You may want to take a look at Borland's Kylix, the linux-native port of Borland's respocted IDE's for C, C++, and Delphi (they claim it supports the same features as windows versions). It comes with a high quality debugger/profiler, and, while not free, a trial version is available.
1) First of all, let us not confuse reliability (which is what database software, with ACID transactions, provides on any machine) with
high availability (which is what mainframes claim
to provide). The problem with spreading servers (and, consequently, data) around is managing data consistency. As others have shown (there have been lots of papers in SOSP, PODC, etc., about issues like consensus, consistency semantics, etc.), there is a tradeoff between availability, consistency, and partitioning/failures (you can only pick two of three). 2) Even to do this tradeoff in the ACID model cleanly, you need true wide-area online replication (as opposed to store-forward, reconcile, log-based replications), which, currently, no database offers (Postgress is planning one for the next major release, but doing it efficiently, and while preserving the ACID properties is very hard). 3) Even with this "back-end" facility, one must start worrying how to maintain
application-level (possibly, even session-level) connectivity semantics, which is another big problem.
Since apple has exclusive license, all it has to do is either create a codec for linux so that it can be used with mplayer or some such (note that mplayer can play qt files as long as the codec used is not sorensen), or open up the specs (or give them under NDA to interested some developers) so that linux codecs can be written without apple's expense.
Plot of News corp's (parent of fox) verus aol. Obviously, aol dropped in value recently, but it appears to me that, long-term, it is more successful than fox. The article contains inaccuracies as well---a company's stock price has only a light relationship with either a writeoff (which only represents the "overpaid" amount in stock) or its bookvalue. Therefore, statements like "
In other words, perhaps as much as $100 billion more could disappear
before the carnage is complete." are completely inaccurate and don't constitute "analysis".
Errr, "contributors shall have". That's any contributor. Not just Lucent. Which is exactly what the GPL provides, no?
Nope! The GPL is structured in such a way that even the original contributor (Lucent, in this case) cannot mix a non-originator's code contribution into a closed product (now or in future), although Lucent is always free to license a closed version that does not include other contributors' code. However, this license implies Lucent is quite free to distribute a closed version of the original code+your modifications to third parties without requiring these third parties, as part of a commercial license for example, to release further contributions and/or modifications to your own code. Of course, parties other than lucent cannot do this because they cannot redistribute the original code under a commercial license, for example. So, the playing field is uneven---lucent can do whatever it wants to with your code, but you cannot.
"You agree to provide the Original Contributor, at its request, with a copy of the complete Source Code version, Object Code version and related documentation for Modifications created or
contributed to by You if distributed in any form, e.g., binary or source. Original Contributor and/or other Contributors shall have unrestricted, nonexclusive, worldwide, perpetual, royalty-free rights,
to use, reproduce, modify, display, perform, sublicense and distribute such Modifications, and to grant third parties the right to do so, including without limitation as a part of or with the Licensed
Software; and Original Contributor and/or other Contributors shall have the right to license or to otherwise transfer to third parties such Modifications without notice, obligation or recourse to You.
You grant to Original Contributor, Contributors and their respective licensees all rights and licenses (including patents) as are necessary to incorporate the Modifications created or contributed
and so distributed by You into the Licensed Software and to use, distribute or otherwise exploit such Licensed Software without payment or accounting to You."
is *very* troublesome because this means that lucent can go closed-source with your modifacations at some later point, and they are not obligated to "pay" your or the community back even if they use your code in proprietary licenses in future. This is the basic problem with asymmetric licenses like Plan9's, and NPL.
Wrong! From this article, msft has about $5.3b in cash and cash equivalents (that can be quickly liquidated), and, "using a slightly more
expansive methodology", about $36b. But, remeber that this includes items like various venture capital investments that a) cannot be liquidated easily, b) probably required to support its stock price and market cap(~$300b with P/E of 45, a relative high). In any case, while the article acknowledges that the $500m that msft spent on xbox marketing won't probably affect it much, the stock market does not necessarily applaud a product unless the investment is recouped somehow in a fairly short interval.
Even though both NT and linux are POSIX compliant, there are enough quirks in the implementations, especially with regard to multi-threading libraries. As long as you use C or C++ (or any language that does not provide both a rich threading interface and good runtime support), consider using the NSPR libraries that are meant to provide a rich set of cross-platform interfaces.
