While OpenOffice is a really good effort it has yet to become a true Office killer. There was a time in the past when Microsoft was not the big name in productivity applications. Take for instance Excel for Windows. It was up against Lotus 123 which was pretty much the speadsheet application.
Companies being presented with Excel had several barriers of entry that Microsoft had to overcome. They had existing 123 spreadsheets that they needed to keep using, Excel required Windows, Excel cost money, and they had people that had been trained on 123. Microsoft responded by giving Excel the capability to read 123 spreadsheets, shipped runtime versions of Windows for free with Excel, gave Lotus switchers discounts, and made Excel friendly to 123 macros and keyboard shortcuts. With every one of these issues Microsoft tackled they got more and more Lotus adherents to switch to Excel. When they added the capability in Excel 4.0 to write 123 files, they eliminated the last thing standing in the way of most switchers. When Excel people could existing transparently in 123 environments organizations could switch to Excel which was packaged with Microsoft's other productivity app, Word.
OpenOffice is slowly following a similar path as Microsoft did in the late 80s. They're doing what they can to make OO.o capable of handling existing Office documents, making sure it runs on Windows, giving it away for free, and presenting users with apps that look and behave similarly to their Microsoft counterparts. As each of these offerings gets better and the whole package becomes "like Office, but better" people will switch from Microsoft Office to OO.o. You can't hit only some of the marks and have everyone switch over just because the package is free. There's a lot more to cost than the initial price of things.
OpenOffice.org is no more an Office killer than Excel 2.0 was a Lotus 123 killer. It took the ability to do everything 123 did and then some for the scales to really tip in Excel's favor. However as Excel's topping of 123 or Word's topping of WordPerfect has shown, just because a product is immensely popular (for whatever reason) it isn't impossible to dethrone it by offering something better.
How about not teaching a particular OS at all? It doesn't matter what computer system people are taught if they understand what they're doing and how it works in general. While you might not necessarily write a web based CRM in BASIC or Logo those languages do teach you the basics of programming arithmetic and flow control.
Word processing is the same way. You can teach people typography on a computer without requiring them to learn a particular application. Typography is typography, using a computer to compose a document simply gives you more flexibility than older mechanical methods. This is what needs to be taught to kids. Sticking three computers in front of them and expecting them to choose the best in their perspective is ridiculous as they aren't likely to magically know the ins and outs of each system to make such a choice. It isn't important that they know Windows or know Linux, it's important that they know how a computer works and what sort of stuff you can do with it.
The iMac can be expandable at the expense of its design or require expensive customer expandsion cards. Let's say Apple managed to stick an upgradable video card in the iMac. They would be able to sell custom made video cards specifically for the iMac or let a third party do so. In either case most iMac buyers would not be upgrading their cards so the production volume would be very low. If they were able to use commodity graphics cards there is no way they would be able to maintain their current form factor. You would then end up with a PowerMac...which you should have bought in the first place if you're concerned about expandability.
It's untrue that your brother's iMac won't be able to use Quartz 2D Extreme. The requirements are either a GeForceFX or Radeon that can handle programmable pixel and vertex shaders (ARB_fragment_program OpenGL extension). The iMacs have a GeForceFX 5200. While this isn't the most powerful graphics card in the world it will handle Q2DE. Complaining that it is "outdated" is a little silly.
Most PCs sold by the likes of Dell and HP aren't much more expandable than the iMac or eMac. Macs are hardly any more disposable than any other manufacturer's computers. The iMacs, even the previous model G5s have Firewire, USB, optical audio, modem and Ethernet ports. An Airport card can be added internally. That doesn't leave a whole lot of need for extra devices to be added to the system. The new ones have Bluetooth built-in as well as gigabit Ethernet.
Lots of features out of the box doesn't make the system disposable, it makes it a good investment. You can sit down at a several year old G3 iMac and be pretty productive on them even running Tiger and Office 2004. They had enough features packed into the box that they're still as useful as they were when they were new. In fact they might even be a bit faster now than they were running previous versions of OSX. Just because you can't stick a new video card in your computer doesn't mean it is disposable.
Broadband + web server + DynDNS = self-hosted website.
There's plenty of people that might like to be able to host their own website. If you want to put up a website for family members and your broadband provider supports servers (may require a "business" account) it isn't terribly difficult to make one. You can also with a little bit of reading set up a fully dynamic company website hosted from your back office.
If say you're in a dorm in college with a publicly accessible IP you can run your blog or gaming news site or whatever from the Mac mini under your desk without needing to be a security guru or really needing all that much experience.
Enterprise communication and RSS make a really excellent combination. Feed readers are a dime a dozen or very simple to build. RSS hosting requires little more than a web server which means damn near everyone with a workgroup server or even a retasked don't-touch-this-workstation-or-else server can get in on the act.
Many readers support SSL and HTTP authentication which means connections to private feeds is relatively secure, moreso than most organizations e-mail systems. Having a small RSS reader running in the background is also a lot more efficient most of the time than the "information managers" (Outlook, Entourage, Evolution) most people run. This is doubly true when you've got memos and announcements coming to the same mailbox as your other intraoffice mail. If you're out and about using your cell phone or dial-up service you don't necessarily want to download upteen thousand mail messages, just the latest news items.
If you look around I think you'll be surprised at the amount of metadata that exists in all sorts of files. I can have Word automagically add my name, company, manager, e-mail, and all sorts of other little tidbits to a document in a meta data section. I just have to add this data once and from then on Word, Excel, and Powerpoint will do the tedious work for me. This metadata can be searched and indexed by Spotlight.
