, Newsreader, etc., but there are many people who actually do it.
I think Widgets are a fine thing, noone has to use them. But if some people do and they like it, it’s a good thing.I actually like the fact, that you may be independent from some companies, that may collect your data with their services. With local widgets there is no problem like that.
Finally: Let’s hope Opera does not go the Emacs way – a good OS lacking a good editor, respective lacking a good browser
(ok, I’m joking at this point)
using
Hmm.. widgets are good. But Opera has got its priority slightly wrong. It would have been better if it had released a API/SDK for developing extensions with v9. Without extensions Opera is gonna have a hard time getting firefox users to switch.
Widgets are usefull, but are so limited!
using
Blah, good comments on widgets, thanks!
Personally, I use only clock and weather widget for now. :p
I gave up with WC 2006 widget because text in widgets looks really bad under Windows XP with or w/o ClearType (Mac OS all Ok).
I didn’t use any other widgets engine before on Windows, so it’s my first experience.
And since I always have Opera running on my PC (and my PC is always on) some widgets can be useful for me.
using
Eddie, Widgets will be able to run on mobile environments and devices too.
This is a great advantage for Opera, since the Opera browser could be installed on devices with screens, such as a TVs or IP phones with screens, and have the screen menu be either a webpage or a widget. This would eliminate the need for developers to write serious code for their UI, they can just implement it as a webpage (or a widget), which is easier, cheaper, and could be easily updated on the fly.
That’s where the true power of widgets lies — on mobile phones and devices.
using
Blah, if you follow the web standards, it shouldn’t be a problem. You could debug it using one of Firefox’s extensions (I use the Web Developer extension), and have it work just fine in Opera. Trust me, this methodology works, I’m a developer myself.
using
Daniel-
as a side note- how come my form is getting filled with the last poster’s name? I’ve seen Pallab and Roc in the “Name (required)” section
using
zielsko, the toolbar from Nontroppo’s OperaWiki is nice, but some of the most important things are marked as “external tool, be patient”. That’s simply unacceptable when Firefox ships with a DOM inspector – it’s not even an extension.
Opera’s custom menu/toolbar WDT is decent, but nowhere near what is available for other browsers.
IMHO, opening Opera to developers should have been very, very, VERY high on the priority list, even since 8.00. I’m surprised they’ve even changed the error console… (which still sucks and needs to be a panel or a tab).
using
Eddie, that’s strange. If you post using a nick, only that nick should be displayed next time (done though cookies). I’ll take a look at the server-side code. Thanks for point it out.
using
Daniel, all browsers have quirks. Firefox and Opera don’t behave the same way sometimes since Opera is more standards-compliant – you should know that
But sometimes, there might be an Opera bug/quirk that can’t be found easily without proper tools.
I swear, things were much easier when tables and font elements ruled the earth
Standards-compliance isn’t the be-all and end-all that automagically saves the day. The worst thing one can do is assume Gecko is always right, which is what happens more and more often.
using
Interesting that “Blah” complains about “lack of vision” at Opera, while he at the same time fails to see the vision Opera has for widgets. Interesting indeed.
Reminds me about when everyone was thinking that WAP would be the next big thing, and looked funny at Opera for going for a mobile web browser instead.
More web developer tools would be nice, but they aren’t exactly “visionary”. Like when WAP browsers ruled the mobile space, Opera is reaching out for something new.
I am also wondering about Dave Gould’s “over-complexity” comment. Has he not noticed that most features are hidden or disabled by default? Especially the ones that are not related to web browsing.
using
Dan-
It seems to be working now. But seriously, I saw Pallab and “pallab.net” in my name and website field and then “Roc” as well. That does sound server side to me.
using
I got someone else’s nick too.
But anyway, word processors, games, calendars, etc. are already running off the web. The web is already an application platform.
“You will never see IRC or BT as a widget – that’s impossible.”
Sonny, there are already lots of web based IRC clients out there. All it takes is to make a widget out of one of them.
BT is another story since it’s all about transferring files to your local PC, so it’s irrelevant and a different kind of application.
I find your lack of vision… disturbing
using
Bleh (we should really get decent nicknames
), would you mind explaining Opera’s vision of widgets? As already stated, they exist in all operating systems, on desktop level, either as 3rd-party software or native, and don’t need a browser to be opened. Most are much more capable, because they can display available drive space, or network load, or something else that doesn’t look like yet another analog clock.