IBM and Hitachi are *merging* their disk business so that IBM gets a 30% stake (and Hitachi, 70%). The story's comment "They plan to sell 70% of the their HD business to Hitachi." seems incorrect to me; IBM is simply estimating that its current disk business is worth 30% of the joint disk business. Also, note that Hitachi has a very strong storage systems business HDS (right behind EMC) that is very profitable (also resold by SUN as Storedge9900 series datacenter/enterprise storage products, I believe), so big blue may have merged their disk business with a view to ensuring future profitability in the overall storage space.
since when did security and privacy cease to be relevant? The previous user was touting flash benefits like easy and powerful programming, and I am trying to make a point that this comes at a high cost not yet examined in detail. A simple example: flash allows a seemingly innocous ad to connect to third-party marketing sites to track the user's browsing habits, the browser configuration (and, for that matter, even what files are on the user's disk). Java, because of its careful namespace handling and "sandboxing" restrictions on mobile code, prohibits such behavior. There are any number of such examples which, even if not directly security-related, can invade user's privacy in the name of "ease".
Yes, but how secure is it? Java is arguably slow because it goes to great lengths to define and manage untrusted code. If any turing-complete language like flash is used (note that javascript uses a restricted function call model, and, therefore, is possibly provably secure), we typically want to see some guarantees. I find it surprising that folks haven't dug deeper into flash vulnerabilities (I am sure there are tons of them; even java, with its well-defined vm+runtime, class, and security definitions had security issues). Sun and others don't seem to push the original safe mobile-code benefits of java anymore probably because java is currently finding its niche in server-side environments where mobile code and byte-interpretor performance are probably not an issue (because one can always use jit's, can trust server's own code, etc.). Currently, flash seems to have found its niche as a replacement for animated gif's, and I am very scary about the security implications.
Does anyone else find it ironical that M$ wants 3-d hardware accelertion for 2-d software and graphics on one hand and soft-wifi for n/w deceleration on the other? Obviously, they are trying to once again encourage the closed specs/driver approach that many manufacturers took for win-modems with the unfortunate result that the hardware is unusable on a lot of non-ms platforms for a long time.
The jpeg2000 page that both links in the story refer to has not been updated in a while, as can be seen by going to the wayback machine page and typing in the url to get the revision history. The current page seems to have been last updated in October 2001. And this last edit seems little more than a book ad and background color change:) How is the statement "These free plug-in's are expected to be available later this year." backed up?
Here is an article that explains why this device may be nothing more than two simple overlaid workspaces but not true stereoscopic 3-D. In particular, it says in bold red:
"For Stereoscopic-3D you'll need special Stereo-3D software in any case, whether it's photography, film, tv, video or computer software. You will never get a real 3D experience out of standard material. There are products which claim to do this, especially pseudo3D-television devices, but those offerings are bogus! - You can't get 3D out of thin air."
From what I have previously heard about stereoscopic vision, and confirmed by what the article says, one needs two slightly different points-of-view of a 3-d object (or simulated points-of-view in case of flat images) for the brain to correctly synthesize the notion of depth.
That is why one typically uses glasses with accurately sync'ed shutters (so that one frame is delivered to one eye and the next frame to the other---there are any number of schematics available on the web to roll your own provided the display hardware/software can support this). Alternate techniques for generating stereo vision include polarization techniques, etc.
Re:Server share data for working sites
on
March Netcraft survey
·
· Score: 2, Interesting
Because of benchmarks like this?
(Note how, ignoring the hardware cost for a moment, the top-of-the-line 16-processor IBM pSeries machine running zeus supports 2.5x more users than the best-available 8-processor IIS server.)
Also, zeus (and, may be, netscape enterprise, etc.) is known to have better single-machine scalability because of serveral interesting I/O techniques it tends to use---these benefits are more pronounced when run on operating systems like solaris that support fine-grained user-level threads to kernel-level thread mappings. On top of the raw performance, many also support application-level clustering and redundancy (may be important for some portal sites that demand underlying data consistency, and, which, therefore, require more app-level work to scale-up/failover than just adding more server instances). However, for the vast majority of the sites out there that serve out mostly static and simple dynamic traffic, I think apache is more than sufficient (these sites tend to be bottlenecked by the n/w, not by the server), and I would pick apache anyday over IIS for simplicity, stability, and security reasons (even the humble tux server almost matches the best-available IIS5.0 on the same hardware in the benchmark above in terms of performance; there is no need to go into security/stability comparisons).
Server share data for working sites
on
March Netcraft survey
·
· Score: 5, Informative
This data for *active* web servers (about 6 million total) seems to give a different picture---while apache lost 0.16% and IIS gained 0.40%, long-term (over the last year) apache grew, while IIS fell. Also, extrapolated future failure and growth rates seems to indicate that one is better off betting on apache than on IIS.
The downstream channel would have to be multiplexed among all simultaneous users of the service, which means they can support roughly 30Mbps total modulated bit rate of the channel/256Kbps per user ~ 120 users...Even if we assume a 10x gain due to statistical packet multiplexing, that is still only 1000 users max. I wonder how they plan to deal with congestion...
