if MS didn't want anyone to debug windows, then 1) would the kernel debugger be a free download ? 2) would the kernel symbols (for use with said debugger) be a free download
furthermore, if disassembly were not allowed: 1) would all versions of Visual Studio allow you to view assmebly dumps ?
i think that your statement might be valid, with some qualification. but as presented, i think its incorrect.
but then, you can't be sure of anyone elses software, either. The advantage of OSS in this facet has been debunked time and time again. Thousands of people that depend on sendmail, bind, tcpdump, libpcap, etc. The source is there for them to look at it. They look at it. They dont find anything because they're not looking hard enough or they're not qualified to do security analysis of software.
Using publicly available tools you can single-step execution of the Windows operating system. You can get the names of all the symbols in the kernel. You can set kernel mode breakpoints on any peice of the network (and rpc/app) stack you want. It's not the same as having the source, but you can isolate exactly what the OS is doing at any time if you want to. And you can just disassemble the body of any function you like, once you've isolated it (which is easy, since Microsoft publishes full symbol information).
An experiment i did in about 30 minutes of poking around (im a novice at kernel debugging) had me to a point where i could make a machine drop to a kernel mode debugger when someone connected to a SMB share on that machine. I could then examine the file they were looking at, what operation they were trying to do, etc etc.
Naturally, this can all be automated. You could write a _kernel mode_ SMB debugger if you wanted to, and were worried about the CIFS implementation.
I'm not even sure if debugging tools like that for linux _even exist_.
And it's a moot point anyway. The overwhelming majority of people that think source availability is the difference between secure and insecure software aren't qualified to look at the source of either side, much less make objective measurements of the security/correctness aspects of the code in question.
I am highly dubious that the person quoted here is smart enough to write any kind of a firewall, much less a ruleset for linux or Windows.... which is all secondary to the point i am going to make:
In W2k and later, the entire network stack is completely pluggable. You can insert any layer you want to that sits between NDIS and a protocol driver, and you can create other layers as required. I'd be very surprised if they couldn't do everything required with windows exactly as it sits today.
I think this is just making political noise, and not based on any shred of technical accuracy.
An excerpt: "He was awarded two Purple Hearts, a Bronze Star, a Soldier's Medal, and the Legion of Merit for his exemplary service in Vietnam"
Let me just summarize for you:
Colin Powell kicks more ass before breakfast than you do all day. I suspect he's stood in the face of more suffering, death, and outright fear than you'll ever know in your entire shit-eating miserable existance. And instead of thanking him for being one of the most distinguished soldiers, politicians, and leaders _ever_, you come up with this lame trite slashdot comment ?
I am all for discourse and critical examination of the american government. It's something that makes america what it is.
However, baseless, senseless, insultation of the people that fight (and often die) for your right to say whatever pissant things your 12 neurons manage to fizzle-off and deliver out of your mouth (or keyboard), is not only embarassing, but makes me wonder what it takes to wet YOUR pants ?
As an indigo^2 hi-impact owner (r4.4k, unfortuneately)...
the I^2 is a lame machine unless you've got impact gfx, at which point its gfx subsystem becomes ok. The phobos ethernet cards in these things cripple the poor EISA bus and slam the cpu. Playing midi on the things is choppy..
in any case, i chose to respond because of the 3.2Gb/s figure. when you say "internal bandwidth", what are you referring to ? The main memory bandwidth on any DDR system is actually about that fast these days. The O2 doesn't have the XIO bus of the octane, and its mostly irrelevant because pc, o2, and Octane all use PCI for the majority of the expansion devices anyhow.
For a while the memory bandwidth of pC's was a big limiting factor, but its actually gotten pretty good lately, especially compared to the desktop stuff of the traditional big iron players (sgi, sun, rs6k).
i concede having my history and my names confused. good catch.
re: "my vision"
the NUMAlink is not magic, it is a very specific high performance link with low latency a goal. Eventually, "off the shelf" stuff will be performance competitive, and at a lower cost.
I realize that using hardware not optimized for low latency in place of the FLASH-derivative chips that the router boards use will hurt. I'm referring more to being able to emulate in hosted software the illusion that the NUMA layer creates - that you have a huge pool of local memory, when infact you have a number of discrete pools of physically separate memory, and some slick hardware and software that makes them "seem" local unless you happen to have an app which breaks the implemented algorithms.
