If internet traffic is such a burden, sell of those assets and move into another line of business. If it's such a loser, get of the business.
They have other businesses and they're incredibly profitable. The problem is that their cash-cow businesses (phone & tv) are facing current (vonage/skype) and future (IPTV) competition from competitors that don't need to physically connect with their customers. Net neutrality isn't about controlling the broadband market. It's about preventing broadband from allowing competition to their major revenue streams.
Don't think that Google/Yahoo and all the other major content providers are simply looking out for your best interests so that you can get your search results as fast as possible. Even a website that hasn't paid its protection racket money to the broadband providers will still be just as fast or faster than it is currently. Just as broadband providers realize it, the content providers also realize that this fight is about next-gen technologies that will require substantially more bandwidth. These companies would be foolish to not be considering forays into the internet-enabled TV or phone markets. We're already seeing actual competition to traditional landlines (Vonage/Skype/etc) and we're already seeing precursors to fully-fledged competition for the currently entrenched cable/satellite providers (Google video and iTMS video).
The argument that Ma Bell wants to censor my little blog is just as much of a red-herring as the argument that content providers want to use the network for free. In the end, net neutrality would be a benefit to the consumer, but it needs to be written correctly to ensure that classes of traffic can be prioritized above other classes of traffic while ensuring that no provider's traffic is prioritized above another's. Allowing TV/Voice/other-real-time-traffic to be prioritized above less time-sensative material makes a lot of sense. Allowing SBC TV traffic to be prioritized above Google TV traffic will mean consumers continue to have little to no choice of providers.
I'm not buying it. Maybe the promotional $19.95 rates are below cost, but $50/mo isn't. If you look at independent broadband providers, they charge ~$50/mo, and they don't have the cash reserves nor the alternative revenue streams that the bigger companies have to allow them to take a loss on service.
The reality is that the consumer broadband market is like many other markets...they oversell their product with the knowledge that most customers will use far less bandwidth than they pay for. Sure, most people reading this are part of the group that's constantly bittorrenting the latest Ubuntu ISO or some other such bandwidth-intensive activity, but there's enough people (think of your parents, grandparents and all those people who you provide begrudging tech support to) out there that only use a few megs a day, but want those few megs to come down faster than dial-up.
Airlines oversell flights knowing that it'll be cheaper to give out free flight vouchers when everyone shows up rather than let seats go empty. Your local police and fire departments don't have nearly the staff to deal with everyone's problems at once. Even webhosting companies offer insane amounts of bandwidth/storage knowing that only a small percentage of customers will use it. My host offers 1TB/mo with 20 GB storage for only $7.95/mo (PLUG!). Those numbers don't come close to adding up, yet I know they're making money.
The money to support this is already being paid. The only reason this is happening is that the telcos will be damned if they let Vonage/Skype eat their lunch over connections they have some measure of control over. Data-only connections are the future, and the bells are finally wising up to this fact and trying to figure out how to maintain their monopoly profits on voice lines in a world where competitors don't have to physically connect with their customers. Prioritized traffic is the obvious answer since it leverages the physical connection to the user against the VoIP providers who they're competing with.
Your post makes it appear that you believe that the record companies are against online distribution. This is not correct: they are against online piracy.
Piracy may be what they say their main motivation is for attacking online distribution methods, but it really isn't their true aim. The record companies care about only 1 thing...monopolizing the distribution chain. They don't care what format music is being sold or who's selling it so long as the music service does not offer unsigned musicians. If such a service became popular enough, artists would never sign the one-sided contracts they're typically coerced into at the moment.
With the cost of a home studio becoming very reasonable, an alternative distribution mechanism would mean the only real service the record companies would perform is marketing. This is what they're fighting against. They're fighting a losing battle to maintain their current position between artists and consumers.
Piracy is a red herring. They realize (correctly) that their profits are closely tied to their control over what you listen to.
While nofollow failed to stop comment spam, that doesn't mean that it failed. On the contrary, it worked quite well. It's only bloggers who still have a comment spam problem that believe it failed. Contrary to the begged question, nofollow was never intended to stop comment spam. It was only intended to stop comment spam from affecting page rank. This is an important distinction.
Comment spam that affects page rank is Google's problem. Comment spam that doesn't have anything to do with page rank is not Google's problem. Google provided a mechanism to bloggers to eliminate a nuisance caused by Google's page rank algorithm. At that point, comment spam is no longer motivated by Google's page rank algorithm.
This blog entry is nothing but sour grapes that Google didn't solve their problem for them.
I'm not going to comment on whether you should switch from C++ to Java...almost every other post in this thread has addressed that issue and only your team can really make the decision on whether your code is broken enough to need an architectural fix or not. What I will do is give you an idea of what your life will be like 2 years from now if you adopt the suggestions of the consultants.
1) You will be cursing the day you ever accepted Hibernate into your development environment. After the initial "ooh...cool" phase where it effectively maps single tables to single model objects and even does a few more complex things than that, you'll start running into obscure bugs and start pulling your hair out trying to a) figure out why the @#$% it is generating certain SQL (usually a hard process) and b) figure out how to get it to generate reasonable SQL (nearly impossible). If you have anything more complicated than simple table mapping and maybe a couple of joins, avoid Hibernate like the plague.
2) Conversely, you'll be quite happy with Spring. Aside from some wierdness in SpringMVC, almost all of Spring enables you to do whatever you want. Provided you've planned ahead so that you're spring config files don't get out of hand (this should be fairly easy considering that you've already built your app once...you should be able to get a fairly close estimate of all your beans up front, so make sure you break it down into spring beans config files that have between 10 and 30 beans apiece (numbers to be a secondary concern to grouping like functionality), and you'll be fine.
3) If you have any need for single sign-on or complex authorization schemes, you'll probably have run into the pain that is Acegi and have navigated its complex array of implementable interfaces to the point where it is working well for you. Relax, and feel good that you've gotten it working and there aren't likely to be any holes in it that need fixing. To be clear, I'm not demeaning Acegi...it's a great product. It's just that the interface is quite confusing and has documentation that is, to be kind, only for advanced users.
4) You will (I hope) have a large battery of unit tests for all your code. Spring's IoC capabilities ensure that you'll almost always be able to test each piece of your code in isolation. Use this to your advantage. Write tests until you believe you've covered everything, then write some more. Use code coverate tools like Emma, Cobertura or Clover and make sure that you've tested everything (well...you can probably leave getters and setters untested).