Where’s the vision in that? Where’s the innovation? In two years, most of the Windows userbase will be using Vista, and will have no need for Opera’s widgets, since they’ll have native ones. That’s tens of thousands of developer-hours in Opera ASA gone down the drain… For nothing. Will MacOS users enjoy Opera’s widgets, when they already have native ones? Uhm, not really.
Web developer accessories aren’t VISIONARY, but they’re NECESSARY for greater acceptance of Opera amongst web developers.
using
Different widget systems exist for most just one platform. Opera even has widgets on mobile phones (Opera Platform). No one else can match that.
Do you really think the current widget implementation is the last word?
You are complaining about Opera’s vision, but you didn’t even bother to actually read the article by Ford
Are Vista’s widgets based on open web standards? Mac OS X widgets? I think not.
using
Huh?, web-based IRC clients are either Java applets with direct socket access, or they’re communicating via HTTP with a server that impersonates them on IRC. You won’t find a true web-based AJAX IRC implementation, simply because AJAX works over the HTTP, and not the IRC protocol. It’s just not going to happen directly in a widget. Using someone’s public server with a web frontend the widget could talk to, maybe, but I don’t know of anyone stupid enough to do such a thing.
(On a sidenote, I don’t get anyone’s names, but I did lose the poll vote a few minutes ago.)
using
You are missing the point. IRC is possible to do with a widget because it is possible to do with a web page. The widget doesn’t have to be more than a front end either.
Heck, widgets could have specific libraries for each platform to make it possible to read system information, etc. The possibilities are endless.
using
Bleh, there are no widgets (yet) on mobile phones. When they do appear, what good will they be? Mobile phones have Tetris, clocks, calendars and currency conversion, though not weather forecast
Yes, I believe the current widget implementation is the last word. They won’t be made able to communicate with pages in the browser and won’t get extension-like.
Open web standards? For widgets? Who cares, man? Opera’s widgets will only be used by Opera, just as Vista’s widgets will only be used by Vista. Standards are irrelevant here.
using
Sure there are widgets for mobile phones. It’s called “Opera Platform”, and it’s actually being deployed by various companies as we speak. Check out Opera’s announcements on the stock exchange.
Again, you are showing a distinct lack of vision regarding widgets… There’s more to widgets than “communicate with pages in the browser”.
And who cares about open standards? Oh gee, let me think… Those who want to be able to do cheaper application development, cross-platform, etc.?
using
Bleh, if it’s possible to do with a web page, why not – gasp! – enter the URL in the browser and do it that way? You won’t even be limited in your choice of a browser. (Of course, such a web page would be DDoS-ed and used for spamming, hence closed in half an hour at most.)
Widgets won’t get specific libraries to access system info, because Opera would need to provide it, and it wouldn’t be platform and OS-independant. If, on the other hand, Opera doesn’t provide such information, but allows widgets direct system access outside their sandbox, you’ve got a gigantic security hole the size of Jupiter right there. Not. Going. To. Happen.
Repeat after me: there are no possibilities
using
Bleh, if it’s possible to do with a web page, why not – gasp! – enter the URL in the browser and do it that way?
Because you are using an application, not a browser. Read my comment again. Read about Opera Platform. Get educated before making statements, for chrissakes.
Repeat after me: there are no possibilities
Repeat after me: Your lack of vision is not Opera’s problem
using
Something went wrong with the formatting… Trying again:
Because you are using an application, not a browser. Read my comment again. Read about Opera Platform. Get educated before making statements, for chrissakes.
Repeat after me: Your lack of vision is not Opera’s problem
using
Bleh, I’ll just write this, and then I’m done talking to you.
As of now, there are NO widgets for the Opera Platform.
If you think there’s a Grand Vision Regarding Widgets, please explain it. “There’s more to it” isn’t going to work and it’s not an argument. Either give some of your thoughts and insights, or don’t say anything at all.
Opera’s widgets are proprietary. They’re not going to run in Fx, they’re not going to run in Safari or IE, nor Galeon, nor Flock, nor Konqueror, nor Epiphany, nor Shiira, nor Camino, nor anything else but Opera 9+. There are NO open standards regarding Opera’s widgets. They’re proprietary development for a browser with 1% market share on the desktop, and they’re not going to increase that market share. There’s even no guarantee they’ll run on Opera Mobile.
using
Widgets are applications. Are you saying that those who are deploying Opera Platform do not have any applications?
This is not about me. It is you who are assuming that there isn’t, and that Opera hasn’t a clue what it is doing.
Perhaps you should actually read what I’m writing then. So far all you have done is to ignore it or dismiss it without any decent arguments.
