SUN's biggest problem is that they employ a ton of chip (second only to Intel) and system designers to design their systems.
I see two alternatives:
1. Through ignorance or oversight, you're not counting IBM, the largest computer company in the world. They maintain PowerPC (750 and 970 series and embedded stuff), Power (eServers and stuff) and mainframes (zSeries -- S/390, more than the riced out Unix that Sun wants to call a mainframe). Additionally, IBM designs ASICs not only for themselves, but for other companies.
2. You truly have an insightful point in how many people are employed by Sun-- which points to how much bloat Sun has. If, in fact, you are correct, you are pointing out how many people Sun needs to cut in order to be competive with IBM, to say nothing if Intel who only maintains a few chipset ASICs, the Pentium III/IV and the new cheap one they're advertising on TV.
Come on. Linux does what most of us want already. It's the Pentium II of OS kernels. For icing only:
Get something going with Xfree86, find out what hooks they want for better performance. Talk with NVidia and ATI and find out what common API that's not DirectX9, but is both open in specificaion and foreward thinking enough to expand to meet tomorrow's needs. Settle on what it takes to require GCC3.X as the compiler so distributions can ship with just one GCC version. Get a command line option to prevent printing out all the copyright notices (the corproate equivalent of teenagers' "shoutouts" -- does the warez community still do "greetz"?) and only print out error messages. Howabout a file system (ext2?) that you can fsck while mounted RW (even if the output is "all is well" instead of actually doing fixes)?
And an uptime counter that gets me past 430 days or however many it resets. That's the one thing BSD has over us.
They're showing that companies can actually listen to their customers. Support them and maybe other companies will take notice.
So, you're suggesting that I, as someone who heard about the tricks Intuit was making with Turbo Tax and pre-emptively moved to H&R Block's TaxCut should congratulate and thank Intuit for seeing the light and give them money for realizing they screwed the pooch?
In 2002 (FY 2001), I used Intuit's Turbo Tax on the web. It sucked. It was slow and I couldn't easily look stuff up 6 months later. I was thankful to Intuit for the service, and vowed to buy their standalone software.
In 2003 (FY 2002), I heard about Intuit's treatment of their customers and, having learned my lesson about using web tax preparation, bought TaxCut. Much more responsive, no DRM problems and I was happy with the results.
In 2004 (FY 2003), I will buy TaxCut again. I appreciate the fact that I have never heard of them treating their customers like criminals, and their product this year has bought my loyalty.
In college, I was the ideal Slackware user. I wanted to learn programming, loved to compile my own stuff and felt that Redhat was only useful in removing the user from my goal of learning *nix. I extolled its virtues at every opportunity. I didn't have to hunt down an RPM just to install a new kernel, and I certainly knew how to compile my own kernel. I didn't know of any other serious distributions in 1996.
When I entered the real world and had a job and non-computer hobbies, I still had the need (perceived need as opposed to life or death need) for a *nix machine for my home mail, DNS and web serving. I no longer had the time, however, for fixing dependencies, applying source code patches and hunting down the minor details that I had arduously learned how to hunt down in the previous years. Debian came to my rescue. Dselect may be rough around the intuitive UI edges, and it's not quick on my 486, but it's consistent and only requires occasional answers to keep my machine well patched. It's been years since I've had to compile my own kernel, let alone wanted to.
I fondly remember Slackware as I remember my first girlfriend. It was a good idea at the time, but that time has since past and I have moved on. I am much happier now, but the lessons and memories will stay with me for a long time.
I was of the understanding that to double the size of a cache would
effectivly half its speed. Mainly due to the way caches are designed.
By removing the addressing overhead. Hence the reason we dont have
whopping great level 2 caches, and only a very small level 1 cache. So
a L2 cache the size of main memory could actually cause performance
loss.
To the first order, yes, doubling the cache size will likely double the latency. A skilled SRAM designer
knows a few tricks which may mitigate this some. However, the
CPU designer needs to know what tradeoffs the area expensive and high off current
SRAM array will have against the application advantages of not having to go hundreds of CPU cycles out to main memory. Don't forget
that doubling the access time of an SRAM array to 20 cycles is far less than multiple hundred to get to main memory.
If one were maximizing performance of RC5 brute forcing, Distributed.net found that 256k of cache was plenty. On the other hand,
SETI found they needed 1 Meg of cache (IIRC)-- the RC5 application would suffer from the effect you point out-- an L2 too large to take advantage of.
A database program, on the other hand, would really probably prefer no datacache penalties if the application is truly random access. Cache misses cost cycles too.
Someone gets it. As an Electrical Engineer-in-training, I was always frustrated with people who got these big bad processors and wondered why their improvement was minimal.
They never quite grasped that the biggest bottleneck is between the processor and memory.
Don't get too frustrated with this. There will always be people who don't understand something fundamental to your training. That's why you're trained, to understand these non-obvious fundamentals. Now that you understand a CPU has to be fed data in order to process it, it's obvious, but a PHB wouldn't necessarily come to that conclusion on his own.
My EE instructor always said that they could improve performance by doing one simple thing: make the interconnects on the motherboard between the motherboard and RAM rounded instead of cornered. You could then increase bus speed as you wouldn't have magnetic loss at the corners like you do now.
You fix that, and you can see a SUBSTANTIAL improvement in performance. The only thing that can be done beyond that is to get a Platypus drive (Solid state "Hard Drive" made from Quikdata made from DDR RAM). Then you reduce your access time to your hard drive from milliseconds to nano/microseconds.
Your EE instructor will tell you lots of things that can help performance. For example, making the L2 cache be the size of main memory. Just because it helps performance doesn't make it worth the price. Rounded edges on the PCB are not easy to accomplish and their benefits may not be outweighed by the added price-- even for exceptionally high end servers. Without looking at the math, I would like to toss "10% performance adder, 50% cost adder" out into the air, and say that most people would rather save the dough. Another factor to consider is reliability. Intuition suggests to me that reliability would go up without sharp edges, but intuition also tells me that modelling board coupling on a 4 layer board would be a real pain in the ass, to say nothing of a server class 6 or 8 (or higher) layer board if you have to model curved structures. You might not find an easy way to capitalize on your wonderful curved wire performance. Not only do you have to worry about your slowest path, but your quickest one can't arrive so quickly that the other chip can't sample the previous output.