My point is that the hardware/software hybrid solution doesn't work magic, its very fast custom hardware with some software that makes it plausible to NOT _have_ to treat what is essentially a cluster like a cluster. There is little inherent advantage to having a single-image OS if you can write a library that your user space HPC apps will use that provides for single-memory and thread/data migration across the nodes in a performant way. And i think cheaper non-custom hardware and intelligent software can bring that layer to traditional clusters.
I'm a huge SGI fan, and i respect the O3k for what it is. (and i respected the O2k and O200 before it). I've read the white papers, even the DASH/FLASH ones from stanford.
Today, the O3k lets many apps scale up to a high cpu count and a large memory size, so that naive implementations can tackle larger problems. In the future, i claim that software and faster interconnect hardware (be it some new ethernet or myrinet or something similar) will let you do the same on commodity hardware.
im saying that it is possible to fake single-image in software just the way O3k fakes it in hardware.
its called "Numa" for a reason - the memory access is non-uniform, and there are perf penalties for thinking that all memory accesses are created equal. If you have an app where the hardware/software implementation of the page migrator is a performance blocker, then you'll need to do some tuning work to bring your app up to speed. And at that ponit, you're smart enough to write an app-specific best-case distributed memory management system (ala a much improved MOSIX, which is why i mentioned it)
The advantage of the O3k is that to many apps, it looks like a single image machine, and it "is", but it doesn't let you always make the same assumptions as a single image machine.
History speaks pretty clearly about what happens to companies that marginalize their business into making 1-offs for infinite-budget DoD contracts and agencies. Eventually, projects get cancelled, line items in budgets get axed, and whole departments are re-orged into something different.
Cray, anyone ? Cray-Research basically went under when the Cray-3 contract was axed. They were counting on that single-machine to keep the afloat. They futzed around with GaAs custom process and never got it qutie working right, and then the cold war ended and with it the justification for subsidizing a maker of 1-off supercomputers.
(Incidentally, the purchase of Cray is what really broke SGI's back. 50% more employees, 2% more market cap, and the O2k/O3k technology came from stanford, not Cray) SGI bought itself into the supercomputing space with the cray acquisition, but their sales reps didn't know what to sell... T3, vector, or Origin. It bled the company pretty badly.
Nobody argues that right now, there are some things for which there simply isn't any other rational choice besides SGI. In the early 90s, that was "anything with video, at all". Look how that market has all but vanished for them. The problem is the number of markets for which SGI is the only choice is shrinking and will continue to shrink. Only the institutions that need to be 1-3 years ahead of the curve will pay the huge markup for it. The big advantage of the O3k system is, as you ponit out, the single-system image. But this is only really advantageous for lazy programmers, and when you're talking 3m for a machine to do scientific or simulatino work, i suspect a lot of the code running on these is very custom, and NOT done by lazy programmers. So the brilliant thinking SGI has put into the hardware can sometimes be beaten by domain-specific software. Eg, lets say that MOSIX and 10Gig ethernet advances to the point that you can build a 1024p 512 node cluster, where the backbone (10Gb ethernet) is constructed in the same hypercube fabric as the numalink cables, and MOSIX can with software emulate the memory/process/thread migration that O3k is doing now....
then will 2.9m for a machine still seem justified ?... a 512 node wintel cluster is cheaper than 2.9m if the node cost is under about 5500. How many x86 boxes do you know of that cost 5500.. even with 2 procs, a few gb of ram, and 4 or 5 10GB ethernet controllers (so that each node is n-way connected in the same hypercube fabric that O2k and o3k provide)
RDP ("remote desktop protocol" - windows terminal services) is significantly more bandwidth-efficient than both X protocol and VNC.
Do this comparison - run an xterm remotely over a 28.8 PPP link. Now run it again through ssh -C (ssh with compression) over the exact same link. See how much faster that is ? Doesn't that tell you something when an unoptimized stream cipher can significantly cut the cruft out of X protocol in real time while not impacting latency ?
OpenGL doesn't necessarily work well over X, either, as the xserver has to have the right OGL extension with the right openGL extensions. And thats only if you're using a well behaved OGL app. Nobody can remotely use the old SGI apps that used the dgl protocol.
1. You can do this by writing a 12 line VB app that embeds the MSHTML COM control on separate tab controls. Some projects already do this. (Yawn)
5. uh, hit ctrl-H in IE6
7,8. Hold control, scroll mouse-wheel
17. IE does this
22. This can be set in IE
31. IE can do this
46. Is this a joke ?
77. I don't buy this. IE is a ship-component of Windows XP, and thus exists in 25 distinct locales.
97. This is just fanboyism. There is no substance here.
101. Got me there, champ.
These are just the things I know are crap off the top of my _head_. Why does fanboy shit like this make it to slashdot on such a consistant basis ?