They’ll run on any platform Opera runs on.
Sure there are. Widgets are normal web pages without browser chrome. Did you even read Ford’s article?
Again, your lack of vision is your problem, not Opera’s.
using
What’s really funny is the way Blah complains about a “lack of vision” on Opera’s part, and then comes up with his super duper solution for Opera:
A web developer toolbar!
Hilarous.
A web developer toolbar. What a visionary!
But wait, there already is one!
Never mind then, ROFL…
using
Daniel—The problems with the wrong posters showing up in the form could be related to caching of some sort. It could be saving the formatted page as provided to the last person to comment. It doesn’t quite fit, since WP-Cache (if you’re using it) splits the cached results up by cookies, so all anonymous users get the same cached copy, but people who have commented or are logged in get their own personalized copies.
But that kind of function would be where I’d look. (Sorry to be off-topic.)
using
Kelson, yes I was thinking of that too when I came home from work. I do in fact use WP-Cache on this site. I will look into that further.
using
@Blah: for JS-debugging see http://my.opera.com/community/dev/jsdebug/
@Blah: Applications for Opera Platform are not necessarily called Widgets. They are applications and they will be delivered by the device manufacturer/seller. You will never know about them if you don’t buy such a device. And as part of concept you maybe even don’t notice them if you are running such an Opera Platform application.
@Blah: Some of the Widgets already work in Firefox. All of 4 tested Widgets were at least shown somehow, althoug most didn’t completely work. leo_widget-1_1c.zip was functional. Except for the chrome features (always on top, managing of widgets etc) Opera Widgets are theoretically able to run in different browsers (as ever maybe with some tweaks).
@Blah: You’re free not to use widgets. Maybe you’ll be surprised some day (I love Opera surprising me with some nice or useful feature I didn’t discover until now). And in the beginning of Opera Widgets I didn’t think I want to use them, but I tried for testing reasons. And then I really enjoyed Pandora and one day after I set up TouchTheWire for my sites I saw them down (due to server changes I had to adapt my .htaccess …). I have to admit I kind of already like some small & fast Opera Widgets (I don’t like system widgets as they tend to slow down my system – for the same reason I don’t use Pandora any more).
@Daniel Goldman: Prefilled fields had been wrong for me and they are correct again now. Thanks.
using
This is just too naive. Turst me, while I follow web standards, Opera sometimes just doesn’t give what I want. I consider myself as an expenienced web programmer that follows the latest web technologies.
using
I think this widget idea is pretty innovative. There are all sorts of web apps out there already. SolutionWatch just covered 75 different web applications just dealing with note-taking and organization:
http://www.solutionwatch.com/368/fifty-ways-to-take-notes/
http://www.solutionwatch.com/450/25-to-do-lists-to-stay-productive/
Imagine allowing all those apps to run natively, without needing to connect to a website. I think Opera has something there, and am looking forward to seeing the results of one of these service providers porting their app as an Opera widget.
Ok, maybe graphics-intensive games will never be ported as Opera widgets, but those are hardly the only kinds of applications out there. Productivity applications, spreadsheets, and other similar apps are the most natural candidates for being wigetized. So there’s a lot of potential there.
I still hope Opera will allow an easy way for extensions to be written for it, in some future version, as I think that’s really what’s driven Firefox’s popularity.
using
Well, this joe-average guy who just uses a browser to browse finds Opera 9 the most ingenious piece of software going. The widgets are great. I’ve tried Yahoo’s offerings and they were a drag in terms of memory and Java conflicts! Opera’ widgets work elegantly and excellently in Opera 9 and the beta version of 9.01 currently available. Vista is not available right now so I can’t compare to its offerings.
As for Opera’s lack of extensions, I find that Opera needs far fewer extensions than Firefox or IE, of course, because it has tons of well-integrated features/applications that just plain work!
Read this for more thought-provoking commentary on extensions: http://my.opera.com/Andrew%20Gregory/blog/show.dml/232399
using
Web 2.0: Mashups and Widgets (Why Opera
mightshould become the most important Web technology company)In the best tradition of self promoting, long winded comments on someone else’s blog:
The success of widget technology in general and Opera Widgets specifically is based on a paradigm switch.
If you use widgets just to reimplement “old” stuff, then widgets are irrelevant. How many clocks are needed? even if they’re cool because they’re done in SVG… it’s a waste of effort.
But if widget technology enables you to accomplish something breath-taking, something previously unattainable, then widget technology becomes as revolutionary as the web browser was in 1994.