All this will leave you far more agile than you are now. You'll be able to make changes quickly since you can re-run all of your tests to ensure that you haven't broken anything. You'll have developed a great appreciation for interfaces and what programming towards them can allow you to accomplish when it comes to pluggability of components.
But will you be better off than you would if you had continued on your current path? That isn't a question I can answer. I can't determine whether your employees would be able to adapt from a C++ environment to a Java environment. If you simply write C++ code in Java, your code base will still be ugly. I also can't answer whether it would be less effort to re-architect your product in C++ and bring with you the code that does work well. These are all questions you have to answer for yourself. Do not let management be sweet talked into dictating your technical path. There is no answer that is right all the time. Make your decision based on how painful development is now and how painful its likely to be given each path you're considering taking. Make your decision personally (or as a team, with your fellow developers), make it clear to management what your decision is and if they mandate any other path, start looking for a new employer.
A co-worker of mine is fond of the saying "There is no magic in programming" (though Spring sometimes comes pretty damn close). And it's true. A well-architected solution in almost any language will be better than a poorly architected solution. Spend your time and money on doing things right no matter which tools you choose.
There is a way to say that Black people, on average, commit more crimes than white people and still be politically correct.
If you were to say that people living below a certain socio-economic threshold (the poverty line, or some other number...it doesn't really matter) are more likely to commit crimes, no one will object. It's a sensible statement that is backed up by numbers. You can also say that Black people, on average, come from a lower socio-economic background. This too will elicit very few objections. It can be backed up by actual numbers and, moreover, many leading minority activists would be happy to bring more focus to this issue. This still asserts your original premise, but it does so in a way that doesn't attribute criminal behavior to the color of one's skin.
And this is an important distinction to make. If we as a society are to find ways to address these issues, it's important to lay blame in such a way that it is not targeted at something that isn't changeable. A black person cannot change the color of their skin. They can, however unrealistic it may be in practice, change their socio-economic status. And it's that separation which allows people to divorce their own ego from the situation. Someone will identify much closer with things like skin color, gender and other innate properties rather than things about themselves that they wish they could change.
It's important that we address these issues in a way that allows people to talk rationally about them rather than reacting with a bruised ego.
One thing that seems to get glossed over in the "SWT is optimized for Windows" argument is that SWT *is* open source.
IBM put a whole lot of time and energy into optimizing for Windows because it allows them to sell WebSphere studio for Windows. Optimizing for other platforms isn't cost effective, so they didn't do it. But what they did do is release the source, so someone with expertise on other platforms can pick up the ball and run with it and, when it passes all the tests, be merged back into the main eclipse source tree. The extensive tests are key in facilitating this process.
If SWT were to be included in the JRE, I think you'd see at least Apple (who distribute their own JRE anyways) spend sigificant effort in making SWT performant on OS X. I don't think its unreasonable to think that such an effort wouldn't be organized for Linux as well. But as long as SWT is just "that Eclipse thing", these efforts won't happen.
My one quibble with what the GP said is the "SWT is what Swing should have been" comment. SWT wasn't designed to be what Swing tries to be. It was designed to be much less OS abstracted. Much of the ugliness of the underlying OS filters through into SWT (widgets requiring parent composites, requiring developers to call dispose, requiring developers to update widgets via syncExec and asyncExec). JFace, on the other hand, is much more comparable to Swing. It burries much of the ugliness of SWT and presents a much cleaner interface to the developer. It's a shame that the people at Eclipse chose to create only an SWT download and an RCP download and not a JFace + SWT download.
What you've described isn't just a small-business phenomenon...
"American industry cannot compete effectively with the rest of the world unless something is done about the great imbalance between health care costs in the United States and nation health care systems in virtually every other country..." -- Lee Iacocca, Former chairman of Chrysler
It's somewhat telling that Diamler acquired Chrysler and not the other way around...
Well, I like digital downloads because it enables me to buy music from artists who wouldn't otherwise be able to afford to get it to me. The "music today sucks" argument tends to ignore that there is plenty of great music these days which goes un-noticed because the major distribution channels only cover a small portion of the musical landscape. Shutting down P2P apps is mostly about the RIAA shutting down alternate distribution mechanisms so that the public is forced to get their music from established distributors.
To give you an example, My roommate is a musician. When he made his first album, he ripped it himself, added a note in the ID3 tags that pointed to his performance calendar on his website, and put the mp3s on Audiogalaxy himself. This worked out great for him...he developed a small following of fans who started showing up at his shows which enabled him to start performing in slightly larger venues and make more (read: some) money performing and selling CDs at his shows. Fast forward to 2006 and he just finished his second album. Without Audiogalaxy (a P2P solution that actually helped users find music they'd like rather than just returning search results), he's now having to physically go around to small, independent record stores to try to get them to sell his CD. It's a ton more work and it's not really working...he's barely made back his outlay to get the CDs produced.
When digital downloads become common enough that every musician has equal access to that distribution channel, it will enable musicians like my roommate to actually make money from their music. But the RIAA makes digital downloads expensive enough that they won't become the main method for purchasing music. And this makes customers only want to buy stuff they think they'll like rather than experimenting. At $.99 per track, people only buy what the want. But imagine that iTunes sold songs by everyone, RIAA label or not, and allowed artists to set the price for their own music. Musicians like my roommate would be able to sell their tracks for a few cents over Apple's base cost and make far more money than they currently do. And users would download almost anything that looked appealing since each song is only a few cents. The only people who would lose out would be the RIAA artists who want to sell their music for $.99 or more, since more people would choose to spend money downloading random tracks for almost nothing rather than spending almost a buck a track.
I know there are a ton of simpler ways to do the simple task that I chose...I explicitly said that it's a trivial example. I was only trying to find a simple example that would show some of the typical unixy text manipulation functions which obscure what's actually happening on the command line. If you want a more complex example, try this: (a variation of something from O'Reilly's UNIX power tools)
ls -l *.jpg | awk '{print $NF}' | sed "s/\(.*\)\.jpg/mv '&' '\1.jpeg'/" | sh
Useful? yes. Readable? not really.
