last updated:21 Jun 2004 16:00 UK time JOS Statistics - Recent Comments (Comments added for week ending Sun 20 Jun 2004) | View Other Weeks HELP! Overcoming procrastination | Sun 20 Jun | Beyonce Im terrible at time management and I want to stop procrastinating ... Its affecting badly my personal software projects and personnal life ... Have you been there ? Is there a cure ? I came across this product : http://www.nightingale.com/tproducts_productdetail.asp?productidn=10940 Is it worth the $69 price tag ?!? Thanks for your help folks! Sun 20 Jun | . | Well you really overcame that 'gotta go spam' procrastination, so I guess it did work. Bitch. Sun 20 Jun | Beyonce | Dear Mr DOT I aint a spammer. A serious software developer Beyonce Sun 20 Jun | muppet from forums.madebymonkeys.net | smells spammy to me. Sun 20 Jun | sgf | I'll give you my answer tomorrow. Sun 20 Jun | Beyonce | Go and buy Rick Chapman latest book ;-) Sun 20 Jun | A Pterodactylus Ate My Baby | Quit doing what the producers tell you to do. Get some REAL songs, not that r'n'b gobledidaink. Dress up in your videos. If you tried not to sell yourself to everyone for a dollar, maybe you'd find time to do some serious software development and strike it big. Sun 20 Jun | No Destiny | Maybe Anthony Robbins could help you out try some of his seminars, books, tapes .... ___________________________________ Sun 20 Jun | RP | Spammy Wammy! Sun 20 Jun | Beyonce | This aint a fuckin spam, you fucking idiots Sun 20 Jun | Beyonce | I got an attitude problem too any book recommendation ? Sun 20 Jun | CF | Ok, Ok, let's assume it's not a spam. In that case, I would recommend against some 'solution' that solves it all. It basically comes down to getting up and doing it. Not all at once, but in small steps. For example, I normally cook dinner in our house and both my wife and I clean. Before, we would always leave the dishes and say that we would get back to them later. But later usually meant the next day. So, we made a pack to start doing the dishes as soon as we were done before we did anything else. And it felt good. So pick something small. Win battles before trying to win the war. And, perhaps most importantly, try to get an accountability partner, someone who is in the same boat as you so you can each check up on the other. The above is basically what you would pick up from one of those seminars. Rather then spend the time going out and buying it, try the steps. And, finally, procrastination is not always a bad thing. There are times when I am doing development or architecture work that I just don't want to do something right then. So I let my mind wander a bit, play with Silly Putty, read JoS, walk outside. Sometimes it pays to take a break. Just don't let it take over everything. Sun 20 Jun | CF | pack->pact Sun 20 Jun | say no to spam | You don't need to spend money to overcome procrastination. I find that I fall into procrastination when I'm not absolutely sure how to proceed or where to start. To overcome this, break down your task into small steps making the first steps as clear as practicable. Then do the first one, this will get the ball rolling. Sun 20 Jun | Shilly Billy | Those audio tapes also increased my penis length by 3 inches, through self-hypnosis! Sun 20 Jun | Matthew Lock | Overcomming procrastination is simple. Just decide to procrastinate. Then you will automatically procrastinate procrastinating and end up being productive. Sun 20 Jun | Staunch Procrastinator | I was going to write a reply here, but never got around to it. VS.NET bugginess negates productivity gains? | Sun 20 Jun | . 1. The HTML reformatting bug. http://weblogs.asp.net/fbouma/archive/2003/05/15/7051.aspx 2. The destroy custom events bug http://support.microsoft.com/default.aspx?scid=kb;en-us;Q314965 In my opinion these are dealbreaker bugs - bugs with no workaround that cause more wasted time fixing than the time saved by using the tool. Is this acceptable? Are we supposed to suck this up and wait for Whidbey? How can the biggest SW corp in the world consistenyl fail to ship a decent HTML editor? Sun 20 Jun | Philo | I wrote an EDI management interface, XML preprocessor, and two websites in VS.Net. I *lived* in that thing for about 18 months. The HTML reformatter was an issue at first, but once I got my head around it, it really was not that big a deal. There were some other 'quirks,' some of which annoyed me more than others (like VS.Net 2002 not giving specific SQL Server errors when saving SP's - this was fixed in 2003). To be honest, I'll take VS.Net over any other IDE I've ever worked with. .Net and the IDE definitely saved me time and effort. BTW, the HTML formatter is one of the VS.Net team's top priorities. [disclaimer: I work for Microsoft] Philo Sun 20 Jun | Aaron Boodman | All you have to do is not use the designer. 'real web devs' would not be using it in the first place. http://discuss.fogcreek.com/joelonsoftware/default.asp?cmd=show&ixPost=153707&ixReplies=12 search: 'As someone who ' Sun 20 Jun | muppet from forums.madebymonkeys.net | Disclaimer:  I'm cooler than you.  Peasant. Sun 20 Jun | Jack | The IDE is one of the reasons .NET (and VB and Delphi) boosts productivity. So, in my opinion a good developer: - should know how to work without the IDE - should use the IDE as much as possible, in order to get things done fast The bugs in .NET are also in the .NET libraries, unfortunately. :-( And yes, they do eat away productivity. People who say that VS .NET is the best IDE obviously haven't used Delphi. I'm not saying Delphi is a lot better, but it's certainly better. Sun 20 Jun | .NET Developer | 'Disclaimer: I'm cooler than you. Peasant. ' Maybe it is that you feel that people who works for Microsoft are cooler than you. Sun 20 Jun | .NET Developer | That should have been 'people who work for Microsoft'. I know, people who know grammar are cooler than me. Sun 20 Jun | Shilly Billy | I used .NET and was able to get prescription V1ÅGRÄ at only $2.50 per dose! Did spyware kill shareware? | Sun 20 Jun | Michael Moser A question to shareware authors. Are people still ready to download and try out shareware? How do you create enough trust so that people would not suspect you of planting another key logger? Is security paranoia much of an issue with regular users, or is that more of a corporate thing (Corporate policy like never ever install anything not approaved by the company) ? Sun 20 Jun | Michael Moser | ... and are there shareware authors around, who choose to write applications in a sandbox (either java or .net) just because of the trust issue? Sun 20 Jun | kc | I'd say that adware, spyware and Open Source together are all killing shareware. Why take a risk on a project/product you don't know and can't know (ie. look at the source) if you can get a comparable Open Source app? Sun 20 Jun | www.marktaw.com | Only programmers care about looking at the source. I still download & try out software on a regular basis, but there's so much out there, that if I can find a freeware program that gets the job done, and I'm sure I can, I'll go for that one. Antivirus, Antispyware, Firewall, I've got it all. Maybe something slips through from time to time, but I'm pretty secure. Sun 20 Jun | hoser | That's a good question. It is interesting that, from my perspective, the majority of shareware download sites are marginalized purveyors of - gakk, something, flashing banners, casinos, warez crack links, porn, pop-ups and who knows what. Scum and villany. Yet somewhere in there is a nugget of software which might be useful. Do I dare touch it? Fahgetaboutit. Another question: Has free software licensing marginalized shareware? For purposes of VxWorks development, I had to install a windows FTP server. Well, no repectable Linux user is going to pay for a protocol as simple as FTP. But, this is a windows environment. What's out there? WS_FTP at $495 per license? Laughable. WarFTP comes with source and an free license. Got it. Quite an interesting world out there these days. Sun 20 Jun | non | While spyware may have something to do with it, I'd say what's really killing shareware is the expectation people have of software: They expect it to be free. Nobody wants to pay for software anymore. You can play free flash games on the web more sophisticated that the games people were paying $50 for ten years ago. The overwhelming glut of free software has spoiled people. I don't think Open Source has much to do with it, because 99% of consumers wouldn't even know what it was. Sun 20 Jun | Seer | Everybody expects free software, and everybody expects to post their questions to the web and have experts answer them for free. This will all change, soon! The IE Team Let out of the Dungeon | Sun 20 Jun | Dennis Forbes http://blogs.msdn.com/dmassy/archive/2004/06/16/157263.aspx Discuss! Sun 20 Jun | Dan G | Someone has been working on IE, the popup blocker is in XP SP2 What are the opportunities for Freelance .NET dev | Sun 20 Jun | Bobba Feet What are the opportunities for Freelance .NET developers for the next 10 years? Sun 20 Jun | Green Pajamas | Whatever you do. Just make sure that you continue learning new things as they appear :) Sun 20 Jun | Sassy | The same as any other language, it depends on what problems you want to solve. Are you a biz programmer? Learn about business domains. Learn about finance, real estate, biotech, operations, supply chains. Learn everything you can. Are you a hard science guy? Learn numerical analysis, oject modeling, software architecture, number crunching algorithms. Learn how to speak, write, and present yourself as a professional. Learn to talk to senior people. Prove you can step in the door and grok a company's business in a week, and you will always find work. Similarly, stand head & shoulders above the pack on technical issues, and you will be in high demand. Contrary to popular belief, the sky isn't falling. The jobs are not all going to India. But the days of jeans and T-Shirts and halitosis and shit-talking to management are over. What's a RSS feed ? | Sun 20 Jun | Bobba Feet I was not able to find a tutorial on RSS feed using google Sun 20 Jun | KayJay | I'm surprised. What keyword did you use? http://www.google.com/search?hl=en&ie=UTF-8&q=what+is+rss+feed&spell=1 http://www.xml.com/pub/a/2002/12/18/dive-into-xml.html Sun 20 Jun | mb | or better (though not obivously so before trying it), 'what is rss'. google defintions will take over and make some suggestions. Programming Meade Telescope | Sun 20 Jun | Evgeny Gesin /Javadesk/ Can you share your experience programming COM port on Meade Telescope and provide links and references. Sun 20 Jun | no name | can you use google first? Sun 20 Jun | AnMFCAndJavaProgrammer | Have you tried the Yahoo! Groups? There are specific groups there for discussion on Meade telescopes, and if memory serves me, for using AutoStar Sun 20 Jun | Alex | A fine example of how Russians don't use articles! ;P Sun 20 Jun | anon to protect the guilty | I have always found poor command of language to be annoying. I can understand how some people have problems with English, but I'd think that the concept of a question would be fairly universal. Or is the rest of the world missing all punctuation but periods? customizing WebBrowser control | Sun 20 Jun | Jurgen How do I control the default appearance of a webbrowser control when used as a file browser? (i.e. browsing to a file folder) The webcontrol-view always starts in tiles-view and I want it to default to list-view. Anyone? Thanx! Sun 20 Jun | www.marktaw.com | Select one folder, and then go into some preferences panel and check off "View all folders like this one" or whatever that option is. Sun 20 Jun | Jurgen | if only it was that simple... I've tried that, but the webcontrol always starts in tiles-view. ... Mixing managed and unmanaged C++ performance | Sun 20 Jun | Jurgen Is there anyone who has experience with severe performance issues as a consequence of mixing managed and unmanaged C++? And if so, what kind of machine (specs) did you use? We ask this, since our team is wondering if we should take the effort to apply the workarounds, mentioned on the web, regarding mixing managed and unmanaged code when we are going to implement. Thanks in advance! Sun 20 Jun | son of parnas | What is unmanaged C++? Sun 20 Jun | Jurgen | 'Traditional' C++ code that is compiled into an executable that doesn't need the .NET runtime and in which you have to clean your own mess you made in memory... Or the C++ you made your MFC apps with.... ..... Get it? Sun 20 Jun | sgf | Maybe "Developer Managed" would be a better term :) Microsoft is ruining the internet | Sat 19 Jun | Mike http://news.netcraft.com/archives/2004/06/16/akamai_attack_highlights_threat_from_bot_networks.html Bot networks screw up akamai, which hoses a lot of other people. Thanks for the swiss cheese. Sun 20 Jun | Alex | Funny thing... There is a link in the article to 'DDoS blackmail schemes' which leads to a story about how an e-commerce firm was threatened with DDoS unless it paid. Sure enough, they don't miss the opportunity for a good ad: 'The Columbus, Ohio company's turnkey order processing system allows customers to open online stores for a $49 setup fee and a percentage of each charge, and pays merchants twice a month. 