So what makes Opera Widget technology revolutionary? I’ll answer that in a few paragraphs but first let’s look at Yahoo Widgets and Apple’s Dashboard widgets and contrast them with Opera Widgets.
Yahoo Widgets is a browser-free widget system. You write code in JavaScript and they’ve implemented a declarative user interface format. So the widget engine reads your UI specification and creates a UI for you. The engine then excites the JavaScript. Since JavaScript has good network libraries, you can call servers to get information. They’ve also put special hooks to call operating system calls so you can query the system to find out the battery like on your laptop. So with Yahoo Widgets, there’s no browser. And it currently only works on Mac and Windows.
Apple’s Dashboard widget technology uses the underlying web browser technology called WebKit– on which Apple built the Safari browser– and provided extensions to HTML and JavaScript to enable optimized 2D/3D graphics to be executed on the widget canvas. Another JavaScript extension allows you to call operating system calls and in fact the native scripting language, Applescript, from your widgets. You can also create widget extensions in Objective-C to basically do anything that the Mac can do.
Both Yahoo and Dashboard widgets run in their own individual encapsulated mini-windows.
Opera Widget technology implements a third approach. Like Apple’s Dashboard and WebKit, Opera Widgets are based on the browser technology used to create the Opera browser. So that means that you have platform independence. Anywhere Opera works Opera Widgets work. Unlike Dashboard though, Opera Widgets can use SVG and therefore has a fundamental graphics advantage. It can also use plug-ins like Java Applets.
To recap, at this stage of our understand Opera Widgets give us the ability to create a dedicated HTML, SVG, or Java applet app. This dedicated micro-site/mini-app in running on your desktop and not in a browser window. This is evolutionary progress and not revolutionary.
So where’s the revolutionary technology? It’s still just a browser and a web page. It might be cool because it’s in SVG and floating on your desktop, but it’s still a web page. And because of web browser security, your JavaScript cannot connect to a different server from which its page was downloaded. It’s that cross-domain issue.
Fortunately, Opera has dared to change the rules. Opera Widgets can contact any number of servers. This is another difference between Dashboard widgets and Opera Widgets.
And with the 9.0 release, Opera has implemented AJAX so now your Opera Widget can use JavaScript to access those servers.
Another difference between web pages and Opera Widgets is that web pages are always re-downloaded to be executed and Opera Widgets are stored locally and can be reloaded from the widget system. You can start and stop them as needed. And, they can run continuously.
Finally, another major problem is that your JavaScript in a traditional web browser cannot explicitly store data. Cookie data is limited by its storage size and time duration.
If you can access different servers and have long life, a widget is going to have to store its data somewhere. Opera Widgets can store data locally.
So from a widget technology perspective, Opera Widgets provide unfettered web access, cool graphics, data persistence, and desktop deployment.
All of these items are evolutionary advances. Evolutionary, even though no other browser can do these things so well. In effect, Opera has created a new development and deployment system exceeding Java technology in many cases.
So, what’s missing? It’s what will allow Opera as a company to go beyond the browser. It’s what will allow Opera Widgets to become an enterprise standard. It’s what will allow Opera Widgets to easily overtake J2ME in phones.
So what is this missing ingredient? What will make Opera Widgets revolutionary? It’s when users can harness the underlying capabilities of Opera Widgets. And that can only happen if and when users can dynamically extract the relevant chunks of web content– and in the enterprise SQL or web services data– and have it displayed on their desktop. It’s about semantically extracted HTML content of a remote web page (or web pages) and have it dynamically transformed and constantly updated on your desktop.
In Web 2.0, this is called a Mashup because you can mix and repurpose content as you want. No need to recreate the Web’s content, let’s just repurpose it.
Opera Widgets allow a Mashup to be deployed to the user on the desktop today and on a mobile phone tomorrow.
Opera just needs a partner to create a point-and-click Mashup tool allowing the repurposing of Web, SQL, or Web Services/XML data. And if this partner could also generated Opera Widgets with all the JavaScript and AJAX source code generated, even better.
Now that would make Opera Widgets revolutionary. Information at a glance. On your desktop. No browser launching. No scrolling. Constant updates.
I am pleased to announce the <alt> Mashup Designer For Opera Widgets was released earlier this week.
Welcome to Web 2.0: <alt> Mashups and Opera Widgets.
using
Where is that taken from?
using
> Unless you guys get your act together RIGHT NOW,
> Opera’s market share will continue to fall.