Re:Making Linux A REAL Alternative
on
Halloween VII
·
· Score: 2
You're right. The alleged document seems to indicate that it's a dead heat between "low TCO" and "ABM".
Internally, theres an acronym called "ABM", meaning "Anything But Microsoft". ABM is the attitude most slashdotters seem to have, and it is a very real problem that MS faces when talking to some customers.
This is interesting because Low TCO is an objective, measurable, business related reason. ABM however is a bit more of an emotional reaction that may or may not make any business sense, and may or may not be the result of due dilligence work.
So of the people surveyed that liked OSS, objectivity and irrationality were about neck and neck. I find that interesting, and honestly, not surprising.
Incidentally, ESR got the letter and the meaning of acronym "BDM" wrong. It's unfortuneate that so many people pay him any creedence, because he's a bit of an over-the-top blowhard.
gcc is the compiler people on linux use because its available, not because its the best.
gcc on MIPS-IRIX is just awful. gcc is the least common denominator in terms of performance and just as bad as the others w.r.t. compilerisms and peculiarties. it just so happens that if you ported your code to gcc, it would _compile_ everywhere and run in a degraded state on non-linux-x86 platforms.
so to review: On Solaris - the sun compiler smokes gcc for c, c++, and fortran code
On IRIX - the SGI C++ compiler is almost a reference for how a good C++ compiler should be done. Oh yeah, its code generation is ideal on MIPS architectures (big surprise). can gcc even emit MIPS4 code yet ?
On Windows - well, msvc is a pretty performant compiler.
microsoft's political lobbying here was essentially non-existant until various legal entities started persuing them.
suddenly when a bunch of fame-seeking money hoarding politicians are asking for a peice of your pie, it looks worthwhile to give them some crumbs to keep them from taking a whole slice.
Well, im not saying openbsd will make inroads into the server market, but regardless, i dont feel that lack of SMP is confining openBSD.
I beleive the largest commercial deployment of *bsd boxes is Yahoo, using freebsd, and i seem to recall most of those are uni-proc machines (but i could be wrong on that)
In any case, like i said, for the type of apps that exist on openbsd _today_, its cheaper to just get multiple uniproc boxes and distribute the load (either manually with a junior admin babysitting, or automagically with some clever automation).
it is highly unlikely that pf, for instance, would get faster with SMP, as that would be an extensive re-gutting of the bsd kernel to make it highly re-entrant (i.e. interrupts being serviced by all cpus, and so on)
look how long its taken linux to try and not be a joke in the smp world. look at the system call latency that solaris and NT take on because of their highly SMP-friendly architectures..
you're right, i shouldn't presume who's done what. I was merely pointing out that i _like_ SMP and i _like_ MT programming, but in the case of OpenBSD, i feel it adds nothing but complicates much. I'd prefer to have a mature SMP implementation (solaris) running whatever MT apps i have behing a mature security product (openbsd)
why are you saying probably ? if you know im wrong, say why. if you dont and you're guessing, don't say anything..
I owned 3 different sun machines before i left high school. There is an SGI sitting on the desk next to me, and a dual proc sparc at home.
preferences are preferences, numbers are another. If you cared, i could tell you _why_ the old sparcs were good at some things (MMU hardware contexts, and register windows were early optimizations that x86 didn't have for instance..)
sun simply hasn't kept up with ChipZilla. Economies of scale have made wintel what it is.. that means that sexier products are marginalized... but the average person gets more computing power for under $1k than could be had at any price just 10 years ago..
point 1 came straight from theo's email a couple years ago
re: point 2: apache is fork/exec only until you get the hybrid MPM for apache 2.0 (which doesn't work)
I'm a huge advocate of pthreads - and i suspect i've written more high performance MT apps than you have. But SMP+a solid user space threads implementation on top of SMP is hard to do, and will complicate openBSD needlessly, given the sort of work its usually employed in..
Can you say 512 cpus in a single system image ? Can you say clustering multiple 512cpu nodes ?
Sun isn't even close. Never has been. The E10k and E15k are basically iterative improvements over the old Cray CS6400. A 10 year old system.
And im not making this shit up. go look at speccpu and specfp scores. Modern athlon and p4s smoke any sparc chip in integer scores. And the itanium 733 smoked it in floating point.
Not all companies are laying people off. Microsoft, for instance, hasn't done any layoffs. Microsoft also isn't playing any short-term quarterly financials games. Fucking yourself for short term gains is what caused the.com fallout to begin with. Sun is killing their future by destroying investments in intellectual capital.