2Checkout says it processed more than $100 million in transactions in 2003, and more than $53 million thus far this year.' There must be lots of e-commerce providers, what better way to stand out? If I had a firm, I'd *pay* to be attacked just to make the news :) Sun 20 Jun | Matthew Lock | > Thanks for the swiss cheese. Bah, it's all about people not patching their machines. If John Q Public had Linux or OSX they still wouldn't patch their machines, and there'd be the same problems. Sun 20 Jun | Clutch Cargo | 'Microsoft is ruining the internet' What is this, Slashdot? Get that weak stuff outta here. Sun 20 Jun | Chris Ormerod | No this isn't slashdot, this http://slashdot.org/article.pl?sid=04/06/15/1427213 is slashdot. At least slashdot had the sense to realise that it was Akamai that is ruining the internet and Microsoft was nothing but a helpless customer. It also affected Yahoo, Google, Fedex, Xerox, and Apple but of course nobody would ever accuse them of any wrongdoing. Of course all of these companies (with Akamais help) can be seen to be ruining the internet as pointed out by the first slashdot poster the fact that it is reducing large portions of the internet to a single point of failure. Sun 20 Jun | somemorone | As I gather the machines that hosted the source code for Win2000 and Halflife 2 were Linux machines. Sun 20 Jun | Chris Ormerod | Some Morone, Im not sure what your point is but you can be pretty sure the win2000 code was on a Winfile server, but according to all reports I have read the HL2 code was in CVS on Linux. Still, I don't see the point of your post to this discussion. Sun 20 Jun | www.marktaw.com | >>'Bah, it's all about people not patching their machines. If John Q Public had Linux or OSX they still wouldn't patch their machines, and there'd be the same problems.' Exactly right. Right now, there's a hundred million clueless users out there who continue to launch executable attachments because they have absolutely no understanding of the most basic concepts - like what an 'executable' file is. Firewall? no clue how to install it or configure it. For years people have been pushing the idea of the computer as an appliance, 'as easy to use as a toaster'. All of the problems with worms/viruses are ultimately a result of this mindset. Sun 20 Jun | Myron A. Semack | Mike, I'm really having a hard time following your logic. I don't see how you can conclude that 'Microsoft is ruining the internet'. Akamai (which MS happens to own a stake in), fell victim to a massive DDoS attack. This was not made possible by any vulnerability in Windows (at least I see no mention of it in the linked-to article). It was a deliberate, coordinated attack. There is no mention of what was used to control the zombie machines. They could have all be Linux machines, or BeOS for that matter. Sure, you can SPECULATE that it was a Windows vulnerability, but you really have no idea. Furthermore, Akamai uses Linux for most of it's web hosting, NOT Windows. http://uptime.netcraft.com/up/graph/?host=akamai.com If anything, I think this nicely underscores a point that I've made before in this forum. If a computer system is connected to the Internet, it is vulnerable to remote attack. It doesn't matter what OS you're using. If you're connected, your vulnerable. I also wrote about this in my weblog, see http://www.semack.net/Articles/PracticalSecurityInMissio.html Sun 20 Jun | MyNameIsSecret(); | Your logic is flawed. Are you a troll? Sun 20 Jun | Eric Debois | Should we blame the matches? should we blame the fire? or the doctor who allowed him to expire? NO! Blame Canada! /South Park When security fails its usually because of several minor failures contributing to a big one. To single out one thing (The OS or the Users) is counter productive and gives you an unrealistic perspective. Its a chain of events, and for the security to fail, EVERY link in the chain must fail. Sun 20 Jun | Mike | 'There is no mention of what was used to control the zombie machines. They could have all be Linux machines, or BeOS for that matter. Sure, you can SPECULATE that it was a Windows vulnerability, but you really have no idea.' Well, what OS is on the spam zombies? I'd bet the same one is on the machines that launched the attacks. I'd be willing to speculate money on that. To be fair Akamai is running some hybred dns of their own that to hear Paul Vixie tell it is a poor architecture as far as fault tolerance. Basically they have a monoculture. Had they architected differently they might be less vulnerable. GMail?  Heck No !! I've Got Redif Mail! | Sat 19 Jun | www.marktaw.com http://mail.rediff.com/learnmore.htm 1 gig of storage, Free. Hosted in India, so you know its good. Sun 20 Jun | Chris Hester | gmail is pretty impressive.  Their UI is really well done. Sun 20 Jun | WTF? | >Hosted in India, so you know it's good. Huh? Sun 20 Jun | Chris Ormerod | I just got a gmail account this morning so I only have 4 emails in it, but even with just that many i can see the ui is pretty good, just can't wait till I get a few hundred in it and see how it handles. But I know I will be stealing some of their ideas for my own app, especially that little red 'loading' bar at the top, that is quite useful. The only concern I would have with Redif is that I had never heard of them before this morning when I read it on slashdot - and it would suck to go and check email one day and its all gone (but even with Gmail I still use my domain email address which forwards to Gmail and my ISP) - but more of a concern is what they might do with your emails, at least Gmail being a US company there will be more backlash if they start selling your messages than if an indian company started doing it. And on the ads on Gmail, only 1 message out of 4 showed ads. Anybody else with Gmail, is this normal? Sun 20 Jun | karthik | <> I used Rediffmail and have never had any problems. Care to be more coherent? Sun 20 Jun | www.marktaw.com | I didn't write that, it's spam. Forum moderators can probably confirm that. Sun 20 Jun | Green Pajamas | ahan I don't get it. Rediff.com is a popular service in India. Why would they need to spam? It's like Yahoo or Google spamming on these boards. Sun 20 Jun | //Jorge | Has anyone tried spymac? They offer 1 gig to everybody right now, web and pop Regards Sun 20 Jun | anon | Why did you post as MarkTAW Sathyaish? Sun 20 Jun | Zahid | How are so many people getting gmail accounts? Am I the only person online without access to the preview release? Sun 20 Jun | KayJay | No. I am one too :-) Sun 20 Jun | anon | I don't know why people don't get it. Size doesn't matter :) I am sure before launching Gmail, Google would have predicted that everybody will start offering 1GB of space. The best things about Gmail are it's UI, google search along with the 1GB space. If Yahoo or Rediff or Spymac want to beat Google then they have to offer a better UI not just more space. Sun 20 Jun | Brad Wilson | Once you get an account, you get invites to bring in friends. In my experience, the invites come at a pretty regular pace, 3 at a time. Sun 20 Jun | Anonypus | GMail accounts are less than $2 on eBay. I'm a sucker so I just ponied up $1.75 to see the new interface. Obviously, I don't know enough people "in the know" so... Sun 20 Jun | Chris Ormerod | I got mine from someguys weblog, he had a post 'Anybody want Gmail post here' and people have been donating invitations to it. So far he had 6 to give away, I am sure there will be more I would post the link if I hadn't forgotten it. Making my own browser | Sat 19 Jun | Laura Gatisun Just wondering how hard it would  be to make my own browser.  Just a basic browser that would download an HTML page and display it... More difficult than it seems eh? Sat 19 Jun | Dewd | This interview by one of the Mozilla developers can give a tip on how difficult it is: http://discuss.fogcreek.com/redirect.asp?http://arstechnica.com/etc/linux/collins-interview-1.html Yes, I think it's very difficult if you intend to use it to browse anything but your tested pages. Sat 19 Jun | anon | Parsing and rendering HTML is not easy although you can just use the WebBrowser control but that would be cheating ;-) Sat 19 Jun | Rhys Keepence | What operating system and what rendering engine should it be based on? For example, in OS X you can create a simple web browser with no coding using the free developer tools. Sat 19 Jun | Green Pajamas | 'Parsing and rendering HTML is not easy although you can just use the WebBrowser control but that would be cheating ;-)' And that's a common practice. On Windows, there are probably only two types of browsers: IE-based and Gecko-based. Sat 19 Jun | Green Pajamas | Okay, I actually meant majority of them. I know Opera is on its own. :) Sat 19 Jun | Benji Smith | I think you should do it. This may sound crazy (actually, it is crazy), but I think you should build a browser. Not a normal browser, mind you. Not a browser that parses and renders HTML pages. (BORING) Instead, I think it would be an interesting and worthwhile project to design a browser spicifically designed for hosting thin-client applications. If you've followed much of the discussion here lately, then you've probably seen the http://oddpost.com website, which emulates ordinary windowing widgets using a clever combination of DHTML and JavaScript. I've put together a few web applications like that (though nowere near as detailed and feature-rich as the oddpost site), and it can be an incredible pain in the ass to develop. When you're developing a rich GUI in a browser, you're not actually working with a windowing toolkit. You're using tricks and hacks with HTML, CSS, and JavaScript to emulate the object-oriented hierarchy of a decent GUI toolkit. But what if you had a specific type of application that could connect to a server (using HTTP, or whatever networking protocol you prefer) and retreive the GUI (and some limited client-side logic) for an application that's primarily based on the server. The application would look like a normal fat client application, with a rich set of object-oriented widgets, but it would act like a web application, sending and retreiving its data to the server for processing/storage. It would be easy to deploy (just like web apps). But it would also be easy to use (like ordinary apps). To be honest with you, I don't think you could finish such a project. But you could do _some_ work on it. And it's an entirely new type of application, so you'd be breaking new ground. That would be interesting and fun, wouldn't it? Even if the application that you end up developing is still somewhat of a prototype. That's what I think you should work on. When you finish, send me a check. Sun 20 Jun | fool for python | or just use the mozilla platform. It is exactly what you just described. It includes a windowing toolkit and all the networking stuff you need as XPCOM (cross platform COM) objects written in c++. And a whole lot more. and it's open source so you can add, change and remove as needed. Sun 20 Jun | has | > For example, in OS X you can create a simple web browser with no coding using the free developer tools. O'Reilly has a three-part 'Build Your Own Browser' tutorial on doing this; see http://www.macdevcenter.com/ Sun 20 Jun | Tom H | WxPython gives you everything you need. The demo program that comes with it even includes a couple of browser examples. Check out wxHtmlWindow and WxHtmlParser here: http://www.wxpython.org/onlinedocs.php Designers vs Programmers | Sat 19 Jun | Christian Ive read Joels API war and related posts, and I have some few points to share. First of all, I am very much involved in DHTML stuff. I talked to various people, searched the web etc... Here is my conclusion. The current HTML and Javascript is not enough for developers to develop robust applications. One issue is that, Joel talks like as if what we have currently can be extended to develop applications, however web designers will be against that, and looking into CSS and HTML it looks like that they have more free time to push their wishes in W3C. CSS and XHTML are designed for Zeldman type of graphic designers. If you put more complexity, graphic designers will complain. Second, neither Mozilla nor Opera are strong enough to convince developers to follow them. So far zealots who hate Microsoft follow them, but you dont see them in the real world much. Productivity is very important, being able to recruit new people, train them, reuse and debug code are very important to developers. Mozilla, which already has a solution called XUL is not the solution developers are looking for. There are no developers tools like visual studio .net that can let you to develop easily and rapidly. Even, now, there is not enough documentation for XUL. It is all hype. Another issue is, are we trying to solve real world problems, or is this about trying to piss of Microsoft in a childish way like many online trolls do. Opera is known to bark at Microsoft in various publicity stunts. Similarly, Mozilla says they are trying to isolate Microsoft and colloborate with Apple, Opera etc... But developers are not concerned with pissing off Microsoft now, we are trying to make things easier and better for us. We will follow whoever can solve our problems. Microsofts upcoming solutions are more thorough than what WHAT currently talks about. If W3C turns into an anti-Microsoft organization and accepts WHATs proposals, that would only hurt W3C, since W3C is relavent today because Microsoft supported it in many levels including sending its own employees to this organizatin. If the idea is to isolate Microsoft, I dont