Something like: (and I've never used MS's PowerShell, so this is just OO pseudo-code, not the actual syntax)
dir.each(file) { if (file.getName().endsWith(".jpg") { file.rename(file.getName().replace("jpg", "jpeg")) }}
For me, despite 7 years of using UNIX as my primary desktop (was Linux, now OS X), the latter is much more readable.
Probably has something to do with the awk/sed/line-noise that tends to make its way between multiple pipes to take string output from one program and pass it to another.
Honestly, what's more readable: (and yes, I realize that this is a trivial example)
ls -la | '{print $9}' | xargs rm
or
del(foreach(dir).getFileName())
The idea of being able to pass objects is a really powerful one that could be extremely useful if it's done with the proper amount of documentation (Unix too would be next to unusable without man find and some of its brethren). However what are the chances that Microsoft got it right?
My guess would be somewhere in the vicinity of `touch hmmm; cat hmmm | wc -l`%
The real problem with Ebert reviews of videogame movies... is that he always assumes that movies based on games are actually based on the games they are based on.
Here's one where he doesn't make that assumption. Sometimes a movie is so badly done that it can seem like they copied the video game because you can't imagine a screenwriter coming up with so bad a plot. Many times the plots of Video Games are horrendously bad, but are intended to facilitate gameplay. These can be hard to distinguish from horrible movie plots which have no such justification. Only someone who has played the game would be able to distinguish between the two. But when a movie takes the basic premise of the game and expands it to become a movie in its own right (see Tomb Raider in the link above), there's no confusion. You really can't blame Ebert for speaking ill of the video game inspirations for these horrendous movies. You can, however, blame the people who take your fond memories of playing the game and destroy them by making such horrible films, simply because they know you'll fork over $10 to see it.
How about a "Theft Protection" program from laptop makers. Offer a service that, for an extra $200 or so, comes with a different version of the BIOS which reads the physical location of the laptop from an embedded GPS chip and looks for a Wi-Fi/ethernet connection or connected phone line to phone home to the computer manufacturers. It could even wirelessly broadcast it's location in a manner that would be picked up by other laptops nearby and relayed by those laptops if they have internet connectivity. The BIOS would then lock the drive from being written or erased if it finds that the laptop has been reported stolen.
Then all you'd have to do is report your laptop as stolen to the manufacturer and the next time it was powered up with some sort of possible internet connection, the manufacturer would contact the local authorities with the actual physical location of the laptop. Police officers like really easy busts...and it doesn't get any easier than, "Go to [address] and arrest the person using a [brand-name] laptop with seral number: [serial-number]."
There are significant privacy concerns, but as long as those are addressed in an upfront manner...something to the effect that until such time as the laptop is reported stolen, all data about the laptop is immediately destroyed (i.e. ignored). If that was specified upfront, I would definitely opt for such as service. And I can't imagine a business that wouldn't choose such a service. What people forget about laptop theft is that the damages to the victim are usually significantly greater than the replacement cost of the laptop. If I were to lose my laptop, it would take at least 2 days of work for me to duplicate my setup on new one, and that doesn't even count re-writing any code I hadn't checked into source control or documents that I haven't had a chance to burn to DVD yet. For a business, the potential value of the data on the computer is much higher.
I've had my car broken into twice, and both times, they smashed the window simply to pop the trunk and take whatever was in it. In my case, it was a set of ratty golf clubs bought at a garage sale for $10, and absolutely nothing the second time. In both cases, the thieves completely ignored almost $60 in bridge tolls that I kept in my car's visor and around $10 in quarters that I keep for parking meters.
I have no idea whether my experiences are typical, but they do show that thieves will, at least sometimes, target the unknown contents of trunks.
Re:Googlemobiles!
on
Google's DNA
·
· Score: 4, Funny
Googlemobile 0.8.12 beta
Release Notes:
- Added new feature that allows the car to integrate with Google's free Wi-Fi networks that would use your current location to find open parking spaces near you and indicate them on the Google Maps interface. - Added new DIY maintenance feature that allows you to search a database of thousands of repairs and receive instructions as well as maps to the closest retailers selling replacement parts. - Added feature to use the hot air generated by Google Fanbois (TM) as an energy source replacement for gasoline. The Googlemobile gets an EPA estimated 280 mi/GPA (Google Product Announcement). - Disabled optional feature to post back to Google that you're currently being cited for speeding or other driving infraction. The google network is no longer able to alert all other Google drivers that there are officers in that vicinity so that they can avoid being ticketed. This feature has been disabled pending the outcome of the class action suit filed on behalf of 10 separate state highway patrol organizations.
End User License Agreement (EULA)
This automobile is provided by Google "as is" and any express or implied warranties, including, but not limited to, the implied warranties of merchantability and fitness for a particular purpose are disclaimed. In no event shall google be liable for any direct, indirect, incedental, exemplary, or consequential damages (including, but not limited to, procurement of substitute goods or services; loss of use, life or profits; or business interruptions) however caused and on any theory of liability, whether in contract, strict liability, or tort (including negligence or otherwise) arising in any way out of the use of this automobile, even if advised of the possibility of such damage.
Why can't more people think like you do? And by think, I mean actually think about it as opposed to just blindly following what they're told by their religion.
What I've never understood about ID is why they believe that God wouldn't be smart enough to use evolution. Compare evolution to what's described in the Bible and evolution is much more "intelligent". It's a system that's capable of adapting to almost any challenge thrown at it without any intervention on the part of God.
Which brings me to what I've always wondered about Christians/Catholics...why do they have such an insistance on believing in a literal interpretation of the Bible? To me, the Bible seems to be more of a historical political document aimed at unifying the Roman empire, rather than an exact historical accounting. As such, the events/stories/wisdom contained within it are delivered in a fashion that facilitates internalizing its messages, lessons, etc. Yet to suggest this to people who are deeply religious usually results in a response equivalent to if you had told them that God does not exist. I've rarely seen anyone capable of separating the bible from their faith in God and Christ.
Can anyone explain why the two are so inexorably linked in most people's minds? Why are most people incapable of believing that there is a God, who created all of us by an ingenious method (evolution) and sent his son to Earth to impart the teachings necessary for us to live together peacefully and with a common morality. That is really the core philosophy of Catholocism/Christianity, not the literal events of the Bible.
(thus endeth the rantings of someone who was raised Christian but could never fully express his faith until he was able to look past the inconsistancies of the bible and recognize that the bible was written by men with agendas and that true faith in God comes from within, not without).