So for instance, I'm working on a project with a few other people in my department and we're sharing files and sending e-mails all over the place. I can bring up Spotlight and type "John Smith" and it will find me e-mails send to and from John, his Address Book contact, calendar events with John associated or mentioned, and importantly Word and Excel documents he has sent to me. All of this data was parsed from information that was already lying around various locations on my computer. Word and Excel were already adding John's name to files that he authored.
Your premise is that metadata sucks because no one bothers adding it to files. It isn't the end users that will be or need to be doing the work, the onus for that is on the developers. Since Spotlight will now be a major feature in the OS, developers will begin going out of their way to not only use Spotlight in their own applications but make sure their file formats are extremely Spotlight friendly. It's only going to get more prevalent from this point on because it is going to become a lot more useful to the person on the street.
But Apple is delivering Spotlight in two weeks to regular Joes. Spotlight's back end has also been around since MacOS 8.5 when Sherlock was introduced. Content indexing has existed in various forms since then and I happen to use it all of the time. Spotlight basically takes content and metadata indexing and puts a swank front end on it.
Dashboard as was mentioned by someone else is a reimplementation of Desktop Accessories. Arlo Rose is a bright guy and Konfabulator is a nice program but it was hardly all that original. There's a significant difference between the two in terms of resource usage and implementation. Dashboard widgets are only invoked when Dashboard is brought to the fore. Konfabulator widgets run constantly because Konfabulator runs constantly.
I don't quite see how including missing functionality in an upgrade is a bad thing. Jaguar and previous versions of OSX did not have an easy way to compress files. Panther added the ability to compress files and folders as Zip files that can be opened on Windows or other Macs. Was Apple trying to kill off Aladdin née Allume's sales of StuffIt? Was Apple ripping off StuffIt by adding a function theretofore missing from the OS? Panther also re-added support for labels in Finder. Uh oh, they must have been ripping off Unsanity's LabelsX. One of the biggest gripes Mac users had about early versions of OSX was the lack of features they had come to enjoy that were in previous versions of the OS, Desktop Accessories and labels being two of these things. Adding these things back into their OS is hardly dumping on third party developers.
While XML isn't the answer to every question it does have its uses. Just transmitting straight database information like I used in my example might not be the best use of it. If I started to add meta data to each data element though, XML would be much more well suited.
With your Perl code putting a CSV into a hash you might have to change that around for every program you write. So eventually you will build a generic module that handles CSV files so anyone can throw a CSV into a function's arguments and have a nice and neat hash of all of the data. Now you're put as much work as an XML parser that does pretty much the same thing. Again it comes down to your uses, if you're writing apps to import Access databases into MySQL using CSV table dumps your CSV module would likely be quicker than XML::Parser. If your database app allowed for direct XML exporting which maintained all sorts of important meta data then the XML::Parser route would probably be more effective.
I use XML sparingly because it tends to be a little heavy on the processor doing the parsing. I wouldn't swear by it but I'm not going to disavow it because parsing can take a few extra cycles here and there. There's plenty of times where providing some XML support can really save some cross-platform/environment/language hassle, especially for lousy programmers.
Except the XML file tells the parser where its own definition is. Each of the XML files inside of an OO.o package tell you how to figure out what they are. A generic XML parser can at least find the URI to the file's type definition. Knowing what all of the elements represent makes it possible to figure out what to do with all of the data and to tell if the XML file itself is even made properly. A comma delimited file doesn't tell the parser what each of the fields means, neither does a straight binary data file. At best they annouce their filetype in a header.
Twenty years from now the dude writing a parser for the XML can do so largely without a manual or any other documentation. The latitude and longitude and second address line aren't necessarily obvious in the CSV example. If the data was much more ambiguous, like the dump from a database, the XML would be even more self explanitory. While XML might not be the absolute best or most space efficient answer to self describing data it does a pretty good job, much better than a lot of the data encoding schemes it is starting to replace.
I think that Enterprise was doomed from the start for a number of reasons. Star Trek fans footing the bill for another season will be dumping their money in the trash. While the third and fouth seasons have been better than the first two they are by no means on par with the other series in the franchise. I see this being the fault of two groups, Rick Berman and Brannon Braga and Paramount/UPN.
Berman and Braga have taken really awesome concept and driven it into the ground. Enterprise was going where no Trek series had been (pun intended), the beginning of the Federation. Enterprise could have been an awesome pre-history of the other series. Instead Berman and Braga have paid little attention to established pre-history of the later series. This wouldn't be so bad if their story arcs didn't require a reset button at the end of the series just for later ones to exist in the mythos. While sticking to Trek canon isn't required for the show to be successful it does help keep existing Trek fans interested. People who grew up watching Romulans and Klingons want to see Romulans and Klingons. New species are nice and all but there's a lot of untold history that Enterprise was well positioned to handle.
UPN is doing just as much damage as stupid story arcs and bad writing. UPN is a joke of a network and Enterprise on its lineup is the punchline. Unless Trek fans are also fans of stupid one-off sitcoms there's little on UPN's lineup to attract them or affiliates. UPN also decided to shuffle Enterprise all over their schedule making it difficult for casual viewers to keep up with the dumb story arcs.
While I appreciate people wanting to see more of their favorite franchise I think the world needs a break from Star Trek for a while. Berman at the helm of the franchise is driving it into the ground. TNG was a successful revival because there was a desire for more Trek that was more than a rehash of the original series. There's been continuous Star Trek on television since the late 80s. A break is required, as is some new blood at the helm.