think we will solve problems that way, we will create more problems. Many trolls will enjoy this situation but as a developer you have to learn two different things. W3C was there not to hurt Microsoft and Mozillas and Operas childish attittude will not help us as developers, if they have their day in W3C. Finally, I think Microsofts new solutions are good. I do hope that they make it more public so that we can use the same technology in different platforms easily. Mono can help us to accomplish that. Otherwise, somebody should put resources and implement something as good as XAML. I know that neither Mozilla nor Opera can do it alone, so only if IBM and/or Sun put their resources to this, then WHAT may become a real contender in this race. Otherwise just look at where SVG is today. Coming up with standards is not enough, you have to implement it and then distribute it. SVG has been started in 1999, maybe before, this year is 2004 and we still dont have SVG default support in browsers. Lets not kid ourselves here, Microsoft can solve this, Mozilla and Opera can not. Opera should at least first try to make its browser as good as Mozilla for DHTML before going into WHAT. As a developer I would be much more happier. Sat 19 Jun | Dewd | The last paragraph contradicts the previous ones. You should decide if Microsoft is the only one capable of establishing standards or not. Sun 20 Jun | anony coward | 'One issue is that, Joel talks like as if what we have currently can be extended to develop applications, however web designers will be against that, and looking into CSS and HTML it looks like that they have more free time to push their wishes in W3C.' Huh? So let me get this straight... | Sat 19 Jun | www.marktaw.com Joel writes an article. If you want to respond, you have to write something up on the web and email it to him. Hes completely removed the Discuss link from the bottom of the articles. Um. You guys havent noticed anything strange about all of this? Sat 19 Jun | www.marktaw.com | Maybe he's trying to preserve the quality of discussion in this forum by preventing the influx of..... anyone. Sat 19 Jun | www.marktaw.com | Or maybe he wants to control the manner in which is ideas can be publicly challenged. Sat 19 Jun | www.marktaw.com | Or maybe he's... Wait, what's that behind me, it's the hand of the moderator come to delete this thread. HELP HELP...... arrgghaw3klj45awl4h6fyr ,z Sat 19 Jun | Bored Bystander | Damn you're smooth, Mark. :) Sat 19 Jun | www.ChristopherHawkins.com | Actually, I can't recall the last time Joel explicitly put a discussion link on *any* article, let alone his most recent ones. Sat 19 Jun | www.marktaw.com | Joel has recently discovered one of life's unpleasant facts: If you say something in public there are lot's of people just waiting to analyze, criticize and disagree with your every word. That's the price you pay for living in a society that puts great value on freedom of speech. And he is obviously too thin-skinned to deal with it. Sat 19 Jun | Y2K | It's not like Joel's new articles are anything to brag about.  More common sense and dreaming than anything else.  I mean everyone asks where's Joel and why isn't he writing any articles.  So Joel feels pressured to write some articles.  I'm not saying they weren't an interesting read it's just old news in a different light.  Take a different approach Joel. Sat 19 Jun | Bored Bystander | Or he could flame away like certain angry coders (tm) and poison the whole environment... Joel's got a nice ecosystem here, say what you will about thin skinned-ness. I have seen an extremely active board gutted by a netcopping administrator who invited everyone to leave - IE, http://www.realrates.com/bbs. Joel is quite liberal, or at least keeps the 'taking things personally' aspect separate from the community. Sun 20 Jun | Clutch Cargo | Real people (everyone other than www.marktaw.com) are interested in content, not if there's a space to comment after stories. Sun 20 Jun | Mike Swieton | I don't really think Joel messed up with the 'put it on the web somewhere' thing, because he said to do that *rather than emailing him*. I think that's the key point. He didn't say to write up a standard rather than discuss, he said he couldn't possibly address the emails people were sending him. The two are manifestly different. Sun 20 Jun | www.marktaw.com | > Joel has recently discovered one of life's unpleasant facts: Another one I didn't write. Someone's spoofing me. :) I must be famous. I did notice a long time ago that he stopped putting 'discuss' links on his articles. Sun 20 Jun | Tom H | I assume he is trying to keep all discussion in this forum. Seems reasonable. It was getting kind of quiet while he was gone, but his last couple of articles have certainly stirred things up again. Sun 20 Jun | Ken McKinney | Did I miss something?  The "discuss" links I remember were links to this forum.  I haven't noticed any lack of discussion about Joel's articles here recently. Perhaps people are able to find there way here without a direct link from the article. Sun 20 Jun | Dennis Forbes | Some people may be able to find their way here, but as the OP says it's more indirect (and is largely the group that's always been here). Joel _did_ previously put a linked "Discuss" at the bottom of articles that lead into this "board" (though I always found it odd that it didn't go to a particular thread, so it would sort of confuse a first time visitor seeing a completely different set of discussions going on). Anyhoo there is a variety of reasons he might have dropped the discuss link, including the non-topical nature of this discussion software (and the inability of a single post to contain a long running thread). Sun 20 Jun | Bill Clinton | >>'Someone's spoofing me. :) I must be famous.' Don't let it go to you're head. You're not really famous. Sun 20 Jun | Colonel Sanders | I'm more famous than all of you. Sun 20 Jun | James T. Kirk | You are not! Sun 20 Jun | G. W. Bush | I'm _in_famous - two more than all of you! Sun 20 Jun | Ronald McDonald | I just got back from seeing 'Super Size Me' and I'm more famous than all of you put together. Sun 20 Jun | Uncle Osama | How about me? Sun 20 Jun | JC | I'm almost as famous as the Beatles. Sun 20 Jun | RP | This thread is just ripe for a mention of those guys who started WWII... Sun 20 Jun | Damian | >>This thread is just ripe for a mention of those guys who started WWII... That was Al Qaeda wasn't it ? Sun 20 Jun | KayJay | This thread is just ripe for a mention of those guys who prefer salad cream over mayonaise. Or was it the other way around? This is JoS, after all! problem with Firefox | Sat 19 Jun | the real world Well, I downloaded Firefox to try it out and am now back to using IE. I like the PopUp blocking, but I read lots of message boards on the net, and for some reason Firefox does not purple the links Ive visited. This causes me to click on the same link again and again which is VERRRRYYY annoying. Its back to IE for me for now. Sat 19 Jun | muppet from forums.madebymonkeys.net | if you're using boards with frames, Firefox won't mark links as read until the frame (page) is loaded again.  This happens on the board on my site, which is why I tend to only use Firefox in brief fits :) Sat 19 Jun | muppet from forums.madebymonkeys.net | also annoying is the fact that you can't reposition the toolbars (apparently) Sat 19 Jun | Koz | You can reposition the toolbars. Right Click on a toolbar -> Customize. Go wild. Sat 19 Jun | www.marktaw.com | I can't click a link in forefox until the page finishes loading. If I try to stop the page from loading, then I still can't click the link, but I can open it in a new window. This sucks on pages with heavy graphics, usually by amateur designers. Sat 19 Jun | muppet from forums.madebymonkeys.net | I clicked on Customize already, and got 80 hojillion options for changing buttons around, but none for moving toolbars. Sat 19 Jun | muppet from forums.madebymonkeys.net | oh wait.. I can drag buttons from one toolbar to another ONE AT A TIME, and then eliminate the extra toolbar.  Why not the ability to drag an entire toolbar? Sat 19 Jun | Koz | There's something you guys can do about this: http://bugzilla.mozilla.org Please, if you have a reproducible bug, file it. If you have a feature request, file it. Mark, I can't reproduce that one, but it certainly sounds annoying. Sat 19 Jun | Green Pajamas | There isn't a decent progress bar. I really get no idea when downloading will be done. :) Plus, there wasn't a default 'Go' button. I had to add it myself. I use it sometimes. Sat 19 Jun | muppet from forums.madebymonkeys.net | what?  there are people who put the Go button in ON PURPOSE? Sun 20 Jun | Green Pajamas | Yeps. Sometimes when I am using only the mouse. :) Sun 20 Jun | Koz | If you're only using the mouse, where does the url come from ..... Sun 20 Jun | mb | the url comes from copy & paste. basically if you have a link which shows up as text instead of a link. the firefox address bar has a bunch of problems that the netscape one has with clicking. or at least i seem to remember that, it's probably been 7-8 years since i used netscape. ie of course has its own set of problems. the middle-click thing annoys me the most. i'm going to look for bugs on it. i use it for scroll, with IE I click blindly, move the mouse to scroll to the end, then spin the wheel to stop. with Firefox, I have to find a place on the page which is not a link, click, move the mouse, wait forever since it doesn't scale as much as IE, then click again to stop the scrolling, since the wheel works in addition to the scroll-mode. (which can be used as a backwards way to accelerate scrolling I guess) Sun 20 Jun | mb | cool. looks like you can vote for bugs. Sun 20 Jun | www.marktaw.com | I know there's bugzilla, but I really don't care about it that much. You're right, it's not always. I don't know why it happens on some pages and not others. Maybe some pages have an onload event that needs to run first. Sun 20 Jun | Zahid | 'with Firefox, I have to find a place on the page which is not a link, click, move the mouse, wait forever since it doesn't scale as much as IE....' - mb I think the scaling is adjustable. I'm not at my machine right now, so I can't verify, but take a look for a setting on this. Ultimate Forum/Forum Master? | Sat 19 Jun | Ravi I was wondering if there was a site which uses web scrapping techniques to access other forums such so that users can benefit from a central forum and post to mulitple forums and query multiple forums Sat 19 Jun | Akilesh Ayyar | Well you can't post to multiple forums, unfortunately. You can search multiple forums, in a very primitive fashion, by using boardreader - http://www.boardreader.com/ - it only covers some forums, and you can't do very sophisticated searches with it. I would love for a Usenet search engine like that for Google Groups ( http://groups.google.com/ ) to come to web forums, but it simply isn't available yet. Sat 19 Jun | Eric Debois | I think most forums would block a service that allowed you to post to multiple forums. It sort of defys the purpose of a forum. Sat 19 Jun | www.marktaw.com | No, but there is software that will spam multiple PHPbb forums for you, I've seen it. This sounds like what you want. Sat 19 Jun | mb | Some forums now have RSS feeds. And there's Usenet of course. Sat 19 Jun | Aussie Chick | Yes, I would think a newsgroup would be what you wanted. Sun 20 Jun | Ori Berger | http://gmane.org - all the public mailing lists you could ask for; And while the web interface is nice, it really shines on the nntp front. Not web scraping, though. Sun 20 Jun | Tom H | Why does the idea of posting to multiple forums sound like spam to me? No thanks, I visit different forums to get a *variety* of opinions; the last thing I want is the same post in multiple places. Sun 20 Jun | www.marktaw.com | Write the software yourself & get RICH RICH RICH...... Actually, I could sort-of understand this. Similar to that service that scrapes all your email & banking accounts (what was the name, it was some weird word? Yodlee?). You could supply it with all your logins for various forums & it would present you with the latest pots without you having to log in to all of them. Posting to multiple forums, however, sounds like a too-easily abused feature. Have I got them right? | Sat 19 Jun | Estudiantin Id be obliged if someone could validate the functions Ive written below from the try-it-yourself excercises of K&R: #include #include #include void DWordToWords(const int DWord, short int* LoWord, short int* HiWord); void MakeDWord(const short LoWord, const short HiWord, int* DWord); unsigned GetBits(unsigned, int, int); unsigned SetBits(unsigned, int, int, int); unsigned Invert(unsigned, int, int); /*int main(void) { int DWord; short LoWord, HiWord; DWord = 1066825727; LoWord = HiWord = 0; DWordToWords(DWord, &LoWord, &HiWord); printf(DWord = %d\nLoWord = %d\nHiWord = %d\n\n, DWord, LoWord, HiWord); DWord = 0; MakeDWord(LoWord, HiWord, &DWord); printf(DWord = %d\nLoWord = %d\nHiWord = %d\n\n, DWord, LoWord, HiWord); }*/ void DWordToWords(const int DWord, short int* LoWord, short int* HiWord) { *LoWord = DWord & ~(~0<<16); *HiWord = DWord>>16; } void MakeDWord(const short LoWord, const short HiWord, int* DWord) { *DWord = HiWord<<16; *DWord |= LoWord; } unsigned int GetBits(unsigned x, int p, int n) { /*gets the right-adjusted n bits of