1) It makes security that much harder. Think/tmp race conditions are bad ? How about race conditions in the kernel ? How about the fact that not even Intel is consistent in their docs on how two x86 chips re-order operations and maintain cache coherence in some situations.
2) 99% of the software on openBSD is fork/exec anyway. You might as well use assymmetric multi-processing, or, better yet, buy 3 uni-proc boxes for the price of a dual proc box, and partition your load accordingly.
while i agree that its a shame stock price doesn't necessarily affect quality of products...
i think its pretty clear that the entire tech sector was _Extremely_ over valued. SUNW lost a lot of ground not because of their bad products, but because of the bad investors that inflated the price so damn high.
Realistically, anyone that had any sort of experience with the markets would have started selling sun short about a year ago. And they'd be filthy rich right now. With double and triple digit P/E ratios, the entire tech sector is a joke.
Look at MSFT - no debt, $40b in cash, record revenues all through an economic downturn (and the tech sector crash). Still, the price has been floating around the same level for about 2 years. Granted, after 1:30PST today there may be a big change in MSFT, but even so, the financials at MS by all indications show that it is one of the strongest companies there is.
Yet its share price has stagnated.
Right now, so much price has dropped out of the tech sector, that all that remains is for the _value_ of those stocks to catch up. Then the stock prices will rise again, and do so at a normal rate (backed by people that know what the fuck they're doing, not idiot day-traders)
Incidentally, SUNW's product offerings are a mixed bag. They cant decide wether they're a hardware or a software company publicly (What is SunONE ? Has anyone seen it ? Why were they telling MSFT that Software didn't matter 1 year ago ?) Also, as nice as the ultrasparc III is, its only nice compared to other sparcs. x86 chips are doing quite a bit better, and itanium crushes it for FP work. SGI systems scale better than Suns, pretty much everyone makes a workstation that outperforms them at the bottom end. IBM's entire line of products is in roughly all ways superior to the equivalent (where they exist) offerings from Sun.
Sun's biggest asset right now is its installed base, and as we've seen, linux/x86 is eroding that because for many things, the port is almost free, and because the hardware is soooo much cheaper (absolute and price/perf)
Sun was the poster child of the.com era. And now they're paying for it.
Sun is not a healthy or a prosperous company. They're laying people off, forcing vacations, etc etc. This isn't because of their stock price. Its because people aren't buying as much sun gear, because it costs too damn much money, and because everyone that bought any in the last 2 years has auctioned it off as they went under, so the market is flooded.
i think thats perfectly fine. Aren't i allowed to hate asians and blacks ? It's not like im shooting them, harming them, or even being impolite to them. But i have the right, if i want to, to choose to be so ignorant as to disqualify a person based solely on skin color as a candidate for friendship, and acquantance
Why should i be forced to do business with someone i dont like ?
Governments that legislate thought do two things
1) alienate some fraction of the citizenry 2) dont change anybodies mind about anything
I think in many places the restrictions placed on business owners are ridiculous. It's _my_ business, i should determine the parameters of how it operates. The market will be my judge. Money will change my mind long before a law will.
Disclaimer: I hope that by virtue of being able to drive a mouse to the "reply" button, you understand that i myself dont hate any races, but i certainly think i should be able to do so, and i think its ok for others to do so.
I'm not a racist, (any more so than anyone else, at least) but I'm anti-PC, im anti-big government, and im pro individual choice, both for humans and for businesses.
It's called price gouging. Perhaps you've heard of it. It's morally repugnant. There is a difference between making a tidy profit, and gouging the consumer.
Your morals have two interesting properties
they're bullshit
they're yours
vote with your dollars. Let the market do its job. If people buy the games at price x, whats the point in charging x ? Why does the EU get a say in how items are priced ?
See, Goku would spend 20 minutes "getting his power up" and making this huge constipated sound with an even more constipated face. Rocks and energy spheres would be flying and exploding all around him, but the power rangers wouldn't so much as get dust in their eye standing 20 feet away.
the power rangers would start hanging curtains all over the place.
Slashdot Mirror
if MS didn't want anyone to debug windows, then
1) would the kernel debugger be a free download ?
2) would the kernel symbols (for use with said debugger) be a free download
furthermore, if disassembly were not allowed:
1) would all versions of Visual Studio allow you to view assmebly dumps ?
i think that your statement might be valid, with some qualification. but as presented, i think its incorrect.
you can't be sure.
but then, you can't be sure of anyone elses software, either. The advantage of OSS in this facet has been debunked time and time again. Thousands of people that depend on sendmail, bind, tcpdump, libpcap, etc. The source is there for them to look at it. They look at it. They dont find anything because they're not looking hard enough or they're not qualified to do security analysis of software.