Take care in your classes when you use the word "only". Taking advantage of our wonderful next generation 64 bit processors and multiple gigs of RAM, we could conceivably copy the contents of the hard drive to main memory (especially if we are only concerned with 1-4 gigs of data in a low cost solution). Here, we get the enhanced bandwidth of main memory instead of having to kludge through the southbridge, PCI controller, IDE/SCSI to RAM interface and back.
There are many things that improve system performance-- and the system is the only thing that matters-- rounded wires and SSDs (solid state drives) are only the beginning. Depending on the application, a larger L3 cache may make more difference, or a wider faster CPU to CPU interface, or a pair of PCI controllers hanging off the southbridge for twice the bandwidth, or integrating the northbridge onto the CPU, or...
The best engineering advice I can give you is that the answer is always, "It depends". You'll spend the next 5-30 years of your life learning how to answer the followon question "Depends on what?". Almost everything has advantages and disadvantages and there are few absolutes.
The "Someone gets it" and "They never quite grasped" attitude may get you in trouble. Being proactive and explaining and educating instead will likely be more effective.
An updated ssh package just hit the Debian security mirrors.
Can you share what version is fixed for the Debian folks? I did my apt-get update&&apt-get install ssh and came back with ssh 1:3.4p1-1, but I don't know how recent this is, or if I got a bad mirror or...
One of the issues not often addressed is the misuse (in my opinion, and some would argue by its original intention) of the Social Security number as a universal identifier in so many public and private functions. It happens for convenience - the SS # is government issued, unique and relatively difficult to spoof, so it's handy.
I'm not certain about all of what you said.
My mother worked in a state university admissions department in the 1960s and 1970s, and was a programmer and operator of their computer. One year, they had two applicants apply under than same social security number. They were able to verify that both people owned the same number! Turned out, the US Government didn't guarantee the uniqueness of the SSN-- it ALONG WITH YOUR NAME AND BIRTHDAY were your taxpayer unique ID. But the university had no way of admitting both students as they wanted to under the same SSN, so they asked one of them to get a new one. It wasn't hard once the Social Security Administration figured out why.
Times have changed and computers have proliferated, and I've only done some casual investigation, but I've never found any guarantee by the US government that the SSN is unique.
SCO has already shown a willingness to sue based upon shaky grounds. I'd bet if they don't win the IBM lawsuit they will go after someone else next.
I contend the two incidents are unrelated. If they don't win against IBM, and they still have money left, sure they'll sue someone else. If they do win against the world's largest computer company, you honestly think they'll stop there?
Maybe not completely unrelated. They'll move on to other big fish if they win against IBM, and just pick on the fellow little guys if they don't.
Here is a study done by an independent research firm which claims that under certain circumstances, it is cheaper to develop applications and enterprise solutions for Windows than for Linux.
My initial instinct was that this was a combination of "absurd" and "special case so specific it's mostly useless". But then I started to think of a Slashdot thread from just a few weeks ago about the big worms that started recently...
The thread discussed how much cheaper it was to hire just any person and have [him|her] maintain the "Windows Server". Of course, an affordable admin in many small business cases would be unable to keep such a server patched well enough to fend off all the attacks and the machine would be compromised. The thread continued to say that if you compare a competant Windows admin with a competant *nix admin, not only are the costs similar but so is the security-- but you could have a Windows box up, running and making money with an incompetant admin.
No offense is intended, by the way, in calling such a person an incompetant admin, just that many small businesses can afford neither a service contract nor a full time "real admin", so someone who does not specialize in such tasks part-times it. This is a rare situation with *nix, where the barrier to entry of a steep learning curve usually causes entry admins to be better than Windows (I have no real evidence to back up this assertion, only personal observation). The theory is that a small business can't afford to keep 100% uptime, but can afford to go down for 12-24 hours.
This makes me wonder about programming on Windows in a general case. I can understand how someone can develop a Visual Basic program for cheaper than a C (or whatever) equivalent on Linux. Instead of comparing.NET to J2EE, as the article does, I'd be interested in seeing a problem solved by a beginning application developer in Windows (would (s)he choose Visual Basic?), another in Linux (C/C++ plus GTK or similar?), and then someone experienced on the two platforms solve the same problem and find out where the added costs present benefits. Can we tell the difference in benefits between the two skilled solutions or the two unskilled solutions? What benefits are gained by keeping one platform but redeveloping with a skilled developer?
From the press release: "For example, the owner can send an email instruction to AIBO asking it to take an unsuspecting picture of the
owner's children and send it back to the mobile telephone."
I don't know about you, but when I was growing up, the kids were the ones who set the VCRs and kept the family computers going. None of my friends' familes were more capable -- anything technical was handed to the kids "who have no fear, but never seem to break it".
So, some suspicous dad is going to be able to get the AIBO to do things to catch the kids doing something? Maybe once-- likely not after that. Most likely, in my opinion, the parent's going to have to ask the kid to show him/her how to do it the first time, then mess up the next several times. That's a warning to the kid to establish countermeasures.
10-15 years ago, my mother had an alarm system installed, and set an option that would make quiet beeps when a door or window opened. I had a Lego model in my late teens that looked "cute" to my mom, but could be used to bypass the window sensor for their home alarm system. Allowed me to come and go as I pleased.
Now that my first daughter is on the way, I'm already wondering how much trust can ever be established with offspring. I figure I have a decade before that decision needs to be made in ernest.
Really, this problem can be solved in two easy steps:
1. Find a religion. Preferably one that allows forgiveness (for step 2.) or reincarnation.
2. Use your own intuition and repair the cabinet in the way that makes the most sense. You took a circuits class in college and know that electrons are electrons. This is no different-- how hard could it be.