integer x starting from position p*/ return (x>>(p-n+1)) & ~(~0<<(p-n)); } unsigned int SetBits(unsigned x, int p, int n, int y) { /*set n bits of integer x starting from position p to the right-most n bits in integer y*/ y&=~(~0< Sat 19 Jun | Estudiantin | And this one as well: unsigned RightRot(unsigned x, int n) { /*Rotates x rightwards by n bit positions*/ int temp=GetBits(x, n-1,n); x>>=n; return SetBits(x,31,n,temp); } Sat 19 Jun | anon | Dear Sathyaish Chakravarthy: Did you make Unit tests for these? That is, did you work them out by hand first using sample numbers and knowledge of the platform and problem domain to come up with the correct answers. Once again let me pose a question to you: 1. Do you have CS degree? Why do you seek attention and approval from the people on this forum? Does your family not give you this? Does your current work environment not recognize you as you believe you deserve to be recognized? What exactly have you done lately to be recognized? Is a paycheck not enough? Do you have the true and false notion that the people on this forum are smarter than you? My friend Sathyaish, please, please, please cure yourself of this disease. I recommend getting laid and take some time off. Sat 19 Jun | A Pterodactylus Ate My Baby | lmao Sat 19 Jun | Sathyaish Chakravarthy | Dear Anon, I would much appreciate it, if you clearly probed into yourself and re-stated with more clarity, your case against me. I am truly puzzled by your post. I hope you are Joel posting in disguise, because if you are not, I would be surprised at your audacity of shooing another poster as yourself for doing exactly what you do. To answer your questions, I am not a Computer Science Student, and I regret that my parents couldn't afford to pay for my education after my schooling, and neither were they informed enough to offer career guidance. I visit this forum because I find the collective personality of the body of this forum an indispensible source of knowledge to tap into. This is by no means the only place I post my queries to. I would want you to know, so you may not unsettle your mind in this regard, that I do not consider myself any higher in stature than a student who wishes to learn some of what he could not, one thing of them being the C language. If I have upset you, I sincerely apologize and humbly I request you not to read my posts, lest I be the cause of your discompeture. Sat 19 Jun | anon | Nice haircut Sathyaish! Sat 19 Jun | no name | While you're telling him not to read your posts, maybe you ought to think about whether or not this is really an appropriate forum to be posting CompSci 101 questions to. Sun 20 Jun | Egor | One thing for sure: K&R wouldn't like your NamingConventions, Sathyanish. Sun 20 Jun | somemorone | Not that I think that you should listen to Anon's remarks but just an FYI: You could signup to the devcpp mailinglist, there a lot of people on it and they know their stuff. Look at the archives to see its something for you http://sourceforge.net/mailarchive/forum.php?forum=dev-cpp-users signup here: http://lists.sourceforge.net/lists/listinfo/dev-cpp-users Sun 20 Jun | anon | It's not that I have anything against you posting to the forum Sathyaish but I do have something against your constant 'nagging'. How many times have you posted this and still not figured it out? It seems to me that you seek attention, friends and recognition. This is normal but you are going about it the wrong way. You want to seek real life friends not people on some stupid forum. Don't spend so much time in front of a computer. You do seem an odd character to me. Do you work in the software industry? What kind of software do you actually write over there in India? If you are a representation of the kind of programmers available in India... well then I think offshoring to India will come to a halt very soon. One time you made a post about making a program that displayed the 'sine wave' of a sound file. There are tons of programs that do that. Are you guys in India writing trivial programs like these just to learn? If so then your company and your economy is a puff ball and will go poof very soon. It will crumble before your eyes. Not having an education is fine, I didn't have a college education either until I was older but I did learn both C and C++ on my own and never posted to forums asking if I was right about certain problems. I learned C and C++ before any other language. You have to take the responsibility upon yourself to determine if what you have written is correct. It is not that difficult to verify. You may think I am being harsh in my criticizm of you but I'm not. You may also think that I hate you but I don't. You may also believe that I do not care about your problems but I do. This is why I am giving you this advice: Figure it out on your own. Sun 20 Jun | Tired of this | 'I find the collective personality of the body of this forum an indispensible source of knowledge to tap into' Except the collective personality of the body of this forum is not interested in teaching you how to program. There are tutorial web sites out there to help you with that. Need a free mp3->Wav converter | Sat 19 Jun | Bella Nero crashes when it tries to burn mp3 directly to an audio CD.  I need an mp3 decoder to do that decode step manually.  Know of a free one?  Sat 19 Jun | Mike Swieton | I'm sure WinAmp can do it through an output plugin. At least, I'd be extremely surprised if it couldn't. Sat 19 Jun | Roose | I had the exact same problem, try this: http://www.dbpoweramp.com/dmc.htm It is free, and there are other tools too that you might find useful. Sat 19 Jun | GinG | CDEx is a CD ripping utility but it can also do batch conversions (MP3->WAV and WAV->MP3.) http://cdexos.sourceforge.net/ Sat 19 Jun | muppet from forums.madebymonkeys.net | or you could just pay for your music. Sat 19 Jun | www.marktaw.com | CDex seconded. Sat 19 Jun | www.marktaw.com | EAC (Exact Audio Copy) may do it, I know it goes from CD -> WAV or MP3, but I don't know if it iwll go from MP3 to WAV. Sun 20 Jun | LollyGagger | WinAmp Ctrl-P Preferences->Plug-ins->Output->Disk Writer Sun 20 Jun | Clutch Cargo | Who still uses audio CDs? Sun 20 Jun | Björn Roth | Or just use iTunes. Sun 20 Jun | www.marktaw.com | HA HA HA, good points. 'I know a good mp3 -> WAV converter, it's called an IPOD.' You know what I want.... A good MP3 -> Vynil converter. Oh the sacrilege. Web - Single Rendering/Executing Vendor | Sat 19 Jun | Michael Sica (michaelsica.com) I dont think we can get these great web applications with more than one vendor rendering and executing the GUI. I have worked in Web Development for the last 2 and a half years. Ive read Zeldmans book on Web Standards, Ive built my personal site in XHTML Trasitional and CSS. Ive programmed in Coldfusion and Im learning Java. The next technology Id like to learn after Java/JSP is Macromedias Flash. Ive seen, used, and written quite a few web applications. And here is what Ive learned: When you have a specification instead of a product, obviously it is with the intention of having multiple groups implement it. The problem with multiple implementations of a specification that is based around what people see is that if one vendor is off by a couple of pixels, or the behavior of the client is slightly different, you have just destroyed the user experience. In my experience as both a user and a developer is that JavaScript is twitchy and doesnt reliably work the same in all browsers. Not to mention, what happens when the user starts seeing GUI elements on the screen and tries to start working, but the .js file hasnt finished downloading! With JavaScript, youre literally leaving it up to all the different browser makers to perform the behavior of your application. And that scares me. If all the vendors implemented the spec exactly the same way, then maybe wed be able to use a future form of HTML/CSS/JavaScript for developing web applications with rich GUIs. The reality of the situation is, Microsoft rules the web browser and the corporate desktop. Every vendor that comes into my place of work and pitches their software say the same thing, “We only support IE. You can try it in other browsers, but its not supported.” Translation, “The product will probably work like crap in anything other than IE.” So if Microsoft doesnt implement the new “standards”, theyll be about as popular as XUL. Not to mention, when you develop a rich GUI for a web app, do you really want to spend time tracking down all the different bugs across all the different platforms? Anyone whos done a lot of CSS work will tell you how much time you can waste making everything look the same in all the browsers. It can be accomplished, but think of how much more work you could have gotten done if the client you are aiming for was just one renderer. When you are talking about a standard web site, one that delivers primarily text content (articles, blogs, product information, etc..) You can get away with not everything being being pixel perfect across all browsers. Heck as long as everything is readable and doesnt look bad in all the browsers, youre golden. I just dont believe the same is true for web applications. I want all the people using my app to see the same thing! And I want it to behave exactly the same in all environments! I believe you need to go one level above web browsers, and into the domain of a single vendors product that runs on all major platforms. -Flash -Java Applets (Does anyone know of any other technology that can give you a rich interface that installs in all major platforms?) Dont confuse the crap youve seen done with these technologies and the quality of these technologies. With Flash and Java Applets, you can write your code once, and test it inside of 1 environment. You wont spend all your time opening all the different browsers (Opera 6&7, Netscape 6.0, 6.1, 7+, IE 5.0, 5.5, 6+, Mozilla , Safari 1.0, 1.2) watching how they look and act, fixing bugs, seeing how they degrade to the older versions, etc.... Youll write to your platform version and thats that. Both of those platforms (Flash / Java Applets) already have most of the rich functionality that a web application should have. Developing a desktop app in native code for one OS is easier than for multiple OSs. It costs less, and you dont have to stretch your support across multiple platforms. With web development of “rich GUI apps” you can be an ass and only support 1 browser or you can increase your QA and support costs by supporting multiple browsers. OR You can develop for one environment, get a ton of “rich”-ness, and have your application work on all the major platforms! The web browser is a delivery medium for text. We need to stop trying to bend HTML and the joke that is JavaScript to do what we want and start developing applications in technologies that are built for it. (I should go apply for the Macromedia cheerleader squad.) Does this answer the Amazon.com question? No. Amazon shouldnt re-write their site with a rich GUI. Its a catalog that remembers what I was last look at and then tries to remind me to buy it, not a web application. Does this answer the web application question? Yes. Think about how amazing Gmail would be in Flash. All that stupid Frames/JavaScript crap they are doing right would be 10x smoother and less buggy in Flash. Start taking advantage of the tools we have available now. If you dont, one day Hotmail will launch with some nifty XAML front end and everyone wil be scrambling to figure out how to catch up with Microsoft. Sat 19 Jun | Tom H | 'if one vendor is off by a couple of pixels, or the behavior of the client is slightly different, you have just destroyed the user experience' We're looking into implementing at least parts of our system with a rich client environment, maybe Flash. But that said, there are not many applications where a 'user experience' is important; for the vast majority of web sites I prefer simple, clean, and fast as possible. And really, how many users are so sensitive that a couple of pixels will destroy their experience? I think you are taking your artwork a bit too seriously. Sat 19 Jun | CF | Some comments: 'I don't think we can get these great web applications with more than one vendor rendering and executing the GUI.' ... 'When you have a specification instead of a product, obviously it is with the intention of having multiple groups implement it. The problem with multiple implementations of a specification that is based around what people see is that if one vendor is off by a couple of pixels, or the behavior of the client is slightly different, you have just destroyed the user experience.' Like the Windows API? And besides, if one vendor is 'off' that means they didn't implement the specification correctly. You are going to run into that wherever you go. Take for example some previous posts on JoS which mentioned the finnagling major companies did with the Windows API. They didn't implement the specification, and so their product didn't work. Microsoft knew they would be the ones that would be blamed and so worked around it, but that's beside the point. 'In my experience as both a user and a developer is that JavaScript is twitchy and doesn't reliably work the same in all browsers.' Granted, I have stayed away from heavy CSS scripting the past 7 years I've been doing web development, but very rarely have I had major issues with Javascript. Except Opera which didn't quite implement JS correctly when it first came out. But, most of the work I did had to degrade gracefully and not rely on users having JS enabled, so it was a different approach. 