You hit upon two very salient truths about the Huygens probe. The main imaging instrument on Huygens is the Descent Imager Spectral Radiometer (DISR). It contains only a single CCD. The various lens assemblies are connected to the main CCD by a series of optical fibers, each shines light on a different region of the CCD. The CCD is 512x520 pixels though about half of that is reserved as a storage section, the left over 256x520 pixel area is used for the imagers and spectrometers. The visual imagers of the DISR intrument are the High Resolution Imager (HRI), Medium Resolution Imager (MRI) and the Side Looking Imager (SLI). There's also two spectrometers, the Upward-Looking Visual Spectrometers (ULVS) and the Downward-Looking Visual Spectrometer (DLVS). Finally there's space reserves for the Solar Aureole camera which is used to measure sunlight streaming through the atmosphere to determine the size of particles within it.
The HRI is 160x254, the MRI is 176x254, and the SLI is 128x254 pixels each. Larger images have to be assembled as mosaics and even these aren't going to be large enough to compete with the megapixel images from the MERs. The Pancams and Navcams on the MERs are 1024x1024 each and have essentially a full range of motion so really nice panoramas are easy to create. The DISR is fixed on Huygen's chassis.
Bandwidth is also a tremendous issue with Huygens. The Huygens probe only hasd a 4800bps datalink to Cassini and has to transmit all of its images within two and a half hours. Even with its limited data rate Huygens was able to transmit 350 images back from Titan which is rather impressive.
So it is a combination of geometry, bandwidth, and limited technology. Also remember that despite these images being relatively stark in comparison to MER images they contain tons of very valuable information. When researches have had more time to process Huygens images they will get prettier. Until then they're going to remain relatively bland to laymans' eyes but terribly exciting to scientists.
While it is disappointing to see ISPs dropping Usenet support doing so will hardly kill it. The awesome part of Usenet is it is a naturally distributed network of systems. It doesn't take much to carry the text-only traffic of Usenet, especially considering the price of processing power and network bandwidth anymore. Binary feeds take quite a bit more but if you want the basics the barrier of entry is relatively low.
While web-based forums have gotten very popular in the past few years they simply do not have the advantages of Usenet groups. A forum is limited by a single server/cluster's capacity in terms of both bandwidth and processing power. An angry admin, hacker, FBI raid, or backhoe can take down even the largest of web forums. It would take a lot of doing to kill a newsgroup. A couple of yahoos with spare Linux boxes could keep a group going without much effort. Forums also fall down when it comes to availability. To access a thread on a forum you need to be connected to the web. A newsgroup's posts can be downloaded once and held onto for as long as you'd like. This is a feature mailing lists also have over web forums, the entire history of the list can be stored in your local mail spool. While a forum is likely to be public accessible the sum of its content is rarely available for anyone to mirror if they have the prerogative.
Programs like Leafnode allow you to create local mirrors of feeds while Usenet-Web can process those spools to make them available to anyone with a web browser. Emoticons and oversized picture signatures are little reason to use web forums in lieu of newsgroups.
...and the Great Satans in the West supporting them. You cannot be taken seriously if you ignore the inconvenient fact that the West, most especially the US, has been supporting those regimes and Islamic theocracies.
In the 1982 during the Iran-Iraq war (1980-88) when Iran appeared to have gained an upper hand in the conflict Reagan and his advisers decided to secretly supply Hussein with military equipment, dual use industrial technology, and even tactical assistance in the form of satellite photos of battlefields. The dual use industrial technology gave Iraq the ability to mass manufacture chemical and biological weapons. The US didn't stop supporting Iraq even after it used chemical weapons against the Iranians.
Even al Qaeda was a US sponsored organization at one point. al Qaeda actually means "the base", the CIA's nickname for the group when they were paying them and the Taliban to battle the Russians during the invasion of Afghanistan. Post-war Afghanistan is still impoverished and nearly as dangerous as pre-war Afghanistan. Pre-war Afghanistan was a dirtball largely because of the Russian invasion and subsequent war that we helped provoke.
Labeling someone as "evil" because they're doing what we taught them and paid them to do is a little ridiculous. A kid blowing himself up on a bus is no more or less evil than a guy dropping thousand pound bombs out of an airplane. They're both doing what they're doing for their own reasons and doing what they believe is the "right" thing to do. Maybe, just maybe, no one ought to be blowing anyone up for any reason. Blowing people up simply entices more people to blow more people up. As Ghandi said, an eye for an eye leaves everyone blind.
I don't think you really understand the need for gravitational assists. Without such assists a probe using current launch technology would not be able to make it anywhere close to the edge of the solar system. We can't just fire these things off in a roughly straight line. They're launched in what are essentially variations of a Hohmanntransfer orbit. Instead of moving from one circular orbit to another though, you're transfering from a circular orbit to a hyperbolic one. The same concept applies, energy needs to be added to increase the altitude of the orbit which is where the gravity assist comes in.
Shooting a probe off perpendicular to the mean plane of the solar system would be shooting yourself in the foot. You'd be entirely missing out on the enormous gravitation boost Jupiter and/or Saturn offer. You would need überamounts of extraordinarily energetic fuel to make the trip without using the gas giants. Even then the rocket (as mentioned) would need to be about 98% fuel. We can at best build ones that are 90% fuel.
There's also a shitload of vaccuum in the planetary plane just as there is perpendicular to the plane. There would really be little use in avoiding the planetary plane as you suggest. Besides the boost the probe would get from the gas giant(s) it could be put on a course where some secondary science could be conducted on the way to the edge of the solar system. You wouldn't need extra equipment to study doppler shifts as the probe entered say Jupiter's gravity well during an assist or watch it's radio signal as it was occulted by the planet for frequency varations to infer magnetic fields or ionized particles in the atmosphere.
