I was excited about javascript news and downloaded your preview. Then I realized the preview was typical Microsoft FUD. The soduku solver you posted DOES NOT SOLVE GAMES FASTER than Chrome 7. Chrome can solve 1000 games in 0.198 seconds and your IE9 preview takes 0.269 seconds for 1000 games making you 30% slower.
Internet Explorer 9 for Mac Sep 18, 2010 - 42 Comments The Internet Explorer 9 beta is out now for testing, and a reader sent in this picture which is sure to disappoint Mac users all over: Internet Explorer 9 is not available to download for Mac, its for Windows Vista and 7 only. It currently is making the forum not even accessible in IE. Will help you get ie9.
When I run 10,000 games you're 0.008 seconds faster. Why don't you let people solve only 1 game? Because the fewer games the more Chrome beats you? If I ever need to solve 100,000,000,000,000,000 soduku games maybe I'll runIE. For playing 1 soduku game which is what people actually do on the web Chrome is better choice.
I second Hugh Isaacs II. CSS transitions seem the most useful of his requests, to me — easy to author, they can save us from writing lots of JavaScript, and they're a great mechanism for progressive enhancement (usually you don't care if old browsers don't show your transitions).
And the focus on JS speed has always been a little wacky but there's certainly a case for focusing on it out of proportion to how important it is to.today's. apps. It opens up whole new application areas (cryptography in JS, image operations, realtime games, larger client-side DBs) and frees us to use programming practices (huge libraries like Ext, compilers in JS like Processing.js or Objective-J) that we're just not doing much of today.
The great thing is precisely that we can't foresee all the use cases! As always, congrats on moving the Web forward and good luck with IE9.next. ? You know Apple and Google aren't standing still, right?
(Also, the blog post text is blurry on Chrome for Mac 9.0.572.1. Not a clue why. May well be Chrome dev branch's fault. Looks normal with the zoom level set high.).
Has anybody had this same problem with the last couple of IE9 Platform Previews? If I go to Page Open, put in a URL and hit Enter, the status bar gives updates as if things are loading correctly, but nothing displays. I have to either switch to a different window and switch back, or grab the window's title bar and move the window slightly before all the content actually shows up.
If I move the mouse around on the blank page, the status bar will change to show any URLs that I'm hoving over, but the page remains blank unless I do one of the two things above. Other than that, the browser's running great.
Obviously this is a big problem for me, though. As others mentioned already, I'd rather like to see IE 9 improve in CSS support. The CSS text-shadow property has been mentioned first in CSS 2.1 back in 2005 but then moved to CSS 3. According to the Microsoft Connect site there are no plans to add it in IE 9 which is really disappointing because all competitors support it (some for many years).
Also, I'd like to see CSS background gradients as those would help immensely to cut usage of background image stuff which not only decreases page load time but also simplifies development (one markup for all, something mentioned on this blog before) as all other major browsers support it for at least a year. While scripting performance is definitely important, especially looking at the poor performance of IE up to v8, it is not the only usual suspect. Web designers have bickered over the poor IE CSS support for years and it seems that IE 9 will only bring marginally improvements (CSS border-radius and box-shadow) which IMHO just doesn't cut it. IMNSHO, it's not only about making IE faster and giving the customers/Web developers some perks like HTML5 video support. By leaving out HTML5 forms (like input type and placeholder support) in IE 9 (again, check Microsoft Connect to confirm that Microsoft has declined to support it in IE 9) will just give us an inconvenient feeling that before IE 12 is hitting the grounds, there will still be a split in Web development for modern Web browsers vs. Microsoft Internet Explorer. Luckily IE is on a steady decline, with less than 50% in Europe and less than 60% in the US which makes it easier to argument against building special workarounds into Websites to make them shine in IE.
Last but not least, if IE would come up with the FileReader API support we finally could easily build multiple file upload features into Websites without quirky Flash or scripting based hacks. Oh well, but I digress.