'Every vendor that comes into my place of work and pitches their software say the same thing, “We only support IE. You can try it in other browsers, but it's not supported.” Translation, “The product will probably work like crap in anything other than IE.”' Actually, the translation should be, 'We were too lazy to try to stick to standards and wanted to rely on Windows-specific features, so we didn't bother trying to make it cross-browser compatible.' Or, 'My boss doesn't get why we would want this to be cross-browser, so our deadlines only allowed us to build for IE. Sorry.' If it doesn't work in anything other than IE, and isn't for Intranet use, more often then not someone is being lazy. A notable exception is accessing printing features from the browsers - before IE 5.5's time there wasn't a pretty way to access print features, except through very browser specifc calls. 'So if Microsoft doesn't implement the new “standards”, they'll be about as popular as XUL.' I'm sure developers would love to bicker with you about the popularity of XUL. But, the biggest issue I see is getting people to upgrade. Even if Microsoft implemented new standards, who is going to upgrade to it? Granted, they could push it out through Windows Update, but the number of PCs being sold doesn't outnumber PC Owners anymore, so it would be a hard sell. You still are going to have to do browser checking like you do today. 'I want all the people using my app to see the same thing! And I want it to behave exactly the same in all environments!' Then write a friggin desktop app. Let it use TCP/IP as a transport mechanism to get the information you need. But I'm sure you want to only have to write the application once, right? 'With Flash and Java Applets, you can write your code once, and test it inside of 1 environment.' Yep. And sorry to say, that is a false statement. Ever worked with Microsoft's Java implementation? PoS. Or wanted Flash to interact with the browser in a meaningful way (hint: it uses Javascript too)? 'We need to stop trying to bend HTML and the joke that is JavaScript to do what we want and start developing applications in technologies that are built for it.' 'If you don't, one day Hotmail will launch with some nifty XAML front end and everyone wil be scrambling to figure out how to catch up with Microsoft.' The technologies I have seen built for it are not the ones that are embedded in a browser. Maybe Java, but it isn't the most fun thing to work on in multiple-OS environments. It's been things like Qt and wxWidgets using various languages as its backend. Flash maybe has promise, as it was one of the client-side technologies I used to do XML DOM parsing and rendering and it did a good job. But even then, the conversation boils down to one thing. If you want cross-platform Rich-client GUI apps, don't plan on using your browser. But, that means that it will be much more difficult to rely on every one having whatever they need to run your app. Which doesn't seem that much different then where we are today. Sat 19 Jun | Michael Sica (michaelsica.com) | Hi guys, you made some good points, but I'd like to say a few things! Tom H: 'And really, how many users are so sensitive that a couple of pixels will destroy their experience? I think you are taking your artwork a bit too seriously.' -I use a web application at work called Unicenter. It's built by CA and is only supported in IE. It works in Mozilla, but the spacing of everything is off (so it looks crappy), and the JavaScript doesn't work exactly the same between browsers. IE is a lot smoother, when I use Mozilla I end up retrying to do things because whatever events they are using just don't translate over to Mozilla in the same manner. Is this the fault of the development team who didn't 'stick to standards'? Well, I don't know. Maybe IE did things in a way they wanted to use, or maybe they just don't know what they are doing. -But it doesn't matter because its built for the most dominant browser. And for the vast majority of the development teams out, that is the only thing they seem to care about. -If the front end of the application was written in Flash, we wouldn't even be having this conversation. Because it would look and act the same in all the browsers I run it in. CF: 'Like the Windows API? And besides, if one vendor is 'off' that means they didn't implement the specification correctly.' -Exactly. The vendors won't ever all be on the same page. And as long as MS has the lock they do on IE, it won't ever change. Too many companies are building IE only apps, and the business units that need them don't care. (they don't know the difference.) -If an app is built in Flash, then it wouldn't matter which browser was being used, because its running in the 1 environment. 'Yep. And sorry to say, that is a false statement. Ever worked with Microsoft's Java implementation? PoS.' -No, and that's a good point. 'Or wanted Flash to interact with the browser in a meaningful way (hint: it uses Javascript too)?'' -But Flash doesn't have to use JavaScript with the browser. It uses its own internal ECMA Script engine called Action Script. You don't need to interact with the browser at all. (Except for maybe to print or something.) 'Flash maybe has promise, as it was one of the client-side technologies I used to do XML DOM parsing and rendering and it did a good job. But even then, the conversation boils down to one thing. If you want cross-platform Rich-client GUI apps, don't plan on using your browser. But, that means that it will be much more difficult to rely on every one having whatever they need to run your app. Which doesn't seem that much different then where we are today.' -I agree. But I can see everyone on the planet having Flash installed before I can see everyone on the planet having IE, mozilla, Windows, Linux, etc... installed. ......And of course you would be locked into Macromedia as your sole Flash vendor.... Blah Blah... ;) Sat 19 Jun | Michael Sica (michaelsica.com) | By the way, I encourage everyone to go read what Flash is now capable of. http://www.macromedia.com/devnet/ria/ And no, you need to use 'Flex' to get a RIA. From what I can tell you can do everything from a GUI perspective that with Flash than you can with Flex. Flex actually compiles into a flash file. (.swf) Sat 19 Jun | CF | Hi Michael, 'But Flash doesn't have to use JavaScript with the browser. It uses its own internal ECMA Script engine called Action Script. You don't need to interact with the browser at all. (Except for maybe to print or something.)' I was thinking more along the lines of being able to interact with elements outside of the movie. For that, the only way I knew it could interact (without Flash Remoting) was through Javascript or posting requests. I've worked a lot with Actionscript at it is very powerful. By the way, excellent link about Flex - I will have to spend some time playing with that. Sat 19 Jun | Eric Debois | The solution is very simple in my humble opinion. Just assign the stylesheets dynamicly and have one for IE/win, one for IE/Mac and one for Mozilla and the rest for Win and Mac. Thats four variations and it adds about two hours of work if you plan ahead. Its basicly the only sane way of getting good font rendering on both platforms anyway. Also, IE5.5/mac is usually the browser to cause the most trouble with JS and CSS so we cant really trust MS to set standards. Sun 20 Jun | Michael Sica (michaelsica.com) | Typo! 'And no, you need to use 'Flex' to get a RIA. ' Should have been, 'And no, you DON'T need to use 'Flex' to get a RIA.' Asmw PC-Optimizer pro | Sat 19 Jun | inspintw Excellent collection of utilities but need some thing :) http://www.asmwsoft.com Sun 20 Jun | www.marktaw.com | 'Remove junk files and zero size files to save valuable disk space;' So on a 100 gig hard drive, how much space would be taken up by a bunch of 'zero size' files? Sun 20 Jun | Dex | Innovative System Optimizer http://www.innovative-sol.com/optimizer/ Non-halting regular expressions? | Sat 19 Jun | T. Norman I was using regular expressions to parse out certain patterns in a large body of code. When the program hit a particular module, it would just hang as if it was stuck in an infinite loop. After lots of frustration I eventually traced it to a call to the find() method in Suns java.util.regex.Matcher class where it was hanging. It was nothing to do with my code. The string that made the regex get stuck was only about 1K long, and the regex wasnt particularly long or complex. It did have back references, which can cause the processing to take longer, but it still got stuck after I removed the back references. Ultimately I had to change the regex to stop it from getting stuck. The modified regex was a bit less exacting that what I really wanted, so I had to add some code to do look at the groups to make sure it matched what I really wanted. From automata theory I remember that all context-free languages are decidable, which would include regular expressions. Although the modern-day regex classes probably do provide matching abilities beyond what a mathematically strict regular expression can do. Are there some combination of strings and regexes that would legitimately fail to halt in a modern-day regex utility? Or is it a bug in Suns Matcher class? Are there are just some strings that can send the regex object into an exponential or high-polynomial loop, even without back references? Sat 19 Jun | Rhys Keepence | Perhaps give it a try in the regex coach: http://www.weitz.de/regex-coach/ Sat 19 Jun | Tom H | What are you parsing? I've seen many people try to use a regular expression when it would be more appropriate to use a parsing engine, they're different tools for different tasks. Sat 19 Jun | no name | Sounds like a bug. Sat 19 Jun | T. Norman | 'I've seen many people try to use a regular expression when it would be more appropriate to use a parsing engine, they're different tools for different tasks.' That's beside the point. The point is whether you can write a regex with today's regex engines and provide it with a short string (under 2K) that it would cause it to loop forever or for some exponentially long time, with such behavior being a necessary mathematical limitation of the engines, rather than being a bug. Sat 19 Jun | Gerard | I don't know how that particular regular expression implementation works - but if it's the process of building an NFA from the regular expression and then constructing a DFA to process the text with, you can *in theory* get a very big DFA, as the states in the latter belong to the power set of the former. So perhaps the NFA to DFA conversion was wallowing in a lot of states (and memory to allocate them in). Have a look at the Dragon book if you want to persue this possibility. This rarely ever happens in practice, and I would believe that most industrial strength implementations would detect this and either report and error or switch to on-the-fly NFA matching (again, see the Dragon book). Any Java implementors care to comment? Regards, Sat 19 Jun | T. Norman | In this case the Pattern and Matcher were built successfully, and it didn't have a problem when other strings were passed to the Matcher.  It just locked up when a particular string was passed to it. Sat 19 Jun | Frustrated | Try checking memory usage as well as time. Also try feeding in initial segments of the string to find where you get the infinite loop (or exponential time). Pathological cases are known. Why don't you post the regexp and maybe someone can spot the problem and or suggest a different way to write the expression. Sat 19 Jun | Julian | Or, for that matter, you could give your example to Sun so they might fix it in a later release. Sun 20 Jun | Ori Berger | Most regex implementations today do not build a DFA / NFA -- especially those that offer backreferences (which are not 'regular' at all). And because they do NOT build DFAs and NFAs, it's very simple to construct pathological cases - e.g., '((a*)(a*))+b' can take exponentially long to decide that aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaac is not in the language using the matching techniques commonly in use. Sun 20 Jun | Gerard | Ori Berger: 'Most regex implementations today do not build a DFA / NFA -- especially those that offer backreferences (which are not 'regular' at all).' Tell us more... Cheers, Sun 20 Jun | T. Norman | Ori Berger, in the words of Donald Trump ... You're Hired! The expression that gave trouble was very much like what you described. It contained two sub-expressions together that were similar, but with one that would match a subset of the other. The real thing was more complex and specific, but in essence it had something like this: (foo[a-z]+)(foo[a-z0-9]+) With the first being a subset of the other, the 'greedy' matching mechanism probably caused them to eat up each other, which caused an exponential loop to backtrack and resolve the conflicts. The problem was resolved by removing the first one and then a couple extra lines of code to look at the eventual match afterwards to see if it satisfied what I was looking for. Killer App(s) for Desktop PCs | Sat 19 Jun | Akilesh Ayyar For years, weve had better and faster hardware for cheaper prices, but in the last five or seven years, it seems to