-rkgmd
Slashdot Mirror
You may want to take a look at Borland's Kylix, the linux-native port of Borland's respocted IDE's for C, C++, and Delphi (they claim it supports the same features as windows versions). It comes with a high quality debugger/profiler, and, while not free, a trial version is available.
1) First of all, let us not confuse reliability (which is what database software, with ACID transactions, provides on any machine) with high availability (which is what mainframes claim to provide). The problem with spreading servers (and, consequently, data) around is managing data consistency. As others have shown (there have been lots of papers in SOSP, PODC, etc., about issues like consensus, consistency semantics, etc.), there is a tradeoff between availability, consistency, and partitioning/failures (you can only pick two of three). 2) Even to do this tradeoff in the ACID model cleanly, you need true wide-area online replication (as opposed to store-forward, reconcile, log-based replications), which, currently, no database offers (Postgress is planning one for the next major release, but doing it efficiently, and while preserving the ACID properties is very hard). 3) Even with this "back-end" facility, one must start worrying how to maintain application-level (possibly, even session-level) connectivity semantics, which is another big problem.
Since apple has exclusive license, all it has to do is either create a codec for linux so that it can be used with mplayer or some such (note that mplayer can play qt files as long as the codec used is not sorensen), or open up the specs (or give them under NDA to interested some developers) so that linux codecs can be written without apple's expense.
Plot of News corp's (parent of fox) verus aol. Obviously, aol dropped in value recently, but it appears to me that, long-term, it is more successful than fox. The article contains inaccuracies as well---a company's stock price has only a light relationship with either a writeoff (which only represents the "overpaid" amount in stock) or its bookvalue. Therefore, statements like " In other words, perhaps as much as $100 billion more could disappear before the carnage is complete." are completely inaccurate and don't constitute "analysis".
Errr, "contributors shall have". That's any contributor. Not just Lucent. Which is exactly what the GPL provides, no? Nope! The GPL is structured in such a way that even the original contributor (Lucent, in this case) cannot mix a non-originator's code contribution into a closed product (now or in future), although Lucent is always free to license a closed version that does not include other contributors' code. However, this license implies Lucent is quite free to distribute a closed version of the original code+your modifications to third parties without requiring these third parties, as part of a commercial license for example, to release further contributions and/or modifications to your own code. Of course, parties other than lucent cannot do this because they cannot redistribute the original code under a commercial license, for example. So, the playing field is uneven---lucent can do whatever it wants to with your code, but you cannot.
"You agree to provide the Original Contributor, at its request, with a copy of the complete Source Code version, Object Code version and related documentation for Modifications created or contributed to by You if distributed in any form, e.g., binary or source. Original Contributor and/or other Contributors shall have unrestricted, nonexclusive, worldwide, perpetual, royalty-free rights, to use, reproduce, modify, display, perform, sublicense and distribute such Modifications, and to grant third parties the right to do so, including without limitation as a part of or with the Licensed Software; and Original Contributor and/or other Contributors shall have the right to license or to otherwise transfer to third parties such Modifications without notice, obligation or recourse to You. You grant to Original Contributor, Contributors and their respective licensees all rights and licenses (including patents) as are necessary to incorporate the Modifications created or contributed and so distributed by You into the Licensed Software and to use, distribute or otherwise exploit such Licensed Software without payment or accounting to You."
is *very* troublesome because this means that lucent can go closed-source with your modifacations at some later point, and they are not obligated to "pay" your or the community back even if they use your code in proprietary licenses in future. This is the basic problem with asymmetric licenses like Plan9's, and NPL.
Wrong! From this article, msft has about $5.3b in cash and cash equivalents (that can be quickly liquidated), and, "using a slightly more expansive methodology", about $36b. But, remeber that this includes items like various venture capital investments that a) cannot be liquidated easily, b) probably required to support its stock price and market cap(~$300b with P/E of 45, a relative high). In any case, while the article acknowledges that the $500m that msft spent on xbox marketing won't probably affect it much, the stock market does not necessarily applaud a product unless the investment is recouped somehow in a fairly short interval.
Here is a link to an article that says how winnebago successfully used bynari on IBM mainframes as an exchange substitute.
Even though both NT and linux are POSIX compliant, there are enough quirks in the implementations, especially with regard to multi-threading libraries. As long as you use C or C++ (or any language that does not provide both a rich threading interface and good runtime support), consider using the NSPR libraries that are meant to provide a rich set of cross-platform interfaces.