Using publicly available tools you can single-step execution of the Windows operating system. You can get the names of all the symbols in the kernel. You can set kernel mode breakpoints on any peice of the network (and rpc/app) stack you want. It's not the same as having the source, but you can isolate exactly what the OS is doing at any time if you want to. And you can just disassemble the body of any function you like, once you've isolated it (which is easy, since Microsoft publishes full symbol information).
An experiment i did in about 30 minutes of poking around (im a novice at kernel debugging) had me to a point where i could make a machine drop to a kernel mode debugger when someone connected to a SMB share on that machine. I could then examine the file they were looking at, what operation they were trying to do, etc etc.
Naturally, this can all be automated. You could write a _kernel mode_ SMB debugger if you wanted to, and were worried about the CIFS implementation.
I'm not even sure if debugging tools like that for linux _even exist_.
And it's a moot point anyway. The overwhelming majority of people that think source availability is the difference between secure and insecure software aren't qualified to look at the source of either side, much less make objective measurements of the security/correctness aspects of the code in question.
I am highly dubious that the person quoted here is smart enough to write any kind of a firewall, much less a ruleset for linux or Windows. ... which is all secondary to the point i am going to make:
In W2k and later, the entire network stack is completely pluggable. You can insert any layer you want to that sits between NDIS and a protocol driver, and you can create other layers as required. I'd be very surprised if they couldn't do everything required with windows exactly as it sits today.
I think this is just making political noise, and not based on any shred of technical accuracy.
this is _almost_ not worth responding to, however, i'm going to anyway.
/ po well_c.htm
g i? p=http%3A//www.historychannel.com/exhibits/blackhi st/0228.html
http://www.state.gov/r/pa/ei/biog/1349.htm
http://www.galegroup.com/free_resources/bhm/bio
http://www.historychannel.com/cgi-bin/frameit.c
An excerpt: "He was awarded two Purple Hearts, a Bronze Star, a Soldier's Medal, and the Legion of Merit for his exemplary service in Vietnam"
Let me just summarize for you:
Colin Powell kicks more ass before breakfast than you do all day. I suspect he's stood in the face of more suffering, death, and outright fear than you'll ever know in your entire shit-eating miserable existance. And instead of thanking him for being one of the most distinguished soldiers, politicians, and leaders _ever_, you come up with this lame trite slashdot comment ?
I am all for discourse and critical examination of the american government. It's something that makes america what it is.
However, baseless, senseless, insultation of the people that fight (and often die) for your right to say whatever pissant things your 12 neurons manage to fizzle-off and deliver out of your mouth (or keyboard), is not only embarassing, but makes me wonder what it takes to wet YOUR pants ?
As an indigo^2 hi-impact owner (r4.4k, unfortuneately)...
the I^2 is a lame machine unless you've got impact gfx, at which point its gfx subsystem becomes ok. The phobos ethernet cards in these things cripple the poor EISA bus and slam the cpu. Playing midi on the things is choppy..
in any case, i chose to respond because of the 3.2Gb/s figure. when you say "internal bandwidth", what are you referring to ? The main memory bandwidth on any DDR system is actually about that fast these days. The O2 doesn't have the XIO bus of the octane, and its mostly irrelevant because pc, o2, and Octane all use PCI for the majority of the expansion devices anyhow.
For a while the memory bandwidth of pC's was a big limiting factor, but its actually gotten pretty good lately, especially compared to the desktop stuff of the traditional big iron players (sgi, sun, rs6k).
i concede having my history and my names confused. good catch.
re: "my vision"
the NUMAlink is not magic, it is a very specific high performance link with low latency a goal. Eventually, "off the shelf" stuff will be performance competitive, and at a lower cost.
I realize that using hardware not optimized for low latency in place of the FLASH-derivative chips that the router boards use will hurt. I'm referring more to being able to emulate in hosted software the illusion that the NUMA layer creates - that you have a huge pool of local memory, when infact you have a number of discrete pools of physically separate memory, and some slick hardware and software that makes them "seem" local unless you happen to have an app which breaks the implemented algorithms.
My point is that the hardware/software hybrid solution doesn't work magic, its very fast custom hardware with some software that makes it plausible to NOT _have_ to treat what is essentially a cluster like a cluster. There is little inherent advantage to having a single-image OS if you can write a library that your user space HPC apps will use that provides for single-memory and thread/data migration across the nodes in a performant way. And i think cheaper non-custom hardware and intelligent software can bring that layer to traditional clusters.