me (and this is no original thought) that there have been no real exciting new applications that make use of this new hardware. Sure, there are games. Sure, theres exotic multimedia stuff like video editing. But where is the new software that revolutionizes how most people interact with their computers on a day-to-day, moment-to-moment basis? Where is the software that makes deep use of 3 Ghz computers? Where is the software that gives us smart, integrated voice and gesture recognition, powerful and startlingly beautiful new interfaces, extraordinary ways of creating new things and dealing with what we already have--in other words, a more intelligent, pleasurable, coherent user experience? I know that internet-based applications have been a fountainhead of innovation. But, as per the topic du jour, what about the power that resides on the desktop computer? Have we really made the most of it -- is this all thats possible? Sat 19 Jun | Mike Swieton | If the concept for the next killer app was that obvious, it wouldn't be the *next* killer app anymore. Sat 19 Jun | Akilesh Ayyar | I see that there's SOME barrier, but is it really a dearth of ideas? What's the real problem here? Is it that the technology is in some kind of trough where it's more powerful than before but not powerful enough for new really cool applications? Is it that software engineering has not advanced enough to create these new apps? Is it that the industry is too occupied with security and similar concerns to want to innovate radically in terms of features anymore? Is Microsoft the problem? What's really stopping us from getting visionary desktop software? Sat 19 Jun | KayJay | Any period of exponential growth is always followed by a lesse period of consolidation. The nineties blew open the internet. Now the industry is in the consolidating mode. Hence Joel's request for 'Web Apps <--> Desktop Apps' and Raymond's penchant for 'backwards compatibility' and Eric Sink's Law #13: 'The Law of Sacrifice is all about saying 'no' to opportunities.' Unless this consolidation occurs, further growth cannot happen, as the 'new' stuff needs some scaffolding to launch itself off. It happens in business, in personal life, and also in software development. Sat 19 Jun | matt | From what I've heard, this is what MS are trying to do with Longhorn. Taking them long enough, but still. I'd say Mac OS X is definately getting there aswell. Sat 19 Jun | has | 'But where is the new software that revolutionizes how most people interact with their computers on a day-to-day, moment-to-moment basis?' Same place it's been the last twenty, thirty or forty years, I expect: stuck in some lab somewhere. I don't think the difficulty is finding new ideas, it's getting folk to adopt them. And most folk hate radical change, thus words like 'innovation' become synonyms for 'not too different to what already exists', as anything more ambitious is likely to be rejected outright. And without a reliable market, there's little motivation for investing the (considerable) resources required to take a truly revolutionary idea and work it up into a shippable form. It's just too risky a proposition. e.g. Take Smalltalk: decades ahead of its time, but most users don't want Smalltalk, they just want a 'better C'. Professional programmers that have spent years becoming expert in C++ or Java are probably going to turn white at the thought of throwing all that aquired knowledge away and start out again on ground level. No amount of potential long-term advantages is going to convince them it's worth paying the very definite - and painful - short-term cost of a radical change. They'd rather stick with what they already know and have, and wait for it to 'get better'. Ditto for Aunt Millie who's wedded to her Windows 95 and Outlook Express, and will hang on to them for dear life no matter how much better off she might be with the latest and greatest. Totally different kind of user, but completely the same kind of problem. Real revolutions do happen now and again, but most progress is slow and incremental so as not to scare the natives away. What's really frustrating is the stagnation and eventual ossification of existing tech at the point of 'just good enough'. For example, the CLI hasn't changed in, what?, thirty or forty years. GUI hasn't moved in twenty. Data storage (fifty years) might just be starting to crack now that the 800-ton gorilla's on its case, but I'm not going to hold my breath just yet. If you want to find out where your interaction revolution's got bogged down, these dinosaur-ridden prehistoric tarpits might be a good place to start your search. (Anyone that can crack them is assured of at least one fan...;) Sat 19 Jun | Alex | In truth, 3 GHz isn't much power. Voice recognition requires more power. 'extraordinary ways of creating new things' requires nothing short of AI, which you can't do with 3 GHz and I'm not sure you can do at all. Same for machine translation. Sat 19 Jun | Mr. Analogy | Next killer app wil be a truly easy to use personal information manager that lets you: 1. Brainstorm 2. Outline 3. Email 4. Store and retrieve all of the above, with intuitive searching. 5. Is EASY to use. There are lots of apps that do PART of this, but few do them well. And NOTHING does them all well. Sat 19 Jun | Mr. Analogy | The problem is that the folks innovating are too far removed from the users: a. CEOs who have assistants, so they dont need to worry about organizing their information. Assistant does it for them. b. Engineers who dont understand that Grandma wants her software to be EASY and INTUIIVE. I's a basic lack of understanding of the PROBLEMS and to make an EASY TO USE program that solves the problem. This is a non-trivial problem: it's very difficult to correctly diagnose a problem and actually come up with a workable solution, much less actually IMPLEMENTING that solution. Sat 19 Jun | John | To me, iChat with iSight is the current Killer App! http://www.apple.com/ichat http://www.apple.com/isight Sat 19 Jun | has | > 'extraordinary ways of creating new things' requires nothing short of AI, which you can't do with 3 GHz and I'm not sure you can do at all. Darn, someone oughta told John McCarthy this. And Doug Engelbart. And Xerox PARC. And Apple. And... ;p Sat 19 Jun | has | > Next killer app wil be a truly easy to use personal information manager that lets you: 1. Brainstorm 2. Outline 3. Email 4. Store and retrieve all of the above, with intuitive searching. 5. Is EASY to use. I think the real answer there is a document-oriented component architecture (e.g. OpenDoc) to tie together nice individual solutions to the first three tasks (plus anything else the user might need while they're at it), plus a relational filesystem (e.g. PARC's Placeless Documents; WinFS) to look after #4. That, plus some kickass HCI designers, should in turn take care of #5. Sorted! Now we just need for someone to develop it, and for everyone else to go use it. :) Sat 19 Jun | son of parnas | Once we have wide spread wireless control of robots we will have a lot of potential killer apps. Sat 19 Jun | They'd like too though | Don't know what it is, but Microsoft won't build it. Sun 20 Jun | http://badblue.com/blog | I'd say Napster qualified as a killer app for a desktop PC, even though it required networking. If, as Joel alludes to, the browser experience keeps getting richer... the web app interfaces (like GMail) keep getting better... well, we don't need much of a killer thick-client app anymore. mozilla future | Fri 18 Jun | fool for python Lots of goodies on the table. From the creator of javascript: http://weblogs.mozillazine.org/roadmap/ Sat 19 Jun | Nigritude Ultramarine Hater | I'm a big Mozilla fan, but their lack of direction is a bit worrying. First they say that Mozilla is going away, to be replaced by Firefox/Thunderbird, but now they're talking about Mozilla 2.0 while at the same time continuing to develop FF/TB. I don't get it. Sat 19 Jun | Eric Debois | I think Mozilla 2.0 is the 'project name' for the next generation of firefox and thunderbird. The bundled mozilla installer is still going away afair. Sat 19 Jun | Simon Lucy | mozilla is the whole schmozzle, Gecko, Mail, XPCOM, etc, etc. Firefox is the browser. Sat 19 Jun | Billy | Can't someone rescue the 'Netscape' name--buy it back from AOL? Just put the name back onto the Mozilla project. Netscape is still recognizable to a lot of people, and they might be more likely to switch back (after a decent PR campaign) than to learn about what Mozilla is or what Firefox is, etc. (I have heard that AOL now might release a new rev of Netscape based on the latest Moz. code.) Sat 19 Jun | Andrew Hurst | AOL won't give up the Netscape name, they're using it for thier new low-end ISP's name: http://www.clickz.com/news/article.php/3367881 Why you'd name an ISP after a web browser, is beyond me. Sat 19 Jun | mb | you use Netscape to download the Internet, of course. Sun 20 Jun | Bob Riemersma | Reminds me of somebody who recently told me she had her son come over and install the Internet on her PC. Then again in the '70s we had people who would call because 'my computer is broken.' Of course all they had were dumb terminals... and no idea that everything wasn't inside of that box on their desks. Do people hear about 'horsepower' and imagine there are all these tiny horses under the hood of their cars too? Distributing Joel's New Browser | Fri 18 Jun | Chris Altmann I dont know if it came up elsewhere, but I have to ask: How this new browser with all of these advanced application development features be distributed, and will its size be anywhere near 25MB? Fri 18 Jun | Almost Anonymous | Firefox with XUL is only 5MB... it's a pretty impressive development environment and it's not out of beta yet. I'd say adding the extra features would probably only add 1MB to that. Sat 19 Jun | Phillip J. Eby | Yep. Download Firefox, run it, then go to this URL: http://mab.mozdev.org/ and click on the 'Start MAB Now' link. That's a sample of what you can do with XUL. Sat 19 Jun | Blues | ** VERY ** impressive. Seems XUL covers a good chunk of what Joel is asking for. Anyone knows if there is any chance XUL will become a "de jure" standard ? (No chance unfortunately it can be a "de facto"). Sat 19 Jun | Koz | It doesn't meet one of Joel's key requirements. Graceful degradation. Sat 19 Jun | Almost Anonymous | 'Graceful degradation.' It would if someone would just write a Mozilla ActiveX plugin for Internet Explorer! Sun 20 Jun | Phillip J. Eby | Ooooh...  that would be totally sick and twisted...  Poetic justice, even.  I love it.  :) Windows Service Pack 2 and You. | Fri 18 Jun | A Pterodactylus Ate My Baby So, I see lots of people complaining that - Sp2 makes the system slow as heck - Sp2 breaks many of my precious apps - Sp2 makes my toaster burn my bread Just a gauge for you guys who DID install it, what do you experience? Are you more unstable now than you were before? Do apps break? Do YOUR apps break? Anyway, just to bring a bit of credibility into the discussion; Ive been running SP2 #2096 for since it was released (3 months ago or so?), and never had a problem. Today I upgraded it to #2149 (most recent, and I think, the official release to come), and have had no problems whatsoever. Okay, so not much have changed for me, though, Ive to say Im satisfied with the new firewall, and the pop-up blocking, more security aware Internet Explorer (even tho I just use to for my online banking + windows updates, for the rest I use Opera). Your thoughts? Fri 18 Jun | Mo Hawke | I haven't tried the 'latest' but I did install the one before that (build 2142?). The only real problem I saw was that my network activity icon disappeared from the system tray. When I tried to get it back, I found that going to Control Panel -> Network Connections gave me a blank screen. Annoying but not a major problem. Is that fixed in the latest build? More interesting, is the subject of whether or not SP2 will install on 'pirated' versions of XP. This subject has been discussed before, here and elsewhere, and just as I suspected, it turns out to be much ado about nothing. Just before Windows XP Service Pack 1 was released in September of 2002, the buzz going around the 'net was that it wouldn't install on 'pirated' copies of Windows XP. There was even this guy who had this website where he proclaimed in BIG BOLD LETTERS that everyone using a 'pirated' copy would be doomed when SP1 was released. What really happened was that Microsoft simply blacklisted two well-known, widely distributed keys, and SP1 wouldn't install on any computer using one of those keys. Anyone with half a brain could see that one coming. And that's why anyone with half a brain had already changed their key -- a key generator was released in early 2002 less than 6 months after Windows XP first hit the market. Now SP2 is almost here. And the dance beings again. First Microsoft says that in the interest of security, SP2 *WILL* install on 'pirated' copies of XP. Then they say it *WON'T*. It will. It won't. It will. No wait, it won't. This time we really mean it. People in various discussion groups debate at length the pros and cons of the issue. But, just as I suspected, it is once again a non-issue for anyone with half a brain. So, I downloaded the latest (at that time) 'Release Candidate' of SP2 and it installs just fine on my 'pirated' copy of XP Pro. Windows Update works just fine too. Why? Because just like before, Microsoft simply blacklisted a few keys *THAT THEY KNOW ABOUT*. Nothing more. You would think a company with $56 Billion in the bank would be a little smarter than that. Fri 18 Jun | A Pterodactylus Ate My Baby | No probs. The build 2149 (which is the final, mark my words) do NOT accept pirate keys. Well, unless if you change it, which is easy enough. It lives happily with my machine, and if they try to track me; sorry, I don't live in the Land of The Free. Good luck, MS. Fri 18 Jun | A Pterodactylus Ate My Baby | BTW, anyone familiar with the FIRST reg key released - by a cracking team called 'The Devils Own' ? :=) If not, google and have a good laugh. Fri 18 Jun | Brad Wilson | Running XP SP2 RC2 on my laptop. ABSOLUTELY LOVE IT. The Firewall and the upgrades to the Wireless networking support alone are worth the download, in my book. Fri 18 Jun | Paul Jensen | I just installed build 2149 on both my desktop and my laptop.  