really. They need to sit down computer noobs in front of their GUI and analyze what users find intuitive and what nonintuitive.
What a programmer assumes is intuitive will often not be the case for non-programmers. And not because non-programmers are stupid! Because they aren't.
It appears Gnome developers are making UI design decisions without consulting the non-developer end users first. Big mistake.
You mean like the fact that for nearly 10 years, apple resisted proportional sized sliders (so you dont have any idea how large the window view you have is in relation to the rest of the document)?
Yeah, some wisdom.:-P
(Generally I consider apple's design guides from the 80's-90's very good and applicable to any UI design, but the non proportional sliders thing was simply retarded.)
the main problem with rdram is that it is EXPENSIVE to produce. doesnt matter how technologically superior your product is if it cant be produced cheaply enough to crack the market.
sort of like how AMD is killing off the Itanium. not because amd64 is better technologically, but because it doesn't cost $2000 for the lowest cost entry level cpu.
rdram might allow for simpler motherboard designs, but the memory is still extremely expensive to produce.
fact is, regardless of whatever technological advantages rdram might claim to have, as long as it remains 2x more expensive than ddr/ddr2, it is never going to compete. economy of scale of ddr/ddr2 will continue to leave rdram in the dust.
Many hospitals use Sparcs running Solaris, HP-PA boxes running HP/UX, and IBMs running AIX. Linux is just another Unix to them, something they are already familiar with.
Large university hospitals (UW, Merced, etc.) tend to have very very good clueful IT folk -- most of these sites have been using computers since the 1960s. (And yes there's the odd exception to the rule, but in general large hospitals are more clued in than most other sectors when it comes to IT.)
Nobody masters a language anymore. You master _techniques_. Languages are just syntaxes to implement those techniques.
To put your analogy to use, there's really only so many ways (techniques) to build a cabinet. The different programming languages would be the different tools you use to apply those techniques to build a cabinet.
The nice thing about learning different languages is it often reveals to you new ways of doing things, which you can then apply to languages you already know.
actually having worked in the healthcare field I can say from experience that healthcare uses M-O for the past 10-20 years and 9-track before that.
anyone using dvd-ram is a very very very recent thing. i cant imagine cash strapped healthcare sites dumping their huge M-O investments for brand spanking new DVD-RAM.
hm, all their other papers didn't make any mention of it. shrug.
still, doesnt matter if your metallized layer lasts 20 trillion years if your dye layer degrades, since its the dye layer which holds your data. phthalocyanine (organic dye) won't last forever.
AFAICT, reading all the available literature from Mitsui on their gold MAM-A discs, the reflective metallized layer _is not actually gold_. It's aluminum.
Mitsui is claiming their _special dye_ is what makes their MAM-A discs last so long, and the dye is what gives their discs their gold color. Not the metallized layer.
And really when you think about it, it doesnt matter how long-lived the reflective layer is, if your dye deteriorates. Since you're recording your data onto the dye layer -- not the reflective layer.
I'd like to know if there are any good compactflash players. After seeing so many portable drives die in all sorts of devices (cameras, laptops, PDAs, mp3 players), I'd prefer solid state storage. And large CF cards can be had relatively cheap (much cheaper than SD/MMC).
I'm not interested in players with only a fixed amount of storage (eg 256m or 512m players with no removeable storage).
It crashes on even the most simplistic applications like xterm.
I guess part of the reason it's uselessly unstable is that it hasn't been updated since 1997. It speaks an ancient dialect of X11 and barfs on anything more recent.
the amd64 performance gains come mainly from the on-chip memory controller and the increased memory bandwidth and 64-bit wide operations. the gains from having extra registers is negligible.
While huge sparc's register sets might have mattered back in the late 80's, it's irrelevant today. Enormous l1 and l2 caches have nullified any advantage huge register sets might have had.
What good does a couple hundred physical registers get you when you can have hundreds of thousands of effective registers with l1 / l2 caches?
Answer: nothing.
On a per dollar basis, x86 blows away sparc by an order of magnitude or more. Doesnt matter how "cool" your architecture is if it isnt cost effective.
you dont need jumbo frames to fill a GbE pipe. the hardware does all the hard work for you, as I stated before. with a 64/66 system you can _easily_ fill a GbE pipe. it is _trivial_.
btw, something is _very_ wrong with both the memory bus and the pci bus implementations on serverworks chipsets. when I did benchmarks with smp p3 systems on an _expensive_ motherboard based on an _expensive_ high end serverworks chipset, i was very disappointed with both pci throughput and memory throughput.
just to make sure I wasn't smoking crack, I talked to someone who designs high end x86 motherboards and he confirmed that there are performance problems with the serverworks chipsets.
The tg3 chipset runs rather hot, the e1000 is tiny and runs cool. I havent noticed a performance difference between either, and both chipsets run fine regardless of whatever PC I put them in.
Motherboards with embedded GbE typically use e1000 (if theyre good), or realtek (if theyre cheap).