These companies were given their spectrum licenses at firesale prices by the FCC. They have since then been allowed to conglomorate into even larger entities by the FCC. They have also lobbied and litigated to kill any and all threats to their control of their markets.
This is protectionism under the guise of capitalism. Here's both of our kick in the nuts.
Many providers, including T-Mobile and AT&T/Cingular in the US, have moved to SMS over GPRS. Most if not all handsets sold now also support SMS over GPRS which has much more bandwidth available than traditional GSM messaging channels. SMS messages are sent via GPRS like any other bit of data. There's little reason for GSM carriers to charge obscene amounts of money for messaging (other than to rip us off).
Charles Parnot at Stanford made a grid for his personal project. He's got more than a hundred people donating some spare cycles to his grid which is pretty impressive for a fairly small project. Daniel Côté started an awesome project to get Xgrid working on non-OSX Unix systems. With a bit of work his Xgrid Agent program could be really robust and reliable enough for getting real work done. Like you I'd like to see this technology proliferate so maybe we can start seeing open grids pop up in various computer user communities.
Now all we need in the US is services that are condusive to actually serving content instead of simply consuming it. I suppose the ridiculous asymmetry of broadband services in the US ought to be expected of a country raised by televisions.
I've got a broadband connection. It's 3Mbps downstream and 256Kbps upstream. While it is decidedly quicker than a 56k dial-up connection in either direction it is definitely not designed let me serve content at reasonable speeds. Many ports are also blocked at the cable company's head end so I can't use standard service ports (80, 21, etc). I also have to pay an obscene amount of money if I want a static IP address that I can point a DNS entry to.
Some people do have residential broadband that offers saner upstream bandwidth, no port blocking, and free static IPs. Unfortunately this is not the norm here. Most of us either have to pay for hosting or a "business" service package from our broadband provider. In either case we're paying a lot of money for services that ought to be provided for all broadband users.
What is not needed is hydrogen powered cars but a viable means to generate the power to make hydrogen (energy) readily available for everyone. Molecular hydrogen as demanded by the "hydrogen economy" is very simply a medium with which energy is physically transfered. Gasoline is as much of an energy transport medium as hydrogen.
Both hydrogen and gasoline can be used to generate electrical energy, gasoline and its hydrocarbon cousins however release the carbon part of their hydrocarbons. Hydrocarbons can be an excellent source of hydrogen and can be manufactured and transported relatively cheaply. If they're originally generated via a renewable process you can get a carbon-neutral system. The carbon pumped into the air will be used in photosynthesis by plants used to make more hydrocarbon fuels.
Two billion dollars is a lot of money but it would get used in a lot of inappropriate ways if it went towards deconstructing our oil economy. There is enormous vested interest in the status quo of the oil industry. Those interests far exceed a relatively paltry two billion. Any research funded by said two billion would be met with ten times that amount in sabotage (FUD, physical sabotage, political chicanery) by groups threatened by such research. The money being spent to further the sum of human knowlege would end up in the long run being more effectively used. The research in robotics and telepresense could be used immediately on the ground to help normal every day people.
Weening our society and economy off oil is going to take a lot more than two billion dollars. It will require not only research dollars but also a great deal of social and cultural change. Meaningful conservation efforts would go a long way to make our current energy generation and transport system more effective and longer lasting without needing fancy new technologies. Fancy new technologies would then serve to make those meaningful conservation efforts even more meaningful and effective.
For college kids scraping pennies together there's Apple credit accounts that work like a credit card. With these you can make minimum payments for as long as you don't have money and then pay down the principal when you do have it. Minimum payments on a ~$1500 computer aren't going to be too ridiculous, if you can save up $300 over a couiple of months you can easily make such payments. These coupled with educational discounts ends up being a decent enough deal if you can't afford a system outright. For high school students there are fewer options they can actualize independently but that's why it pays to keep a good relationship with your parents/family.
Even at $600 an "iSlab" is not a very profitable idea for Apple. The target audience for such a system would be...cheapskates, people wanted an "intro" Mac, and offices wanting a bunch of one-off desktops.
For offices an all-in-one iMac or eMac would be a better deal than a headless system. AppleCare covers the machine and monitor for up to three years, that's essentially an upgrade cycle for many offices. With an AIO systems there's less for the IT people to worry about. There's far fewer connectors to get stressed from squeezing the system into odd places and the systems are much easier in general to position in the first place. The G5 iMacs take up almost no space at all and don't weigh very much and will easily last in an office environment for a number of years.
For the people wanting an intro Mac a headless system is going to be far more hassle than its worth at a $600 price point. The eMac's monitor easily rivals the "free" 17" CRTs that come with some Dell and HP systems after rebates and such. The LCD of the iMacs is really nice and competes well with any third party displays you'd probably hook up to an "iSlab". If you've got a monitor that "you can just use" and want an eMac there's very likely some schools or other such places that would love a free monitor that you can write off on your taxes.
There might be a market for a $600 "iSlab" but it isn't necessarily one worth tackling for Apple. If people buy these and no other Mac software it doesn't entice anyone to make more Mac software. If people want a cheaper system to try out OSX there's a really large used/refurbished Mac market in existance. With a moderate amount of effort you could find a nice Mac for $600 or less that would run OSX nicely (which itself can bre resold to go towards a new Mac).
Slashdot Mirror
While OpenOffice is a really good effort it has yet to become a true Office killer. There was a time in the past when Microsoft was not the big name in productivity applications. Take for instance Excel for Windows. It was up against Lotus 123 which was pretty much the speadsheet application.