My problem isn't what goes into making you want it, but what goes into preventing you from wanting anything else.
Music's impact on the human mind is quite profound. It can alter mood, attention span and many other brain functions. For many people there is a primal, subconscious need for music. This can even be seen in cultures that are not exposed to the marketing hype that consumer-driven societies are. You can point to the marketing (MTV, movie tie-ins, etc), but what I believe is more important is that all of this is preventing us from hearing non-corporate music. There's two ways to shut people up, either shut them up (in a free society, this can be quite hard) or you can simply yell louder than they're capable of talking (in free societies, this tends to be quite easy, since money translates quite well into elevating the volume of your message). The marketing, P2P squashing and everything else is not so much about getting people to pay for their product, it's about drowning out any chance for independent music to inform us of its existance.
They know we'll actively seek music...we're conditioned as humans to do so. They're only concern is that they be the only type of music that we're able to find. That's where the DRM/DMCA/typical/. rants come in. These technologies are tools being used by the established media corporations to monopolize the channel between themselves and consumers in order to maintain the illusion that there is nothing else out there.
Saying employers want to pay for healthcare directly is like saying people want to file their taxes online. Yes, compared to the alternative of paying employees extra money, they can realize a savings by purchasing in bulk. But what employers really want is for the government to step up and provide nationalized health care. Lee Iacocca was very vocal about this issue when he was head of Chrystler. His argument was that employee health care costs were a major obstacle to being able to compete with automakers in Japan, Germany and other countries that have nationalized health care. I forget the exact dollar amount, but he claimed that for each car that rolled off the assembly line, he started around $1000 in the whole compared to his foreign competitors. So in order to stay price competitive, US auto makers were forced to use cheaper parts and cut other corners resulting in the general perception that US-made cars are of lower quality.
Nationalized health care isn't just a leftist pipe dream...there are a lot of companies who would benefit greatly from such a program.
To be honest, election officials are not the type of insider I'd be most worried about. How do we know that there isn't some uber-complicated escape key sequence which drops the voter into an interface which shows them all the votes that have already been cast and allows them to modify those results. Given that voting booths basically guarantee complete privacy, we'd have no way of knowing. If all we have to verify the results is that the total number recorded is equal to the total number of people who voted, it would be trivial to modify those results. The whole concept of a voting machine whose design and code is not open to the public makes it way too easy to compromise election results.
In Diebold's defense, Black Box Voting should have videotaped their investigation of the machine including keeping logs of every keystroke the entered into any interface. At a minimum, it would have shown their belief that everything relating to voting should be handled with no possible deception, but it also would have allowed Diebold to verify the integrity of the machines remotely and would give them important information about how someone determined to compromise the security of a voting machine would go about doing so.
Open Laszlo is also a nice product. It still mostly requires flash on the client-side (they have an alpha-ish support for rendering in DHTML), but the server-side is licensed under the CPL (a.k.a. what Eclipse uses).
Yes, but is it necessary to constantly tout the total number of downloads as if that really means anything? Might it not be better to find some blog entry online that tries to argue that the total number of downloads might be a spurious metric of browser popularity and try to make a Slashdot story out of it? You can even include blatant propaganda like, "It's a fact that Internet Explorer is inferior to Firefox with its extensive collection of extensions and ability to support qualified web standards."
That way, hundreds of thousands of people will see it and you will have still mentioned the incredibly high number of downloads without simply sounding like a broken record. You'll have also enumerated the features of Firefox that make it most appealing to you, and you might even get that warm feeling that comes from having had something posted to Slashdot.
Now granted that most of the hundreds of thousands of people who view your advertisement will simply dismiss it as Slashdot story-spam, but some people will actually be dumb enough to reply to it. Even more amazing, some even dumber people will respond to those people (the dumbest of which will adopt a sarcastic tone and write a long, drawn-out post that could have easily have been summerized in a single sentence).
So...back to the matter at hand, wouldn't that be better than just simply announcing each new milestone in the number of Firefox downloads?
I'd say that JBuilder was increasingly marginalized because there wasn't significant effort to make it good enough, not because there were free IDEs that matched its abilities. Look at IntelliJ...it's a for-profit product that is still doing quite well because they've put a lot of effort into making it a great product. Not to knock Eclipse (it's actually my primary IDE), but the developer experience isn't always intuitive. The plug-in nature means that you really have to get to know it before you know all the menus and preferences. The people at IntelliJ have spent a lot of time ensuring that you get a consistant experience using their IDE. It doesn't have all the bells and whistles of Eclipse, but what it does, it does well.
JBuilder's maintainers never put forth the kind of effort that IntelliJ's did. And that's why it became marginalized. IDEs are actually one kind of software where you'll always be able to charge a lot if it's for a quality product. Developers (well, the companies that employ them) will shell out cash for something that makes them more productive. JBuilder just stopped being such a product.
when you have to write extremely high-performance applications that needs to scale to thousands of users, both EJBs and JMSs become godsends.
I'm afraid we'll have to disagree on this point. I have had to write code for very high performance environments (> 1000 req/sec) and I'm convinced that EJB just won't scale the way a properly designed POJO system will. And it's not just performance either. EJB development is far more prone to errors since it's much more difficult to unit test. I've done it the EJB way and I'll never do it that way again.
You're right about JMS though...it's almost essential in a lot of situations (though if I can help it, I like to use an ESB on top of it in case I need some extra flexibility at a later date).
Not that I've hidden it particularly well, but I'm definitely an adherant to the Spring philosophy. If you're at all curious, take a look at Expert One-on-One J2EE Development without EJB. They make a really good case for avoiding EJB completely (backed up by a strong resume of successful large-scale projects).
You know, you're right...the last time I used Rails it required a tight coupling of Models to Controllers but it seems that they've remedied that. If you can inject stubs in, then you're getting much closer to being able to write actual unit tests. It still seems like Ruby is wanting for a good Mock package (unless I'm missing something...if so, please point one out), but you can work around that with stubs.
...but the whole framework is an idealogically good fit for Ruby; and it's not for Java.