IE, even v9, will not be the browser of my choice because it is lacking too much features compared to other modern Web browsers. Sorry, Microsoft but you failed to catch up with your competitors and you should expect IE to drop more market share, even with IE 9 hitting the grounds. I also see the IE sudoku tests execute faster in Chrome 8 than IE Preview 7. Maybe it depends on the processor. I have a developer class Dell i5 running x64 Windows Ultimate. Microsoft only tests on a dual core machine so maybe V8 does better on quad core than Chakra? I agree what matters is javascript performance on actual web applications and not benchmarks but running a few tests here I see Chrome 8 and V8 10% to 40% faster for my sites.
I'm just dropping PIT markers when I enter and exit script and comparing the difference. I only see a 20% improvement in IE Preview 7 over IE8.
Still much work to do on Chakra. It sounds normal that IE beats everyone else at the SunSpider benchmark if some parts of the code are analyzed as dead code and then not run. That's not cheating. However, you have to agree that it's unfair to compare the result with other browsers. You're taking advantage of the fact that the benchmark is not well-designed. Moreover, if you keep saying that you want to be the fastest at 'Real world web patterns', why do you keep running and showing results on a microbenchmark?
Create your own 'Real world web patterns' benchmark! Show how quicker IE is to run actual current websites! In real world patterns, what is the proportion of dead code? How does this impact performance?
Everyone get the qualitative idea ('reducing the size of the compiled program in memory and running the program faster'), but could you quantify this for 'Real world web patterns'? May be because IE team didn't optimize while thinking about this particular test.
If they really wanted to 'cheat', they'd have made sure that some minor variant of this code also triggers DCE to shut everyone up. And it seems this is the first 'iteration' of DCE in Chakra. So it's obviously not perfect (and you can't achieve perfect DCE; at least not with loose js).
'We continue to tune this for IE9, and bugs reported via Microsoft Connect are examples of where the optimization could do more. We continue to tune these and other optimizations for the final release.' If IE9 had a deadcode analysis engine, everyone would be rejoicing to some extent. The issue lies in that IE9 has something much closer to a 'SunSpider benchmark detection' engine. If all code was optimized by this deadcode analysis, i think this would be a benchmark for all browser vendors to take note of. However, every browser vendor is capable of optimizing on a per-benchmark basis.
That is far from stellar and frowned upon. If you change SunSpider slightly (which perhaps makes it unrecognizable as a SunSpider test?), you can then bring IE9 back in line with the benchmarks we have come to expect. IE9 is leaps, bounds and backflips beyond IE8. However, It is still catching up to other browsers in many aspects. Javascript performance is one of those. It probably won't do much but I'll post it anyway. For those asking for new features.
Constantly putting it in blog post comments wont really help. If you really want to get a feature added, the best course of action is to open up a connect suggestion, then post the link somewhere and get all of your friends and like minded people to vote for it. Right now this is the best way to get the IE team to actually listen.
The problem with blog comments is that they are very hard to take seriously, since in my experience, the people who post this kind of comment are the minority who just shout the loudest. This can be even more problematic since the software don't require you to log in with an email account to post comments, meaning you could be the same person posting multiple times.
So use connect constructively, get a really high vote for these features to show a lot of people actually want them and then you will have a higher chance of seeing them. The Chakra DCE 'optimization' handles but not –but not. In short, it handles exactly the set of operators found in math-cordic, and not others. This makes no sense if it's intended to be a general-purpose optimization. Furthermore, Rob also points out that the optimization makes assumptions that aren't true for JS, leading to bugs where it incorrectly eliminates code that is not, in fact, dead. Fixing those bugs will be difficult; they require whole-program analysis or introduction of dynamic guards, both of which will slow the engine down.
HTML5: great work in html5 section elements: JavaScript: it's not my job to rate it. We're speaking about milliseconds! ? Faster than an eye blink. CSS3: I can't believe there's still no multi-column support. And even the new coming Opera will support for CSS3 multi-column.
Please do your best and implement it in the final release with or without vendor prefix, because it's urgent! Nowadays a screen resolution at 1024.768 means a very outnumbered user group. 1280.1024/ 720p is now the general, and 1080p will became general soon. So there two cases: – Fix layout: you see an about 900px wide block surrounded by 1000px wide blank space.