Companies being presented with Excel had several barriers of entry that Microsoft had to overcome. They had existing 123 spreadsheets that they needed to keep using, Excel required Windows, Excel cost money, and they had people that had been trained on 123. Microsoft responded by giving Excel the capability to read 123 spreadsheets, shipped runtime versions of Windows for free with Excel, gave Lotus switchers discounts, and made Excel friendly to 123 macros and keyboard shortcuts. With every one of these issues Microsoft tackled they got more and more Lotus adherents to switch to Excel. When they added the capability in Excel 4.0 to write 123 files, they eliminated the last thing standing in the way of most switchers. When Excel people could existing transparently in 123 environments organizations could switch to Excel which was packaged with Microsoft's other productivity app, Word.
OpenOffice is slowly following a similar path as Microsoft did in the late 80s. They're doing what they can to make OO.o capable of handling existing Office documents, making sure it runs on Windows, giving it away for free, and presenting users with apps that look and behave similarly to their Microsoft counterparts. As each of these offerings gets better and the whole package becomes "like Office, but better" people will switch from Microsoft Office to OO.o. You can't hit only some of the marks and have everyone switch over just because the package is free. There's a lot more to cost than the initial price of things.
OpenOffice.org is no more an Office killer than Excel 2.0 was a Lotus 123 killer. It took the ability to do everything 123 did and then some for the scales to really tip in Excel's favor. However as Excel's topping of 123 or Word's topping of WordPerfect has shown, just because a product is immensely popular (for whatever reason) it isn't impossible to dethrone it by offering something better.
How about not teaching a particular OS at all? It doesn't matter what computer system people are taught if they understand what they're doing and how it works in general. While you might not necessarily write a web based CRM in BASIC or Logo those languages do teach you the basics of programming arithmetic and flow control.
Word processing is the same way. You can teach people typography on a computer without requiring them to learn a particular application. Typography is typography, using a computer to compose a document simply gives you more flexibility than older mechanical methods. This is what needs to be taught to kids. Sticking three computers in front of them and expecting them to choose the best in their perspective is ridiculous as they aren't likely to magically know the ins and outs of each system to make such a choice. It isn't important that they know Windows or know Linux, it's important that they know how a computer works and what sort of stuff you can do with it.
The iMac can be expandable at the expense of its design or require expensive customer expandsion cards. Let's say Apple managed to stick an upgradable video card in the iMac. They would be able to sell custom made video cards specifically for the iMac or let a third party do so. In either case most iMac buyers would not be upgrading their cards so the production volume would be very low. If they were able to use commodity graphics cards there is no way they would be able to maintain their current form factor. You would then end up with a PowerMac...which you should have bought in the first place if you're concerned about expandability.
It's untrue that your brother's iMac won't be able to use Quartz 2D Extreme. The requirements are either a GeForceFX or Radeon that can handle programmable pixel and vertex shaders (ARB_fragment_program OpenGL extension). The iMacs have a GeForceFX 5200. While this isn't the most powerful graphics card in the world it will handle Q2DE. Complaining that it is "outdated" is a little silly.
Most PCs sold by the likes of Dell and HP aren't much more expandable than the iMac or eMac. Macs are hardly any more disposable than any other manufacturer's computers. The iMacs, even the previous model G5s have Firewire, USB, optical audio, modem and Ethernet ports. An Airport card can be added internally. That doesn't leave a whole lot of need for extra devices to be added to the system. The new ones have Bluetooth built-in as well as gigabit Ethernet.
Lots of features out of the box doesn't make the system disposable, it makes it a good investment. You can sit down at a several year old G3 iMac and be pretty productive on them even running Tiger and Office 2004. They had enough features packed into the box that they're still as useful as they were when they were new. In fact they might even be a bit faster now than they were running previous versions of OSX. Just because you can't stick a new video card in your computer doesn't mean it is disposable.
Broadband + web server + DynDNS = self-hosted website.
There's plenty of people that might like to be able to host their own website. If you want to put up a website for family members and your broadband provider supports servers (may require a "business" account) it isn't terribly difficult to make one. You can also with a little bit of reading set up a fully dynamic company website hosted from your back office.
If say you're in a dorm in college with a publicly accessible IP you can run your blog or gaming news site or whatever from the Mac mini under your desk without needing to be a security guru or really needing all that much experience.
Enterprise communication and RSS make a really excellent combination. Feed readers are a dime a dozen or very simple to build. RSS hosting requires little more than a web server which means damn near everyone with a workgroup server or even a retasked don't-touch-this-workstation-or-else server can get in on the act.
Many readers support SSL and HTTP authentication which means connections to private feeds is relatively secure, moreso than most organizations e-mail systems. Having a small RSS reader running in the background is also a lot more efficient most of the time than the "information managers" (Outlook, Entourage, Evolution) most people run. This is doubly true when you've got memos and announcements coming to the same mailbox as your other intraoffice mail. If you're out and about using your cell phone or dial-up service you don't necessarily want to download upteen thousand mail messages, just the latest news items.
If you look around I think you'll be surprised at the amount of metadata that exists in all sorts of files. I can have Word automagically add my name, company, manager, e-mail, and all sorts of other little tidbits to a document in a meta data section. I just have to add this data once and from then on Word, Excel, and Powerpoint will do the tedious work for me. This metadata can be searched and indexed by Spotlight.
So for instance, I'm working on a project with a few other people in my department and we're sharing files and sending e-mails all over the place. I can bring up Spotlight and type "John Smith" and it will find me e-mails send to and from John, his Address Book contact, calendar events with John associated or mentioned, and importantly Word and Excel documents he has sent to me. All of this data was parsed from information that was already lying around various locations on my computer. Word and Excel were already adding John's name to files that he authored.
