If I remember correctly, the speedup of turning on this numa-awareness on an opteron multiprocessor system is negligible even on benchmarks which should stress it. It's a neat feature but doesn't appearantly really help the opteron (partially because the HT bus is very fast and the number of CPU's involved isn't so great).
This isn't news about a point release - I've never heard of VCF before this so to me this is actually new. And it's actually vaguely interesting - as in something to consider using as opposed to many other announcements on slashdot. Obviously a disclaimer would be nice, but it's interesting enough to warrant some attention, and for me therefore leniency:-).
And - nobody's reading slashdot with the expectation of well-filtered news anyhow.
A hypocrite is a person who professes beliefs and opinions that he does not hold. I don't believe that's the case here... rather; it's a bit rich to here about media consilidation evils coming from such a man - But the more Ted Turner represents the very media consolidation he now says are so detrimental, the more you should take his arguments seriously. I can hardly image a more appropriate preacher for this gospel:-).
--Eamon
The article you're referring to is probably one of the worst hardware-review/benchmark articles in a long long time. Taking a game engine which was so obviously immature and pretending that the benchmark results it provides have any real meaning whatsoever isn't very honest - especially considering that ATI was partnered with valve at the time. HL2 was at the time the first game to really use such new features as well meaning that it's very very likely that future driver releases would have changed the picture dramatically.
If you want to by ATI that's fine. Frankly that's not a good idea: not because nVidia's cards are so much better but because you should look at the cards individually. For example, the high-end NV30 cards (5800 and related) performed badly. Don't buy those then! However it's also true that these new nVidia cards (esp. the 6800 GT) price/perform very nicely all around. Slightly lower end, the 9800Pro still looks very good for all games currently on the market; including doom 3 and far cry...
Anyways - that article suggests conclusions that, based on the facts it presents, are completely off base.
Not that it's in any livable price-range; but two 30" displays connected to a card like that running MacOS-X, can you image that? The beauty, the beauty...;-)
I don't believe that source code theft is really such a problems for such companies - I really really doubt microsoft would use much of altavista's code even if they legally could! (It's so unbelievably much work to figure out someone elses mature code....)
However, employee education leakage is far more important. The raison d'etre for some of those architectural choices, or experiences with certain emergent pattern in large scale systems, and similarly complex issues are very, very valuable.
So really - feel sorry for microsoft... this just gives them bad PR, potentially opens them up for lawsuits (however unfounded), and generally doesn't do them any good..
I wonder where the claim of 5000 dollars damage comes from? The article says he claims he was curious about the progression of the product (which honestly, however illegal, I sympathize with - you put so much of yourself in these systems and then all of a sudden you're not allowed to know anything about them... arg!), so maybe it's all just much ado about nothing.
Why would social and financial systems be safe even if they took "default user behaviour" into account? And safe for whom?
I don't think things are quite that simple:-)
Back on topic; The ethics of microsoft's (comprehensibly lacking) user education are further muddied by their being a monopoly...
Requiring a goody-goody warm and fuzzy company is obviously neither realistic nor particularly useful. However, with power comes responsibility; and microsoft, being a monopoly, certainly has power in the market, whereby some responsibility for that market is assumed.
Having said that, at a more emotional level I reject this deplorable reduction of such issues to mere market mechanics and capitalism (oh how we pray to the yadda yadda yadda). I just want to live in a nice society - and having big bullies around which rationalize their actions (i.e. distributing IE with windows) by essentially saying "just because we can", is pretty damn lame.
Maybe it's fair that the schoolyard hulk that has trained and learned how to fight well always beats other people up. Fair in some darwinian sense... certainly it's not ideal for me!
I don't exactly know which spikes the parent poster is referring too; but I'm quite sure they're not the spread-spectrum spikes. Those are frequency (not temporal) spikes and can increase EMI. Spread spectrum reduces the sharpness of the frequency spike by modulating the FSB - reducing, not increasing the stability.
