IEBlog

PingBack from http://asp-net-hosting.simplynetdev.com/internet-explorer-8-is-now-available-in-20-additional-languages/

Could you please let us know when it would be available for Windows 7?

I hit the submit button too early :-). I meant the "Telugu" version of IE 8 for Windows 7.

Undesired Behavior in IE8 - How to fix?

My web application presents a floating "context" toolbar (similar to the latest MS Office) when a user slects text in a certain portion of my app.

I just checked out our application in IE8 after getting some bug reports and saw that IE overlays its own blue arrow icon for Accellerators? in the same general location where my context toolbar appears both obscuring my toolbar and making clicking certain icons impossible.

How do I programatically disable this feature for a portion of my page? or the whole page?

I know I can physicallly stop the text selection but that doesn't help me since I want the user to select the text in order to see the available actions on the toolbar.

Is there an MSDN article or KB article on how to disable this so my users can contine using our application in IE?

When IE8 (incl MUI Packs and compability Updates) will be deployed to WSUS ?

O time de Internet Explorer anunciou a disponibilidade do IE8 em mais 20 idiomas ontem. Com isso, o IE

Is XP IE so different from Vista IE, that MS cannot provide specific language version for both of them at the same time?

@Edward Chapros: No markup or attribute is available to turn off the Accelerator icon within IE8 for a specific section of the page.  

Users may individually turn off the Accelerator icon using the option inside Tools / Internet Options / Advanced / "Display the Accelerator button on selection."

@Polski Sklep: Is XP IE so different from Vista IE, that MS cannot provide specific language version for both of them at the same time?

We've been shipping XP and Vista language versions of IE8 at the same time for most of the IE8 languages to date.

The 5 Indic languages with "Coming soon" in the table above, carried with them significant changes between localized XP and Vista versions (updates to fonts, usp10.dll, font shaping engines). These differences required us to break Vista, XP releases for these languages.

Please stay tuned for this upcoming release.

@EricLaw - are you serious that there is no way to turn off accelerators within a page? That seems like incredibly short sighted design indeed!

The Autocomplete feature has a way to disable it so that you can tie your own much more advanced autocomplete mechanism in but the accelerators are an all or nothing feature?

Wow that really sucks!

Looks like there is a need for a much larger test group so that mistakes like this don't get made when releasing a new browser version.

I have an application that will be affected by this too (to a smaller extent). I guess for IE8 users I'll have to offer up some inline help indicating that they need to turn the feature off or upgrade to Chrome, Safari or Firefox.

The number of issues that I have to deal with just for IE is astonishing.  Can I send my support call costs to Microsoft for re-reimbursement?

noel

@NoelG: Yes, I'm serious.  I'm usually serious, and when I'm not, I'll add a helpful little smiley. :-)

The design here was explicit and thoughtful.  To keep this comment stream on-topic, I'll move our conversation over to my blog in a few minutes: http://blogs.msdn.com/ieinternals/archive/2009/06/03/slowing-down-disabling-accelerators.aspx

@EricLaw - thanks for the write up. I tried to log in to your site to add a comment but it doesn't like my ID.

I noted the bit about how AutoComplete doesn't work and I have to agree.  Not because it is conceptually flawed, but because Microsoft failed to implement it correctly and has thus far refused to fix it.

There is a great write up about the messed up implementation and a workaround for those developers that would like Autocomplete to actually work in IE.

"bug 137 - IE AutoComplete hardly ever stores data"

http://webbugtrack.blogspot.com/2007/08/bug-137-ie-autocomplete-hardly-ever.html

Of course the funny thing is that it works fine in other browsers.  Only IE managed to mess this one up.

On a side note though, welcome back to regular blogging! Its good to hear from you!

...which, in a way, is good.

Now, after 6 months of shoulder patting, could we please have a glimpse on IE 8.5 or IE 9, which will (hopefully) properly implement XHTML, like Seamonkey, Opera, Firefox have done for quite a while and now Chrome and Safari also do?

Because properly supporting XHTML would come along with proper DOM support (and we MIGHT get proper HTML DOM support along with XHTML DOM), and I catch myself hoping that we'll also get proper DOM events support!

Alongside that stuff, me might also get, through properly supported plugins, MathML and SVG support.

One day.

Remember that during those six months when you congratulated yourselves on the (admittedly better than before) IE 8 release, your competition didn't stand still, and not only released equivalently feature-packed browsers, but also started (and advanced) development of faster, even more complete, even more standards-abiding browsers?