Cabling: Hand crimped cat5e. Works fine. One interesting note about GbE, you no longer have to worry about crossover cables -- the GbE spec requires that devices autodetect crossover. You can make all your GbE cables "straight through" cables.
Do pay careful attention to following strict T568 wiring code though. You can no longer get away with incorrectly wired cables which just happened to work for 100bt. Since all pairs are now used in GbE, your wiring order must be 100% spec.
Here's some wiring guides: http://www.lanshack.com/make-cat5E.asp h ttp://yoda.uvi.edu/InfoTech/rj45.htm
Don't worry about them. Only very, very expensive systems will be able to take advantage of them.
If you have 32/33 pci, you arent going to get max throughput from GbE anyway. I've managed to get around 90mbyte/sec using ttcp, which is about 750mbit/s.
Because the hardware does all the work for you (hardware checksum, interrupt mitigation, etc). the cpu usage is very low even at that rate. And thanks to polling, the interrupt rate isnt an issue either.
Your bottleneck will be your PCI bus, plain and simple. You arent going to get the full 132mbyte/s from 32/33 pci, period.
Unfortunately 64bit/66mhz PCI motherboards are somewhat expensive and 64/66 cards are 3-4x the cost of 32/33 ones.
is a link to the actual text of the indictment anywhere? without it we won't know exactly what the claims are, and only have his version of the story to go on.
IIRC it's NZ which has the highest lawyers per-capita in the entire world. but in terms of litigation germany has the highest per-capita, iirc it is mainly business lawsuits.
Slashdot Mirror
it's only still images taken from video footage.
there's no video footage to download from the page though.
ok, a really really big one actually.
smallpox is a virus , not a bacteria .
really. They need to sit down computer noobs in front of their GUI and analyze what users find intuitive and what nonintuitive.
What a programmer assumes is intuitive will often not be the case for non-programmers. And not because non-programmers are stupid! Because they aren't.
It appears Gnome developers are making UI design decisions without consulting the non-developer end users first. Big mistake.
You mean like the fact that for nearly 10 years, apple resisted proportional sized sliders (so you dont have any idea how large the window view you have is in relation to the rest of the document)?
:-P
Yeah, some wisdom.
(Generally I consider apple's design guides from the 80's-90's very good and applicable to any UI design, but the non proportional sliders thing was simply retarded.)
let's see you make a better one then.
OTOH tribes / tribes2 network protocol blows.
Mispredicts, jerkiness up the wazoo, even when there's little going on.
quake3's network protocol is much better.
Just some datapoints to counter yours:
We (ISP) had about 5 major data loss disasters with ext3, 3 with XFS and only 1 with reiserfs.
And we use far more reiserfs than ext3 or xfs. So from a reliability standpoint for us, reiserfs is by far more reliable than ext3 or xfs.
http://www.st.northropgrumman.com/media/SiteFiles/ mediagallery/video/MTHEL_m.wmv
the main problem with rdram is that it is EXPENSIVE to produce. doesnt matter how technologically superior your product is if it cant be produced cheaply enough to crack the market.
sort of like how AMD is killing off the Itanium. not because amd64 is better technologically, but because it doesn't cost $2000 for the lowest cost entry level cpu.
rdram might allow for simpler motherboard designs, but the memory is still extremely expensive to produce.
fact is, regardless of whatever technological advantages rdram might claim to have, as long as it remains 2x more expensive than ddr/ddr2, it is never going to compete. economy of scale of ddr/ddr2 will continue to leave rdram in the dust.
This is not so suprising.
Many hospitals use Sparcs running Solaris, HP-PA boxes running HP/UX, and IBMs running AIX. Linux is just another Unix to them, something they are already familiar with.
Large university hospitals (UW, Merced, etc.) tend to have very very good clueful IT folk -- most of these sites have been using computers since the 1960s. (And yes there's the odd exception to the rule, but in general large hospitals are more clued in than most other sectors when it comes to IT.)
Nobody masters a language anymore. You master _techniques_. Languages are just syntaxes to implement those techniques.
To put your analogy to use, there's really only so many ways (techniques) to build a cabinet. The different programming languages would be the different tools you use to apply those techniques to build a cabinet.
The nice thing about learning different languages is it often reveals to you new ways of doing things, which you can then apply to languages you already know.
You'd do well to learn Perl or PHP, btw.
actually having worked in the healthcare field I can say from experience that healthcare uses M-O for the past 10-20 years and 9-track before that.
anyone using dvd-ram is a very very very recent thing. i cant imagine cash strapped healthcare sites dumping their huge M-O investments for brand spanking new DVD-RAM.
hm, all their other papers didn't make any mention of it. shrug.
still, doesnt matter if your metallized layer lasts 20 trillion years if your dye layer degrades, since its the dye layer which holds your data. phthalocyanine (organic dye) won't last forever.
i suspect M-O media will last longer.