All spikes are equal - but some are more equal than others;-)
But I think it would be even nicer if the time-stamp were less prominent (for example by placing it behind as opposed to in front of the title using a z-index).
I especially like the way the underline on hover breaks across lines:-).
The rendering differences might be solved by using the "Standard Compliance mode" in IE and gecko, using a valid xhtml doctype is the simplest way of achieving that but...
http://gutfeldt.ch/matthias/articles/doctypeswit ch.html...gives an in depth overview. The standards mode generally works more consistantly and predicably especially with complicated effects, and you get a bit of cross-browser advantages for free:-).
However... Just authenticating a site is really important! It's more likely that you'll be fooled by a hoaxed web site than that someone manages to sniff your packets or man-in-the-middle you. I don't know about most of you; but I'm on a reasonably trusted network (i.e. I know there aren't any sniffers on the LAN and once the connection's being routed you'ld have to sniff all kinds of gunk and manage to somehow, transparently get into that connection between the various routers - much harder than on a lan.)
So if I had to choose between authentication and encryption - well; I know what I would choose...
I run gentoo; my suggestion is just not to worry about it... I've compiled with some rather experimental flags in my time and they often cause problems in things like xine or mplayer or whatnot; and the speed difference isn't that great... (not perceivably so anyhow)
Anyhow; just use stage3 if you feel like it - you can always recompile portions with new CFLAGS by
emerge --oneshot [my-package-here]
I run ~x86 and for me the real advantage of portage is the access to up-to-date packages.
Some people are really snobbish about this; pretending that updating other systems is even remotely as easy (it's not) or as up-to-date (witness debian)
With so many developments happening in linux-land is nice to be able to actually get those features.
For example kde 3.2 (in my opinion full of real world improvements over 3.1.x) was available for gentoo right away (the same day...) - not so for other distros. Also gentoo has ebuilds for pretty much anything under the sun... almost all programs you'll want to install are emerge-able.
Gentoo may not be so bug-free; it may not be as fast (at installing); but it sure as hell has a great forum and bug-support system. It's really a great way to get help and to post some help to others that suffer from the same problems.
So it has costs (more hassle to set up) but benefits too (more up-to-date, great support, good documentation).
I.e. don't worry about the compiling from source bit... that's not even so important.
This discussion obviously boils down to quasi-religious beliefs; so lets not get too carried away with things (not that you did, I just hate flamewars).
To take your example about the get-away car and the robbery-plans, I think it wouldn't harm to institute a rule allowing the government a lot of leeway in such investigations; however simultaneously guaranteeing that the government provides for replacement (or more likely: remuneration) for taken goods including lost business/opportunities.
Why?
To me it seems that the issue isn't so much guilt; rather such searches really do cause costs to the people involved. By making them explicit; they can better be taken account of in policymaking. Also, this kind of thing causes costs that aren't so transparent; namely risk costs: even people but especially corporations never effected by such a large and serious warrant might spend time and effort ( = money, money) to plan for such an event, a cost to society that the government really doesn't need to encourage!
Furthermore, I think if people weren't so frightened of these things (and if there wouldn't be any grevious losses I don't think people would be), perhaps more friendly cooperation would be encouraged. For instance; if I saw reasonable evidence of a computer crime spread over machines on a network I'm a stakeholder in (say as a user or administrator), then I'ld think twice before informing any law enforcement given the possibilty of such serious personal costs.
The above are just my personal beliefs... I'm sure you can find fault somewhere;-).
Sure AGP is a sort of bottleneck - similarly to how Ultra-ATA133 can be a bottleneck and serial ATA can perform better - it's not just bandwidth that matters.
In PCI-Express's case (I haven't read up on this for months so it's not so fresh) some of the improved things are cache-control, which can now be driver controlled, error detection and handling (PCI doesn't have this, can be an annoying cause for hidden data-corruption), p2p in the sense that devices can send info directly to each other... Another important issue is buffer management and relaxed ordering and just general smartness and better QoS / data-type-awareness (this should improve effective streaming bandwidth, and improve choices in case bandwidth is low, etc.)