Actually Opera’s market share is rising.
using
I see a browser as a tool to browse the web. Not that I think the widgets are in my way, but in no manner do they help me browse the web.
using
@Leo Kennis: Once upon a time people used Gopher to browse the web. Images and multi media content was just something to waste bandwidth. We’ll wait and see and maybe some day Widgets will help you to browse the web.
using
Most of the comments seems to be related to Opera for desktop, and the use of widgets here. I agree that this seems to be a limited innovation from Operas side.
Everyone should, however, keep in mind that Opera for desktop is only a small part of Opera Software as a company. What Opera Software is really about is a browser that runs on mobile phones, tv set top boxed and gameing machines like Nintendo DS and Wii. Opera has revealed that the next version of opera for devices will use widgets (http://www.opera.com/pressreleases/en/2006/05/30/).
Think about is… it makes perfect sense… Event insignificant widgets like a small clock will be a break-through when run thorugh Opera on a set to box or maybe even directly on a TV.
So… forget about wigdets for Opera desktop. It is all about widgets for devices. Widgets designed for the desktop will also run directly on your TV through Nintend Wii (maybe) or through Set-top boxes (for sure) in a years time…
using
I don’t care about widgets, I want create and send html emails in Opera.
using
YtseJam, what do you mean? Where was widgets taken from?
using
Joar, you’re aboslutely correct. This is in fact how Widgets came along for the desktop. It was initially created for the Opera Platform, which runs on mobile phones and devices.
using
Daniel-
Ytsejam (I believe) was asking where Zaid referenced that long entry.
I too would like to know where Zaid referenced that from.
using
Yep, I’m with Eddie.
using
There’s no point OperaSoft investing time in elegant technology if no-one’s going to use it. We’ve seen this happen thousands of times.
OperaSoft need to focus on marketing & usability.
Usability includes being able to go to any webpage that works in IE and have it work in Opera.
Usability includes wanting a feature and being able to find it in Opera eg a “How do I?” search box on the personal bar.
Here’s a slick marketing gimmick, a feature called “Can Your Browser do This?” that lets Opera record your own demo or provide pre-recorded demos that you can put your own soundtrack over & email to friends.
If Widgets are glorified webpages, shouldn’t they act like other Opera webpages ie tabs in the Tab Bar?
Interesting discussion. Glad I checked back.
using
Good explanation, Zaid!
using
> There’s no point OperaSoft investing time in
> elegant technology if no-one’s going to use it.
No-one’s going to use it? Some of the widgets were downloaded over 100 000 times already.
using
I referenced the product documentation for our Mashups Designer for Opera Widgets:
There is a group of like minded companies aligning on Web 2.0 much like in 1993 a few companies commercialized the Web 1.0 technologies. There were 2 browsers, a web server, and a few hundred html documents. But it was the visionary venture capitalist that brought all the technology pieces together to create Mosaic/Netscape Corporation.
Yahoo bought its Widget technology to put their web services on the desktop. Likewise Google put an API into their home page to allow Web 2.0 Mashups. They’ve created special Google-only server-side JavaScript libraries to enable Mashups. On the Google Desktop, they’ve provided an API to access web services and search technologies. At Microsoft’s last developer conference just a few months ago, Sir Bill himself was talking about the next phase of the web being about Mashups or in MS-speak the web Mix and how the next release of ASP.NET and Atlas– the new AJAX developer tool– supports Mashups.
The BBC and Washington Post are allocating their best developers to enable Web 2.0 Mashups of their respective content.
So just about every player senses a dramatic shift in web technology. The Opera CEO in the quick interview posted to this site, was evangelizing this paradigm shift in his own demure way.
As I mentioned in my previous post and the Opera CEO mentions in the interview, Opera Widgets have the potential to change the web landscape. It does this by the creation of a new development paradigm: allowing end users to create their own apps.
And our ability to extract any chuck of web data and place it into an Opera Widget with auto update/refresh seals the deal. It’s about the long reach of technology to data.
Web 1.0 allowed users to view data that did not originate from a corporate database. It allowed non-programmers to create applications. It allowed users to publish data using HTML.
The volume of web data became unmanageable for the horizontal portals and the users rebelled.
In the Web 2.0 world, everyone does not need to create their own web sites to express themselves. You have blogs. And we don’t need a portal to aggregate your content. You can bring to bear semantic user defined taxonomies and “tag” a page to bring attention to it.