Your premise is that metadata sucks because no one bothers adding it to files. It isn't the end users that will be or need to be doing the work, the onus for that is on the developers. Since Spotlight will now be a major feature in the OS, developers will begin going out of their way to not only use Spotlight in their own applications but make sure their file formats are extremely Spotlight friendly. It's only going to get more prevalent from this point on because it is going to become a lot more useful to the person on the street.
But Apple is delivering Spotlight in two weeks to regular Joes. Spotlight's back end has also been around since MacOS 8.5 when Sherlock was introduced. Content indexing has existed in various forms since then and I happen to use it all of the time. Spotlight basically takes content and metadata indexing and puts a swank front end on it.
Dashboard as was mentioned by someone else is a reimplementation of Desktop Accessories. Arlo Rose is a bright guy and Konfabulator is a nice program but it was hardly all that original. There's a significant difference between the two in terms of resource usage and implementation. Dashboard widgets are only invoked when Dashboard is brought to the fore. Konfabulator widgets run constantly because Konfabulator runs constantly.
I don't quite see how including missing functionality in an upgrade is a bad thing. Jaguar and previous versions of OSX did not have an easy way to compress files. Panther added the ability to compress files and folders as Zip files that can be opened on Windows or other Macs. Was Apple trying to kill off Aladdin née Allume's sales of StuffIt? Was Apple ripping off StuffIt by adding a function theretofore missing from the OS? Panther also re-added support for labels in Finder. Uh oh, they must have been ripping off Unsanity's LabelsX. One of the biggest gripes Mac users had about early versions of OSX was the lack of features they had come to enjoy that were in previous versions of the OS, Desktop Accessories and labels being two of these things. Adding these things back into their OS is hardly dumping on third party developers.
No, it isn't. In fact GCC isn't even installed by default. Even when you install Xcode you need to enable distcc inside of Xcode's configuration.
This is obvious.
While XML isn't the answer to every question it does have its uses. Just transmitting straight database information like I used in my example might not be the best use of it. If I started to add meta data to each data element though, XML would be much more well suited.
With your Perl code putting a CSV into a hash you might have to change that around for every program you write. So eventually you will build a generic module that handles CSV files so anyone can throw a CSV into a function's arguments and have a nice and neat hash of all of the data. Now you're put as much work as an XML parser that does pretty much the same thing. Again it comes down to your uses, if you're writing apps to import Access databases into MySQL using CSV table dumps your CSV module would likely be quicker than XML::Parser. If your database app allowed for direct XML exporting which maintained all sorts of important meta data then the XML::Parser route would probably be more effective.
I use XML sparingly because it tends to be a little heavy on the processor doing the parsing. I wouldn't swear by it but I'm not going to disavow it because parsing can take a few extra cycles here and there. There's plenty of times where providing some XML support can really save some cross-platform/environment/language hassle, especially for lousy programmers.
Except the XML file tells the parser where its own definition is. Each of the XML files inside of an OO.o package tell you how to figure out what they are. A generic XML parser can at least find the URI to the file's type definition. Knowing what all of the elements represent makes it possible to figure out what to do with all of the data and to tell if the XML file itself is even made properly. A comma delimited file doesn't tell the parser what each of the fields means, neither does a straight binary data file. At best they annouce their filetype in a header.
Joe,Slashdort,,124 Anyplace,CO: Mom,Los Angeles,CA,34,118
versus
<slashdork>
<firstname>Joe</firstname>
<lastname>Slashdort</lastname>
<nickname></nickname>
<address1>123 Anyplace</address1>
<address2>CO: Mom</address2>
<city>Los Angeles</city>
<state>CA</state>
<latitude>34</latitude>
<longitude>118</longitude>
</slashdork>
Twenty years from now the dude writing a parser for the XML can do so largely without a manual or any other documentation. The latitude and longitude and second address line aren't necessarily obvious in the CSV example. If the data was much more ambiguous, like the dump from a database, the XML would be even more self explanitory. While XML might not be the absolute best or most space efficient answer to self describing data it does a pretty good job, much better than a lot of the data encoding schemes it is starting to replace.
I think that Enterprise was doomed from the start for a number of reasons. Star Trek fans footing the bill for another season will be dumping their money in the trash. While the third and fouth seasons have been better than the first two they are by no means on par with the other series in the franchise. I see this being the fault of two groups, Rick Berman and Brannon Braga and Paramount/UPN.
Berman and Braga have taken really awesome concept and driven it into the ground. Enterprise was going where no Trek series had been (pun intended), the beginning of the Federation. Enterprise could have been an awesome pre-history of the other series. Instead Berman and Braga have paid little attention to established pre-history of the later series. This wouldn't be so bad if their story arcs didn't require a reset button at the end of the series just for later ones to exist in the mythos. While sticking to Trek canon isn't required for the show to be successful it does help keep existing Trek fans interested. People who grew up watching Romulans and Klingons want to see Romulans and Klingons. New species are nice and all but there's a lot of untold history that Enterprise was well positioned to handle.
UPN is doing just as much damage as stupid story arcs and bad writing. UPN is a joke of a network and Enterprise on its lineup is the punchline. Unless Trek fans are also fans of stupid one-off sitcoms there's little on UPN's lineup to attract them or affiliates. UPN also decided to shuffle Enterprise all over their schedule making it difficult for casual viewers to keep up with the dumb story arcs.
While I appreciate people wanting to see more of their favorite franchise I think the world needs a break from Star Trek for a while. Berman at the helm of the franchise is driving it into the ground. TNG was a successful revival because there was a desire for more Trek that was more than a rehash of the original series. There's been continuous Star Trek on television since the late 80s. A break is required, as is some new blood at the helm.