Looks awful, and tiring to read. – Flexible layout: you see a 1200+ px wide block. It's horror to read. Solution for both: /. column-count: 3; — if you need fix number of colums./ column-width: 300px; column-gap: 10px; column-rule: 1px solid black; Simple, elegant, comfortable. Please think about it.
Hi, You're right, If the test was made by me, I will put the November 17 snapshoot of each browser, but as you point this, for sure others will say me why don't pick 4.0b7 or November 15 snapshot that was 4% faster. Only wanna point on the fact they don't get the latest pre-release versions of browser.
IE - Stable 8, Beta IE9 Beta, Platform Preview: IE9PP7 Firefox - Stable: 3.6.12, Beta 4.0b7, Pre-Release: 4.0b8pre. Chrome - Stable: 7.x, Beta 8.x, Developer: 9.x IE has made huge improvements on his browser, compared with IE8 is awesome. If IE9PP7 is better that Firefox, Opera and Chrome Betas it's very good, and with these numbers, it's much possible that it's faster than latest browser snapshoots for sure. The problem is 'benchmark' optimizations, if there are, and we should find a 'real World' popular Benchmark. I really wonder to which situation the term 'real world' refers to. On the one hand, in real world websites there is much dead code chakra has to eliminate.
This implicates, that 'real world' refers to web sites created by medium or low skilled developers. On the other hand it is stated that 'micro benchmarks fail to represent the real world web'. And: 'Real world sites would actually use the much faster and CPU-optimized functions already available in JavaScript engines.' Ah, would they? Isn't it a inconsistency in the article.
I think, micro benchmarks perfectly represent the real world web by using not the best solutions for what they are trying to achieve. And what or where are the 'other worlds'?
I agree, that dead code elimination is absolutely necessary and a good approach. As a professional developer I have been involved in many proprietary closed source but also some open source projects over the last 15 years. I could only see one world, but two types of developers: The ones that nearly never miss deadlines, not being interested in code quality, most often looking for a solution that is 'good enough' to fix a problem, also often using proof of concept code in production releases. And the others, to whom code quality and good robust solutions are more important than deadlines, always trying to improve their skills.
Basically, developers of second type slowly are getting better software engineers. They would be interested to get feedback from chakra to remove or edit dead code before analysis of the code can take too much time and reduces the responsiveness of the page. Is it (or would be be) possible to have a debug mode in chakra that does extensive code analysis and reports it to interested developers? The others are served by optimized code analysis during runtime. That would enable developers to take active part in improve responsiveness of their pages, but doesn't force them to. When working with Visual Studio, I'm using additional tools that help me optimize my code (and not leaving dead code in production releases).
I would be very interested in getting feedback by the original compiler how I can support him to optimize code. @ Hugo, 'And the others, to whom code quality and good robust solutions are more important than deadlines, always trying to improve their skills. Basically, developers of second type slowly are getting better software engineers. ' Or quickly get themselves kicked out of any business. Really, in business, nothing is more important than deadlines. If you don't meet your deadline, well, you may get away the first time, but you'll surely get fired the second time. Or if it's your own business, then you'll lose your customers to competitors who meet deadlines with lower quality code.
Which means, unless you have plenty of wealth to sustain yourself and your family over long periods of unemployment, most people cannot afford to be this 'second type'. The best practice is to produce high quality code within the deadline, but if you are force between a choice of code quality and deadline, well, working code before the deadline good quality code overdue, if you still want the job that is. Those sheep who don't read the entire articles entitled 'IE 9 Caught Cheating?' If you read the entire article, you will find out that the answer to the question is 'NO, it's Not cheating.' Second, this all the more proves that the SunSpider Benchmark is Not a reliable Benchmark, because it's Not fool-proof. You have to fake math-cordic tests such that results will favor your favorite browser.
And you're fine with Google's V8 JavaScript Benchmark test favoring Google Chrome? Well I think they are cheating, too. Those sheep who don't read the entire articles entitled 'IE 9 Caught Cheating?'
If you read the entire article, you will find out that the answer to the question is 'NO, it's Not cheating.' Second, this all the more proves that the SunSpider Benchmark is Not a reliable Benchmark, because it's Not fool-proof. You have to fake math-cordic tests such that results will favor your favorite browser. And you're fine with Google's V8 JavaScript Benchmark test favoring Google Chrome? Well I think they're the ones who are cheating.