ALT Mobile, my company, brings to the Web 2.0 party a point-and-click extraction and transformation capability from any web site or SQL database or Web Service thereby allowing developers to create new content by repurposing other people’s data. Until a few days ago, the deployment targets were usually web browsers or WAP phones. Once Opera released its Widget technology we added commercial support to our Mashup tools to deal with all the technical issues of Opera Widget creation: the XML configuration file, the CSS, JavaScript, AJAX, and even the Opera look and feel issues.
So while Opera Widgets can be used to recreate clocks and the such, we feel that the Opera Widget technologies allow a user to view and interact with the Web in a fundamentally different way. It’s about that Web 1.0 famous concept: content is king. It’s now Mashedup content or just the semantically important bits of a site that you want.
Semantic extraction of web data (and potentially its amalgamation with other data sources) into a new site that is uniquely relevant to a specific group (or social network in Web 2.0-speak) is where the Web has moved.
Yes, we can deploy Mashups to a browser: WAP, SVG, or HTML. And that solves 2 of 3 problems with Web 1.0 data. It solves the content relevance aspect in that you choose which data you want. And the developer empowerment aspect is also solved because we provide point-and-click tools to do all the selction and extraction. The XSL and Java code then gets generated. Opera Widgets solves the 3rd aspect: usability and ease of access because the specific content is in a floating mini-window as big or as small as you want to contain that data. It’s the important data and only that data. It probably came from an HTML table or a malformed HTML list. There’s an irrelevant logo on the top next to a banner ad. Both of which were not extracted because they have no semantic purpose.
Let’s give an example appropriate to every IT organization in every company in the world. In fact, it’s even more relevant because just about every Opera user needs it. From the Opera CEO to those hip Opera testers: Choose any aspect of the release of the Opera 9 and it’s important to someone. The CEO wants to know when his managers say that the bug count is acceptable and the features are in place. A special web page is probably generated with this. The engineers and testers monitor a special web page which is generated from the builds. We monitor a special page with the stable build download link. The press and this blog watch the Opera home page for that “It’s Available” announcement.
Some of this is sent in email or part of an RSS feed. But if someone spent 10 clicks with our tools then we (this Opera 9 social network) would all have dedicated Widgets constantly telling us status without navigating, scrolling or refreshing. An executive information system even though most of us aren’t executives but need simplified access to various data just as much as an executive does.
Right now, I’m tracking a bug in the loading of the Widget config.xml file. I have to look in 5 different places and do a lot of scrolling. The bug exists on some Opera platforms but not all of them. So at least 20 minutes is allocated on launching and scrolling. Web 1.0 gives me a bookmark to the top and I want to drill deep through links. Our semantic extraction technology solves most of this and Opera Widgets solves the resulting visualization, refresh, and update issues.
On the consumer side, here in the US, most schools have a status web page. It tells you if your child’s school is closed due to bad weather. How many parents would benefit from an Opera Widget that displays the current school status. 1 Widget for each child’s school. 3 children mean 3 Widgets. 3 Widgets that are constantly updated means that I don’t have launch 3 browser windows (in IE) or 3 tabs and sequentially view the results. All the data is just there on my desktop waiting for me to gaze. Gaze at the screen and not click and then ctrl-control and then scroll.
So rather than paying us to build widget clocks, let’s pay Opera engineers and managers to create Mashup Widgets which give us insight into the status of Opera releases. My dream Widget would combine the bug status and a URL link to download the fix.
using
Opera’s widgets also are very accessible – did you try to zoom widget yet? Try!
Opera widgets TODO: make images interpolate same way as in normal browser window – I mean smoth.
using
Minhong:
“This is just too naive. Turst me, while I follow web standards, Opera sometimes just doesn’t give what I want. I consider myself as an expenienced web programmer that follows the latest web technologies.”
Funny, that´s what usually happens whith Firefox. As an example, you cannot style generated content as a block: http://www.howtocreate.co.uk/wrongWithIE/eightpointbox.html but, hey, fanboyism is your choice so go ahead…
Way to go “expenienced web programmers”!!!
using
To Minghong and (! Experienced Web Programer): It depends on your starting point.
If you start with Firefox and test in Opera, you’ll sometimes find Opera doesn’t do what you want.
if you start with Opera and test in Firefox, you’ll sometimes find Firefox doesn’t do what you want.
This is because standards compliance is more than just a percentage—it’s a matter of satisfying hundreds of individual conditions. If Firefox implements feature A and Opera implements feature B, the feature you expect to be available will determine which browser you think is deficient.