I'm a huge SGI fan, and i respect the O3k for what it is. (and i respected the O2k and O200 before it). I've read the white papers, even the DASH/FLASH ones from stanford.
Today, the O3k lets many apps scale up to a high cpu count and a large memory size, so that naive implementations can tackle larger problems. In the future, i claim that software and faster interconnect hardware (be it some new ethernet or myrinet or something similar) will let you do the same on commodity hardware.
im saying that it is possible to fake single-image in software just the way O3k fakes it in hardware.
its called "Numa" for a reason - the memory access is non-uniform, and there are perf penalties for thinking that all memory accesses are created equal. If you have an app where the hardware/software implementation of the page migrator is a performance blocker, then you'll need to do some tuning work to bring your app up to speed. And at that ponit, you're smart enough to write an app-specific best-case distributed memory management system (ala a much improved MOSIX, which is why i mentioned it)
The advantage of the O3k is that to many apps, it looks like a single image machine, and it "is", but it doesn't let you always make the same assumptions as a single image machine.
They're not out of the woods by any means.
History speaks pretty clearly about what happens to companies that marginalize their business into making 1-offs for infinite-budget DoD contracts and agencies. Eventually, projects get cancelled, line items in budgets get axed, and whole departments are re-orged into something different.
Cray, anyone ? Cray-Research basically went under when the Cray-3 contract was axed. They were counting on that single-machine to keep the afloat. They futzed around with GaAs custom process and never got it qutie working right, and then the cold war ended and with it the justification for subsidizing a maker of 1-off supercomputers.
(Incidentally, the purchase of Cray is what really broke SGI's back. 50% more employees, 2% more market cap, and the O2k/O3k technology came from stanford, not Cray) SGI bought itself into the supercomputing space with the cray acquisition, but their sales reps didn't know what to sell... T3, vector, or Origin. It bled the company pretty badly.
Nobody argues that right now, there are some things for which there simply isn't any other rational choice besides SGI. In the early 90s, that was "anything with video, at all". Look how that market has all but vanished for them.
The problem is the number of markets for which SGI is the only choice is shrinking and will continue to shrink. Only the institutions that need to be 1-3 years ahead of the curve will pay the huge markup for it. The big advantage of the O3k system is, as you ponit out, the single-system image. But this is only really advantageous for lazy programmers, and when you're talking 3m for a machine to do scientific or simulatino work, i suspect a lot of the code running on these is very custom, and NOT done by lazy programmers. So the brilliant thinking SGI has put into the hardware can sometimes be beaten by domain-specific software. Eg, lets say that MOSIX and 10Gig ethernet advances to the point that you can build a 1024p 512 node cluster, where the backbone (10Gb ethernet) is constructed in the same hypercube fabric as the numalink cables, and MOSIX can with software emulate the memory/process/thread migration that O3k is doing now....
then will 2.9m for a machine still seem justified ?... a 512 node wintel cluster is cheaper than 2.9m if the node cost is under about 5500. How many x86 boxes do you know of that cost 5500.. even with 2 procs, a few gb of ram, and 4 or 5 10GB ethernet controllers (so that each node is n-way connected in the same hypercube fabric that O2k and o3k provide)
nice bit of bullshit there at the end.
RDP ("remote desktop protocol" - windows terminal services) is significantly more bandwidth-efficient than both X protocol and VNC.
Do this comparison - run an xterm remotely over a 28.8 PPP link. Now run it again through ssh -C (ssh with compression) over the exact same link. See how much faster that is ? Doesn't that tell you something when an unoptimized stream cipher can significantly cut the cruft out of X protocol in real time while not impacting latency ?
OpenGL doesn't necessarily work well over X, either, as the xserver has to have the right OGL extension with the right openGL extensions. And thats only if you're using a well behaved OGL app. Nobody can remotely use the old SGI apps that used the dgl protocol.
1. You can do this by writing a 12 line VB app that embeds the MSHTML COM control on separate tab controls. Some projects already do this. (Yawn)
5. uh, hit ctrl-H in IE6
7,8. Hold control, scroll mouse-wheel
17. IE does this
22. This can be set in IE
31. IE can do this
46. Is this a joke ?
77. I don't buy this. IE is a ship-component of Windows XP, and thus exists in 25 distinct locales.
97. This is just fanboyism. There is no substance here.
101. Got me there, champ.
These are just the things I know are crap off the top of my _head_. Why does fanboy shit like this make it to slashdot on such a consistant basis ?