Just curious, but what do you mean by this? Does any of that change with the addition of Groovy to Java 6.0? Why does language even matter to a framework? One of the points brought up at the recent SpringExperience conference was that it looked like Spring would be enabling users to IoC in components written in Ruby and other scripting languages that have a BSF implementation. Also remember that many in the Java community are shying away from EJB and other parts of J2EE that are just too bloated. If you look at what BEA has been up to recently, it looks like they are positioning their real-time Weblogic product, combined with Spring, as a high-performance replacement for the full J2EE specified environment. I can see why some of the Rails philosophy is in contrast to traditional J2EE development, but given the current trend towards small, light-weight components and IoC/AoP to wire everything together, why isn't the Rails philosophy appropriate for this developing methodology?
Slashdot Mirror
Don't think that Google/Yahoo and all the other major content providers are simply looking out for your best interests so that you can get your search results as fast as possible. Even a website that hasn't paid its protection racket money to the broadband providers will still be just as fast or faster than it is currently. Just as broadband providers realize it, the content providers also realize that this fight is about next-gen technologies that will require substantially more bandwidth. These companies would be foolish to not be considering forays into the internet-enabled TV or phone markets. We're already seeing actual competition to traditional landlines (Vonage/Skype/etc) and we're already seeing precursors to fully-fledged competition for the currently entrenched cable/satellite providers (Google video and iTMS video).
The argument that Ma Bell wants to censor my little blog is just as much of a red-herring as the argument that content providers want to use the network for free. In the end, net neutrality would be a benefit to the consumer, but it needs to be written correctly to ensure that classes of traffic can be prioritized above other classes of traffic while ensuring that no provider's traffic is prioritized above another's. Allowing TV/Voice/other-real-time-traffic to be prioritized above less time-sensative material makes a lot of sense. Allowing SBC TV traffic to be prioritized above Google TV traffic will mean consumers continue to have little to no choice of providers.
I'm not buying it. Maybe the promotional $19.95 rates are below cost, but $50/mo isn't. If you look at independent broadband providers, they charge ~$50/mo, and they don't have the cash reserves nor the alternative revenue streams that the bigger companies have to allow them to take a loss on service.
The reality is that the consumer broadband market is like many other markets...they oversell their product with the knowledge that most customers will use far less bandwidth than they pay for. Sure, most people reading this are part of the group that's constantly bittorrenting the latest Ubuntu ISO or some other such bandwidth-intensive activity, but there's enough people (think of your parents, grandparents and all those people who you provide begrudging tech support to) out there that only use a few megs a day, but want those few megs to come down faster than dial-up.
Airlines oversell flights knowing that it'll be cheaper to give out free flight vouchers when everyone shows up rather than let seats go empty. Your local police and fire departments don't have nearly the staff to deal with everyone's problems at once. Even webhosting companies offer insane amounts of bandwidth/storage knowing that only a small percentage of customers will use it. My host offers 1TB/mo with 20 GB storage for only $7.95/mo (PLUG!). Those numbers don't come close to adding up, yet I know they're making money.
The money to support this is already being paid. The only reason this is happening is that the telcos will be damned if they let Vonage/Skype eat their lunch over connections they have some measure of control over. Data-only connections are the future, and the bells are finally wising up to this fact and trying to figure out how to maintain their monopoly profits on voice lines in a world where competitors don't have to physically connect with their customers. Prioritized traffic is the obvious answer since it leverages the physical connection to the user against the VoIP providers who they're competing with.
Your post makes it appear that you believe that the record companies are against online distribution. This is not correct: they are against online piracy.
Piracy may be what they say their main motivation is for attacking online distribution methods, but it really isn't their true aim. The record companies care about only 1 thing...monopolizing the distribution chain. They don't care what format music is being sold or who's selling it so long as the music service does not offer unsigned musicians. If such a service became popular enough, artists would never sign the one-sided contracts they're typically coerced into at the moment.
With the cost of a home studio becoming very reasonable, an alternative distribution mechanism would mean the only real service the record companies would perform is marketing. This is what they're fighting against. They're fighting a losing battle to maintain their current position between artists and consumers.
Piracy is a red herring. They realize (correctly) that their profits are closely tied to their control over what you listen to.
While nofollow failed to stop comment spam, that doesn't mean that it failed. On the contrary, it worked quite well. It's only bloggers who still have a comment spam problem that believe it failed. Contrary to the begged question, nofollow was never intended to stop comment spam. It was only intended to stop comment spam from affecting page rank. This is an important distinction.
Comment spam that affects page rank is Google's problem. Comment spam that doesn't have anything to do with page rank is not Google's problem. Google provided a mechanism to bloggers to eliminate a nuisance caused by Google's page rank algorithm. At that point, comment spam is no longer motivated by Google's page rank algorithm.
This blog entry is nothing but sour grapes that Google didn't solve their problem for them.
I'm not going to comment on whether you should switch from C++ to Java...almost every other post in this thread has addressed that issue and only your team can really make the decision on whether your code is broken enough to need an architectural fix or not. What I will do is give you an idea of what your life will be like 2 years from now if you adopt the suggestions of the consultants.
1) You will be cursing the day you ever accepted Hibernate into your development environment. After the initial "ooh...cool" phase where it effectively maps single tables to single model objects and even does a few more complex things than that, you'll start running into obscure bugs and start pulling your hair out trying to a) figure out why the @#$% it is generating certain SQL (usually a hard process) and b) figure out how to get it to generate reasonable SQL (nearly impossible). If you have anything more complicated than simple table mapping and maybe a couple of joins, avoid Hibernate like the plague.
2) Conversely, you'll be quite happy with Spring. Aside from some wierdness in SpringMVC, almost all of Spring enables you to do whatever you want. Provided you've planned ahead so that you're spring config files don't get out of hand (this should be fairly easy considering that you've already built your app once...you should be able to get a fairly close estimate of all your beans up front, so make sure you break it down into spring beans config files that have between 10 and 30 beans apiece (numbers to be a secondary concern to grouping like functionality), and you'll be fine.
3) If you have any need for single sign-on or complex authorization schemes, you'll probably have run into the pain that is Acegi and have navigated its complex array of implementable interfaces to the point where it is working well for you. Relax, and feel good that you've gotten it working and there aren't likely to be any holes in it that need fixing. To be clear, I'm not demeaning Acegi...it's a great product. It's just that the interface is quite confusing and has documentation that is, to be kind, only for advanced users.