I realize that lots of people like Ask Slashdot because they feel that they're pretty similar to the average Slashdot reader, but there is a wider audience out there. Epinions is a site where you can search for many parameters, like brand or price, or features, and see what fellow consumers have recommended.
I wrote a little description of the TV I bought a while ago here.
As with anything where you ask for people's opinions, I find it helpful to find the harshest critics and decide if you side with them or discount their ideas. Many times, in epinions and in real life, the people with the worst criticism complain about features something doesn't have as if they didn't read the product description, or how unreliable it is when it's treated abnormally ("Sony VCRs suck because one broke when my son dropped it off my roof" kind of things). If, however, you find people complaining about how the product doesn't perform as advertised, or how it fails in normal circumstances, you know it really is bad.
Personally I find SCO's management style refreshing.
No wishy-washyness. It's damn the torpedoes, and full speed ahead. Never a moment of doubt that they may be making a huge mistake. No second guessing themselves. We know what we want, and we know where we're going. And we'll be damned if ANYTHING is going to dissuade us. Full court press, lads.
Thanks! I knew I'd seen that before! [Tasteless_Humor]The last time I heard about damning torpedoes and proceeding without second guessing, the Kursk sunk.[/Tasteless_Humor]
Honestly, SCO appears to be proceeding very quickly, but not very decisively. Yes, they've been proceeding to sue lots of companies, but they've also had very strange public faces. At first, they hadn't talked to Linus and only had a beef with IBM, no end user would notice. By now, they talked to Linus way back in December, who told SCO to screw themselves and SCO now says end-users need to cough up lots of money in exchange for GPLed code they can download from SCO this very moment.
I prefer IBM's style. 300k+ employees with a single voice, saying, "We believe our customers have no need to fear and are pursuing SCO as necessary". SCO barely has 300 employees, if that, and they still can't keep a consistant story.
What about speed issues? Isn't photoshop+wine a lot slower than running it in native win32? I can hardly run mirc with wine on a 1ghz computer (only a test, I don't really use mirc;)
Best I can say is "Your mileage may vary". I don't know about real benchmarks, only personal experience like you. I used to run Wine regularly (until I got a Mac), and I found some applications that were pretty slow compared to Windows' implementation, some that were about the same, and one that "felt faster" (Starcraft -- the screen updates were smoother, scrolling was faster, etc, not that the game actually moved faster).
I'd honestly like to see some public benchmarking here. I expect that GCC (Wine) is going to be slower in most things, but I have this sneaking suspicion that the Windows API is so screwed up and kludged together that it takes processor cycles to get all the intracacies right-- and some of those intracacies are what causes Wine not to have 100% compatibility. By skipping some of those details, can Wine be faster despite a less optimized compiler?
the users don't own their machines - the company does. if they want to piss around with _any_ os, let them do it on their own time, on their own network, and on their own equipment.
I certainly see your point. The company pays to maintain my laptop for me, the company pays for the support and pays for me not to have to worry about it. They pay you to do all that for me.
I see my employer as someone who pays me to do a job. I'm not that good at windows. I can do many things with *nix better or faster or both. If I asked my manager if I could modify my laptop and my productivity would increase by 10%, she'd approve it. Personally, I can keep a Linux install up to date, well maintained and all the appropriate patches on it. Certainly better than the company can do that to my windows equipped laptop while they tiptoe around taking me down at inconvenient times.
The company pays you to keep me out of trouble. The company pays me to be efficient. If I can be more efficient and keep myself out of trouble, why should you care that you have one less Windows machine to maintain (to say nothing of the grumpy luser you have to deal with)? Of course, if I get 0wned, you need to come down on me hard and make sure that my manager knows you're here to keep that from happening to me but I didn't let you.
As far as license management goes, maybe you could work with a rogue Linux user and find out how to satiate your needs and his / hers?
You are suggesting that in the current climate, the US Gov. will encourage people to build rockets with warheads and fire them in the US?
There's a difference between a gaggle of people out in the middle of Nowhere, South Dakota huddling up in shacks in the middle of the woods with their automatic weapons just in case the IRS comes and a gaggle of rocket geeks paving some land for a good rocket pad. I'm not sure how to tell the difference, aside from the fact that the idiots trying to secede from the union have the right to bear arms and the geeks with the rocket are insisting that their manned vehicle isn't an arm but they still have rights.
I'm interested in private space flight, but I think the legislative obstacles are severe on mainland US, whereas an abandoned oil rig in the middle of the south pacific or something would be ideal. Sure, getting there may be fun, but even if it goes haywire and doesn't explode, killing a sperm whale is different from levelling a town of 100.
Personally, I'd trust John Carmack (whom I've given probably $150 at this point, from Quake III, Quake I and Doom II) with the rocket more than the tax evading guys enforcing their own rights.
Language precision, assumptions
on
RAID for Zero-G?
·
· Score: 1
and I need to it work in zero-G (but not in a vacuum)
First off, this states that you explicitly need it not to work, i.e. "fail" in a vacuum. I doubt this is your intent. If you need something to fail in a vacuum, one idea may be to have a membrane holding back some sort of acid at air pressure, but which bursts and destroys your media at a sufficiently low pressure. If you merely don't care about failing in a vacuum, then leave that out of all future problem statements.
Secondly, to all the posters before me, it appears as though everyone assumes this person is going into space. While Cujo is from John Hopkins' Applied Physics Lab (looking at the e-mail address) which is apparently sending a craft to Mercury, it is possible that Cujo is working on a different project which is terrestrial. Perhaps a roller coaster type environment or sky diving type airplane bourne freefall. I'm not sure why you need 150GB of storage on a craft that will forever have a 9600 baud connection home (if even that fast)-- so I suspect that the poster has less lofty goals. But maybe 150GB is necessary because of the volume of data to record and the craft will return.