You're right. The alleged document seems to indicate that it's a dead heat between "low TCO" and "ABM".
Internally, theres an acronym called "ABM", meaning "Anything But Microsoft". ABM is the attitude most slashdotters seem to have, and it is a very real problem that MS faces when talking to some customers.
This is interesting because Low TCO is an objective, measurable, business related reason. ABM however is a bit more of an emotional reaction that may or may not make any business sense, and may or may not be the result of due dilligence work.
So of the people surveyed that liked OSS, objectivity and irrationality were about neck and neck. I find that interesting, and honestly, not surprising.
Incidentally, ESR got the letter and the meaning of acronym "BDM" wrong. It's unfortuneate that so many people pay him any creedence, because he's a bit of an over-the-top blowhard.
gcc is the compiler people on linux use because its available, not because its the best.
gcc on MIPS-IRIX is just awful. gcc is the least common denominator in terms of performance and just as bad as the others w.r.t. compilerisms and peculiarties. it just so happens that if you ported your code to gcc, it would _compile_ everywhere and run in a degraded state on non-linux-x86 platforms.
so to review:
On Solaris - the sun compiler smokes gcc for c, c++, and fortran code
On IRIX - the SGI C++ compiler is almost a reference for how a good C++ compiler should be done. Oh yeah, its code generation is ideal on MIPS architectures (big surprise). can gcc even emit MIPS4 code yet ?
On Windows - well, msvc is a pretty performant compiler.
microsoft's political lobbying here was essentially non-existant until various legal entities started persuing them.
suddenly when a bunch of fame-seeking money hoarding politicians are asking for a peice of your pie, it looks worthwhile to give them some crumbs to keep them from taking a whole slice.
Well, im not saying openbsd will make inroads into the server market, but regardless, i dont feel that lack of SMP is confining openBSD.
I beleive the largest commercial deployment of *bsd boxes is Yahoo, using freebsd, and i seem to recall most of those are uni-proc machines (but i could be wrong on that)
In any case, like i said, for the type of apps that exist on openbsd _today_, its cheaper to just get multiple uniproc boxes and distribute the load (either manually with a junior admin babysitting, or automagically with some clever automation).
it is highly unlikely that pf, for instance, would get faster with SMP, as that would be an extensive re-gutting of the bsd kernel to make it highly re-entrant (i.e. interrupts being serviced by all cpus, and so on)
look how long its taken linux to try and not be a joke in the smp world. look at the system call latency that solaris and NT take on because of their highly SMP-friendly architectures..
you're right, i shouldn't presume who's done what. I was merely pointing out that i _like_ SMP and i _like_ MT programming, but in the case of OpenBSD, i feel it adds nothing but complicates much. I'd prefer to have a mature SMP implementation (solaris) running whatever MT apps i have behing a mature security product (openbsd)
why are you saying probably ? if you know im wrong, say why. if you dont and you're guessing, don't say anything..
I owned 3 different sun machines before i left high school. There is an SGI sitting on the desk next to me, and a dual proc sparc at home.
preferences are preferences, numbers are another. If you cared, i could tell you _why_ the old sparcs were good at some things (MMU hardware contexts, and register windows were early optimizations that x86 didn't have for instance..)
sun simply hasn't kept up with ChipZilla. Economies of scale have made wintel what it is..
that means that sexier products are marginalized... but the average person gets more computing power for under $1k than could be had at any price just 10 years ago..
it gets modded up because its right
point 1 came straight from theo's email a couple years ago
re: point 2:
apache is fork/exec only until you get the hybrid MPM for apache 2.0 (which doesn't work)
I'm a huge advocate of pthreads - and i suspect i've written more high performance MT apps than you have. But SMP+a solid user space threads implementation on top of SMP is hard to do, and will complicate openBSD needlessly, given the sort of work its usually employed in..
Read up on the SGI Origin 3800.
.com fallout to begin with. Sun is killing their future by destroying investments in intellectual capital.
Can you say 512 cpus in a single system image ? Can you say clustering multiple 512cpu nodes ?
Sun isn't even close. Never has been. The E10k and E15k are basically iterative improvements over the old Cray CS6400. A 10 year old system.
And im not making this shit up. go look at speccpu and specfp scores. Modern athlon and p4s smoke any sparc chip in integer scores. And the itanium 733 smoked it in floating point.