Hi there, you wrote: 'We’ve focused on improving real world site performance. We’ve made progress on some microbenchmarks as a side effect. Focusing on another subsystem microbenchmark is not very useful. ' Sure that is true, but wait, why all the other webbrowsers out there are faster on the other benchmarks AND fast on real world scenarios and websites?! Your competitors aka Google Chrome, Apple Safari, Opera Softwares Opera and of course Mozilla Firefox, all are much faster in Mozilla Kraken:, Google V8 benchmark:, Mozilla Dromaeo: (very important because of much tests incl. DOM.!) and Futuremarks Peacekeeper: (very important because of much tests.)!
You don't have to be only fast in real-world scenarios, but in the benches, too! Neither only in microbenches being fast, but not in real-world scenarios, nor being fast only in real-world scenarios, but not in microbenches is good! Regards, iNsuRRecTiON from Germany. @insurrection It important to note that kraken was made by Mozilla, the developers of firefox, V8 was made by Google, the developers of Chrome, and Dromaeo was made by Mozilla, the developers of Firefox. Peacekeeper is unfortunate in that although the one impartial benchmark, is just plain terribly written.
It does reveal some weaknesses in IE, however it is critically flawed in that it contains tests which only execute in Internet Explorer but are skipped otherwise, This bug lowers IE's score dramaticly. @iNsuRRecTiON 'Neither only in microbenches being fast, but not in real-world scenarios, nor being fast only in real-world scenarios, but not in microbenches is good!' I don't see why you think that. Being fast only in real-world scenarios, but not in microbenchmarks is exactly as good as being fast in real-world scenarios and in microbenchmarks. There is no intrinsic value to microbenchmark performance whatsoever.
The purpose of a web browser is to browse the web, not to run sunspider or kraken or whatever the fastest. Microbenchmarks are an attempt to simulate real-world scenarios — ultimately an inherently flawed attempt. The only value in a microbenchmark is in how well in reflects real-world scenarios, so by definition, a browser cannot be good at real-world scenarios and bad in a good microbenchmark. Either the microbenchmark is bad, or the browser is also bad at real-world scenarios. @ giuseppe, '@wetran: you are part of the problem' nope, you are the problem since you don't realize that people need money and must feed themselves. Unless you can convince all the businesses out there to delay deadlines for highly quality code, where most people can't, then you can't do anything by sticking with 'higher quality code deadline' ideology except starving yourself and your children to death. The truth is, 99% of the web developers out there won't be able to find any job around their place if they tell their boss 'code quality deadline', welcome to reality.
High quality code is nice, but not nice enough to get yourself and your children dying for. @wetran You're out of your depth here. You may be feeding your family for a week because the first few gullable customers shell out for you.
Then the negative reviews pour in and your sales will fall. Then you whip up the next revision at lightning speed (based on the bug-ridden release of course, and charging for this Software(tm) 2.0); you'll continue this scheme until a smart competitor sees how you botch things up and.does. fill the market's need with a quality prosuct, making a handsome profit and gaining the public trust whereas you had been running the threadmill for minimum wages and no loyal customer base at all. IE 9 preview 7 result of the Sputnik JavaScript conformance test its getting better and better in every preview but still fail on 76 out of 5246 test. @anonym – hmmm, Microsoft have excelled at releasing average code post deadline and still getting away with it. Microsoft is an Excellent marketing organisation who have built an absolute empire with their uncanny ability to distract the digital heard from all the failings of it's software. Even now, some 30+ years post incorporation Microsoft still hasn't seen a steady decline in profitability, merely a gradual (read slight) decline in profits.
Even the last quarter (year?) produced a sizeable boost in profit taking turning around this apparent degradation (what, 16.2 billion or up by 25% last quarter??). Them, Thar's a lotta short-term grub tuh feed the famuhly. @anonym, '@wetran You're out of your depth here. You may be feeding your family for a week because the first few gullable customers shell out for you. Then the negative reviews pour in and your sales will fall.
