I agree with everything you say up to that last part where you say it's bad. So what if some people with the desire or ability to educate themselves about the GPL don't use it? If they DID use it, they would more than likely just screw up and do someting stupid like Linksys did and get themselves sued, which is a huge hassle for the GPL people as much as it would be for the company.
Tough shit. Part of operating a buisness is dealing with customer support issues. If they didn't want to do that, then they can stop making flashable firmware or they can stop selling to customers directly and have other people rebrand it and deal with the support issues.
You didn't read very carfully, then. Go read it again. Pay careful attention to the reason why the build is failing. And read the section about symbol exports.
There's a GPL variant called the LGPL exactly for people who feel this way. If you feel the GPL is too harsh, then don't use it - find an LGPL, or BSD, or whatever licensed project to build off of instead. Or build it from scratch. This ain't a hard concept, and I wonder why people get so worked up about it.
You may be a great guy, but you're breaking the law - it's not legal for you to distribute GPLed software that way. Period. On the other hand, it's quite possible that AcmeSoft is violating the GPL by distributing those binary-only drivers to you, so might have some relief there.
zMuds decent, but it's only "good" in comparison with all the other available ones. And Zugg as a developer is a fucking moron, pardon my french. There's no professionalism there at all. I'll put up with crap like non-backwards compatible script syntax changes in minor version releases (with nothing more than a small note in the changelog) from a free application, but not one you're trying to sell.
zMud is a load of cruft that has a decent core and has slowly been layered on and layered on in the worst tradition of software development. New features seem to be added in a "whatever Zugg read about in Delphi Journal this week" manner, with no real consistency in interface. The database module is an excellent example - Delphi has a fantasticly easy database access libarary - it's famous for it. So what does Zugg do? He writes his own really crappy and slow one. Features are added when he feels like it and are left alone when he gets bored of them, even if they're broken, like the COM support.
Now, I have to be fair - zMud is far and away the most feature rich and the most technically complex mud client. It's got a decent UI if you ignore the crappy layout libraries and how slow the settings manager is. It's certainly quite fast, especially when you take the complexity of the parsing and trigger engines into consideration. It's easy to use. It's "good enough" for almost everyone. Basically, it's kinda like Windows.
Well, if you're going to do serious Windows development, you have to spend money. I'd recommend the Intel compiler if you've got the 700 bucks - it's argument compatible with GCC, so you can just drop it in and your current build scripts (should) work.
The main problem is that mingw32s API headers aren't up to date and in places are broken, and, at least in my experience, it can't handle the Microsoft-isms in the official SDK headers. Try the Borland C++ command line compiler, if you can.
If these are NT based machines (maybe 9x too, I'm not sure), theres an API for recieveing file/directory change notifications, without polling.
Also, real Windows development you need to compile & run on Windows - cross compiling doesn't really cut it except for trivial stuff - holds true for any cross platform development, really.
Low level modem comm does suck ass, especially if it's not high enough level for you to use stuff like TAPI.
More than that - he's claiming, at least unless theres unquoted context there, although I can't imagine what it would be, that the alleged copyright claims they have over Linux not only exist, but are so overwhelming as to trump the copyrights of every other person who has ever contributed to Linux - including Linus himself.
Hypothesis : The price of a stock isn't some defined value - it's how much it's being sold for - basically, however much the people willing to buy the stock are willing to pay for it.
Speculate that a couple people in collusion hold a large amount of shares (not hard to believe, since roughly 60% of the stock is held by insiders & big firms). If one of those people sells large amounts of his shares to the other person for slowly increasing amounts (and then back again), the stock price will rise.
Now, I'm not a broker either, and this seems pretty obvious to me so you'd think that the SEC would be all over that - so maybe theres levels of indirection or whatever.
I'm amazingly skeptical of the stock market as any indication of anything since it's all just a big collective hallucination anyway.
Simple logic dictates that capitalism, if unregulated (all those Free Market doofs out there), will erode democracy - or any other form of government, for that matter.