4) You will (I hope) have a large battery of unit tests for all your code. Spring's IoC capabilities ensure that you'll almost always be able to test each piece of your code in isolation. Use this to your advantage. Write tests until you believe you've covered everything, then write some more. Use code coverate tools like Emma, Cobertura or Clover and make sure that you've tested everything (well...you can probably leave getters and setters untested).
All this will leave you far more agile than you are now. You'll be able to make changes quickly since you can re-run all of your tests to ensure that you haven't broken anything. You'll have developed a great appreciation for interfaces and what programming towards them can allow you to accomplish when it comes to pluggability of components.
But will you be better off than you would if you had continued on your current path? That isn't a question I can answer. I can't determine whether your employees would be able to adapt from a C++ environment to a Java environment. If you simply write C++ code in Java, your code base will still be ugly. I also can't answer whether it would be less effort to re-architect your product in C++ and bring with you the code that does work well. These are all questions you have to answer for yourself. Do not let management be sweet talked into dictating your technical path. There is no answer that is right all the time. Make your decision based on how painful development is now and how painful its likely to be given each path you're considering taking. Make your decision personally (or as a team, with your fellow developers), make it clear to management what your decision is and if they mandate any other path, start looking for a new employer.
A co-worker of mine is fond of the saying "There is no magic in programming" (though Spring sometimes comes pretty damn close). And it's true. A well-architected solution in almost any language will be better than a poorly architected solution. Spend your time and money on doing things right no matter which tools you choose.
There is a way to say that Black people, on average, commit more crimes than white people and still be politically correct.
If you were to say that people living below a certain socio-economic threshold (the poverty line, or some other number...it doesn't really matter) are more likely to commit crimes, no one will object. It's a sensible statement that is backed up by numbers. You can also say that Black people, on average, come from a lower socio-economic background. This too will elicit very few objections. It can be backed up by actual numbers and, moreover, many leading minority activists would be happy to bring more focus to this issue. This still asserts your original premise, but it does so in a way that doesn't attribute criminal behavior to the color of one's skin.
And this is an important distinction to make. If we as a society are to find ways to address these issues, it's important to lay blame in such a way that it is not targeted at something that isn't changeable. A black person cannot change the color of their skin. They can, however unrealistic it may be in practice, change their socio-economic status. And it's that separation which allows people to divorce their own ego from the situation. Someone will identify much closer with things like skin color, gender and other innate properties rather than things about themselves that they wish they could change.
It's important that we address these issues in a way that allows people to talk rationally about them rather than reacting with a bruised ego.
One thing that seems to get glossed over in the "SWT is optimized for Windows" argument is that SWT *is* open source.
IBM put a whole lot of time and energy into optimizing for Windows because it allows them to sell WebSphere studio for Windows. Optimizing for other platforms isn't cost effective, so they didn't do it. But what they did do is release the source, so someone with expertise on other platforms can pick up the ball and run with it and, when it passes all the tests, be merged back into the main eclipse source tree. The extensive tests are key in facilitating this process.
If SWT were to be included in the JRE, I think you'd see at least Apple (who distribute their own JRE anyways) spend sigificant effort in making SWT performant on OS X. I don't think its unreasonable to think that such an effort wouldn't be organized for Linux as well. But as long as SWT is just "that Eclipse thing", these efforts won't happen.
My one quibble with what the GP said is the "SWT is what Swing should have been" comment. SWT wasn't designed to be what Swing tries to be. It was designed to be much less OS abstracted. Much of the ugliness of the underlying OS filters through into SWT (widgets requiring parent composites, requiring developers to call dispose, requiring developers to update widgets via syncExec and asyncExec). JFace, on the other hand, is much more comparable to Swing. It burries much of the ugliness of SWT and presents a much cleaner interface to the developer. It's a shame that the people at Eclipse chose to create only an SWT download and an RCP download and not a JFace + SWT download.
What you've described isn't just a small-business phenomenon...
"American industry cannot compete effectively with the rest of the world unless something is done about the great imbalance between health care costs in the United States and nation health care systems in virtually every other country..."
-- Lee Iacocca, Former chairman of Chrysler
It's somewhat telling that Diamler acquired Chrysler and not the other way around...
Well, I like digital downloads because it enables me to buy music from artists who wouldn't otherwise be able to afford to get it to me. The "music today sucks" argument tends to ignore that there is plenty of great music these days which goes un-noticed because the major distribution channels only cover a small portion of the musical landscape. Shutting down P2P apps is mostly about the RIAA shutting down alternate distribution mechanisms so that the public is forced to get their music from established distributors.
To give you an example, My roommate is a musician. When he made his first album, he ripped it himself, added a note in the ID3 tags that pointed to his performance calendar on his website, and put the mp3s on Audiogalaxy himself. This worked out great for him...he developed a small following of fans who started showing up at his shows which enabled him to start performing in slightly larger venues and make more (read: some) money performing and selling CDs at his shows. Fast forward to 2006 and he just finished his second album. Without Audiogalaxy (a P2P solution that actually helped users find music they'd like rather than just returning search results), he's now having to physically go around to small, independent record stores to try to get them to sell his CD. It's a ton more work and it's not really working...he's barely made back his outlay to get the CDs produced.
When digital downloads become common enough that every musician has equal access to that distribution channel, it will enable musicians like my roommate to actually make money from their music. But the RIAA makes digital downloads expensive enough that they won't become the main method for purchasing music. And this makes customers only want to buy stuff they think they'll like rather than experimenting. At $.99 per track, people only buy what the want. But imagine that iTunes sold songs by everyone, RIAA label or not, and allowed artists to set the price for their own music. Musicians like my roommate would be able to sell their tracks for a few cents over Apple's base cost and make far more money than they currently do. And users would download almost anything that looked appealing since each song is only a few cents. The only people who would lose out would be the RIAA artists who want to sell their music for $.99 or more, since more people would choose to spend money downloading random tracks for almost nothing rather than spending almost a buck a track.
Something like: (and I've never used MS's PowerShell, so this is just OO pseudo-code, not the actual syntax)For me, despite 7 years of using UNIX as my primary desktop (was Linux, now OS X), the latter is much more readable.
Honestly, what's more readable: (and yes, I realize that this is a trivial example)orThe idea of being able to pass objects is a really powerful one that could be extremely useful if it's done with the proper amount of documentation (Unix too would be next to unusable without man find and some of its brethren). However what are the chances that Microsoft got it right?
