Great article. I actually sat and listened to the budget and there was a lot of really positive stuff in there for businesses, it's just a shame so much it was overlooked because of the dust storm kicked up by the cut in the 50p top tax rate band
I wish the dust storm was about the top rate of income tax! At least we could have enjoyed a few minutes of "sane" class hatred.
No, the real topic for media reporters was a tax on pasties (which is not even a real tax, but rather a realignment of some food categories in the tax system).
F# appears to have been started in 2002. Scala was in 2003 and Clojure around 2005 with the first public release in... 2007? This announcement isn't about the release of F#, but its productization.
F#, like OCaml, is also multi-paradigm. Maybe not as much as Common Lisp, but it does have language-level support for OOP and imperative programming. It's not a purely functional language like Haskell.
Doing all the rendering on the server side is certainly an interesting approach - however, I wonder if the speed comparison will be as favourable once that server is doing the rendering for millions of users!
Opera Mini has become the world's most popular mobile browser. Since its worldwide launch in 2006, more than 44 million people have downloaded and used Opera Mini. More than 11.9 million people used Opera Mini in March 2008.
Adding another proxy is adding another "line of attack" though.
Also where are the Opera servers located (and Opera as an entity etc.)? could be that it provides another country the opportunity to subpoena your records from them etc.
Interesting privacy concerns as well perhaps; I can imagine a few people would prefer not to use it because of that.
My privacy has never really worried me yet but this makes me stop and think for a second (ultimately I think I'll be fine with it, but it needs considering)
Whilst this is interesting technology wise and may open impressive possibilities... so what? Without a solid install base why would I spend my time writing impressive software that is tied to this machine? Profit? No install base = no customers. Hubris? No install base = no one to impress.
I'm fully aware this is a chicken and egg situation...
(I say all of the above as a total Amiga fanboy from back in the day!)
I think it's designed for the number crunchers out there. A GPU can crush a CPU for some applications but there is a huge middle ground between a GPU and a CPU.
Why do you think it's designed for the number crunchers? I want one (I'm an ex-Amiga user, and still miss it...), but I certainly don't think a dual core PowerPC + a 400 MIPS co-processor is going to be cost/performance competitive with a suitable x86-64 based server at this stage. In fact, I'd be surprised if it was much faster than my current $600 laptop.
The XCORE ("Xena") co-processor they've added is exciting, but because of the hard-realtime capabilities / low latency and IO lines, not for raw performance.
I was thinking of the people that crunch numbers not the programs. I don't think it's designed for a production environment; rather it's a dev box that lets you explore programmable processors. I think they avoided making an x86 with coprocessor because they want to own the dev environment for such things, but the innovation is all about that coprocessor.
PS: 1x Xorro slot with a card that's 102400 MIPS is fast. An example of number crunching would be real time encoding of super? HD (4x 1080p). Or a custom high bandwidth router etc.
I don't consider it a dev box as much as a box intended to satisfy Amiga hobbyists, some of which may use it for development, but a lot of which just want a non-x86 box running AmigaOS at decent speeds... Keep in mind that the people behind this are long time Amiga supporters - they're not in this for the sake of the XCore.
For the people who are interested in the XCore, it's cheap to buy USB attached dev-board, so this only really matters if you want to play with a system that may have significantly tighter integration (we don't really know yet how tight).
For the Amiga community, an x86 box would've been a non-starter. The people in the most likely buying segment for this box are people that have stuck to ancient "classic" Amigas, or at best upgraded to sub 1GHz single-core PowerPC boards running newer versions of AmigaOS, that they've paid more for than what you'd pay for a quad core 2GHz+ x86 box. Most of them would've been happy with "just" another PowerPC machine, but the XCore adds some extra excitement and may entice a few ex-Amigans like me to have a go for the fun too.
> PS: 1x Xorro slot with a card that's 102400 MIPS is fast. An example of number crunching would be real time encoding of super? HD (4x 1080p). Or a custom high bandwidth router etc.
The 102400 MIPS is an example based on a hypothetical card that doesn't exist, though.
For comparison, XMOS themselves sells a roughly 25000 MIPS experimental board for $1500, so while you can make it fast, it'd also be fairly expensive and it does require a new product.
I am excited about having the XCORE chip there, and I may buy an X1000 (who am I kidding, I will buy one unless the price is absolutely outrageous), but I am an ex-Amiga guy and would buy it in part because I still really love the whole Amiga experience, without any illusion that I couldn't do most of the same things on a PC much cheaper in most instances.
When it comes to the XCORE what excites me about it, rather than performance, is that it will be standard. In other words, (the few) people who write Amiga software can soon reasonably rely on having it available in the computers of most users of their software. People are already thinking up all kinds of bizarre uses for it.
"We believe that with this easy gateway to the world of 'Software Defined Silicon' and a path to massive parallelism, the X1000 will once more make the AmigaOS platform the best choice for truly creative and unique applications. For custom hardware control from robotics to theatrical lighting, for hobbyist creativity, for hardware hacking and for a multitude of applications we haven't even imagined yet, the X1000 is a dream platform - and therein lies another meaning of 'X', the unknown. It is you, not us, who will define the future."
I thought it was a great way to attract hobbyist developers. A quad core 2+ GHz x86 chip can do an insane amount of computation quickly and cheaply and even a mid range GPU takes that to a whole other level. I don't know how many flops you can get per MIPS on XCore, but if you are handcrafting at that level you can do things for less power than a CPU while having more flexibility than a GPU.
It never really occurred to me that people where still using the Amiga OS. From that standpoint it's much more evolutionary than revolutionary, but still cool.
PS: In my mind computer hobbyists and developer often mean just about the same thing. But, I realize developer makes people think of Software Developer not a HW dev.
My take is that it is less for GPU-like number crunching and more for interfacing with real-world devices; perhaps writing relatively simple scripts in ARexx for working with serial, parallel, etc. devices.
I completely agree with this. It then leads to what you charge for your product - if 1% at $10 per month breaks even you'd need $20 at .5% and so on.
But it's not even that simple - because demand and supply kicks in - if you charge more you'll be reducing the size of the market willing to pay the higher price - so you'd need >.5% of the original market to get your original break even point!
"puzzling because Windows 7 is one of the few Microsoft products that actually is somewhat fresh and not just a minor version bump"
Interesting... granted I don't particularly follow Windows development, but my perception was that it is exactly that - a minor version bump over Vista fixing the problems that shouldn't have been there in the first place...
The money an MBA costs would be better spent as seed capital to get something released! Even if your start up fails, it still could be cheaper than what you'd spend on the MBA and you'd have REAL experience.
To be precise Interface Builder doesn't actually generate ANY code at all. You are manipulating in-memory objects and setting their properties. When you save your UI it's effectively just a serialized form of the objects you were playing with.
The key point that I was trying to make is that IB is fundamentally separate from the code that is running in the application. In traditional RAD tools like VB, Delphi and Powerbuilder the interface and application logic are one and the same, you literally have code behind each form and each UI element.
Apologies - I read "in that the code doesn't live directly behind the GUI elements" as an inference that it lived somewhere else. I just wanted to add a clarification in case others read it the same way.