In a capitalist economy, the only thing that matters is capital - the buying and selling of goods and/or services. Access to votes is just another service. So is access to voters, for that matter. And the information, as we see alot these days - accurate information is a valuable commodity. Therefore, not everyone has access to it, which means that a company who controls access to information can manipulate markets. The ability to manipulate markets is just another commodity to be bought and sold on the open market.
This is an excellent point, and just as with the patriot act, sticking a rider on this bill would be a great way to get your pet legislation passed, as the reponses in this article show.
Slashdotters may remember a couple years ago when a town in California rejected a wi-fi network because they were afraid of the microwaves. That'd be in Mendocino county. They're gonna have a chore getting touchscreens into there:P
I was really excited when I read this article because using an RDBMS for persistent data storage has always pissed me off - it seems like a useless waste of cycles to write code that converts my in-memory binary data format to SQL code, so that the DB engine can parse the SQL (expensive) and turn it into an in-memory binary representation. A direct object to table mapping without the SQL interface is what I've been looking for.
Prevalence might even be usefull (but I'm not a Java programmer), but the fact that they're selling it as a replacement for an RDMBS is just silly. It's (arguably, but very likely) a better solution for object persistence than using MySQL or whatever. It certainly won't hold up for massive data storage though, and I'm sure that's not even what it's really for.
I'm a government employee too, and my feeling is that any data that we store should be in a standard format and accessible with public, standard tools. Being beholden to a private company to keep running (which we are, right now) is a ridiculous breach of the public trust.
I don't care what the software used to manipuate the data is, but the storage should be transparent. Yes, I know this means we couldn't use Oracle or SQL Server. Maybe we could funnel some of the tax dollars we spend on licensing feels to MySQL or Postgresql development.
Having worked in retail, I can say that people do in fact do this (mostly old ladies, for some reason), but not really that many and stores just eat the loss, because customer satisfaction from the liberal return policy is worth it. I think the problem here is that in "normal" retail, theres a long history of customer service and satisfaction that stores can't reasonably break without alot of ill will. In the games, and especially the computer industry, people are so used to things not working as described (shipping games that don't run), people not taking returns, and all the consumer-unfriendly stuff in general that people accept is as normal - look at the number of posts doing so.
I tend to agree with you - a liberal return policy certainly would be abused, but almost certainly not en masse. I certainly would buy more games if I could return them easily.
That said, my one major game return experience was at CompUSA. I had bought a game, and when I got home I found the box was empty - someone had nicked the CD out of it, and I hadn't noticed the cut tape. I expected a huge problem returning it, but it took 30 seconds at customer service and I wasn't even questioned. So there you go.
a) I hate the Swing look & feel. Thats purely subjective, though, so I won't make an issue of that.
b) It's slow. I don't care what anyone says about how it's not anymore. On my workstation, which is aging but perfectly capable of running any of the normal, non-Java applications I ask it too (but no gaming, please), it's slow. It's slow to start, it's cumbersome to use. I can't stand it. There's varying levels of crappiness, with the low end being, say, Oracle's admin tools, and the high end being Eclipse, but they're all crappy. I like menus that show up right when I click on them, dropdowns that populate right when I drop them, and windows that move right when I move them.
Now, thats my personal experience and thats in my personal situation, but thats why I avoid Java applications. I'll take a well written (or even a decently written) native application instead, thanks.
Hah, I shoulda double checked. It's www.wxwindows.org, of course. wxWindows won't give you binary compatability, of course. But source compatability is pretty straightforward. Thats good enough for me. And I like my gui with the native look & feel, without the ass-like slowness of Swing (Yes, I know it can be fast(er). It's not in the average case, and certainly not on hardware which is sufficent for any other purpose.)
wxWindows is, indeed, a widget lib, but it's also an OS abstraction layer. It includes things like threads, timers, filesystem, and IO classes to minimize the platform specfic code you'll need - it's very possible to write major applications that use no platform specific code whatsoever. There's a couple database libraries, but wxOTL (a wxWindows wrapper around the OTL library) is the one I prefer. It's not as mature as JDMC is, though - wxWindows is mainy used for writing desktop applications (obviously), not server side code.
wxWindows is free (it's license is similiar to the LGPL). I'm not 100% sure about the others I mentioned, but most of the mature toolkits are LGPL or something similiar. Qt is a notable exception.