To combine RAID type reliability and resistance to vibration and probably some attention to power dissipation, solid state is a good possibility, especially if you have few write cycles. You can apparently get 512 MB of compact flash for $128 but someone might give you a discount if you buy 300 of them. If not, that brings you to less than $40k, which still leaves you with ~$12k to pay a student to integrate them into a smart FPGA based controller.
For pre-existing solid state solutions, google for SSD or solid state drives.
You don't have to worry about crossing the country borders, but the comments are interesting.
Personally, I left my computer in my college town (at a you-store-it kind of place, not environmentally controlled, just a few feet off the concrete ground) even when I was gone. Get a used POS for your summer needs. It's not worth the trouble to ship your baby just for the summer.
I would love an article that deserved the title "Building A (Serious) Home Network From Scratch". This isn't it.
I want to see how to open my wall with the least destruction. How to snake it up to the attic. How to snake it up to the plywood under the carpet in the second floor. Then, how to patch drywall.
I want to see how to effectively route wires from all over the house to a punchdown block in a closet. I don't need to see ethernet cable connectors every step of the way, we're talking about a serious network. PUNCH DOWN BLOCKS! There is, of course, a discussion of conduit. People who have done it and found a need, along with a few sentences about people who have spent (how long does it take?) a few hours to do it and found it to be a waste of time. Eventually, because it's only a Home Network, we probably just go with wires and no conduit, but we've at least thought about conduit and feel better for it.
This Serious Home Network probably also has jacks to tie in the neighbors for bandwidth sharing or LAN gaming, or possibly some other fancy stuff, but we're not wasting our page views on how to cut down a rack. We can figure that stuff out.
Another useful fact is that the merchant contract prevents the store from requiring seperate ID beyond the card itself. However, they are allowed to ask for ID, they just aren't allowed to require ID unless they have reason to believe that the transaction is fraudulent (note, blanket policy of requiring ID from all CC users is not sufficient, that is the equivalent of saying, "if you shop at our store, we think you are trying to commit fraud" and thus is not acceptable under their merchant contract).
This is not the case for American Express. I just called. They protect me against fraud for 100% of the money (looked at my agreement), and the merchant website says they protect merchants against fraudulant activity for up to 48 hours after the fact.
When I called AmEx a little earlier today, I was read the riot act about why I would want to be accused a thief. With identiy theft so high these days, they'd be fools not to ask me. Of course, I'd be furious when I got the bill.
Thing is, I'm not furious when I get a charge I disgree with. I've charged back over $500 worth of money at this point, and it's no big deal. I just call 'em up, tell them I want more information on a charge because I don't think it's legit, it gets taken off, and that's the last I hear. I've done that with 4 credit cards over the past 10 years (total of $500 over those 4). Once Cheker Auto Parts took two tries to sell me an alternator, and both of the attempts went through-- first one gave a confusing message. Second time, a movie theatre computer charged me for seeing like $150 worth of movies in one day from a kiosk I'd seen misbehave before. After that, 3 and 4 were not memorable, I just remember vaguely how much money happened each time. An identity thief, however, will have a fake ID on them to be able to "prove" they're me. How does handing over my driver's license with my real address and birthdate on it protect me in any way? I've seen merchants xerox my ID, write down information, enter it in a computer, and even disappear into the back room. I don't hand it over any more.
I've seen people forget to pick up their credit cards after conducting transactions. Heck, I've served customers (when I did retail) who left them on the counter with the recipt when I returned them (and run after them in the parking lot as soon as I realized, caught all but one who came back later that day). I carry multiple credit cards, partially in case I do that because I don't want to end up screwed later in the day when I conduct the next transaction. I do not carry redundant driver's licenses, so if I later decided to buy beer or had a police officer ask for my identification, I would be screwed simply because a vendor asked for my ID and I may have been absent minded. I have more respect for myself than that, and I HAVE taken my business elsewhere because of it. My less geeky wife has started doing the same thing.
American Express policy dictates that merchants can ask for an ID and refuse a transaction when I refuse. In Texas, I have the right to refuse to fork over my ID. The merchant has the right to refuse to accept money from me.
Where i work we are very seriously working towards ridding our machine room of SCO forever.
To this end, I'm taking suggestions as to innovative and torturous ways to take a SCO Unixware box down.
Note; we will be putting Linux on the boxes after SCO is removed, so please, no suggestions that involve damage to the hardware.
fdisk and mkfs. Simple, quick, to the point. Any ceremony you add increases the time you're running SCO. Depenending on where your office is, this is likely faster than even incinerating the hardware (starting the clock on where you poweroff the SCO install).
IBM *has* a copy of the offending code. IBM has had a copy of the System V source code for years now. Anyone with a copy of both Linux and System V can easily find which lines they have in common.
That's true only in the strictest sense of the word. SCO is alleging that IBM has made simple modifications to System V code and imported it to Linux. If this is the case, then a massive grepping party at IBM wouldn't reveal the offending code. You'd have to have an army of people sifting through tens or hundreds of megabytes code in order to find out what SCO is talking about. And how similar are we talking here? Where is the line between similar code that's similar because of illicit activity and similar code that's similar because it's the best approach drawn? If the scheduler is similar, perhaps that's because that's the best way to write a scheduler. It started out as very straightforward and based on academic works.
I believe that SCO needs to be specific with the request, and any judgement against IBM needs to consider intent and practice. If a different team came up with the code, IBM shouldn't be liable. If SCO won't tell IBM what specific code is in question, IBM shouldn't be liable. If IBM legally
dual-licensed the code that IBM wrote, IBM shouldn't be liable (key here is, does the SCO / IBM contract allow dual licensing?).
SCO notified IBM on March 6, 2003 that it intended to terminate in 100 days, as required
under the Software Agreement, as modified by a side letter, if IBM did not correct certain actions that
violate the agreement. As of the deadline -- 12:00 midnight, June 13, 2003 -- IBM had not complied with
SCO's request, which triggered the termination. The termination is self-effectuating.