Firefox 3.5 has experimental HTML5 video and audio support. Chromium does accelerated 3D in canvas. Opera 10 has better animated SVG.

All of these are now able to use Web technologies (HTML, ECMAscript, DOM, SVG) to do real time rendering of vector graphics, physics, with user input, on even puny systems.

I really hope that during those 6 months of self-congratulations, the IE team didn't stand still. Throw us developers a bone, please.

Mitch 74

Mitch, keep in mind that you won't be taken seriously when you're deliberately obnoxious.

Sharing information about improvements isn't "shoulder patting"-- it's something that web developers have been asking for for a long time.  As for this "6 months" nonsense, IE8 shipped less than 3 months ago.

Your suggestion that the IE team is just "sitting around" is silly; they've already stated (and demonstrated, see Connect) that this isn't true.  Yes, they could blog about all of the experimental non-standard stuff they're trying out (like the features you mentioned) but they've stated before that their plan is to listen to feedback and build an IE9 that meets actual developer needs.  Trying to ship code before the standards are actually finalized is a recipe for encumbering the standards-based web.

@Mitch, have you had a chance to fill out the IE9 survey on Connect yet?  

@Lex, thanks for the note. I've tweaked the settings a bit, so perhaps you can comment on the post now?  Of course, keep in mind that the post really isn't about AutoComplete at all-- it's about user-control vs. webdev-control.

(I didn't realize I'd ever "left" regular blogging, although my stuff does end up scattered across about 4 different blogs, wikis, and groups.  :-)

@Dan

"As for this "6 months" nonsense, IE8 shipped less than 3 months ago."

You're right... but on the other hand, the shoulder patting actually started back in December 2007, long before the release.

Stifu, borrow the cluestick from Mitch and whack yourself with it. Then, rather than complaining that the IE team is making improvements in the browser and talking about them, either be grateful, or take your trolling elsewhere.

@Dan - cluestick whacking? and you are calling others trolls?

Dan you need to relax.  Your comments on this blog have always been extremely MS-biased and laced with a total lack of respect for others on this blog.

Are developers a little tired of hearing about features that shipped in IE8? yeah probably.  Are there some that they may not know of or have overlooked, likely.

However they also know that software development is a never ending process.  You reach milestones, but you never reach completion.

IE has a long way to go to regain its competitive edge and more importantly the respect of web developers that gave up on IE after the IE6 hiatus.  While MS stood silently the other browsers were working away accepting bug reports, patches and code contributions.  MS decided to remain playing the proprietary game and thats fine... but there is a cost and they've seen it first hand.  If you ask developers if they develop in IE - they laugh.  If you ask developers which browsers they'd love to drop support for? they all start with "IE".

The developer tools, extensions and market have moved over to Firefox, Chrome, Safari and Opera.  In this Internet age there is no room for standing still.

MS has learned hard lessons in the last decade.  The power they once had due to size and market monopoly quickly eroded to the point where they can't recover.  Do you search the web on Live? not a chance, you Google (at a 100:1 ratio according to my web stats).  Do you surf in IE? nope! not unless you are stuck in a corporate environment that doesn't allow users access to the best tools available.  Is Wave going to stomp something fierce when it gets out? you bet!  The buzz on the internet is everywhere for web based software but just take a look at the screen shots and web casts none are done in IE.

If you want to give straight up facts to rebut comments on this blog by all means do so.  If you want to add your own perspective and views - go for it.  However there is no need to call names on other people.  They are just speaking their mind in the only forum available to them.

There's a lot we'd like to say about IE, how it is developed, the lack of transparency, speed, progress, standards... the list goes on and on.  We just hope that one day Microsoft listens.

Harold, "trolling" is a verb, not a name.

Many people who comment here, including you, don't bother to do even minimal research and post your fantasies as if they're true.

All statistically credible surveys show worldwide IE marketshare exceeds all of their competitors by at least 2x (>66%+ IE, <33% others).  Anyone who thinks IE is anything other than the most relevant browser in today's world is either delusional or isn't paying attention.  

... And that is why professional web developers (and there are many) are so keenly interested in IE and read this blog to begin with.  Of course, most readers won't bother to read the comments, which represent basically a cesspool of rants and whining.

I'm not sure where "your" web stats come from, but worldwide marketshare for search are Yahoo >20%, Google ~64%, and Live/MSN ~8%.  Sure, good numbers for Google, but far from 100-1.