Seriously - give wxWindows a look. It's very mature, it's free, it's heavily developed and well supported, and it's just fun to work with.
I don't know why you think that GTK is the only cross platform toolkit, because it's about the crappiest one (no, GTK on Windows isn't any less horrible), but good ones make cross platform programming a breeze. C++ rather than C tends to be the language of choice. I prefer wxWindows, but theres a number of others, including Qt, FLTK, and FOX.
Well, in at least one of the copies I recieved, the virus exe was a big scary looking demon head in my email client (no, not outlook). You'd think someone who spends the time crafting an email like this wouldn't put a demon head icon in the exe, but whatever.
Slashdot Mirror
I agree with everything you say up to that last part where you say it's bad. So what if some people with the desire or ability to educate themselves about the GPL don't use it? If they DID use it, they would more than likely just screw up and do someting stupid like Linksys did and get themselves sued, which is a huge hassle for the GPL people as much as it would be for the company.
Tough shit. Part of operating a buisness is dealing with customer support issues. If they didn't want to do that, then they can stop making flashable firmware or they can stop selling to customers directly and have other people rebrand it and deal with the support issues.
You didn't read very carfully, then. Go read it again. Pay careful attention to the reason why the build is failing. And read the section about symbol exports.
There's a GPL variant called the LGPL exactly for people who feel this way. If you feel the GPL is too harsh, then don't use it - find an LGPL, or BSD, or whatever licensed project to build off of instead. Or build it from scratch. This ain't a hard concept, and I wonder why people get so worked up about it.
You may be a great guy, but you're breaking the law - it's not legal for you to distribute GPLed software that way. Period. On the other hand, it's quite possible that AcmeSoft is violating the GPL by distributing those binary-only drivers to you, so might have some relief there.
You know thats like 15% of the profit margin on these home user routers, right?
zMud is a load of cruft that has a decent core and has slowly been layered on and layered on in the worst tradition of software development. New features seem to be added in a "whatever Zugg read about in Delphi Journal this week" manner, with no real consistency in interface. The database module is an excellent example - Delphi has a fantasticly easy database access libarary - it's famous for it. So what does Zugg do? He writes his own really crappy and slow one. Features are added when he feels like it and are left alone when he gets bored of them, even if they're broken, like the COM support.
Now, I have to be fair - zMud is far and away the most feature rich and the most technically complex mud client. It's got a decent UI if you ignore the crappy layout libraries and how slow the settings manager is. It's certainly quite fast, especially when you take the complexity of the parsing and trigger engines into consideration. It's easy to use. It's "good enough" for almost everyone. Basically, it's kinda like Windows.
The main problem is that mingw32s API headers aren't up to date and in places are broken, and, at least in my experience, it can't handle the Microsoft-isms in the official SDK headers. Try the Borland C++ command line compiler, if you can.
If these are NT based machines (maybe 9x too, I'm not sure), theres an API for recieveing file/directory change notifications, without polling.
Also, real Windows development you need to compile & run on Windows - cross compiling doesn't really cut it except for trivial stuff - holds true for any cross platform development, really.
Low level modem comm does suck ass, especially if it's not high enough level for you to use stuff like TAPI.
You know that of all the compilers commonly used on Windows, mingw32 is about the worst one? Maybe you need to re-evaluate some of your assumptions.
More than that - he's claiming, at least unless theres unquoted context there, although I can't imagine what it would be, that the alleged copyright claims they have over Linux not only exist, but are so overwhelming as to trump the copyrights of every other person who has ever contributed to Linux - including Linus himself.
Speculate that a couple people in collusion hold a large amount of shares (not hard to believe, since roughly 60% of the stock is held by insiders & big firms). If one of those people sells large amounts of his shares to the other person for slowly increasing amounts (and then back again), the stock price will rise.