Picture two overlapping circles. One is Firefox’s standards support. The other is Opera’s. One circle might be bigger than the other, but each covers some area that the other does not.
The fact of the matter is that Mozilla and Opera (and later KDE+Apple) have continuously put work into improving their standards support. They’ve never stopped. Most of the differences in CSS support between Firefox and Opera are trivial*, especially in comparison to what Microsoft offers — and they’re still working toward the same goal (i.e. the specs).
And that’s why, starting with standards-compliant code, it’s a lot easier to adapt that code to work across Opera, Firefox, and Safari than it is to get it to work in IE6 as well.
What this has to do with widgets, I couldn’t tell you.
* Unless you’re working with something like aural styles, which only Opera supports.
using
How about writing a widget, that enables gmail, yahoo or any other web mail application? You could write HTML mails in Opera then.
using
Zaid,
Thanks for that. That’s the utility I’m looking for out of widgets.
Detractors will likely argue that you can simply do this same stuff in the browser chrome, and we’re not going to win them over with that.
What needs to be said about this is that is *a step* towards moving these applications like gmail, writely, all these mashups we’re talking about, banking,… away from the “browser” user interface (address bar, back/forward/home) and more into the “application” UI.
This way things like (for example) “Never break the back button” can be re-evaluated as our complex web applications continue to mature.
Nobody expects a “back” button on a clock widget. What does it mean? What’s it for? But we can all imagine a clock on our cell phone that can take advantage of being net-enabled: example: your cell phone has an (analog) clock widget on it that gets weather information from weather.com and then uses the appropriate weather icon as the background of the clock face. Further- this clock interfaces with google calendar to highlight somehow when your next appoint is… etc.
..and this brings home my thoughts here and in every other widget based thread here on OperaWatch- Economy of UI may not mean too much on the desktop, but in mobiles or devices, it’s crucial to leverage this ability in conjunction all the benefits of rich internet apps.
using
Ugh- that last comment to Zaid was mine.
Still have the problem with the forms being auto-filled in with someone else’s info.
Sorry ReWiz
using
Blah, have You tried the new error console?
I tend to really like it, it gives You very detailed information on all kinds of issues with JavaScript, HTML, CSS and so on, including names of files and line numbers.
At JavaScript errors You can open the faulty file directly from the error console.
Personally, I prefer to work on my code directly rather than editing the temp copy in the browser and then having to find all alterations and port them back to the originals.
Thus, the error console of Opera does it for me completely.
As for the much demanded extensions, I am glad that Opera has not gone down that road so far, since they post a lot of security dangers, such as the activex extensions of IE.
As for standards compliance, to my knowledge, Opera is the only browser that passed Acid 2, so with sticking to standards gives You a really good start.
I too develop on standards, I don’t experience many problems with Opera, some more issues with Mozilla and then IE doubles development time ….
using
Eddie, I added you name to comment #61. Sorry about the mixup.
using
When Konfabulator came out originally, I had a blast downloading widgets. I even wrote a few. But I got tired of having to move windows or close applications or enter the special Widget Mode to get to them. Eventually, I turned them off and never went back. They were never that useful.
My biggest regret about Opera embracing widgets is that the very under-used panel function in Opera is now going to be even less important. I’m a panel power-user. I have all kinds of online tools resting in panels, from color pickers, to weather, resources for writing and web design and even an internet radio panel of my own devising. Panels provide one-button, inside-the-browser access to very useful tools. Widgets provide inconvenient access to (mostly) fluff.
Widgets are toys, not the precursor to the future of in-browser applications. Only the largest of screens will allow them to co-habitate with your main browser window in any useful or meaningful way. Even on my 21″ monitor I have to minimize my main browser window to interact with most of the widgets I tried. Its just not worth it.
Opera Developers: While you are spending all this time creating cool widgets, take a little extra time to place them in a form so that they can live in panels too!
using
@Kelson, that’s what I mean exactly. I didn’t say Mozilla’s implementation is prefect.
@(! Experienced Web Programer), generated content can be displayed as a block. See my webpage’s CSS (search for “body:before”):
http://minghong.f2g.net/skin/default/screen.css
Maybe there are some bugs though…
using
Daniel Goldman, the last days forms have been filled with the name of the last post. After one refresh I always saw my name again.
No problem today, no reload needed (with O9 final, posted with last weekly).
using
For now, widgets in Opera are less than a web page, not many widgets are useful.
using
“For now, extensions in Firefox are less than a web page, not many extensions are useful.”