Then you whip up the next revision at lightning speed (based on the bug-ridden release of course, and charging for this Software(tm) 2.0); you'll continue this scheme until a smart competitor sees how you botch things up and.does. fill the market's need with a quality prosuct, making a handsome profit and gaining the public trust whereas you had been running the threadmill for minimum wages and no loyal customer base at all.' Clearly you are out of the reality here. What's the use of your high quality code when your family all died starving because you can't feed them for weeks? And it's the customers who set deadlines, and if you can't meet THEIR deadlines, you are out of job already. Clearly you haven't tried to convince your customers and/or bosses to push back deadlines set by THEM so you can produce better quality code, I have tried many times before, and I can tell you it's IMPOSSIBLE to delay deadlines. The customer will threaten to ditch you, and your boss will threaten to fire you, and as a lowly developer, there's nothing you can do.
I could care less how much dough the customer shell out, since they do not shell out for me, but for my boss, or the boss of my boss, or the boss of the boss of my boss. You should go talking to all those bosses and customers who set extremely straining deadlines and force the developers to meet the deadlines by whatever means. I'm talking as one of the developers out there, who have to suffer through all these things, and if you go up to your bosses and tell them you think code quality deadline, you'll not be able to get a job at all. Similarly, a business won't get any customers if it can't meet the deadlines set by the customers. Your 'smart competitor' who produce overdue quality product won't get any customer at all, and will starve to death in no time. The only choices in the real world are either 1) produce a quality product by the deadline, or 2) produce a working product by the deadline.
![Http://osxdaily.com/2010/09/18/internet-explorer-9-for-mac/ Http://osxdaily.com/2010/09/18/internet-explorer-9-for-mac/](/uploads/1/2/5/4/125428576/850728605.jpg)
Producing a quality product without meeting the deadline is NOT a choice, since you'll be kicked out by your bosses/customers when the deadline passes before you finish the product. Maybe you can feed your family with ideologies, but as real people, we need our jobs to feed ourselves and our families, and every business out there are the same regarding deadlines, deadlines code quality for them, and a lowly developer like me and 99% of the developers out there can't do nothing about it, despite how much we wanted and how hard we tried, the only we developers can do is to try produce as good quality code as possible within the extremely short time given to us before the deadline. You really should wake up and face the cold, harsh reality now. @Stifu: 'Good developers don't advocate bad coding practices.' It may be true. Unfortunately, good developers rarely share about their misfortunes. All they talk only about their successes, and not their failures.
I sometimes see missing balances in their stories. Perhaps, it's good at motivating others, but sometimes it's not the whole story to tell the whole truth. Writing a high quality code is a good advice and it's initially a good hope as well.
Unfortunately in this real world, it can be a reality or it can't be. It's good if it could be a reality. But when it becomes INFEASIBLE due to constraints (time, budget, people, other resources, etc.), then one can't deny that he/she can't achive his/her hope. One may hope for the best, but.
can happen as well. So, good hopes and advices are generally wise, but the truths are wiser and undeniable. Nope, I think it is impossible to get fast on all 'real' websites nor to test this. But you can test all microbenchmarks, if you're not good or much slower compared to your competitors (like IE9 PP7 is.!), then something is not ok in your browser/subsystem/JS-engine, etc.
There must somethink wrong, if all other webbrowser are so much faster in these benches, but IE9 is much much slower. They have to be in both real world and microbenches fast as possible, or something is wrong. @Tom Unfortunately no, no web workers, nor web sockets support, I want this, too! And yes, IE9 have to be fully compliant and pass the Google Sputnik Javascript test, as Ali already mentioned! Although improving javascript performance isn't a bad thing, you really need to implement CSS3 transitions. CSS transitions are much more important than a spinning earth or solving a sodukio games 1000 times. Let's think of real world scenarios.
People could actively use CSS transitions in sites to improve user interaction and clarity. If you wanting to solve a soduko game 1000 times why wouldn't you just do that server or better yet who would need to continue to the soduko game after solving it once? @ iNsuRRecTiON @ Ali Do do realized IE9 has the second best score on the sputnik conformance test right?