In order for IBM to be able to comply with certain actions, as I understand it, IBM would have to either:
A) Stop selling AIX.
B) Remove the offending code from Linux.
In order to do A), well, IBM would have to give up. In order to do B), IBM would have to have a copy of what
SCO thinks is the offending code, review it, engineer suitable replacements, and submit patches to Linus. I don't
think Linus would necessarily have to accept it for IBM to prove that it has done all it could. But, I believe
we've read before, SCO didn't want to share its violated code until last week or so. If IBM didn't have access to
that until last week, SCO was asking IBM to take their word for it. Doesn't sound very legal to me.
I've seen IBM work. Sometimes it's slow, but sometimes they can move a staff of 300k people so quickly the
earth spins the other way. I've got to think that IBM has enough talent to replace many 60 line blocks and have
them tested before 100 days had expired, if given a fair chance.
Last night, I had convinced myself that I thought it was reasonable for IBM to be dual licensing code they had
written. I'm still not sure SCO has proven IBM has liberated code, but if it had, and it was originally IBM's,
why not allow it?
By stating "IBM has clearly demonstrated its misuse of UNIX source code..."
by "using UNIX methods to accelerate and improve Linux as a free operating system", is SCO saying that even if
a completely disparate group of Unix virgin IBMers couldn't work on Linux without undermining the contract?
That sounds awefully strict.
I work at the local EB, and we've been having more and more customers come in with problems related to copy protection schemes in the last year or so...We've got people with strange hardware configurations that aren't allowed by SafeDisk and its friends. All these things in an attempt to prevent piracy....and all they do is make things difficult for the paying customers.
One reason that I havn't bought a lot of games lately is because of your statement combined with very restrictive return policies that the stores have. If I bought a game at Fry's, and suffered from the phenomenon you describe, I couldn't return it except for a new copy of the same game. Well, if my computer isn't SafeDisk compatible, am I out $60? How does it work at ElBo? Do you apologetically give the money back and ask them to write to the publisher? Or do you stiff them (or slightly better, only give store credit)?
Slashdot Mirror
I see two alternatives:
1. Through ignorance or oversight, you're not counting IBM, the largest computer company in the world. They maintain PowerPC (750 and 970 series and embedded stuff), Power (eServers and stuff) and mainframes (zSeries -- S/390, more than the riced out Unix that Sun wants to call a mainframe). Additionally, IBM designs ASICs not only for themselves, but for other companies.
2. You truly have an insightful point in how many people are employed by Sun-- which points to how much bloat Sun has. If, in fact, you are correct, you are pointing out how many people Sun needs to cut in order to be competive with IBM, to say nothing if Intel who only maintains a few chipset ASICs, the Pentium III/IV and the new cheap one they're advertising on TV.
1) Reduced code bloat. /sarcasm
2) Reduced memory footprint.
3) More features.
.
Come on. Linux does what most of us want already. It's the Pentium II of OS kernels. For icing only:
Get something going with Xfree86, find out what hooks they want for better performance.
Talk with NVidia and ATI and find out what common API that's not DirectX9, but is both open in specificaion and foreward thinking enough to expand to meet tomorrow's needs.
Settle on what it takes to require GCC3.X as the compiler so distributions can ship with just one GCC version.
Get a command line option to prevent printing out all the copyright notices (the corproate equivalent of teenagers' "shoutouts" -- does the warez community still do "greetz"?) and only print out error messages.
Howabout a file system (ext2?) that you can fsck while mounted RW (even if the output is "all is well" instead of actually doing fixes)?
And an uptime counter that gets me past 430 days or however many it resets. That's the one thing BSD has over us.
So, you're suggesting that I, as someone who heard about the tricks Intuit was making with Turbo Tax and pre-emptively moved to H&R Block's TaxCut should congratulate and thank Intuit for seeing the light and give them money for realizing they screwed the pooch?
In 2002 (FY 2001), I used Intuit's Turbo Tax on the web. It sucked. It was slow and I couldn't easily look stuff up 6 months later. I was thankful to Intuit for the service, and vowed to buy their standalone software.
In 2003 (FY 2002), I heard about Intuit's treatment of their customers and, having learned my lesson about using web tax preparation, bought TaxCut. Much more responsive, no DRM problems and I was happy with the results.
In 2004 (FY 2003), I will buy TaxCut again. I appreciate the fact that I have never heard of them treating their customers like criminals, and their product this year has bought my loyalty.
In college, I was the ideal Slackware user. I wanted to learn programming, loved to compile my own stuff and felt that Redhat was only useful in removing the user from my goal of learning *nix. I extolled its virtues at every opportunity. I didn't have to hunt down an RPM just to install a new kernel, and I certainly knew how to compile my own kernel. I didn't know of any other serious distributions in 1996.
When I entered the real world and had a job and non-computer hobbies, I still had the need (perceived need as opposed to life or death need) for a *nix machine for my home mail, DNS and web serving. I no longer had the time, however, for fixing dependencies, applying source code patches and hunting down the minor details that I had arduously learned how to hunt down in the previous years. Debian came to my rescue. Dselect may be rough around the intuitive UI edges, and it's not quick on my 486, but it's consistent and only requires occasional answers to keep my machine well patched. It's been years since I've had to compile my own kernel, let alone wanted to.
I fondly remember Slackware as I remember my first girlfriend. It was a good idea at the time, but that time has since past and I have moved on. I am much happier now, but the lessons and memories will stay with me for a long time.
To the first order, yes, doubling the cache size will likely double the latency. A skilled SRAM designer knows a few tricks which may mitigate this some. However, the CPU designer needs to know what tradeoffs the area expensive and high off current SRAM array will have against the application advantages of not having to go hundreds of CPU cycles out to main memory. Don't forget that doubling the access time of an SRAM array to 20 cycles is far less than multiple hundred to get to main memory.