"web based software" does have buzz, true, but you're ignoring the fact that MS Office demolishes all web-based competitors (wannabes?) by share, and Yahoo & Microsoft have massive leads in providing the most broadly used web applications (YMail and Hotmail, respectively).

@Noel G

Mayby it would be best to allow websites to shut down other browser addons as well.

Get rid of those pesky ad blockers.

Finally Indonesian :)

Sate Ajam!

@Dan - you don't give up do you?

Trolling is the verb - correct, however one who trolls, is a Troll - noun.  When you accuse others of trolling you are in turn accusing them of being trolls.

You also seem to have skipped the part about suggesting that someone should get whacked with a cluestick.  Put that in any context you want - it was a direct insult, you were caught and someone pointed it out.  The correct action would be to apologize for the comment.

As per your other stats: Yes indeed IE has the biggest browser market share but it is shrinking every day.  Developers hate dealing with IE bugs and although Microsoft has made huge improvements developers still need to develop for IE6 and IE7 and supporting those 2 browsers alone takes a lot of time.

Search stats? Everyone will have different values but @Harold quoted his stats (I'm going to presume he is referring to his incomming stats on his web site(s) or application(s) )  I'm presuming this because after I just checked my stats I found them to be similar about 95:1 Google:anything else.

As for email I'd be curious to see the full stats of YMail/Hotmail/Gmail. I for example have an account in all 3 however my Ymail and Hotmail accounts only exist so that I could get an ID for Y! IM and MSN IM chat clients.

What I do know though is that this blog is the primary online interaction zone for talking to Microsoft about IE development (good or bad).

Accept that there are many frustrated users and developers out there that believe IE can be much better.  They may give feedback in a critical way some times but there is nothing stopping a smart project manager from reading in the complaints to determine the most appropriate fixes and new features.

@EricLaw: up until last year, I got plain ASP server errors every time I tried to register to Connect. Now, I merely get "page not found" when I fill the personal information page and submit it.

Then, when I manage to log in (it looks like it) and I use the embedded Live search engine to find a survey, I constantly get an error message: "an error occurred during your search" or "page doesn't exist".

Is it a Live Search error? Is it a Connect error? Is it a Live account error? What? All I get are cryptic numbers and letters arrays that I must submit back to get an answer within 3 days - no idea what the error actually is, but I need to wait 3 days to get an answer, that my previous experience (several of them) with MS staff tells me will be technically correct, yet completely useless.

So, to answer your question: no, I didn't fill the IE9 survey, I couldn't get to it. Reason: unknown, no explicit error message from MS servers.

@Dan: IE8 was RTW 3 months ago, right: code freeze occurred several weeks before that though, and as far as I know Release Candidate means feature and UI freeze. So, ever since RC release, we got a stream of IEblog posts about this and that feature found in IE 8.

- They are informative.

- They are quite thorough.

- They open many discussions that give great insight into what made IE 8 what it is.

But, and that's why I'm complaining, while IE 8 is a great improvement over IE 7, when IE 8 reached RTW, most competing browsers had already implemented several of its new features on top of their own, while fixing and improving their Web standards support.

We are getting advanced betas showing off these new features, that allow rather easy development of exciting new stuff inside our pages.

But we don't get any communication from the IE team about 'future IE', apart from 'have you filled the survey' that require registering (which, in my case, fails) even though developers have submitted for years test cases, wishes etc. that were simply archived when IE 8 came out, and now have to be submitted again.

Chris Wilson said, before IE 7 came out, that XHTML support wouldn't appear in IE before it was actually well supported; that was in 2005, if memory serves me. How far along is it? Has this goal been discarded? Are SVG and MathML and whatever-XML-namespace-gets-your-fancy considered at all? What features are actually being discussed, and what technical hurdles have to be overcome? Will there be some progress on ages-old bugs like constant queries for absent favicons, or '*/*' supported MIME types, or lack of support for navigator.plugins/collections, blank spaces not taken into account in DOM... All those documented bugs, some of which date back to IE 4, some of which require fixing to support XHTML (and HTML5 as XML), don't get a word. No communication. The last post that said what was going to be fixed were the list of CSS bugs the same Chris Wilson compiled for IE 7 beta 1.

