Princess Ruruna is stressed out. With the king and queen away, she has to manage the Kingdom of Kod's humongous fruit-selling empire. Overseas departments, scads of inventory, conflicting prices, and so many customers! It's all such a confusing mess. But a mysterious book and a helpful fairy promise to solve her organizational problemsâ"with the practical magic of databases.
In The Manga Guide to Databases, Tico the fairy teaches the Princess how to simplify her data management. We follow along as they design a relational database, understand the entity-relationship model, perform basic database operations, and delve into more advanced topics. Once the Princess is familiar with transactions and basic SQL statements, she can keep her data timely and accurate for the entire kingdom. Finally, Tico explains ways to make the database more efficient and secure, and they discuss methods for concurrency and replication.
There have been times where I've ignored my homework in order to research other things: ham radio projects, RSA public key cryptography, etc. So for me the internet has been a distraction, albeit an instructive one.
I should add, though, that Altera's Nios and Xilinx's Microblaze do represent a form of vendor lock-in. They're heavily marketed and the results are starting to show, I guess.
Yeah, all FPGA toolchains are pretty much the same. I went through a course at Berkeley that used Xilinx, and at work we're a Xilinx shop, but I've tried out Altera and Lattice toolsets and found them to be very similar. The big differences are when you get down to nitty-gritty timing and area constraints, but those are tied to low-level architecture, and in a intro class that's not something you need to worry about.
All FPGA manufacturers offer a free version of their tools, so the price argument is moot.
And, just to threadjack my own post, I just-as-quickly forgot about emacs, and allowed myself to be beat about the head and shoulders by vi until now, to the point that I won't go anywhere near emacs.;-)
vi, preferred by masochists everywhere. (Yeah, me too.)
Last news I read on the subject was that there had been a shortage a couple of years ago, mainly due to rapidly increasing US popularity, but then agave farmers overplanted and now there's an oversupply (so you're seeing things like agave syrup at Whole Paycheck) and farmers are looking at tearing it out and going to other crops.
That's interesting. So far though, tequila prices have remained the same, unfortunately. I've been buying more scotch lately anyway:).
By law, something called "tequila" has to be 51% blue agave. The remainder can be made up of cheap reducing sugar, with the practical result of producing hangovers.
The good stuff is the 100% blue agave tequila. Unfortunately, since there's been an agave shortage in Mexico, prices have gone up in recent years. You have to pay $20 and up for (750mL) 100% blue agave these days.
Thanks for the offer. However, after I posted I realized that I have an old friend who's living in the Netherlands (working in architecture, no less). I've asked him to get one for me.
Well, most of these companies take a reference design from the networking chipset manufacturer (Atheros, Marvell, Ralink, etc.), put new plastic around it, and rebrand the drivers. High volume, low margins: not much effort put into support.
Dual graphics GPUs are not an Apple innovation. This is something that is being pushed by Nvidia and AMD - a discrete GPU that can be turned on or off as needed.
I feel kind of silly for giving this some serious consideration, but here goes:
The DDR memory interface uses stub-series termination logic (SSTL). This means that besides Vdd, there's a termination voltage, Vtt = Vdd/2. So in order to keep things symmetrical around Vtt, using diodes on the output lines doesn't cut it -- you should also drop Vtt. If you put red LEDs (2V) on the output lines, and 1V zeners on Vtt, that would be about right.
I'll add a deal-killer, though. (Besides the fact that capacitances on the diodes are probably enough to swamp >1GHz signals.) The data lines on RAM are bi-directional. If you put diodes on them, then the processor won't be able to send data to the RAM. This is essentially the difficulty with mixing voltages in bidirectional chip-to-chip communications, and it's only compounded by the high frequencies involved. For less demanding applications you can use a bus switch or transceiver. For high frequencies / high performance you'd better avoid kludges.
Slashdot Mirror
The description made me laugh out loud.
There have been times where I've ignored my homework in order to research other things: ham radio projects, RSA public key cryptography, etc. So for me the internet has been a distraction, albeit an instructive one.
I should add, though, that Altera's Nios and Xilinx's Microblaze do represent a form of vendor lock-in. They're heavily marketed and the results are starting to show, I guess.