Both work fine.  Haven't had any problems yet.  Two of the features in the service pack were actually features that I had been hoping for for a long time: the ability to choose "never trust ActiveX controls from company blah blah" and the ability to disable wireless networks in the preferred networks list.  I think the stuff about "Windows is helping protect your PC" sounds a little silly, but otherwise, good job, Microsoft. Sat 19 Jun | Martin Schultz | Whats fun about the Devils Own key? I tried googleling it but it didn't turn up anything fun. Sat 19 Jun | A Pterodactylus Ate My Baby | You're right, it isn't funny. Dunno what I was thinking (them beers, them beers). Anyway, the 'Devil's Own' is the Windows XP beta reg-code that was leaked before the official release. Apparently, people believe it was Microsoft themselves who leaked it, to keep tabs on how much xp would be pirated... So hence the name. I guess it's only funny for slashdotters and drunk people. ;) Sat 19 Jun | Dan G | the firewall in XP SP2 is useless. It cant block outbound connections. Applications added to the allow list can be spoofed as there is no hash checking Anyone logged in as admin (so, like, every home user) has the rights to programmatically alter the firewall with no windows dialog boxes, meaning someone can release a virus that disables the firewall without the user knowing This is the big issue MS should be tackling. The local admin issue. Sun 20 Jun | Myron A. Semack | 'the firewall in XP SP2 is useless.' Untrue. If everyone had the firewall turned on, it would have prevented the Sasser worm. While I agree that the integrated XP firewall isn't a good as it could be, calling it 'useless' is simply false. By your definition, the firewall in most broadband routers is also 'useless'. Sun 20 Jun | Dan G | The fact that there is no way to filter any outbound traffic what so ever, makes it nothing more than a toy Sun 20 Jun | Brad Wilson | If you say something over and over again, does it become more true? :-p Yes, it's not the best firewall, but the fact that it's on by default (and that the lack of anti-virus is an immediate warning that people will want to remedy) will cure 99.99% of the problems out there today. Why does Windows slow down over time? | Fri 18 Jun | Alex I might be wrong, but Im sensing a slight decrease in the speed of my system after about 6 months of running Windows XP. Most notably, games are a tad bit slower. What is the definitive truth about this? I mean besides Dude, M$ Windoze needs to be reinstalled once in a while, thats part of the cycle. In my case: There are no more resident (so to speak) programs than on the day of installation. I did install/reinstall/uninstall a couple of dozen programs, could the Registry have been thrashed around? (Still, Im not talking about slower start-up, but slower *operation* i.e. framerate and such.) Is it the fragmentation? The pagefile is in one piece (not fragmented), besides I have a gig of RAM so swapping shouldnt be really happening. I defragged with the Windows utility, no change. S&D Search & Destroy reports all is clean. Ive used Firebird since day one, so no spyware. Im just curious about the theoretical explanation. Exactly which part of the system degrades over time? Fri 18 Jun | Quad | 'I did install/reinstall/uninstall a couple of dozen programs, could the Registry have been thrashed around?' Guranteed. But really, this is Windows rot. It is STILL uncured. Fri 18 Jun | Alex | Ok, but the question was pragmatic. What *is* Windows rot? Fri 18 Jun | old_timer | 1) MS has a free download registry scanner/fixer program. It can clean up a lot of garbage. There's also a freeware thing called RegSeeker which does similar things. 2) Do a defrag every 6 months. Even if it says the percent of fragmentation is low. A few critical files your games may use frequently could be badly fragged. 3) Get both Ad-Aware and Spybot S&D and use them to clean out any spyware that came from Kazaa or BearShare and the like. 4) Run a virus checker at regularly scheduled intervals. Fri 18 Jun | matt | Aagh. Yeah I know what you mean. Windows just slows down and screws up gradually, unless you're super super careful about everything you install it just seems to gradually mire the computer down and the only way to speed it back up again is to reinstall. XP seems a bit better about this than, say, Windows 98, but it's still not ideal. And I have no idea either. I suspect something to do with DLLs or something. Haha. The internals of windows are just such an ugly mess I don't think I'll ever be motivated to learn much about them. Hopefully when they release Longhorn they'll get the chance to throw most of the DLL Hell / messy backwards-compatible APIs out of the window, as mentioned in a recent Joel article. And I hope to god they nuke the fucking Registry too and store it in XML files or something, what a horrific mess. Give me /etc any day. Fri 18 Jun | Windows Registry | Then take your /etc and leave me alone! Fri 18 Jun | A Pterodactylus Ate My Baby | A few tips, although Windows slowing down is something we have to accept, for now. - Run adaware, spybot search & destroy regularly. Once a month or so will do: you'll be amazed of what crap sneaks in when you're not looking. - Run antivirus. Stay off the pig that's Norton, try something like Avast or kaspersky antivirus (both free for personal use) - Get TweakXP (free evaluation works for 30 tries, plenty). There's a ton of little fiddles you can play to max out performance. - Set pagefile to 1.5x amount of memory, both min and max must be the same to avoid resizing of the pagefile. - Run windows defrag on your windows partition in SAFE mode (press F8 when windows boots). This helps defrag a lot of stuff that's normally in use. - Use, as suggested above, RegCleaner or similar to clear out the registry for old and invalid entries. - Install all security updates from windowsupdate.microsoft.com, or get the latest Service Pack beta (build 2149) - it helps a bit + all kinds of security fixes. - Keep all your manually installed applications in a separate partition - it's cleaner, and you know what to get rid off once you uninstall one (wtf is it with uninstallers - why don't they get rid of EVERYTHING they installed in the first place?) Others may have additional notes - hope it helps. Fri 18 Jun | A Pterodactylus Ate My Baby | I forgot to mention: Control Panel->Administrative Tools->Services. Disable any 'started' services you don't need, like Fax Support, Remote Registry, etc. etc. Be a bit careful, though, and read the descriptions before you go berzerk. Fri 18 Jun | Brad Wilson | It isn't so much that Windows slows down, it's that the continued dance of install/uninstall and writing data causes you to need to deframent hard disks and clean out the registry. What's worse is, of course, PC manufacturers ship us PCs chock full of crap, and no original Windows CD, so "paving" a box for most people is hardly any help at all, since it comes with 54,903 applications already installed. Fri 18 Jun | A Pterodactylus Ate My Baby | And another... :) Turn off Indexing service for 'fast file searches' rightclicking your drives. Especially your windows partition. How fast do you need to find etc/hosts anyway? Fri 18 Jun | hoser | Huh, that's funny - I was wondering the same thing this morning. Those are some good comments however, (the last 2). Thanks, Fri 18 Jun | A Pterodactylus Ate My Baby | Ok I'll shut up after this, (well so you hope): If you got some wierd ghost installation CD of Windows XP with your 'puter, then you're a legal owner, and can download a fresh windows xp install off the net. I won't tell you where, but you got a legal registration code: http://www.magicaljellybean.com/keyfinder.shtml Jot down the key and install a fresh XP on your computer. It's perfectly legal in MY country at least. If you're American, blindfold yourself and don't read this post, your laws may be dubious. Fri 18 Jun | Chris Nahr | I can't believe that people are still recommending RegClean. What century was it again? http://support.microsoft.com/default.aspx?scid=kb;en-us;299958 'RegClean is not compatible with the products listed at the beginning of this article.' [Office 2000, XP, 2003] 'The RegClean utility is no longer supported and has been removed from all Microsoft download sites.' And that was a long time ago. This article dates from 9/2003, and RegClean was last updated in 2002. Even then MS warned against using this tool on Windows 2000 or later IIRC. Fri 18 Jun | Mr Fancypants | I am well aware of how Windows historically suffered from 'Windows rot', but I haven't seen this sort of thing really happen since switching to XP and I've had some systems in daily development use for 2+ years now. With XP, I find that as long as you keep the drive defragmented, things tend to stay pretty snappy. If you are experiencing this sort of thing with XP and defragging hasn't helped, I'd definately suggest triple checking for adware programs using some of the tools listed above. In my experience those are by far the biggest culprits when it comes to unexplained slowdowns on current systems and modern versions of Windows. Fri 18 Jun | Mike Aldred | Give Diskeeper a go, it is far better than the default Windows defragmenter.  With the added bonus of a "Set and forget" option, it'll sit in the background and check the HD when you're computer is idle and deframent if needed. Sat 19 Jun | Albert D. Kallal | And, consider turning off system restore..... Sat 19 Jun | Myron A. Semack | The biggesst reason your ocmputer seems slow is that you've gotten used to it. When you first install Windows on your brand new 17GHz PentiumVIII with a terabyte of RAM, you say 'wow this is awesomely fast'. After you've used it for a few months (or years), you learn to get used to the speed. Now, it just seems too damn slow. Another example: When you first go broadband, how fast did it seem? How fast does it feel now? Probably not as fast. The actual speed of your broadband probably didn't change, you just got used to it. It's amazing how much this affects things. Human beings get used to anything. Here's a good experiment: - Think about what you feel is slow on your computer now. Make a list of those tasks that seem slow. - Measure how long it takes the computer to complete thos tasks with a stopwatch. Repeat the test a couple of times (at least three). Average the results. - Defragment your hard drive. - Re-run the tests with a stopwatch and record the values. You now have an idea of how much faster defragging your hard drive makes the computer. - Nuke your hard drive and do a complete reinstall of Windows and your applications. - Re-run those tests on your fresh install. You now know how fast your computer was when it was new. - Compare the numbers. Sat 19 Jun | Alex | You know, Myron, you might be on to something. I got this brand new Microsoft keyboard and it 'just doesn't feel the same, the old keyboard was so much better, etc etc.' until one night I had to type a lot so I decided to pull out the old keyboard. I was absolutely shocked. The *old* keyboard was *awful* now. Sat 19 Jun | Alex | Thanks all for the answers. I did deactivate a boatload of services, indexing and system restore, and defragged with O&O Complete/Name. Booting does seem faster, but I didn't use a stopwatch. Sat 19 Jun | Myron A. Semack | Don't get me wrong, your system may very well have slowed down. There could be something going on that we don't know about. All I'm syaing is that before you try fixing something, make sure it's really broken. Without any real measurements, you have no way of knowing if your system really IS slower. Sat 19 Jun | Brad Wilson | 'The biggesst reason your ocmputer seems slow is that you've gotten used to it. When you first install Windows on your brand new 17GHz PentiumVIII with a terabyte of RAM, you say 'wow this is awesomely fast'. After you've used it for a few months (or years), you learn to get used to the speed. Now, it just seems too damn slow.' That's not my experience at all. Take a Windows machine and basically leave it alone (like, a server), and it persists at a constant level of performance. Take a Windows machine and use it regularly (install, uninstall, move data around) and it degrades. Take the same machine and pave it, and it's fast again, until it degrades from regular use. Sun 20 Jun | Myron A. Semack | And how have you measured this? Do you have some timing numbers with a stopwatch? Do