PCI Express is definitely a step forwards - however, that it will really matter too much for video cards initially, given that the drivers will need to be adapted make use of the new features, is unlikely.
There were some figures on power consumption that suggested Prescott runs quite a bit warmer than Northwood and the same clockspeed.
That's not really surprising either: A feature-size shrink (such as the move now from 130nm to 90nm) lowers the power per transistor, however, it increases the power per unit area (IIRC proportionally to the sqrt of the shrink).
Of course - a different processor layout could completely counteract these effects, and a change to 31 stages instead of 21 stages would qualify as such I suppose:-).
http://www.aceshardware.com/read.jsp?id=60000317 does list some power data, and rates a 3.2 GHz prescott as having a 103W _typical_ power consumption...
Conclusion the prescott has a power consumption problem; hopefully later revisions will fix this...
I agree. It's simply to frequently to similar to previous episodes to remain interesting.
And perhaps it wouldn't damage the franchise to concentrate on books somewhat. Although I don't think they're that great usually (meaning the star trek books simply aren't that nice), books are much longer and more detailed than an episode. Perhaps a more vivid series can emerge from a period of pure book-ness.
This is probably slightly off-topic; but I'm curious about this too. On my machine Qt is much much faster than gtk in window-resizing; this is both in cpu-time user in X-server, and is cpu-time used in the program. This despite the fact that Qt actually shows many many more frames while resizing with content.
Of course I do have an athlon XP 2000+ as well. Using prelink means startup times are generally great; konq starts up and displays home-dir in about 1/2 a second; mozilla and firebird take around 1 second to start up and display a simple home page. I'm using an nvidia card with XRender acceleration on, and normal widget interactions are slightly less smooth (say about twice as slow) as win2k (which is completely smooth resizing simple windows at 85Hz) when using Qt, but when using Gtk+ is around a further 10 times slower than qt (really).
Mozilla (or even more clearly firebird) is a great example for this; it resizes with content quite smoothly in windows as long as the HTML page is simple enough, while even with a blank page it'll never get much beyond 5 fps with GTK+ under linux.
Am I missing some obviously GTK+ configuration step?
(I realize that window resizing isn't the only measure of worth; but it's a noticable, simple and irritating one)
So seriously - could you compare this to your experience with GTK+? Basically I just want to get GTK+ programs to run as smoothly as possible but I was under the impression that GTK simply is much worse at relayouting than Qt.
That's not so stupid at all. Although i have no experience with 2.3; I've heard it said that the general stability of 2.6-test is quite beyond that of the early 2.4-test - so much so that it's quite useable.
Furthermore; Before a kernel is "stable" it's going to have to be stable on most arches that it's to run on; support all the hardware correctly etc etc etc. For a distro specifically targetted at one arch, it can be much simpler to target a good stability because the problematic hardware interactions are far simpler.
Finally, it seems entirely normal, and indeed the opposite rather "questionable", for a _beta_ distribution to include that software that they intend to ship with. 2.6 Is definitly nearly at that point; as such it's the obvious choice to use. By the time AMD64 under linux is ready for prime time, i bet that 2.6 will be too.
Thanks for the update. I haven't heard of any of the problems you're referring to. As to those out-of-date coding practices, they sort of sound like my out of date c++ practice - I haven't (seriously) used it for years. I came across the mozilla guidelines recently and concluded I wasn't up for that anytime soon:-). Did you read the article? D, the language with templates and garbage collections sounds interesting.
Server side tasks are just great for java; your arguments don't add up.
Any platform I've heard of for server stuff will run java. C/C++ may have a wider distribution; but not amongst relevant platforms. Irrelevant to the discussion are portable/client side/ small / light platforms, and the computation heavy things. Servers are about "service" - the naming is not a coincidence.
Most server processes run forever, or as close as possibly; not very short lived at all. Starting a new process (aka application) for every client is generally not best practice - those are threads, and don't require constant re-JIT-ing or other JVM/CLR overhead.