If one were maximizing performance of RC5 brute forcing, Distributed.net found that 256k of cache was plenty. On the other hand, SETI found they needed 1 Meg of cache (IIRC)-- the RC5 application would suffer from the effect you point out-- an L2 too large to take advantage of. A database program, on the other hand, would really probably prefer no datacache penalties if the application is truly random access. Cache misses cost cycles too.
Don't get too frustrated with this. There will always be people who don't understand something fundamental to your training. That's why you're trained, to understand these non-obvious fundamentals. Now that you understand a CPU has to be fed data in order to process it, it's obvious, but a PHB wouldn't necessarily come to that conclusion on his own.
Your EE instructor will tell you lots of things that can help performance. For example, making the L2 cache be the size of main memory. Just because it helps performance doesn't make it worth the price. Rounded edges on the PCB are not easy to accomplish and their benefits may not be outweighed by the added price-- even for exceptionally high end servers. Without looking at the math, I would like to toss "10% performance adder, 50% cost adder" out into the air, and say that most people would rather save the dough. Another factor to consider is reliability. Intuition suggests to me that reliability would go up without sharp edges, but intuition also tells me that modelling board coupling on a 4 layer board would be a real pain in the ass, to say nothing of a server class 6 or 8 (or higher) layer board if you have to model curved structures. You might not find an easy way to capitalize on your wonderful curved wire performance. Not only do you have to worry about your slowest path, but your quickest one can't arrive so quickly that the other chip can't sample the previous output.
Take care in your classes when you use the word "only". Taking advantage of our wonderful next generation 64 bit processors and multiple gigs of RAM, we could conceivably copy the contents of the hard drive to main memory (especially if we are only concerned with 1-4 gigs of data in a low cost solution). Here, we get the enhanced bandwidth of main memory instead of having to kludge through the southbridge, PCI controller, IDE/SCSI to RAM interface and back.
There are many things that improve system performance-- and the system is the only thing that matters-- rounded wires and SSDs (solid state drives) are only the beginning. Depending on the application, a larger L3 cache may make more difference, or a wider faster CPU to CPU interface, or a pair of PCI controllers hanging off the southbridge for twice the bandwidth, or integrating the northbridge onto the CPU, or ...
The best engineering advice I can give you is that the answer is always, "It depends". You'll spend the next 5-30 years of your life learning how to answer the followon question "Depends on what?". Almost everything has advantages and disadvantages and there are few absolutes.
The "Someone gets it" and "They never quite grasped" attitude may get you in trouble. Being proactive and explaining and educating instead will likely be more effective.
Can you share what version is fixed for the Debian folks? I did my apt-get update&&apt-get install ssh and came back with ssh 1:3.4p1-1, but I don't know how recent this is, or if I got a bad mirror or...
I'm not certain about all of what you said.
My mother worked in a state university admissions department in the 1960s and 1970s, and was a programmer and operator of their computer. One year, they had two applicants apply under than same social security number. They were able to verify that both people owned the same number! Turned out, the US Government didn't guarantee the uniqueness of the SSN-- it ALONG WITH YOUR NAME AND BIRTHDAY were your taxpayer unique ID. But the university had no way of admitting both students as they wanted to under the same SSN, so they asked one of them to get a new one. It wasn't hard once the Social Security Administration figured out why.
Times have changed and computers have proliferated, and I've only done some casual investigation, but I've never found any guarantee by the US government that the SSN is unique.
I contend the two incidents are unrelated. If they don't win against IBM, and they still have money left, sure they'll sue someone else. If they do win against the world's largest computer company, you honestly think they'll stop there?
Maybe not completely unrelated. They'll move on to other big fish if they win against IBM, and just pick on the fellow little guys if they don't.
My initial instinct was that this was a combination of "absurd" and "special case so specific it's mostly useless". But then I started to think of a Slashdot thread from just a few weeks ago about the big worms that started recently...
The thread discussed how much cheaper it was to hire just any person and have [him|her] maintain the "Windows Server". Of course, an affordable admin in many small business cases would be unable to keep such a server patched well enough to fend off all the attacks and the machine would be compromised. The thread continued to say that if you compare a competant Windows admin with a competant *nix admin, not only are the costs similar but so is the security-- but you could have a Windows box up, running and making money with an incompetant admin.
No offense is intended, by the way, in calling such a person an incompetant admin, just that many small businesses can afford neither a service contract nor a full time "real admin", so someone who does not specialize in such tasks part-times it. This is a rare situation with *nix, where the barrier to entry of a steep learning curve usually causes entry admins to be better than Windows (I have no real evidence to back up this assertion, only personal observation). The theory is that a small business can't afford to keep 100% uptime, but can afford to go down for 12-24 hours.
This makes me wonder about programming on Windows in a general case. I can understand how someone can develop a Visual Basic program for cheaper than a C (or whatever) equivalent on Linux. Instead of comparing .NET to J2EE, as the article does, I'd be interested in seeing a problem solved by a beginning application developer in Windows (would (s)he choose Visual Basic?), another in Linux (C/C++ plus GTK or similar?), and then someone experienced on the two platforms solve the same problem and find out where the added costs present benefits. Can we tell the difference in benefits between the two skilled solutions or the two unskilled solutions? What benefits are gained by keeping one platform but redeveloping with a skilled developer?
I don't know about you, but when I was growing up, the kids were the ones who set the VCRs and kept the family computers going. None of my friends' familes were more capable -- anything technical was handed to the kids "who have no fear, but never seem to break it".
So, some suspicous dad is going to be able to get the AIBO to do things to catch the kids doing something? Maybe once-- likely not after that. Most likely, in my opinion, the parent's going to have to ask the kid to show him/her how to do it the first time, then mess up the next several times. That's a warning to the kid to establish countermeasures.
10-15 years ago, my mother had an alarm system installed, and set an option that would make quiet beeps when a door or window opened. I had a Lego model in my late teens that looked "cute" to my mom, but could be used to bypass the window sensor for their home alarm system. Allowed me to come and go as I pleased.
Now that my first daughter is on the way, I'm already wondering how much trust can ever be established with offspring. I figure I have a decade before that decision needs to be made in ernest.