you have some benchamrks with WinStone or Sandra? How do you really KNOW that it's slower? Getting "out of the dark" | Fri 18 Jun | wanna be a programmer when I grow up Right now Im just learning ASP with VBScript (bleh; Im used to C syntax). Ive worked with PHP/MySQL somewhat in the past, but the current project uses ASP/VB, so thats what Im learning/working with. Im still in the middle of the learning curve, and I would love to hear your advice for which direction to go in. Heres the problem: I have solid understanding of programming basics (variables, loops etc., functions), but little-to-no understanding of more complex things (methods, classes, objects). These are little more than buzzwords to me. So far, my plan has been to forge ahead, Googling as necessary to get the job done. However, Id really like to get out of the dark and feel as comfortable with all those buzzwords as I am with, say, While loops. But I can only think of two options: 1. Keep forging ahead, trusting that Ill learn these things as necessary. Pro: Keeps me on task, spending time doing the work rather than learning what I might not immediately need. Con: feels like stumbling around in the dark when I need something more complex than what I really know. 2. Buy a good book or two and digest them. Pro: more in-depth explanation of topics. Con: Book may not be talking about whatever Im dealing with, making the learning academic and inefficient. Theres probably a third (or fourth) option Im not thinking of. Is there a best way to learn these advanced concepts apart from the university? How did you get out of the dark? Thanks again for your thoughts. Fri 18 Jun | Sassy | practive makes perfect. Take a class, read a book, write programs. You just have to do. Fri 18 Jun | Patrick FitzGerald | Try having a little look into Bruce Eckel's Thinking In Java. It is online for free at : http://www.mindview.net/Books/TIJ/ From the analysis side on the way in, have a look at Martin Fowler's Analysis Patterns. Some free samples online at: http://www.martinfowler.com/articles.html#ap Great example to start on would be: http://www.martinfowler.com/apsupp/recurring.pdf Just keep reading. It is Greek at first, then French then it starts getting clearer - the more you do, try and read. Fri 18 Jun | old_timer | > practive makes perfect. Yeah, like typing? Fri 18 Jun | Ian Olsen | Definitely both. And get used to the feeling that you had no clue a year ago, every year, until you retire. I've been a projessional developer for seven years, and I'm constantly amazed how I ever accomplished anything before I really understood whatever I figured out last week. That's obviously hyperbole, but my point is this: as a developer, you're learning or you're losing. You're never completely 'out of the dark.' Fri 18 Jun | EAW | 'I'm constantly amazed how I ever accomplished anything before I really understood whatever I figured out last week. ' I retrospect, I've experienced the exact same feeling in different areas, so I suppose it's foolish to expect a magic path toward learning programming. Patrick, thanks for the suggestions. If anyone can recommend sources that really helped them learn, I appreciate it. Thanks again. Fri 18 Jun | Dr. Real PC | Programming is actually the same, whether object-oriented or not. Methods, objects and classes are not more complex that loops and variables, they are just on a different, higher level. or organization. Classes are what you put your programs in, objects are what these classes will generate as needed, and methods are simply functions. Once you know it's really the same, and objects are just a better way of organizing programs, it doesn't seem complicated. Sun 20 Jun | MacSqueeb | Once you feel you have a working understanding of the concepts you mentioned, try some of these books: http://www.joelonsoftware.com/navLinks/fog0000000262.html I'm currently reading The Pragmatic Programmer, and already it has helped me improve. And keep reading. For a long time I only did guerrilla reading like you describe above -- Google for everything; rarely read more than 5-10 pages; get in, get what I need and get out -- no books in their entirety. That is obviously a survival skill you need as a programmer, but it's only part of the equation. I'm recommitting myself to reading on a broader range of subjects, both development and non-development related. Don't just read about the specific thing you're working on right now. If you're on an ASP/VBScript project at work, read a book about usability or database design outside of work. The interesting thing is, this will help you on that work project right now. Don't just read technical material. Software is about people, and people live in a world of ideas, so read from the entire range of ideas. By the way, credit for this advice goes to the authors of The Pragmatic Programmer, not me. I want to be programmer when I grow up too. ERP | Fri 18 Jun | Cecilia Loureiro If you were asked to buy/develop an ERP for a small-middle sized company. What system and DBMS would you recommend? Is it a software developed with VB, C#, Java or C++? or something like powerbuilder? And from the point of view of maintainability, scalability? What offers better productivity at responding to users demands... you know.... I want this report now! or I need this new procedure for yesterday!!! Fri 18 Jun | kc | It depends on your requirements. * What sort of systems do your tech guys manage now? * What sort of integration do you expect with other applications? * Are you looking to be able to customize it? * Do you envision other systems interacting with the data? * What is the lifetime of the application? Fri 18 Jun | Quad | I think I'd look at something like Gupta. Fri 18 Jun | Cecilia Loureiro | I appreciate your comments. I used Centura before along with Lotus Notes. We kept both platforms separated although we sometimes needed to integrate some functionality. We could read the same databases though (DB2 and Oracle). We also used Microfocus Cobol for some sort of batch processes... which however needed to be triggered by users. We could not use centura because it crashed. Another problem with centura was its system of libraries containing classes and so. We had many teams developing libraries grouped by business functionality. Those libraries were used by other developers who integrated them with the “GUI”. All libraries were shared among all the groups on development time… so you could be using a library for your gui while other developer was modifying it. We tried to use Centura Team Developer but it did fit our needs. We had a lot of version problems. Another problem was that libraries contained functionality for different applications. If you needed only one class – function within the library you had to “include” the entire library, and after compiling the resulting EXE (which included all the library) file could be huge. Actually centura exe files did not need any dll on clients … that was one of the advantages as the providers explained. I don’t have any experience with languages like C++ which I’ve read allows memory management not even with VB or C# which also could be used along versioning systems and which I believe are still at a lower level… compared with gupta, powerbuilder and similar. My experience goes more with 4gl like centura, visual foxpro, lotus notes and also with business intelligence tools… but that’s another story. That was the purpose of my question… so I can compare ERP development in both sort of platforms… Fri 18 Jun | Philo | If you were asked to buy a car for a small/medium sized company, what kind of car would you buy? My sarcastic way of suggesting that maybe a wee bit more on the requirements definition is required than 'an ERP system' Since you mentioned MS technologies, here's an overview of MS ERP solutions that might give you a starting point: http://www.directionsonmicrosoft.com/sample/DOMIS/update/2004/03mar/0304bsr2_sb1.htm Philo Sat 19 Jun | Espen Antonsen | Developing a ERP-system is a huge task. For 99,9% i don't even see developing a ERP-system an option. We have developed a web-based ERP-solution - see http://www.24SevenOffice.com - so I have experience with how complex ERP solution's are. There are a lot of ERP vendors on the market. Make sure you research the ones that suit your need and try to get a clear overview of the costs. ERP implementations have a way of becoming a lot more costly than what the ERP vendor says. In our case, a web-based ERP solution, we charge clients with a monthly fee. No implementation costs (except for standard start-up fees, additional training and custom features). This really helps client to see the actual future costs with the product, no suprises. Email me for more info: ea at the doimain above. Sun 20 Jun | Bobba Feet | Espen : How long will it take for your ERP to post 800 sales Invoices containing 20 Sales Invoice Line ? I always thought that processing time was a big issue on web based ERP The ocean is already boiling Joel | Fri 18 Jun | David Cornelson Nice try Joel, but youre already too late. The .NET Framework is spreading. Maybe only within corporate america for now, but eventually it will filter down to public Windows machines. Once people see the benefits of smart clients over web apps, things will change drastically. The web is a great idea for simple applications, but it will never be cost-effective or appropriate for most complex software applications. Fri 18 Jun | mpm | I once heard Bezos say exactly the same thing. Wait, what? Fri 18 Jun | Woodentongue | Simple applications represent 99.999% of all applications, 99.99999999% if you count a website as a simple system. The small systems cost between 0$ (kids playing) and a few grand. The big systems cost millions, but the global value of the little fishes and the market they create is orders of magnitude bigger than the big fishes, even if they make most noise. Fri 18 Jun | kc | One government client that we have wanted to host one of our ASP apps at their location instead of us doing it. Make sense, but they were the first ones with this request, so my boss came up with a contract and got everything signed. He *thought* he was being creative when he charged them a price of $X dollars for the current version and then added a clause to require them to buy the next version (.Net-based) in one year for the same price. As soon as the contract was signed and they had the app, they immediately filed suit against us for the clause because an 'upgrade to the .Net archtiecture would require massive infrastructure upgrades in terms of both hardware and software'. Then they pointed out how it was not cost effective at all for the taxpayers. Our guys realized that they had just been screwed because it's hard (no, impossible) to argue that a government agency should not be effective with public funds. Well, duh. Many government agencies has already outlawed the usage of IE due to security issues. Many others are discussing switching away from Windows anything for hosting due to security issues. Therefore, their contractors will have to have the same/similar systems to ensure compatibility. Then, those who interact with the bigger players (Lockheed, Boeing, etc) will have to do the same. This has the potential to be the single biggest change agent and it's already happening. Fri 18 Jun | . | 'The .NET Framework is spreading. Maybe only within corporate america for now, but eventually it will filter down to public Windows machines. Once people see the benefits of smart clients over web apps, things will change drastically.' This same sort of theme has regurgitated quite a few times in here, and it really astounds me that people can be so intentionally dense. Here's a hint - despite the apparent zombie-like belief in pushing the One True Cause, .NET did not invent rich clients (in fact .NET is a massive step back for rich clients while we wait for the GUI component vendors to catch up to what Delphi, for instance, has had for years) - we've had these crazy things called Win32 apps for years, all full of rich functionality (along with rich Java apps, Flash, etc). The absolutely minimal features like 'One click deploy' most corporations have had for years by ignoring Windows design guidelines (Quake 3 is a great app because I can just copy the folder around -- it has no external dependencies, registry entries, etc) and just dropping the build in a network share. REGARDLESS many firms have migrated to 'simple' web apps. .NET/Microsoft fanatics are far more embarrassing than Linux fanatics. Absolutely disgusting. Fri 18 Jun | JWA | 'As soon as the contract was signed and they had the app, they immediately filed suit against us for the clause because an 'upgrade to the .Net archtiecture would require massive infrastructure upgrades in terms of both hardware and software'. Then they pointed out how it was not cost effective at all for the taxpayers.' That is clearly being misconstrued for the purpose of spreading FUD. Any action such as that would have zero legal basis if the claim was based on the stated requirements of the deal. It's fairly obvious that the real issue was that the public (or someone else within the layers of government) looked at the commitments made in the deal and balked that they had to pay your fee (plus whatever installing the framework may have added) for minimal additional benefit and only one year later. This is a political financial oversight pro