AFAICT, reading all the available literature from Mitsui on their gold MAM-A discs, the reflective metallized layer _is not actually gold_. It's aluminum.
Mitsui is claiming their _special dye_ is what makes their MAM-A discs last so long, and the dye is what gives their discs their gold color. Not the metallized layer.
And really when you think about it, it doesnt matter how long-lived the reflective layer is, if your dye deteriorates. Since you're recording your data onto the dye layer -- not the reflective layer.
I'd like to know if there are any good compactflash players. After seeing so many portable drives die in all sorts of devices (cameras, laptops, PDAs, mp3 players), I'd prefer solid state storage. And large CF cards can be had relatively cheap (much cheaper than SD/MMC).
I'm not interested in players with only a fixed amount of storage (eg 256m or 512m players with no removeable storage).
...that it doesnt work.
It crashes on even the most simplistic applications like xterm.
I guess part of the reason it's uselessly unstable is that it hasn't been updated since 1997 . It speaks an ancient dialect of X11 and barfs on anything more recent.
the amd64 performance gains come mainly from the on-chip memory controller and the increased memory bandwidth and 64-bit wide operations. the gains from having extra registers is negligible.
While huge sparc's register sets might have mattered back in the late 80's, it's irrelevant today. Enormous l1 and l2 caches have nullified any advantage huge register sets might have had.
What good does a couple hundred physical registers get you when you can have hundreds of thousands of effective registers with l1 / l2 caches?
Answer: nothing.
On a per dollar basis, x86 blows away sparc by an order of magnitude or more. Doesnt matter how "cool" your architecture is if it isnt cost effective.
typical sparc apologist drivel.
the sparc _needs_ hardware contexts and register windows because it has a zillion registers to save and reload.
the x86 on the other hand has very few registers, so saving and restoring them on context switches is very cheap.
and since x86 cpus are so much faster than sparc now, sparc gets left in the dust.
you think wrong :-)
you dont need jumbo frames to fill a GbE pipe. the hardware does all the hard work for you, as I stated before. with a 64/66 system you can _easily_ fill a GbE pipe. it is _trivial_.
btw, something is _very_ wrong with both the memory bus and the pci bus implementations on serverworks chipsets. when I did benchmarks with smp p3 systems on an _expensive_ motherboard based on an _expensive_ high end serverworks chipset, i was very disappointed with both pci throughput and memory throughput.
just to make sure I wasn't smoking crack, I talked to someone who designs high end x86 motherboards and he confirmed that there are performance problems with the serverworks chipsets.
I've been using GbE for home LAN for about a year now. Here's the hardware I use:
h ttp://yoda.uvi.edu/InfoTech/rj45.htm
Switch:
Linksys Instant Gigabit 10/100/1000 8-port switch
I think I paid ~$200 for this.
Cards:
Intel PRO/1000 MT Desktop Adapter (~$50 ea)
Use the e1000 driver in 2.4.x or 2.6.x.
Netgear GA302T Copper Gigabit Adapter (~50 ea)
Use the tg3 driver in 2.4.x or 2.6.x
The tg3 chipset runs rather hot, the e1000 is tiny and runs cool. I havent noticed a performance difference between either, and both chipsets run fine regardless of whatever PC I put them in.
Motherboards with embedded GbE typically use e1000 (if theyre good), or realtek (if theyre cheap).
Jumbo frames:
See my post on that here.
Cabling:
Hand crimped cat5e. Works fine. One interesting note about GbE, you no longer have to worry about crossover cables -- the GbE spec requires that devices autodetect crossover. You can make all your GbE cables "straight through" cables.
Do pay careful attention to following strict T568 wiring code though. You can no longer get away with incorrectly wired cables which just happened to work for 100bt. Since all pairs are now used in GbE, your wiring order must be 100% spec.
Here's some wiring guides:
http://www.lanshack.com/make-cat5E.asp
Ok, here's the deal with jumbo frames.
Don't worry about them. Only very, very expensive systems will be able to take advantage of them.
If you have 32/33 pci, you arent going to get max throughput from GbE anyway. I've managed to get around 90mbyte/sec using ttcp, which is about 750mbit/s.
Because the hardware does all the work for you (hardware checksum, interrupt mitigation, etc). the cpu usage is very low even at that rate. And thanks to polling, the interrupt rate isnt an issue either.
Your bottleneck will be your PCI bus, plain and simple. You arent going to get the full 132mbyte/s from 32/33 pci, period.
Unfortunately 64bit/66mhz PCI motherboards are somewhat expensive and 64/66 cards are 3-4x the cost of 32/33 ones.
...then it didnt deserve to become the standard.
of course, I don't really believe WMA is any more secure, and it will just be a matter of time before they crack that too.
is a link to the actual text of the indictment anywhere? without it we won't know exactly what the claims are, and only have his version of the story to go on.
IIRC it's NZ which has the highest lawyers per-capita in the entire world. but in terms of litigation germany has the highest per-capita, iirc it is mainly business lawsuits.