Really, this problem can be solved in two easy steps:
1. Find a religion. Preferably one that allows forgiveness (for step 2.) or reincarnation.
2. Use your own intuition and repair the cabinet in the way that makes the most sense. You took a circuits class in college and know that electrons are electrons. This is no different-- how hard could it be.
Did you look at epinions.com?
I realize that lots of people like Ask Slashdot because they feel that they're pretty similar to the average Slashdot reader, but there is a wider audience out there. Epinions is a site where you can search for many parameters, like brand or price, or features, and see what fellow consumers have recommended.
I wrote a little description of the TV I bought a while ago here.
As with anything where you ask for people's opinions, I find it helpful to find the harshest critics and decide if you side with them or discount their ideas. Many times, in epinions and in real life, the people with the worst criticism complain about features something doesn't have as if they didn't read the product description, or how unreliable it is when it's treated abnormally ("Sony VCRs suck because one broke when my son dropped it off my roof" kind of things). If, however, you find people complaining about how the product doesn't perform as advertised, or how it fails in normal circumstances, you know it really is bad.
Thanks! I knew I'd seen that before! [Tasteless_Humor]The last time I heard about damning torpedoes and proceeding without second guessing, the Kursk sunk.[/Tasteless_Humor]
Honestly, SCO appears to be proceeding very quickly, but not very decisively. Yes, they've been proceeding to sue lots of companies, but they've also had very strange public faces. At first, they hadn't talked to Linus and only had a beef with IBM, no end user would notice. By now, they talked to Linus way back in December, who told SCO to screw themselves and SCO now says end-users need to cough up lots of money in exchange for GPLed code they can download from SCO this very moment.
I prefer IBM's style. 300k+ employees with a single voice, saying, "We believe our customers have no need to fear and are pursuing SCO as necessary". SCO barely has 300 employees, if that, and they still can't keep a consistant story.
Best I can say is "Your mileage may vary". I don't know about real benchmarks, only personal experience like you. I used to run Wine regularly (until I got a Mac), and I found some applications that were pretty slow compared to Windows' implementation, some that were about the same, and one that "felt faster" (Starcraft -- the screen updates were smoother, scrolling was faster, etc, not that the game actually moved faster).
I'd honestly like to see some public benchmarking here. I expect that GCC (Wine) is going to be slower in most things, but I have this sneaking suspicion that the Windows API is so screwed up and kludged together that it takes processor cycles to get all the intracacies right-- and some of those intracacies are what causes Wine not to have 100% compatibility. By skipping some of those details, can Wine be faster despite a less optimized compiler?
I certainly see your point. The company pays to maintain my laptop for me, the company pays for the support and pays for me not to have to worry about it. They pay you to do all that for me.
I see my employer as someone who pays me to do a job. I'm not that good at windows. I can do many things with *nix better or faster or both. If I asked my manager if I could modify my laptop and my productivity would increase by 10%, she'd approve it. Personally, I can keep a Linux install up to date, well maintained and all the appropriate patches on it. Certainly better than the company can do that to my windows equipped laptop while they tiptoe around taking me down at inconvenient times.
The company pays you to keep me out of trouble. The company pays me to be efficient. If I can be more efficient and keep myself out of trouble, why should you care that you have one less Windows machine to maintain (to say nothing of the grumpy luser you have to deal with)? Of course, if I get 0wned, you need to come down on me hard and make sure that my manager knows you're here to keep that from happening to me but I didn't let you.
As far as license management goes, maybe you could work with a rogue Linux user and find out how to satiate your needs and his / hers?
There's a difference between a gaggle of people out in the middle of Nowhere, South Dakota huddling up in shacks in the middle of the woods with their automatic weapons just in case the IRS comes and a gaggle of rocket geeks paving some land for a good rocket pad. I'm not sure how to tell the difference, aside from the fact that the idiots trying to secede from the union have the right to bear arms and the geeks with the rocket are insisting that their manned vehicle isn't an arm but they still have rights.
I'm interested in private space flight, but I think the legislative obstacles are severe on mainland US, whereas an abandoned oil rig in the middle of the south pacific or something would be ideal. Sure, getting there may be fun, but even if it goes haywire and doesn't explode, killing a sperm whale is different from levelling a town of 100.
Personally, I'd trust John Carmack (whom I've given probably $150 at this point, from Quake III, Quake I and Doom II) with the rocket more than the tax evading guys enforcing their own rights.
First off, this states that you explicitly need it not to work, i.e. "fail" in a vacuum. I doubt this is your intent. If you need something to fail in a vacuum, one idea may be to have a membrane holding back some sort of acid at air pressure, but which bursts and destroys your media at a sufficiently low pressure. If you merely don't care about failing in a vacuum, then leave that out of all future problem statements.
Secondly, to all the posters before me, it appears as though everyone assumes this person is going into space. While Cujo is from John Hopkins' Applied Physics Lab (looking at the e-mail address) which is apparently sending a craft to Mercury, it is possible that Cujo is working on a different project which is terrestrial. Perhaps a roller coaster type environment or sky diving type airplane bourne freefall. I'm not sure why you need 150GB of storage on a craft that will forever have a 9600 baud connection home (if even that fast)-- so I suspect that the poster has less lofty goals. But maybe 150GB is necessary because of the volume of data to record and the craft will return.
To combine RAID type reliability and resistance to vibration and probably some attention to power dissipation, solid state is a good possibility, especially if you have few write cycles. You can apparently get 512 MB of compact flash for $128 but someone might give you a discount if you buy 300 of them. If not, that brings you to less than $40k, which still leaves you with ~$12k to pay a student to integrate them into a smart FPGA based controller.
For pre-existing solid state solutions, google for SSD or solid state drives.
Read this.
You don't have to worry about crossing the country borders, but the comments are interesting.