So, while I don't mind and sometimes appreciate posts about a great IE 8 feature, I'd like something to come about IE 9 in an open fashion: something that would be open to discussion, where those that use the tool (web developers) can ask those that make it (the IE team) for things they want to know or discuss, and where the IE team can submit new stuff to those that will use it (developers) so that it gets discussed, refined, modified, improved, changed - or canceled.

Yes, I'm being hard; yes, I'm not gushing over an IE8 I don't use outside of testing because it's extremely slow in the bare VM I run it in (while Firefox 3.5 beta4 in that same VM zips along as if running on actual hardware). Yes, I keep asking the same questions to the IE team.

And I still can't get a clear answer, not even the beginning of a hint, and there is no discussion that gets started on it. Am I being obnoxious? Probably. Are my questions unfair and made especially to make IE (and its dev team) look bad? No, I ask legitimate questions any developer on any development platform would ask the platform developers: will feature X be supported on the platform? Will feature Y, that would be very useful and can be found on insert-similar-platform-name-here, ever be implemented?

Ask a C++ developer if, when he tests a Visual Studio beta that can't build "Hello World!", he doesn't ask the reason why at the VS dev team? With this kind of failure, wouldn't a C++ developer consider switching to, say, Eclipse + GCC if the bug isn't fixed by RTM?

Try to load some Javascript code in a webpage when using application/ecmascript as type. IE fails.

This bug is 10 years old: IE can't execute application/ecmascript scripts (proper type), only text/javascript (obsolete). It has been fixed in all other browsers. It still isn't fixed in IE 8 although IE's Jscript engine is ECMAscript compliant, and would perfectly execute most ECMAscripts out there.

Source: NWG, RFC 4329, 2006, Scripting Media Types

It doesn't break any other browser out there, even when they use different engines - so don't get on my case for "browser sniffing". Why is IE different? Does anybody have a constructive answer? What's the rationale for not fixing this bug? And there are many more like this one in IE.

When these questions are asked in chats, they are carefully avoided and no constructive answer appears. Now, this flood of articles on IEblog makes me feel like the IE team wants to brush IE's remaining failures under a metaphorical carpet.

Other browsers, when they reach RTW, usually announce the following in their fora: "Browser version X released! Trunk open to modifications again." Following, usually, the patch sets marked for integration in further nightlies, so that they can be (at least) discussed. Yes, even at Opera's.

Not so with IE.

Mitch 74

> if, when he tests a Visual Studio beta that

> can't build "Hello World!", he doesn't ask

> the reason why at the VS dev team

I don't think anyone is saying that questions shouldn't be asked of the IE team.  I think people are saying that such questions usually leave out important details or convey the wrong impression.  As in your remark, for instance, you imply that "Hello World" doesn't work in IE, which is of course an utterly absurd comparison, given that millions of web developers make IE say "hello world" and the equivalent every single day.  

A better analogy would be someone goes to the VS blog and says "Hey, I can't get Hello World to compile."  And someone replies: "Okay, tell me more."  And the first guy says "Well, I am writing my code using C++0X syntax, which is not supported by VS.  It is supported by some other competitive product which is used by far fewer people and doesn't support tons of things VS does.  But I want to use this syntax because it's shiny and new and I think I should be able to."  And he gets a reply "VS doesn't support that at this time."  And so he rants and rants and insults the VS team and suggests that somehow they're irrelevant and the worst product ever.  

Sure, he's got free speech and all, he can say whatever he wants, but he's not going to get taken seriously.  And if this continues enough, the signal to noise ratio in the fora gets so low that no one bothers to participate in the comments any more.

>Why is IE different?

Because it's been around longer?  Because at the time it was originally developed, standards were nascent at best, and compatibility with Netscape (the market leader) was the most important factor.  "Real web" compatibility obviously still is king, for IE, given the things they invest in.

>Does anybody have a constructive answer?

What are you looking for?  Who gets to decide what's "constructive"?  Is asking questions about arcana "constructive", or does it take away from focusing discussion on the critical issues?

>What's the rationale for not fixing this bug?

Because, obviously, there are many more important bugs to fix.  You've got "text/javascript" which works in EVERY script-capable browser ever, and then you've got some other string which doesn't work in 70%+ of browsers.  Why make a change?  

On the other hand, IE doesn't support addEventListener, a clearly superior syntax supported by everyone else.  Here, in contrast, it makes plenty of sense to push for a change.

