...and yet it isn't - certainly I haven't seen anything much that has the potential for providing rich information based on data relations using SQL. I'm not a massive database guy, but it seems to me that generating graph data structures and using them in the generic fashion we do is a pain in the neck using SQL (compared to RDF/RDQL).
That said, all RDF queries can be accomplished in SQL - RDQL maps onto SQL. 3Store actually uses MySQL as a backend.
RDF. It's ontology-agnostic, although if you have a triple store that can handle ontologies they can be used to provide you with more interesting results.
hm. I haven't spoken on slashdot for a while, and I appear to have forgotten simple things such as correct formatting. My apologies for the lack of paragraphs.
I'm a developer on the project, and the commentry is appreciated. I'm frankly somewhat surprised at the level of hype the project has generated at this stage of the game. While I find the results that you can get with the current implementation very interesting, there's a lot of work yet to do before it's truly revolutionary.
With that said, I'd ask you to consider the possiblities offered by expansions to the original idea. We're working on converting the system as a whole to a web service, allowing any kind of client to access the information in a sensical manner, and linking mspaces together. This, for example, would allow you to hook together information on localality and, say, restaurants. You might be looking for restaurants within a certain area of where you are now. Once you get that information, you could select information on those restaurants in a powerful manner - you could select restaurants that offer vegetarian meals and meals containing low carbohydrates and without gluten, for prices under £10.00. With those results, you may decide to further filter it down by selecting only italian or american style meals.
Largely, the power behind the existing concept comes with the ability to construct your own dynamic hierarchies. I posted further down about it while forgetting to log in - with a film database, for example, you might find out about russian actors who acted in american films during the mccarthy era. This is the sort of obscure information that people are unlikely to have written extensively about, so collating the information would be difficult. With the system we have, that sort of information is contained within the relationships for you to discover for yourself. For me, a lot of the potential of this idea is contained in the fact that the google is awesome for discovering information on reasonably common things. For more obscure information, what if nobody has written a page about what you want to know? The information is out there, but has never been collated properly. mSpaces can give you that sort of information, without having to explicitly generate it.
But...lets face it, if Apple wasn't so anti-competitive there would be lots of nice cheap apples floating around...but there aren't. It has to therefore be valid to compare the lower end pcs as apple has no alternative. I'd love to be able to buy a nice cheap mac, but since they don't exist I buy PCs.
And comparing apple to premium PC brands it has to be said that for equivalent performance pcs still come out cheaper.
Please tell me you're joking...
The XEON is intel's server chip with vast quantites of incredibly expensive fast on board cache. Try comparing it against a dual 1600MP athlon and you'd probably get a very fair comparison, particularly given the athlon's support for decent memory.
Also, as I understand it the Fire is a pretty expensive opengl card. You'd be much better off using a Radeon 8500 as a basis of comparison, particularly given that the Geforce 4 in the new mac is the slow MX version.
Hopefully that lot would equate reasonably to the apple.
But to be honest I'm not all that thrilled. Some of the numbers are pretty nice, but 115fps in QIII? My single processor athlon gets to around 160 with the same settings, and that cost me under 1000. Don't get me wrong, I'm not a mac hater, but I fail to see why apple is showing such a meagre result as one of the most impressive benchmarks.
Also, it's just too damn expensive. $3000 is way too much, particularly when you consider that the graphics card is in fact the gf4 MX, rather than the full version. If it was the full version I would start saving right now;) Apple has been relatively price competitive in recent times against x86 based systems, but consider what SMP athlon systems cost. I can guarantee that an athlon system costing $3000 would absolutely mash the dual G$ 1ghz.
Unfortunately, with a dual athlon I couldnt experience the joy of OS X....;)
Slashdot Mirror
...and yet it isn't - certainly I haven't seen anything much that has the potential for providing rich information based on data relations using SQL. I'm not a massive database guy, but it seems to me that generating graph data structures and using them in the generic fashion we do is a pain in the neck using SQL (compared to RDF/RDQL). That said, all RDF queries can be accomplished in SQL - RDQL maps onto SQL. 3Store actually uses MySQL as a backend.
RDF. It's ontology-agnostic, although if you have a triple store that can handle ontologies they can be used to provide you with more interesting results.
hm. I haven't spoken on slashdot for a while, and I appear to have forgotten simple things such as correct formatting. My apologies for the lack of paragraphs.
I'm a developer on the project, and the commentry is appreciated. I'm frankly somewhat surprised at the level of hype the project has generated at this stage of the game. While I find the results that you can get with the current implementation very interesting, there's a lot of work yet to do before it's truly revolutionary. With that said, I'd ask you to consider the possiblities offered by expansions to the original idea. We're working on converting the system as a whole to a web service, allowing any kind of client to access the information in a sensical manner, and linking mspaces together. This, for example, would allow you to hook together information on localality and, say, restaurants. You might be looking for restaurants within a certain area of where you are now. Once you get that information, you could select information on those restaurants in a powerful manner - you could select restaurants that offer vegetarian meals and meals containing low carbohydrates and without gluten, for prices under £10.00. With those results, you may decide to further filter it down by selecting only italian or american style meals. Largely, the power behind the existing concept comes with the ability to construct your own dynamic hierarchies. I posted further down about it while forgetting to log in - with a film database, for example, you might find out about russian actors who acted in american films during the mccarthy era. This is the sort of obscure information that people are unlikely to have written extensively about, so collating the information would be difficult. With the system we have, that sort of information is contained within the relationships for you to discover for yourself. For me, a lot of the potential of this idea is contained in the fact that the google is awesome for discovering information on reasonably common things. For more obscure information, what if nobody has written a page about what you want to know? The information is out there, but has never been collated properly. mSpaces can give you that sort of information, without having to explicitly generate it.
But...lets face it, if Apple wasn't so anti-competitive there would be lots of nice cheap apples floating around...but there aren't. It has to therefore be valid to compare the lower end pcs as apple has no alternative. I'd love to be able to buy a nice cheap mac, but since they don't exist I buy PCs.
And comparing apple to premium PC brands it has to be said that for equivalent performance pcs still come out cheaper.
Please tell me you're joking... The XEON is intel's server chip with vast quantites of incredibly expensive fast on board cache. Try comparing it against a dual 1600MP athlon and you'd probably get a very fair comparison, particularly given the athlon's support for decent memory. Also, as I understand it the Fire is a pretty expensive opengl card. You'd be much better off using a Radeon 8500 as a basis of comparison, particularly given that the Geforce 4 in the new mac is the slow MX version. Hopefully that lot would equate reasonably to the apple.
But to be honest I'm not all that thrilled. Some of the numbers are pretty nice, but 115fps in QIII? My single processor athlon gets to around 160 with the same settings, and that cost me under 1000. Don't get me wrong, I'm not a mac hater, but I fail to see why apple is showing such a meagre result as one of the most impressive benchmarks. Also, it's just too damn expensive. $3000 is way too much, particularly when you consider that the graphics card is in fact the gf4 MX, rather than the full version. If it was the full version I would start saving right now ;) Apple has been relatively price competitive in recent times against x86 based systems, but consider what SMP athlon systems cost. I can guarantee that an athlon system costing $3000 would absolutely mash the dual G$ 1ghz.
Unfortunately, with a dual athlon I couldnt experience the joy of OS X.... ;)