Gotta love sweeping statements on behalf of everyone else…
using
General comment on widgets:
A widget engine proves as a perfect platform for all these tiny games like snake, tetris, sokoban etc. Coding such games as independent apps is just overkill.
Whether or not Opera has a chance against other widget engines (like those built-in with the OS) depends on different things:
- Does the Opera engine work better than other engines? A first plus is the small installation! A minus is the lag that I sometimes get.
- Is it easier to code an Opera widget than to code a Vista/Yahoo/MacOSX widget? (honestly can’t say)
- Are there enough opera-widgets around? There should be more, but the competition is a good start!
- Marketing
My personal reason to like widgets is the honeycomb snake!
using
how ’bout a slider that zooms as a toolbar button??
using
I was hoping an Opera engineer would have commented on this issue but it’s important to clarify some of the user interface issues that you mentioned. It’s crucial to understand the human factors related other widget systems and how Opera Widgets differ.
Apple’s Dashboard widgets run in a separate layer. This means that you have to enter “widget view” to view/dismiss the widgets. This would be problematic from a usability perspective except Apple enables quick access to the widget layer through hot spots/corners via the Expose technology. In my case, I’ve mapped Dashboard to the bottom left corner so a quick mouse move and widgets appear. I can glance or interact with my widgets and then move the cursor back into the left corner and the widget layer disappeas. Rapid and efficient. Furthermore, the widget layer is trasparent enabling me to click on an app and the app is activated concurrent with the widget layer disappering. Quick and efficient.
Opera Widgets in beta 2 also had a quick access mechanism to the widget layer via a “Opera Widget” tab that would protrude from the top center of the desktop. It provided access to all the widgets as does the Apple technology and was transparent to the underlying desktop.
During a community vote on the Widget forum, this easy access mechanism was voted off and in its place we have the current implementation.
The current implementation provides a “Widget Manager” which is accessed via the panels as are transfers and links. But the funamental change was that each Widget is self-contained and no longer part of a widget layer. This allows you to use the normal windowing system to cycle through your widgets. For example, with Apple-Tab or ALT-Tab. Each Opera Widget may define its own icon and name which enables an Opera Widget to act just like a normal app. So they act like normal windows but they have special capabilities: an Opera Widget can be float above normal apps or be below them.
Clearly, the Opera engineers are attempting to provide innovations in the user interface. Their current approach to Widget invocation is valid and provides an alternative to Dashboard’s integration with Expose. I believe that the engineers will provide both approaches as well as other integration points such as the system tray. These issues will come to light once Opera Widgets are considered a “core” technology. Once Widgets are marketed and adopted not because they’re “fun” but because they provide crucial “information at a glance”.
It is for this purpose we extended our Web 2.0 Mashup technologies to the world of Opera Widgets. We can extract any chunk of web content– combining it with other pieces if needed– and display it in a Widget. Since the Opera Widgets support JavaScript and AJAX without the constraints of XSS we’ve built support for auto-refresh Widgets and the such.
Make no mistake, Opera Widgets on the desktop have the potential to break into the enterprise as the deployment target for a good number of Web Services reporting apps. For internet users, I belive that Widgets will shine once they’re used by the Web 2.0 crowd to view vertical web data. Or in Web 2.0 speak, customized to a specific social network.
For example, a user would use a web browser to view general information on F1 racing but an Opera Widget customized for each one of the 18/19 tracks. Each Widget therefore has a shelf life of about 2 weeks but during those days, that Widget is always up-to-date and the user will consider it as unubtrusive as one does when checking the clock or your cell phone for new text messages: information at a glance.
–Zaid
using
Opera Widgets have the following unique strengths:
1. Uses web development technologies such as HTML, CSS, and JavaScript. Provides XSS-free AJAX support and SVG.
2. Available on all Opera supported platforms.
3. Customizable Widget behavior. Strikes a middle ground between native windowing integration and a dedicated widget layer.
4. The only commercial widget development tools. Provided by a 3rd party (in this case my company http://altmobile.com). Apple will soon release “DashCode”, their own Mac developer tools and we’ll probably extend it to provide Mashup support.
–Zaid
using
I want to copy 100 widgets from my office to my home PC. do I have to reinstall them??? can I just copy them all from my office? where is the widget folder?
using
Just my two cents:
There are numerous Winamp plugins that let you play games – and how many people do you know who actualy plays that games? Interesting thing for someone to code, isnt always interesting for someone another to use.
So maybe widgets will be useful in some distant future, but surely not for moronic clocks or weather forecasts.