Not all companies are laying people off. Microsoft, for instance, hasn't done any layoffs. Microsoft also isn't playing any short-term quarterly financials games. Fucking yourself for short term gains is what caused the
There's little reason for SMP in openbsd
/tmp race conditions are bad ? How about race conditions in the kernel ? How about the fact that not even Intel is consistent in their docs on how two x86 chips re-order operations and maintain cache coherence in some situations.
1) It makes security that much harder. Think
2) 99% of the software on openBSD is fork/exec anyway. You might as well use assymmetric multi-processing, or, better yet, buy 3 uni-proc boxes for the price of a dual proc box, and partition your load accordingly.
have you visited sun ? (or do you work there ?)
:)
otherwise, i wouldn't be so sure theres no reliance on MS software at Sun
any x86 today will either cost less, be faster, or, usually, both.
the reasons you want this are because
1) you need sparc/solaris machine(s)
2) you need one thats portable
while i agree that its a shame stock price doesn't necessarily affect quality of products...
.com era. And now they're paying for it.
i think its pretty clear that the entire tech sector was _Extremely_ over valued. SUNW lost a lot of ground not because of their bad products, but because of the bad investors that inflated the price so damn high.
Realistically, anyone that had any sort of experience with the markets would have started selling sun short about a year ago. And they'd be filthy rich right now. With double and triple digit P/E ratios, the entire tech sector is a joke.
Look at MSFT - no debt, $40b in cash, record revenues all through an economic downturn (and the tech sector crash). Still, the price has been floating around the same level for about 2 years. Granted, after 1:30PST today there may be a big change in MSFT, but even so, the financials at MS by all indications show that it is one of the strongest companies there is.
Yet its share price has stagnated.
Right now, so much price has dropped out of the tech sector, that all that remains is for the _value_ of those stocks to catch up. Then the stock prices will rise again, and do so at a normal rate (backed by people that know what the fuck they're doing, not idiot day-traders)
Incidentally, SUNW's product offerings are a mixed bag. They cant decide wether they're a hardware or a software company publicly (What is SunONE ? Has anyone seen it ? Why were they telling MSFT that Software didn't matter 1 year ago ?) Also, as nice as the ultrasparc III is, its only nice compared to other sparcs. x86 chips are doing quite a bit better, and itanium crushes it for FP work. SGI systems scale better than Suns, pretty much everyone makes a workstation that outperforms them at the bottom end. IBM's entire line of products is in roughly all ways superior to the equivalent (where they exist) offerings from Sun.
Sun's biggest asset right now is its installed base, and as we've seen, linux/x86 is eroding that because for many things, the port is almost free, and because the hardware is soooo much cheaper (absolute and price/perf)
Sun was the poster child of the
Sun is not a healthy or a prosperous company. They're laying people off, forcing vacations, etc etc. This isn't because of their stock price. Its because people aren't buying as much sun gear, because it costs too damn much money, and because everyone that bought any in the last 2 years has auctioned it off as they went under, so the market is flooded.
YOu are wrong.
There has been a UNIX layout and a PC layout keyboard in type4, type5, type5c, and presumably type6.
All my sparc keyboards are UNIX layout - the way god intended.
and my w2k / xp machines at work ? They run ctrl2cap - from sysinternals.com
The battle hasn't been lost. The weak have given up.
i think thats perfectly fine. Aren't i allowed to hate asians and blacks ? It's not like im shooting them, harming them, or even being impolite to them. But i have the right, if i want to, to choose to be so ignorant as to disqualify a person based solely on skin color as a candidate for friendship, and acquantance
Why should i be forced to do business with someone i dont like ?
Governments that legislate thought do two things
1) alienate some fraction of the citizenry
2) dont change anybodies mind about anything
I think in many places the restrictions placed on business owners are ridiculous. It's _my_ business, i should determine the parameters of how it operates. The market will be my judge. Money will change my mind long before a law will.
Disclaimer: I hope that by virtue of being able to drive a mouse to the "reply" button, you understand that i myself dont hate any races, but i certainly think i should be able to do so, and i think its ok for others to do so.
I'm not a racist, (any more so than anyone else, at least) but I'm anti-PC, im anti-big government, and im pro individual choice, both for humans and for businesses.
Your morals have two interesting properties
vote with your dollars. Let the market do its job. If people buy the games at price x, whats the point in charging x ? Why does the EU get a say in how items are priced ?
i think it'd be a fair fight.
See, Goku would spend 20 minutes "getting his power up" and making this huge constipated sound with an even more constipated face. Rocks and energy spheres would be flying and exploding all around him, but the power rangers wouldn't so much as get dust in their eye standing 20 feet away.
the power rangers would start hanging curtains all over the place.
The two sides would bore each other to death.