I will agree that the nice thing about Mozilla's bug DB is that you can go there and look up all their unfixed arcane 5-year-old bugs and add your own "me too" when people complain about them.  Not that there's really any real value in doing so, but it's cheaper than therapy, and at least the ranting is semi-organized.

@SteveP

"I will agree that the nice thing about Mozilla's bug DB is that you can go there and look up all their unfixed arcane 5-year-old bugs and add your own "me too" when people complain about them."

Doing stuff like that is actually frowned upon. :p

Unless you have something constructive to add to the bug report, simply vote for it.

So when is the IE Blog going to post an appology for the IE6 default search engine being locked on Bing.com?

http://thenextweb.com/2009/06/02/microsoft-forcing-bing-ie6-users/

oh and over here

http://www.techcrunch.com/2009/06/02/oops-bing-is-now-your-default-search-engine-on-ie6-whether-you-like-it-or-not/

Microsoft admits that it happened:

http://www.seroundtable.com/archives/020134.html

But why no news here? where IE development is the #1 topic?

A clear declaration of the oops should be disclosed if thats what it truly was.

And details of when the fix will be pushed out so that we can get rid of Chandler.

@Jane: As you noted, this is a blog about IE development.  The issue you described was not related to an IE bug, but rather a server configuration problem, and it was fixed yesterday (as your last link indicates).

IE7 introduced (and IE8 shares) a new search mechanism which was not affected by the configuration problem.

Of course, if you're really still using IE6, I strongly urge you to upgrade.

I am not a web developer. I am a C++ developer and I don't know much about web development. As a end user, I keep wondering how Opera is a much faster browser than IE8. I just installed Opera 10 Beta and it crawls on my Windows 7. Don't complain of the compatability problems, I am running it on compatible mode for Windows XP.

Just so that I understand this correctly: since one comment mentioned that the five "coming soon" languages take a little longer due to requiring uniscribe and font updates: does that mean, that these language packs will actually provide a uniscribe update for IE? This would be most helpful indeed.

Secondly: Although I really dislike the tone in which some have asked about the development of IE Next, I think, some of the points are valid nonetheless.

To touch on a few points:

- has the bulk of the former IE8 team started to mainly focus on development the next version of IE, are they still busy with IE8-related things (e.g. supporting the creation of these many MUI packs and of the IE version for Windows 7), or have they been switched over to other projects? From outside it's impossible to know. I can well imagine that IE8 still takes quite a bit of attention and it would be interesting to know what is going on.

- are there intended fixes for various long-standing DOM-related and javascript related gripes in the next version of IE?

- if you don't want to implement against unfinished specs - is any support for XHTML, SVG or MathML at least seriously considered for the next version of IE?

- dou you plan an implementation of ECMAScript 3.1 for the next version of IE.

- in case you at least consider implementing against unfinished specs, which parts of the HTML5 draft do you consider the most interesting ones for the next version IE (none is an answer, too)

It was mentioned that there is a survey on connect. Sadly, I could not locate it. I created a connect user but the only thing that I could find was an IE beta feedback program, which is only availalable through an application. Is that what you refer to? If so, then it seems that admittance into the beta program has been rather restrictive and that should limit the amount of feedback that you will get from this.

If not, could someone please provide a link, since I did not see an announcement for such a survey here on the blog?

You have done very solid work on CSS support in IE8. But the question is whether comparable work is going to happen in other areas, where Microsoft's strategic interest might rather be to block development, as with SVG or Canvas. I won't hold my breath, but it would make things a lot less painful for developers.

Sorry for making the comment that long.

Best Regards

Christian

@Jagannath, Opera's new "turbo" mode is designed for slow connections; it sends your network requests to Opera's servers instead of the original server.  Opera's server then super-compresses the response (which makes big images pretty ugly) before sending it down to you.  If you're on a reasonably fast connection, the feature isn't used, so there is no performance change.

Opera 10's script engine was made a tiny bit faster than the old one.

@Christian, in the last chat, IE said they've got a large and growing team working on IENext and finishing IE8 on Win7.  Since Win7's IE8 only has a handful of new features (like touch, which they already put in RC) this suggests that IENext is where they're spending most of their time.  But I don't think they can/will tell you that directly, until Win7 is done for reasons-- which are pretty obvious when you think about it.

@Jagannath: XP mode in Windows 7 is based on a 'legacy' Windows XP install running inside a virtual machine; virtual machines will run slow. If Opera hits a very slow piece of code in the VM, then it will run very slowly.