You hit upon two very salient truths about the Huygens probe. The main imaging instrument on Huygens is the Descent Imager Spectral Radiometer (DISR). It contains only a single CCD. The various lens assemblies are connected to the main CCD by a series of optical fibers, each shines light on a different region of the CCD. The CCD is 512x520 pixels though about half of that is reserved as a storage section, the left over 256x520 pixel area is used for the imagers and spectrometers. The visual imagers of the DISR intrument are the High Resolution Imager (HRI), Medium Resolution Imager (MRI) and the Side Looking Imager (SLI). There's also two spectrometers, the Upward-Looking Visual Spectrometers (ULVS) and the Downward-Looking Visual Spectrometer (DLVS). Finally there's space reserves for the Solar Aureole camera which is used to measure sunlight streaming through the atmosphere to determine the size of particles within it.
The HRI is 160x254, the MRI is 176x254, and the SLI is 128x254 pixels each. Larger images have to be assembled as mosaics and even these aren't going to be large enough to compete with the megapixel images from the MERs. The Pancams and Navcams on the MERs are 1024x1024 each and have essentially a full range of motion so really nice panoramas are easy to create. The DISR is fixed on Huygen's chassis.
Bandwidth is also a tremendous issue with Huygens. The Huygens probe only hasd a 4800bps datalink to Cassini and has to transmit all of its images within two and a half hours. Even with its limited data rate Huygens was able to transmit 350 images back from Titan which is rather impressive.
So it is a combination of geometry, bandwidth, and limited technology. Also remember that despite these images being relatively stark in comparison to MER images they contain tons of very valuable information. When researches have had more time to process Huygens images they will get prettier. Until then they're going to remain relatively bland to laymans' eyes but terribly exciting to scientists.
While it is disappointing to see ISPs dropping Usenet support doing so will hardly kill it. The awesome part of Usenet is it is a naturally distributed network of systems. It doesn't take much to carry the text-only traffic of Usenet, especially considering the price of processing power and network bandwidth anymore. Binary feeds take quite a bit more but if you want the basics the barrier of entry is relatively low.
While web-based forums have gotten very popular in the past few years they simply do not have the advantages of Usenet groups. A forum is limited by a single server/cluster's capacity in terms of both bandwidth and processing power. An angry admin, hacker, FBI raid, or backhoe can take down even the largest of web forums. It would take a lot of doing to kill a newsgroup. A couple of yahoos with spare Linux boxes could keep a group going without much effort. Forums also fall down when it comes to availability. To access a thread on a forum you need to be connected to the web. A newsgroup's posts can be downloaded once and held onto for as long as you'd like. This is a feature mailing lists also have over web forums, the entire history of the list can be stored in your local mail spool. While a forum is likely to be public accessible the sum of its content is rarely available for anyone to mirror if they have the prerogative.
Programs like Leafnode allow you to create local mirrors of feeds while Usenet-Web can process those spools to make them available to anyone with a web browser. Emoticons and oversized picture signatures are little reason to use web forums in lieu of newsgroups.
...and the Great Satans in the West supporting them. You cannot be taken seriously if you ignore the inconvenient fact that the West, most especially the US, has been supporting those regimes and Islamic theocracies.
In the 1982 during the Iran-Iraq war (1980-88) when Iran appeared to have gained an upper hand in the conflict Reagan and his advisers decided to secretly supply Hussein with military equipment, dual use industrial technology, and even tactical assistance in the form of satellite photos of battlefields. The dual use industrial technology gave Iraq the ability to mass manufacture chemical and biological weapons. The US didn't stop supporting Iraq even after it used chemical weapons against the Iranians.
Even al Qaeda was a US sponsored organization at one point. al Qaeda actually means "the base", the CIA's nickname for the group when they were paying them and the Taliban to battle the Russians during the invasion of Afghanistan. Post-war Afghanistan is still impoverished and nearly as dangerous as pre-war Afghanistan. Pre-war Afghanistan was a dirtball largely because of the Russian invasion and subsequent war that we helped provoke.
Labeling someone as "evil" because they're doing what we taught them and paid them to do is a little ridiculous. A kid blowing himself up on a bus is no more or less evil than a guy dropping thousand pound bombs out of an airplane. They're both doing what they're doing for their own reasons and doing what they believe is the "right" thing to do. Maybe, just maybe, no one ought to be blowing anyone up for any reason. Blowing people up simply entices more people to blow more people up. As Ghandi said, an eye for an eye leaves everyone blind.
I don't think you really understand the need for gravitational assists. Without such assists a probe using current launch technology would not be able to make it anywhere close to the edge of the solar system. We can't just fire these things off in a roughly straight line. They're launched in what are essentially variations of a Hohmann transfer orbit. Instead of moving from one circular orbit to another though, you're transfering from a circular orbit to a hyperbolic one. The same concept applies, energy needs to be added to increase the altitude of the orbit which is where the gravity assist comes in.
Shooting a probe off perpendicular to the mean plane of the solar system would be shooting yourself in the foot. You'd be entirely missing out on the enormous gravitation boost Jupiter and/or Saturn offer. You would need überamounts of extraordinarily energetic fuel to make the trip without using the gas giants. Even then the rocket (as mentioned) would need to be about 98% fuel. We can at best build ones that are 90% fuel.
There's also a shitload of vaccuum in the planetary plane just as there is perpendicular to the plane. There would really be little use in avoiding the planetary plane as you suggest. Besides the boost the probe would get from the gas giant(s) it could be put on a course where some secondary science could be conducted on the way to the edge of the solar system. You wouldn't need extra equipment to study doppler shifts as the probe entered say Jupiter's gravity well during an assist or watch it's radio signal as it was occulted by the planet for frequency varations to infer magnetic fields or ionized particles in the atmosphere.