IBM and Hitachi are *merging* their disk business so that IBM gets a 30% stake (and Hitachi, 70%). The story's comment "They plan to sell 70% of the their HD business to Hitachi." seems incorrect to me; IBM is simply estimating that its current disk business is worth 30% of the joint disk business. Also, note that Hitachi has a very strong storage systems business HDS (right behind EMC) that is very profitable (also resold by SUN as Storedge9900 series datacenter/enterprise storage products, I believe), so big blue may have merged their disk business with a view to ensuring future profitability in the overall storage space.
since when did security and privacy cease to be relevant? The previous user was touting flash benefits like easy and powerful programming, and I am trying to make a point that this comes at a high cost not yet examined in detail. A simple example: flash allows a seemingly innocous ad to connect to third-party marketing sites to track the user's browsing habits, the browser configuration (and, for that matter, even what files are on the user's disk). Java, because of its careful namespace handling and "sandboxing" restrictions on mobile code, prohibits such behavior. There are any number of such examples which, even if not directly security-related, can invade user's privacy in the name of "ease".
Yes, but how secure is it? Java is arguably slow because it goes to great lengths to define and manage untrusted code. If any turing-complete language like flash is used (note that javascript uses a restricted function call model, and, therefore, is possibly provably secure), we typically want to see some guarantees. I find it surprising that folks haven't dug deeper into flash vulnerabilities (I am sure there are tons of them; even java, with its well-defined vm+runtime, class, and security definitions had security issues). Sun and others don't seem to push the original safe mobile-code benefits of java anymore probably because java is currently finding its niche in server-side environments where mobile code and byte-interpretor performance are probably not an issue (because one can always use jit's, can trust server's own code, etc.). Currently, flash seems to have found its niche as a replacement for animated gif's, and I am very scary about the security implications.
Does anyone else find it ironical that M$ wants 3-d hardware accelertion for 2-d software and graphics on one hand and soft-wifi for n/w deceleration on the other? Obviously, they are trying to once again encourage the closed specs/driver approach that many manufacturers took for win-modems with the unfortunate result that the hardware is unusable on a lot of non-ms platforms for a long time.
The jpeg2000 page that both links in the story refer to has not been updated in a while, as can be seen by going to the wayback machine page and typing in the url to get the revision history. The current page seems to have been last updated in October 2001. And this last edit seems little more than a book ad and background color change:) How is the statement "These free plug-in's are expected to be available later this year." backed up?
Here is an article that explains why this device may be nothing more than two simple overlaid workspaces but not true stereoscopic 3-D. In particular, it says in bold red: "For Stereoscopic-3D you'll need special Stereo-3D software in any case, whether it's photography, film, tv, video or computer software. You will never get a real 3D experience out of standard material. There are products which claim to do this, especially pseudo3D-television devices, but those offerings are bogus! - You can't get 3D out of thin air." From what I have previously heard about stereoscopic vision, and confirmed by what the article says, one needs two slightly different points-of-view of a 3-d object (or simulated points-of-view in case of flat images) for the brain to correctly synthesize the notion of depth. That is why one typically uses glasses with accurately sync'ed shutters (so that one frame is delivered to one eye and the next frame to the other---there are any number of schematics available on the web to roll your own provided the display hardware/software can support this). Alternate techniques for generating stereo vision include polarization techniques, etc.
Because of benchmarks like this? (Note how, ignoring the hardware cost for a moment, the top-of-the-line 16-processor IBM pSeries machine running zeus supports 2.5x more users than the best-available 8-processor IIS server.) Also, zeus (and, may be, netscape enterprise, etc.) is known to have better single-machine scalability because of serveral interesting I/O techniques it tends to use---these benefits are more pronounced when run on operating systems like solaris that support fine-grained user-level threads to kernel-level thread mappings. On top of the raw performance, many also support application-level clustering and redundancy (may be important for some portal sites that demand underlying data consistency, and, which, therefore, require more app-level work to scale-up/failover than just adding more server instances). However, for the vast majority of the sites out there that serve out mostly static and simple dynamic traffic, I think apache is more than sufficient (these sites tend to be bottlenecked by the n/w, not by the server), and I would pick apache anyday over IIS for simplicity, stability, and security reasons (even the humble tux server almost matches the best-available IIS5.0 on the same hardware in the benchmark above in terms of performance; there is no need to go into security/stability comparisons).
This data for *active* web servers (about 6 million total) seems to give a different picture---while apache lost 0.16% and IIS gained 0.40%, long-term (over the last year) apache grew, while IIS fell. Also, extrapolated future failure and growth rates seems to indicate that one is better off betting on apache than on IIS.
http://znet.net/~schester/facts/database_sizes.htm l
Apparently, walmart's is 24TB, and the entire www index as of 1999 was only 6TB.
The downstream channel would have to be multiplexed among all simultaneous users of the service, which means they can support roughly 30Mbps total modulated bit rate of the channel/256Kbps per user ~ 120 users...Even if we assume a 10x gain due to statistical packet multiplexing, that is still only 1000 users max. I wonder how they plan to deal with congestion... -rkgmd