You should try running the beta directly under Win7.

For example, I guess that IE 8 must hit on a very slow part of Qemu (the VM I use) because it's horribly slow - while it runs quite nicely on a real machine that has more or less the same specs the VM has (same RAM, same video, half CPU clock).

Now, Opera is a bit of a strange case: it's very lightweight (made to run on handhelds), it's quite standards-compliant (Opera Software's head honcho is co-chair at the CSS working group or something), and it runs scripts quite fast.

Its code is also highly portable.

However, for some reason, it seems to hit a ceiling in speed: once you reach a certain amount of CPU power, the browser doesn't feel any faster, and may even feel sluggish compared with the rest of the system. No idea why.

Akhirnya, ada bahasa Indonesianya juga. Kalo bisa ada plug-in seperti Firefox mungkin lebih bagus lagi.

Thanks for the great stuff. Internet Explorer 8 is now available in a total of 63 languages!

I hit the submit button too early :) :)

SteveP, the thing is that Visual Studio 10 Beta 1 actually does fail to compile "Hello World". It is not a hypothetical example.

http://virtualdub.org/blog/pivot/entry.php?id=252

@Mitch 74

I am not running it on XP mode in Windows 7. I installed it on Windows 7 with the compatability mode of XP. You gave a very good reason(probably) for the ceiling of the speed in Opera. But, for a second don't you think, the jist of the answer was that Opera is apparently not as fast as one would suggest? Why do people need to know if the scripts are loaded fast/slow/medium? At the end of the day, does the browser loads the page faster or not matters. Ofcourse for developers it matters how convinient it is to write for the browsers. That I can understand. I stay in India, and could open a site in my local language and see for the difference?

www.eenadu.net . All the browsers pick the right font except Opera. For some reason, I cannot understand why and how people like Opera. Forget IE, but Opera is a joke for me.

Don, thank you for nicely proving my point by providing an example of exactly the sort of disingenuous complaint I'm referring to.

If you don't actually know how to write "Hello, World", you shouldn't be blaming the tool for your own failures.

Bug report for the developers.

IE 8 does not display images in print preview if the image is from an object HTML tag. It does display the image, however, if the object tag is in a page displayed in quirks mode (no doctype). Then the image does display in print preview.

This is a great improvement to IE. Last time I complained about IE8 not adding major African languages like Swahili, Yoruba, Igbo and Hausa to the language pack. This new addition gives me reason to hope that African languages is not entirely forgotten or overlooked by Microsoft.

@Andrew H - that looks like a significant bug.  Do you have a site or url we can look at to reproduce it?  There are other known regressions in IE8 with regards to printing (very slow, form elements, PNG alpha where mixed with filter: based opacity etc.) but I haven't seen your issue yet (since I don't use Object tags).

Be careful!

Safari 4 & Chrome are strong opponents!

In the functional aspects ,I do not see any distinguish;  

however, their performance significantly better than your product.

"safari": Just because you don't "see" any difference doesn't mean there isn't one:

http://www.theregister.co.uk/2009/06/09/apple_security_suggestions/

"Based on a variety of sources, we know that Apple does not have a formal security program, and as such fails to catch vulnerabilities that would otherwise be prevented before product releases," writes Rich Mogull, founder of security firm Securosis and a self-described owner of seven Macs. "To address this lack, Apple should integrate secure software development into all internal development efforts."

@Jagannah: Opera, contrary to other browsers on Windows, doesn't default to Arial for sans-serif font: it defaults to Verdana (like Arial, it is a Helvetica font derivative, but it uses different metrics).

Yeah, it had me miffed too - at first. You can correct this in 2 ways:

- as an Opera user, go into options and change the default sans-serif font to Arial. It should fix it. For you.

- as a Web author, specify your fonts in CSS as such: Arial,Helvetica,sans-serif on all elements that require it.

It's not a bug, it's not an error: Opera may run on devices that don't provide system fonts. By using its own by default, Opera ensures reliable and consistent rendering on most platforms.

About the speed: 'far as I know, Opera draws its own GUI and doesn't use system elements. On the other hand, while Windows 7 got all around faster, a few changes targeted at saving on resources (no more front buffer duplication, for example) make a few operations (such as bitmap blits, I think) much slower than in other OSes; if Opera draws its GUI and hits on a slowed down feature, then it's slowed down by the system.

Solution: Opera programmers will need to change their code: either use system widgets or use different drawing instructions.