My guess would be somewhere in the vicinity of `touch hmmm; cat hmmm | wc -l`%
The real problem with Ebert reviews of videogame movies ... is that he always assumes that movies based on games are actually based on the games they are based on.
Here's one where he doesn't make that assumption. Sometimes a movie is so badly done that it can seem like they copied the video game because you can't imagine a screenwriter coming up with so bad a plot. Many times the plots of Video Games are horrendously bad, but are intended to facilitate gameplay. These can be hard to distinguish from horrible movie plots which have no such justification. Only someone who has played the game would be able to distinguish between the two. But when a movie takes the basic premise of the game and expands it to become a movie in its own right (see Tomb Raider in the link above), there's no confusion. You really can't blame Ebert for speaking ill of the video game inspirations for these horrendous movies. You can, however, blame the people who take your fond memories of playing the game and destroy them by making such horrible films, simply because they know you'll fork over $10 to see it.
How about a "Theft Protection" program from laptop makers. Offer a service that, for an extra $200 or so, comes with a different version of the BIOS which reads the physical location of the laptop from an embedded GPS chip and looks for a Wi-Fi/ethernet connection or connected phone line to phone home to the computer manufacturers. It could even wirelessly broadcast it's location in a manner that would be picked up by other laptops nearby and relayed by those laptops if they have internet connectivity. The BIOS would then lock the drive from being written or erased if it finds that the laptop has been reported stolen.
Then all you'd have to do is report your laptop as stolen to the manufacturer and the next time it was powered up with some sort of possible internet connection, the manufacturer would contact the local authorities with the actual physical location of the laptop. Police officers like really easy busts...and it doesn't get any easier than, "Go to [address] and arrest the person using a [brand-name] laptop with seral number: [serial-number]."
There are significant privacy concerns, but as long as those are addressed in an upfront manner...something to the effect that until such time as the laptop is reported stolen, all data about the laptop is immediately destroyed (i.e. ignored). If that was specified upfront, I would definitely opt for such as service. And I can't imagine a business that wouldn't choose such a service. What people forget about laptop theft is that the damages to the victim are usually significantly greater than the replacement cost of the laptop. If I were to lose my laptop, it would take at least 2 days of work for me to duplicate my setup on new one, and that doesn't even count re-writing any code I hadn't checked into source control or documents that I haven't had a chance to burn to DVD yet. For a business, the potential value of the data on the computer is much higher.
At least in SF, the trunk isn't a good option...
I've had my car broken into twice, and both times, they smashed the window simply to pop the trunk and take whatever was in it. In my case, it was a set of ratty golf clubs bought at a garage sale for $10, and absolutely nothing the second time. In both cases, the thieves completely ignored almost $60 in bridge tolls that I kept in my car's visor and around $10 in quarters that I keep for parking meters.
I have no idea whether my experiences are typical, but they do show that thieves will, at least sometimes, target the unknown contents of trunks.
Googlemobile 0.8.12 beta
Release Notes:
- Added new feature that allows the car to integrate with Google's free Wi-Fi networks that would use your current location to find open parking spaces near you and indicate them on the Google Maps interface.
- Added new DIY maintenance feature that allows you to search a database of thousands of repairs and receive instructions as well as maps to the closest retailers selling replacement parts.
- Added feature to use the hot air generated by Google Fanbois (TM) as an energy source replacement for gasoline. The Googlemobile gets an EPA estimated 280 mi/GPA (Google Product Announcement).
- Disabled optional feature to post back to Google that you're currently being cited for speeding or other driving infraction. The google network is no longer able to alert all other Google drivers that there are officers in that vicinity so that they can avoid being ticketed. This feature has been disabled pending the outcome of the class action suit filed on behalf of 10 separate state highway patrol organizations.
End User License Agreement (EULA)
This automobile is provided by Google "as is" and any express or implied warranties, including, but not limited to, the implied warranties of merchantability and fitness for a particular purpose are disclaimed. In no event shall google be liable for any direct, indirect, incedental, exemplary, or consequential damages (including, but not limited to, procurement of substitute goods or services; loss of use, life or profits; or business interruptions) however caused and on any theory of liability, whether in contract, strict liability, or tort (including negligence or otherwise) arising in any way out of the use of this automobile, even if advised of the possibility of such damage.
Why can't more people think like you do? And by think, I mean actually think about it as opposed to just blindly following what they're told by their religion.
What I've never understood about ID is why they believe that God wouldn't be smart enough to use evolution. Compare evolution to what's described in the Bible and evolution is much more "intelligent". It's a system that's capable of adapting to almost any challenge thrown at it without any intervention on the part of God.
Which brings me to what I've always wondered about Christians/Catholics...why do they have such an insistance on believing in a literal interpretation of the Bible? To me, the Bible seems to be more of a historical political document aimed at unifying the Roman empire, rather than an exact historical accounting. As such, the events/stories/wisdom contained within it are delivered in a fashion that facilitates internalizing its messages, lessons, etc. Yet to suggest this to people who are deeply religious usually results in a response equivalent to if you had told them that God does not exist. I've rarely seen anyone capable of separating the bible from their faith in God and Christ.
Can anyone explain why the two are so inexorably linked in most people's minds? Why are most people incapable of believing that there is a God, who created all of us by an ingenious method (evolution) and sent his son to Earth to impart the teachings necessary for us to live together peacefully and with a common morality. That is really the core philosophy of Catholocism/Christianity, not the literal events of the Bible.
(thus endeth the rantings of someone who was raised Christian but could never fully express his faith until he was able to look past the inconsistancies of the bible and recognize that the bible was written by men with agendas and that true faith in God comes from within, not without).
My problem isn't what goes into making you want it, but what goes into preventing you from wanting anything else.
/. rants come in. These technologies are tools being used by the established media corporations to monopolize the channel between themselves and consumers in order to maintain the illusion that there is nothing else out there.
Music's impact on the human mind is quite profound. It can alter mood, attention span and many other brain functions. For many people there is a primal, subconscious need for music. This can even be seen in cultures that are not exposed to the marketing hype that consumer-driven societies are. You can point to the marketing (MTV, movie tie-ins, etc), but what I believe is more important is that all of this is preventing us from hearing non-corporate music. There's two ways to shut people up, either shut them up (in a free society, this can be quite hard) or you can simply yell louder than they're capable of talking (in free societies, this tends to be quite easy, since money translates quite well into elevating the volume of your message). The marketing, P2P squashing and everything else is not so much about getting people to pay for their product, it's about drowning out any chance for independent music to inform us of its existance.