Oh snap! That took me five minutes dude, you've got to come up with a better argument.
These companies were given their spectrum licenses at firesale prices by the FCC. They have since then been allowed to conglomorate into even larger entities by the FCC. They have also lobbied and litigated to kill any and all threats to their control of their markets.
This is protectionism under the guise of capitalism. Here's both of our kick in the nuts.
Many providers, including T-Mobile and AT&T/Cingular in the US, have moved to SMS over GPRS. Most if not all handsets sold now also support SMS over GPRS which has much more bandwidth available than traditional GSM messaging channels. SMS messages are sent via GPRS like any other bit of data. There's little reason for GSM carriers to charge obscene amounts of money for messaging (other than to rip us off).
Charles Parnot at Stanford made a grid for his personal project. He's got more than a hundred people donating some spare cycles to his grid which is pretty impressive for a fairly small project. Daniel Côté started an awesome project to get Xgrid working on non-OSX Unix systems. With a bit of work his Xgrid Agent program could be really robust and reliable enough for getting real work done. Like you I'd like to see this technology proliferate so maybe we can start seeing open grids pop up in various computer user communities.
Someone already figured out how to do this with Xgrid.
Now all we need in the US is services that are condusive to actually serving content instead of simply consuming it. I suppose the ridiculous asymmetry of broadband services in the US ought to be expected of a country raised by televisions.
I've got a broadband connection. It's 3Mbps downstream and 256Kbps upstream. While it is decidedly quicker than a 56k dial-up connection in either direction it is definitely not designed let me serve content at reasonable speeds. Many ports are also blocked at the cable company's head end so I can't use standard service ports (80, 21, etc). I also have to pay an obscene amount of money if I want a static IP address that I can point a DNS entry to.
Some people do have residential broadband that offers saner upstream bandwidth, no port blocking, and free static IPs. Unfortunately this is not the norm here. Most of us either have to pay for hosting or a "business" service package from our broadband provider. In either case we're paying a lot of money for services that ought to be provided for all broadband users.
You might need to run make configure to set up header and file locations then.
You did install the Developer Tools right? DT installs the GNU C libraries of which endian.h is a part.
What is not needed is hydrogen powered cars but a viable means to generate the power to make hydrogen (energy) readily available for everyone. Molecular hydrogen as demanded by the "hydrogen economy" is very simply a medium with which energy is physically transfered. Gasoline is as much of an energy transport medium as hydrogen.
Both hydrogen and gasoline can be used to generate electrical energy, gasoline and its hydrocarbon cousins however release the carbon part of their hydrocarbons. Hydrocarbons can be an excellent source of hydrogen and can be manufactured and transported relatively cheaply. If they're originally generated via a renewable process you can get a carbon-neutral system. The carbon pumped into the air will be used in photosynthesis by plants used to make more hydrocarbon fuels.
Two billion dollars is a lot of money but it would get used in a lot of inappropriate ways if it went towards deconstructing our oil economy. There is enormous vested interest in the status quo of the oil industry. Those interests far exceed a relatively paltry two billion. Any research funded by said two billion would be met with ten times that amount in sabotage (FUD, physical sabotage, political chicanery) by groups threatened by such research. The money being spent to further the sum of human knowlege would end up in the long run being more effectively used. The research in robotics and telepresense could be used immediately on the ground to help normal every day people.
Weening our society and economy off oil is going to take a lot more than two billion dollars. It will require not only research dollars but also a great deal of social and cultural change. Meaningful conservation efforts would go a long way to make our current energy generation and transport system more effective and longer lasting without needing fancy new technologies. Fancy new technologies would then serve to make those meaningful conservation efforts even more meaningful and effective.
For college kids scraping pennies together there's Apple credit accounts that work like a credit card. With these you can make minimum payments for as long as you don't have money and then pay down the principal when you do have it. Minimum payments on a ~$1500 computer aren't going to be too ridiculous, if you can save up $300 over a couiple of months you can easily make such payments. These coupled with educational discounts ends up being a decent enough deal if you can't afford a system outright. For high school students there are fewer options they can actualize independently but that's why it pays to keep a good relationship with your parents/family.
Even at $600 an "iSlab" is not a very profitable idea for Apple. The target audience for such a system would be...cheapskates, people wanted an "intro" Mac, and offices wanting a bunch of one-off desktops.
For offices an all-in-one iMac or eMac would be a better deal than a headless system. AppleCare covers the machine and monitor for up to three years, that's essentially an upgrade cycle for many offices. With an AIO systems there's less for the IT people to worry about. There's far fewer connectors to get stressed from squeezing the system into odd places and the systems are much easier in general to position in the first place. The G5 iMacs take up almost no space at all and don't weigh very much and will easily last in an office environment for a number of years.
For the people wanting an intro Mac a headless system is going to be far more hassle than its worth at a $600 price point. The eMac's monitor easily rivals the "free" 17" CRTs that come with some Dell and HP systems after rebates and such. The LCD of the iMacs is really nice and competes well with any third party displays you'd probably hook up to an "iSlab". If you've got a monitor that "you can just use" and want an eMac there's very likely some schools or other such places that would love a free monitor that you can write off on your taxes.
There might be a market for a $600 "iSlab" but it isn't necessarily one worth tackling for Apple. If people buy these and no other Mac software it doesn't entice anyone to make more Mac software. If people want a cheaper system to try out OSX there's a really large used/refurbished Mac market in existance. With a moderate amount of effort you could find a nice Mac for $600 or less that would run OSX nicely (which itself can bre resold to go towards a new Mac).