The latest firefox nightly fails 178 test Safari fails 170 tests The latest chrome nightly fails 130 tests IE9 pp7 fails only 76 tests Only Opera bests PP7, at 74 tests failed @Beat You might get web workers if enough people keep asking for it, however websockets is too unstable for Microsoft to implement, see for a brief discussion lamenting the state of development. The way it's looking, no browser's implementation will be compatible with the new standard when it is updated. @ Stifu '@wetran: it doesn't take more time to write good code than to write bad code, quite the opposite. Unless you have a whole shitty program or site to rewrite or improve, in which case that's because it was badly-coded in the first place.'
Then you shouldn't need to delay the deadline in the first place. Then you should really tell that to Hugo, giuseppe and anonym, and I don't know why you are replying to me here, since they are the ones who keep insisting on the idea of 'missing deadlines for better code quality'.
@Stifu '@wetran: either you're a slow worker, or you're working for unreasonable people (in which case you should grow balls and find a better job). Good developers don't advocate bad coding practices.' Good luck on finding 'reasonable' bosses and customers in this world.
![Http://osxdaily.com/2010/09/18/internet-explorer-9-for-mac/ Http://osxdaily.com/2010/09/18/internet-explorer-9-for-mac/](http://images.macworld.com/images/article/2012/06/15big-283819.jpg)
You can grow balls all you want and you still won't find any job that don't have 'unreasonable people' as your bosses and/or customers. For the past ten years, I've worked for all kinds of IT companies, ranging from small private company with only a dozen people, to the local branch of Google and IBM with thousands of people, and they are all the same. And it's not just me, I'm surely no slow worker since all my colleagues face the same situation. Deadline is king. For example, when I worked in Google, they don't check on work attendance, they don't care whatever you do during your work time, but the only thing they require is completing your assigned work by the deadline. Yes you may see them delaying projects at times, but that's a privilege of the higher-ups, the 'worker ant' developers (which counts for more than 99% of the developers out there) who wrote the bulk of the code have no choice but to meet deadlines assigned to them. If you think you should 'find a better job', then no matter how large your balls are, you will end up with no job at all.
I don't advocate bad coding practices, I'm just telling some people that their ideological view can't really hold in this cold, sad reality. WOW: so many comments! Wetran wrote: then you should really tell that to Hugo, giuseppe and anonym, and I don't know why you are replying to me here, since they are the ones who keep insisting on the idea of 'missing deadlines for better code quality'.
I don't know where you're from, but here (good) software developers are highly sought after professionals. Any boss will cherish (good) developers here. Good developers give feedback early in the project about (un)realistic deadlines. Bosses take us seriously. Good code, development process, structural analyis, etc. (e.g.: the tools any good developer would use) ensure the deadlines are met.
Announcing lateness two days before the deadline is not appreciated, and does not define a good developer. You say that the reason that you're releasing Platform Previews is because you wish to 'evaluate browser performance with real-world scenarios' and 'the point of a browser is to run actual websites, not just benchmarks or samples that are hardwired for one browser'. Why then, do you not update the IE9 beta on a regular basis, perhaps the same frequency as the Platform Previews? You have to know that people downloading the Platform Previews aren't, for the most part, evaluating browser performance with real-world scenarios, and that they're mostly running 'benchmarks or samples hardwired for one browser' (ie the Test Drive suite). The people who ARE evaluating IE9 in real-world scenarios, and using the browser to run actual websites, are the people who have downloaded the full-fledged IE9 beta. However, these are the same people who are being snubbed by MS developers, where instead of receiving updated beta releases for their browser, MS is releasing crippled 'Platform Previews' for people who just want to look at fishbowls, or play pinball or sudoku. Releasing a bunch of updated Platform Previews without releasing the same updates into the Beta tells me that MS is doing this mostly for show and good PR, rather than trying to release a quality product.