They know we'll actively seek music...we're conditioned as humans to do so. They're only concern is that they be the only type of music that we're able to find. That's where the DRM/DMCA/typical
Actually, like any prettly little thing about to turn 30, Apple will start lying about its age. This will be the second of many 29th birthdays...
Saying employers want to pay for healthcare directly is like saying people want to file their taxes online. Yes, compared to the alternative of paying employees extra money, they can realize a savings by purchasing in bulk. But what employers really want is for the government to step up and provide nationalized health care. Lee Iacocca was very vocal about this issue when he was head of Chrystler. His argument was that employee health care costs were a major obstacle to being able to compete with automakers in Japan, Germany and other countries that have nationalized health care. I forget the exact dollar amount, but he claimed that for each car that rolled off the assembly line, he started around $1000 in the whole compared to his foreign competitors. So in order to stay price competitive, US auto makers were forced to use cheaper parts and cut other corners resulting in the general perception that US-made cars are of lower quality.
Nationalized health care isn't just a leftist pipe dream...there are a lot of companies who would benefit greatly from such a program.
To be honest, election officials are not the type of insider I'd be most worried about. How do we know that there isn't some uber-complicated escape key sequence which drops the voter into an interface which shows them all the votes that have already been cast and allows them to modify those results. Given that voting booths basically guarantee complete privacy, we'd have no way of knowing. If all we have to verify the results is that the total number recorded is equal to the total number of people who voted, it would be trivial to modify those results. The whole concept of a voting machine whose design and code is not open to the public makes it way too easy to compromise election results.
In Diebold's defense, Black Box Voting should have videotaped their investigation of the machine including keeping logs of every keystroke the entered into any interface. At a minimum, it would have shown their belief that everything relating to voting should be handled with no possible deception, but it also would have allowed Diebold to verify the integrity of the machines remotely and would give them important information about how someone determined to compromise the security of a voting machine would go about doing so.
Open Laszlo is also a nice product. It still mostly requires flash on the client-side (they have an alpha-ish support for rendering in DHTML), but the server-side is licensed under the CPL (a.k.a. what Eclipse uses).
Yes, but is it necessary to constantly tout the total number of downloads as if that really means anything? Might it not be better to find some blog entry online that tries to argue that the total number of downloads might be a spurious metric of browser popularity and try to make a Slashdot story out of it? You can even include blatant propaganda like, "It's a fact that Internet Explorer is inferior to Firefox with its extensive collection of extensions and ability to support qualified web standards."
That way, hundreds of thousands of people will see it and you will have still mentioned the incredibly high number of downloads without simply sounding like a broken record. You'll have also enumerated the features of Firefox that make it most appealing to you, and you might even get that warm feeling that comes from having had something posted to Slashdot.
Now granted that most of the hundreds of thousands of people who view your advertisement will simply dismiss it as Slashdot story-spam, but some people will actually be dumb enough to reply to it. Even more amazing, some even dumber people will respond to those people (the dumbest of which will adopt a sarcastic tone and write a long, drawn-out post that could have easily have been summerized in a single sentence).
So...back to the matter at hand, wouldn't that be better than just simply announcing each new milestone in the number of Firefox downloads?
I'd say that JBuilder was increasingly marginalized because there wasn't significant effort to make it good enough, not because there were free IDEs that matched its abilities. Look at IntelliJ...it's a for-profit product that is still doing quite well because they've put a lot of effort into making it a great product. Not to knock Eclipse (it's actually my primary IDE), but the developer experience isn't always intuitive. The plug-in nature means that you really have to get to know it before you know all the menus and preferences. The people at IntelliJ have spent a lot of time ensuring that you get a consistant experience using their IDE. It doesn't have all the bells and whistles of Eclipse, but what it does, it does well.
JBuilder's maintainers never put forth the kind of effort that IntelliJ's did. And that's why it became marginalized. IDEs are actually one kind of software where you'll always be able to charge a lot if it's for a quality product. Developers (well, the companies that employ them) will shell out cash for something that makes them more productive. JBuilder just stopped being such a product.
when you have to write extremely high-performance applications that needs to scale to thousands of users, both EJBs and JMSs become godsends.
I'm afraid we'll have to disagree on this point. I have had to write code for very high performance environments (> 1000 req/sec) and I'm convinced that EJB just won't scale the way a properly designed POJO system will. And it's not just performance either. EJB development is far more prone to errors since it's much more difficult to unit test. I've done it the EJB way and I'll never do it that way again.
You're right about JMS though...it's almost essential in a lot of situations (though if I can help it, I like to use an ESB on top of it in case I need some extra flexibility at a later date).
Not that I've hidden it particularly well, but I'm definitely an adherant to the Spring philosophy. If you're at all curious, take a look at Expert One-on-One J2EE Development without EJB. They make a really good case for avoiding EJB completely (backed up by a strong resume of successful large-scale projects).
You know, you're right...the last time I used Rails it required a tight coupling of Models to Controllers but it seems that they've remedied that. If you can inject stubs in, then you're getting much closer to being able to write actual unit tests. It still seems like Ruby is wanting for a good Mock package (unless I'm missing something...if so, please point one out), but you can work around that with stubs.
...but the whole framework is an idealogically good fit for Ruby; and it's not for Java.
Just curious, but what do you mean by this? Does any of that change with the addition of Groovy to Java 6.0? Why does language even matter to a framework? One of the points brought up at the recent SpringExperience conference was that it looked like Spring would be enabling users to IoC in components written in Ruby and other scripting languages that have a BSF implementation. Also remember that many in the Java community are shying away from EJB and other parts of J2EE that are just too bloated. If you look at what BEA has been up to recently, it looks like they are positioning their real-time Weblogic product, combined with Spring, as a high-performance replacement for the full J2EE specified environment. I can see why some of the Rails philosophy is in contrast to traditional J2EE development, but given the current trend towards small, light-weight components and IoC/AoP to wire everything together, why isn't the Rails philosophy appropriate for this developing methodology?