Personally, I left my computer in my college town (at a you-store-it kind of place, not environmentally controlled, just a few feet off the concrete ground) even when I was gone. Get a used POS for your summer needs. It's not worth the trouble to ship your baby just for the summer.
I would love an article that deserved the title "Building A (Serious) Home Network From Scratch". This isn't it.
I want to see how to open my wall with the least destruction. How to snake it up to the attic. How to snake it up to the plywood under the carpet in the second floor. Then, how to patch drywall.
I want to see how to effectively route wires from all over the house to a punchdown block in a closet. I don't need to see ethernet cable connectors every step of the way, we're talking about a serious network. PUNCH DOWN BLOCKS! There is, of course, a discussion of conduit. People who have done it and found a need, along with a few sentences about people who have spent (how long does it take?) a few hours to do it and found it to be a waste of time. Eventually, because it's only a Home Network, we probably just go with wires and no conduit, but we've at least thought about conduit and feel better for it.
This Serious Home Network probably also has jacks to tie in the neighbors for bandwidth sharing or LAN gaming, or possibly some other fancy stuff, but we're not wasting our page views on how to cut down a rack. We can figure that stuff out.
This is not the case for American Express. I just called. They protect me against fraud for 100% of the money (looked at my agreement), and the merchant website says they protect merchants against fraudulant activity for up to 48 hours after the fact.
When I called AmEx a little earlier today, I was read the riot act about why I would want to be accused a thief. With identiy theft so high these days, they'd be fools not to ask me. Of course, I'd be furious when I got the bill.
Thing is, I'm not furious when I get a charge I disgree with. I've charged back over $500 worth of money at this point, and it's no big deal. I just call 'em up, tell them I want more information on a charge because I don't think it's legit, it gets taken off, and that's the last I hear. I've done that with 4 credit cards over the past 10 years (total of $500 over those 4). Once Cheker Auto Parts took two tries to sell me an alternator, and both of the attempts went through-- first one gave a confusing message. Second time, a movie theatre computer charged me for seeing like $150 worth of movies in one day from a kiosk I'd seen misbehave before. After that, 3 and 4 were not memorable, I just remember vaguely how much money happened each time. An identity thief, however, will have a fake ID on them to be able to "prove" they're me. How does handing over my driver's license with my real address and birthdate on it protect me in any way? I've seen merchants xerox my ID, write down information, enter it in a computer, and even disappear into the back room. I don't hand it over any more.
I've seen people forget to pick up their credit cards after conducting transactions. Heck, I've served customers (when I did retail) who left them on the counter with the recipt when I returned them (and run after them in the parking lot as soon as I realized, caught all but one who came back later that day). I carry multiple credit cards, partially in case I do that because I don't want to end up screwed later in the day when I conduct the next transaction. I do not carry redundant driver's licenses, so if I later decided to buy beer or had a police officer ask for my identification, I would be screwed simply because a vendor asked for my ID and I may have been absent minded. I have more respect for myself than that, and I HAVE taken my business elsewhere because of it. My less geeky wife has started doing the same thing.
American Express policy dictates that merchants can ask for an ID and refuse a transaction when I refuse. In Texas, I have the right to refuse to fork over my ID. The merchant has the right to refuse to accept money from me.
fdisk and mkfs. Simple, quick, to the point. Any ceremony you add increases the time you're running SCO. Depenending on where your office is, this is likely faster than even incinerating the hardware (starting the clock on where you poweroff the SCO install).
That's true only in the strictest sense of the word. SCO is alleging that IBM has made simple modifications to System V code and imported it to Linux. If this is the case, then a massive grepping party at IBM wouldn't reveal the offending code. You'd have to have an army of people sifting through tens or hundreds of megabytes code in order to find out what SCO is talking about. And how similar are we talking here? Where is the line between similar code that's similar because of illicit activity and similar code that's similar because it's the best approach drawn? If the scheduler is similar, perhaps that's because that's the best way to write a scheduler. It started out as very straightforward and based on academic works.
I believe that SCO needs to be specific with the request, and any judgement against IBM needs to consider intent and practice. If a different team came up with the code, IBM shouldn't be liable. If SCO won't tell IBM what specific code is in question, IBM shouldn't be liable. If IBM legally
dual-licensed the code that IBM wrote, IBM shouldn't be liable (key here is, does the SCO / IBM contract allow dual licensing?).In order for IBM to be able to comply with certain actions, as I understand it, IBM would have to either:
A) Stop selling AIX.
B) Remove the offending code from Linux.
In order to do A), well, IBM would have to give up. In order to do B), IBM would have to have a copy of what SCO thinks is the offending code, review it, engineer suitable replacements, and submit patches to Linus. I don't think Linus would necessarily have to accept it for IBM to prove that it has done all it could. But, I believe we've read before, SCO didn't want to share its violated code until last week or so. If IBM didn't have access to that until last week, SCO was asking IBM to take their word for it. Doesn't sound very legal to me.
I've seen IBM work. Sometimes it's slow, but sometimes they can move a staff of 300k people so quickly the earth spins the other way. I've got to think that IBM has enough talent to replace many 60 line blocks and have them tested before 100 days had expired, if given a fair chance.
Last night, I had convinced myself that I thought it was reasonable for IBM to be dual licensing code they had written. I'm still not sure SCO has proven IBM has liberated code, but if it had, and it was originally IBM's, why not allow it?
By stating "IBM has clearly demonstrated its misuse of UNIX source code..." by "using UNIX methods to accelerate and improve Linux as a free operating system", is SCO saying that even if a completely disparate group of Unix virgin IBMers couldn't work on Linux without undermining the contract? That sounds awefully strict.
One reason that I havn't bought a lot of games lately is because of your statement combined with very restrictive return policies that the stores have. If I bought a game at Fry's, and suffered from the phenomenon you describe, I couldn't return it except for a new copy of the same game. Well, if my computer isn't SafeDisk compatible, am I out $60? How does it work at ElBo? Do you apologetically give the money back and ask them to write to the publisher? Or do you stiff them (or slightly better, only give store credit)?