In the release notes of those platform previews, MS has announced that those are meant not only for the freaking game players or hypnotized JS performance evaluators, but for the 'web developers' who will test the browser performance with and beyond IE-test-drive samples and ideally send them feedback. If you analyze their bit-changed approach of developing IE9 more critically, you will understand that they are not wasting time on QA by themselves but they are asking for user feedback and make the fixes accordingly in those nightly-build PPs (extreme-programming). Earlier they used to release beta and then keep updating the beta as per the user feedback. But now that interim PP is not a right choice to judge performance as a current-release of IE9 because unlike beta, PP is not even a semi-stable release. Surely, MS will update the beta (Beta2 or RC1) after 'achieving' & 'testing' certain milestones (via PPs) they have planned for the next semi-stable or stable release. In fact, the guy at Mozilla who pointed out that so called 'cheating in Chakra' was supposed to book a feedback but his (cry-wolf) way of sending feedback exhibits hatred and professional jealousy that is understandable.
(remember how mad the non-MS AV guys are these days due to the outstanding performance of security-essentials at free of cost!). @anonym, 'I don't know where you're from, but here (good) software developers are highly sought after professionals. Any boss will cherish (good) developers here. Good developers give feedback early in the project about (un)realistic deadlines.
Bosses take us seriously. Good code, development process, structural analyis, etc. (e.g.: the tools any good developer would use) ensure the deadlines are met.
Announcing lateness two days before the deadline is not appreciated, and does not define a good developer.' I don't know where you're from, but giving feedback early in the project is part of.setting. the deadline, not.changing. the deadline. Once the deadline is set, it's next to impossible to change unless it's from the client's request, or from the higher-ups.
Also most developers working in large teams don't have a say in setting the deadlines, for example in Google, the ones who can have any say in setting the deadline are PMs and program leads, not those 'worker ant' developers who do the bulk of actual coding. So if you have actually read my post, I was specifically replying to Hugo's 'to whom code quality and good robust solutions are more important than deadlines', and my point is exactly that, you can't say 'code quality and good robust solutions are more important than deadlines', you can say a good developer/PM should help setting good deadlines, but once it's set, the developers must do whatever they can to meet the deadlines, they can't delay the deadline by saying 'code quality and good robust solutions are more important than deadlines', that'll just get them kicked out of the business. Also a lot of time, even if you are a 'good developer' and you want to have the deadline set for a more reasonable time, you can't do that since the client has some 'hard requirement' about the deadline. For example there are often clients who asks to have a website set up in two months from scratch, because they'll have some event in two months and need the website by that time. And they'll just find those who promise to meet that deadline to do the job, they'll not listen to anything about 'good code quality', they'll just put down the cash and let anyone who can do the task to get it, and bosses never refuse good cash, they'll just tell you to get it done and don't mind the 'code quality'.
I don't know how many projects have you done, but in my ten year's experience ranging from small private companies to big ones like Google and IBM, yes as a head developer you can argue for more time when.setting. the deadline, but more often than not you'll not be able to have any decisive say in it, and even when you manage to get some extra time, it's often far less than what you asked.
And if you are one of the 'worker ant' developers, then you won't be able to have any say at all. For smaller companies, you may be more important, but the client's need is king, and even your boss has to comply to them.
For larger teams, you'd have to be the PM to really have any influence. Tabbed browsing is an integral part of Web-browsing and Without tab management IE is not good for tabbed Browsing Opera recently implemented Tabstacking in their v 11 beta.
Its a good feature. I think it would be great if IE implemented it too albeit with some improvements like colored stacks etc 2. Also please provide some sort of visual feed back to show if the page has loaded or not.
This is a must in 3rd world countries where the net connections are not so stable Please look into these matters and thanks for showing the transfer rate in the download manager ?. One thing: Galactic is fast as all get-out, but it would have been nice if the author had mentioned DirectX as the reason for the graphics performance. That lack was a little too 'propaganda' in my opinion.
Other than that, this seems to be coming along nicely. For all those who whine about not having CSS-this or geolocation-that, give it a rest. These things will be along presently. @wetran: Aye, and double aye. Here's a story from when I was a project lead: the PM for a project came to me and asked me when I could deliver the app update for customer testing. I told her Thursday.
She told me they wanted it on Tuesday. Thursday, I said, and not a day sooner, because that was the original schedule. Tuesday, she repeated.
So I told her, and I quote, 'then tell them to tear Tuesday off the calendar and paste it over Thursday, because that's when they're getting it.' We delivered on Thursday and it was on spec, no bug send-backs. That PM learned not to argue with me again. The customer learned not to try to accelerate the schedule.