Yeah, all FPGA toolchains are pretty much the same. I went through a course at Berkeley that used Xilinx, and at work we're a Xilinx shop, but I've tried out Altera and Lattice toolsets and found them to be very similar. The big differences are when you get down to nitty-gritty timing and area constraints, but those are tied to low-level architecture, and in a intro class that's not something you need to worry about.
All FPGA manufacturers offer a free version of their tools, so the price argument is moot.
How about this?
http://www.nrc.gov/about-nrc/radiation.html
You could look at a nuclear engineering text if you wanted to know more.
http://www.nuc.berkeley.edu/NE-39
http://www.nuc.berkeley.edu/NE-101
http://www.nuc.berkeley.edu/NE-124
I've only ever interacted with George Smoot; he seems like a really nice guy to me.
As for Stephen Chu, I saw him going to his car once, underneath building 50. Can't draw much of a conclusion from that experience.
That Hi-Chew stuff is seriously addictive though.
And, just to threadjack my own post, I just-as-quickly forgot about emacs, and allowed myself to be beat about the head and shoulders by vi until now, to the point that I won't go anywhere near emacs. ;-)
vi, preferred by masochists everywhere. (Yeah, me too.)
That's a pretty cool word.
Last news I read on the subject was that there had been a shortage a couple of years ago, mainly due to rapidly increasing US popularity, but then agave farmers overplanted and now there's an oversupply (so you're seeing things like agave syrup at Whole Paycheck) and farmers are looking at tearing it out and going to other crops.
That's interesting. So far though, tequila prices have remained the same, unfortunately. I've been buying more scotch lately anyway :).
Giancoli isn't very good. Meh.
World of Goo isn't what I'd call a minigame. It's a puzzle game, and I thought the price was right.
By law, something called "tequila" has to be 51% blue agave. The remainder can be made up of cheap reducing sugar, with the practical result of producing hangovers.
The good stuff is the 100% blue agave tequila. Unfortunately, since there's been an agave shortage in Mexico, prices have gone up in recent years. You have to pay $20 and up for (750mL) 100% blue agave these days.
Thanks for the offer. However, after I posted I realized that I have an old friend who's living in the Netherlands (working in architecture, no less). I've asked him to get one for me.
Well, most of these companies take a reference design from the networking chipset manufacturer (Atheros, Marvell, Ralink, etc.), put new plastic around it, and rebrand the drivers. High volume, low margins: not much effort put into support.
I'm amused to note that the author list stretches over three pages, which I gather is common for this sort of paper.
The really interesting thing about this paper is that not everybody in the collaboration signed off on it -- usually the author list is even longer.
Agreed, I'm not buying BD until the DRM is de-fanged.
I'd to love buy a commemorative edition, but the Royal Dutch Mint appears not to ship outside the EU.
That sounds right. According to Wikipedia, the natural abundance of Si28 is 92%.
Googling got me an interesting paper:
http://www.crystalresearch.com/crt/ab35/1023_a.pdf
Not really related, but pretty pictures of silicon:
http://www.periodictable.com/Elements/014/index.html
We're still on 10.5, for whatever reason.
Dual graphics GPUs are not an Apple innovation. This is something that is being pushed by Nvidia and AMD - a discrete GPU that can be turned on or off as needed.
You've got to set the right breakpoints, duh.
I feel kind of silly for giving this some serious consideration, but here goes:
The DDR memory interface uses stub-series termination logic (SSTL). This means that besides Vdd, there's a termination voltage, Vtt = Vdd/2. So in order to keep things symmetrical around Vtt, using diodes on the output lines doesn't cut it -- you should also drop Vtt. If you put red LEDs (2V) on the output lines, and 1V zeners on Vtt, that would be about right.
I'll add a deal-killer, though. (Besides the fact that capacitances on the diodes are probably enough to swamp >1GHz signals.) The data lines on RAM are bi-directional. If you put diodes on them, then the processor won't be able to send data to the RAM. This is essentially the difficulty with mixing voltages in bidirectional chip-to-chip communications, and it's only compounded by the high frequencies involved. For less demanding applications you can use a bus switch or transceiver. For high frequencies / high performance you'd better avoid kludges.
I doubt the CPU core and memory controller run at the same voltage. You're not going to run a 45nm processor on 1.5 volts.
I seem to recall this issue coming up with the AthlonX2/Opteron, when they switched a stepping or a process or something. So this is not new.