Now, I'm not a broker either, and this seems pretty obvious to me so you'd think that the SEC would be all over that - so maybe theres levels of indirection or whatever.
I'm amazingly skeptical of the stock market as any indication of anything since it's all just a big collective hallucination anyway.
In a capitalist economy, the only thing that matters is capital - the buying and selling of goods and/or services. Access to votes is just another service. So is access to voters, for that matter. And the information, as we see alot these days - accurate information is a valuable commodity. Therefore, not everyone has access to it, which means that a company who controls access to information can manipulate markets. The ability to manipulate markets is just another commodity to be bought and sold on the open market.
This is an excellent point, and just as with the patriot act, sticking a rider on this bill would be a great way to get your pet legislation passed, as the reponses in this article show.
It might very well be slower than 2.4, but "feel" faster. The low latency stuff can improve responsiveness at the expense of performance.
*sniff* I kinda miss that place.
Prevalence might even be usefull (but I'm not a Java programmer), but the fact that they're selling it as a replacement for an RDMBS is just silly. It's (arguably, but very likely) a better solution for object persistence than using MySQL or whatever. It certainly won't hold up for massive data storage though, and I'm sure that's not even what it's really for.
I don't care what the software used to manipuate the data is, but the storage should be transparent. Yes, I know this means we couldn't use Oracle or SQL Server. Maybe we could funnel some of the tax dollars we spend on licensing feels to MySQL or Postgresql development.
I tend to agree with you - a liberal return policy certainly would be abused, but almost certainly not en masse. I certainly would buy more games if I could return them easily.
That said, my one major game return experience was at CompUSA. I had bought a game, and when I got home I found the box was empty - someone had nicked the CD out of it, and I hadn't noticed the cut tape. I expected a huge problem returning it, but it took 30 seconds at customer service and I wasn't even questioned. So there you go.
b) It's slow. I don't care what anyone says about how it's not anymore. On my workstation, which is aging but perfectly capable of running any of the normal, non-Java applications I ask it too (but no gaming, please), it's slow. It's slow to start, it's cumbersome to use. I can't stand it. There's varying levels of crappiness, with the low end being, say, Oracle's admin tools, and the high end being Eclipse, but they're all crappy. I like menus that show up right when I click on them, dropdowns that populate right when I drop them, and windows that move right when I move them.
Now, thats my personal experience and thats in my personal situation, but thats why I avoid Java applications. I'll take a well written (or even a decently written) native application instead, thanks.
Hah, I shoulda double checked. It's www.wxwindows.org, of course. wxWindows won't give you binary compatability, of course. But source compatability is pretty straightforward. Thats good enough for me. And I like my gui with the native look & feel, without the ass-like slowness of Swing (Yes, I know it can be fast(er). It's not in the average case, and certainly not on hardware which is sufficent for any other purpose.)
wxWindows is, indeed, a widget lib, but it's also an OS abstraction layer. It includes things like threads, timers, filesystem, and IO classes to minimize the platform specfic code you'll need - it's very possible to write major applications that use no platform specific code whatsoever. There's a couple database libraries, but wxOTL (a wxWindows wrapper around the OTL library) is the one I prefer. It's not as mature as JDMC is, though - wxWindows is mainy used for writing desktop applications (obviously), not server side code.
Seriously - give wxWindows a look. It's very mature, it's free, it's heavily developed and well supported, and it's just fun to work with.
I don't know why you think that GTK is the only cross platform toolkit, because it's about the crappiest one (no, GTK on Windows isn't any less horrible), but good ones make cross platform programming a breeze. C++ rather than C tends to be the language of choice. I prefer wxWindows, but theres a number of others, including Qt, FLTK, and FOX.
Well, in at least one of the copies I recieved, the virus exe was a big scary looking demon head in my email client (no, not outlook). You'd think someone who spends the time crafting an email like this wouldn't put a demon head icon in the exe, but whatever.
I just want to know where you found a woman willing to let you show her boobies to everyone on slashdot :P