Rather, the larger the system, and the more data, generally the less code in relation to that data, and even less JVM in relation to data. So for the quintessential server with tons of data gunk, the c/c++ advantage is much smaller than in the GUI.
Furthermore, it is a mistake to compare the portability of java to that of c++ in the manner you do. c++ implementations aren't generally compatible. Take a look at the mozilla coding guidlines for portable c++: http://mozilla.org/hacking/portable-cpp.html
c++ isn't portable, normally. C++-- might be though. Then again, it might not be. The java language is very standardized; and in case you shouldn't have a compiler on that platform, the bytecode is too!
In conclusion, if you're using c++ for a server-side task you should consider using java instead. As a matter of fact, most scripting languages are probably better suited than c++, I can hardly image a worse fit.
The confession needs to be notarized. That makes it much harder to "fake out" and also makes it quite expensive otherwise - seeing as that notary will charge you:-). And if you're forging notarized documents, you're playing a new game...
Your counter argument isn't convincing. First of all, the issue isn't what congress can or can't due; it's what should be done. I also don't believe that this really "can't" be done - generally, the interplay between te various government arms, the legal system, and the populace has been very sucessfull at doing this clearly beyond any initial boundaries - consider the closed military trials, and the various other crackdowns on civil liberties enacted and enforced after 9/11. The law isn't a religious text; it serves merely to support our values, morals (and religion depending on where you're from).
The federal government already does subsidize and tax at a large scale; doing is necessary for a functioning free market (without being great on the terminology: public goods are/cause market imperfections). Additionally, it's often a tool for a policy.
Given current subsidies in agriculture, social security, and education (which is almost knowledge), it's definitely possible to subsidize this.
I guess this is slightly offtopic; but with all this talk going on not only about the RIAA but also the software patents now in europe and DMCA etc etc etc, it's becomes hard not to notice the big pile of dung that copyrights et al seem to be causing. And for what? There's so many cool things one could do with a more relaxed information environment but instead, copyrights not only prevent this, but often, one of the original motivations behind copyright (namely that things get published at all) is rather side stepped. You can't learn anything from a compiled binary; yet nevertheless it enjoys copyright protection (effectively does in any case).
I don't think the right to exchange information is holy or somehow a human right which you're suggesting here. Consider slander, spam, or malicious information. Malicious information is for instance a virus, or even something as simple as telling a very gullible person that to cure his headache he merely needs to jump off that tower there...
Given the obvious advantages of free information flow (it is for instance the underpinning of a free market, and necessary also for a "democratic" society), I'ld say information should not be needlessly restricted unless there is a very good reason for it.
Supposedly, copyrights/patents are a required to encourage the production of new knowledge.
I would say it's clear that they do encourage some creation of knowledge. By their very nature, however, they also limit it's applicability and extension, therefore also discouraging the creation of such knowledge. Furthermore, I think a better system could be instituted.
Given that copyrights use market dynamics to encourage creation, whilst those dynamics work only in situations of scarcity, and that information itself (not the distribution thereof!) is not scarce, we can conclude that a system that tries to encourage new knowledge without enforcing scarcity would be optimal, as doing so would bring encouragement without destroying the actual point of the knowledge in the first place.
People regularly comment on the fact that communism (specifically in Russia) collapsed because it (it being the abstract administrative process that is communism) is a fundamentally bad match in the real world (in which resources are scarce). Generally it's not so widely noted that the same could be said of our current Intellectual Property mess.
Fortunately, we already have a mechanism to support non-scarce goods (aka social goods) in our society! Subsidizing knowledge production is a far superior solution... and we already do it to some extent with schools, art grants, universities, etc etc etc.
The question then becomes: how to divide such grants? I don't have an easy answer to that but a model ala de references by academic papers (or for that matter hyperlinks in the net) comes to mind.
To draw an analogy: in our current situation, knowledge is exclusively controlled by it's creator, which is comparable to how a completely "closed" internet portal would control its content and display information and news depending mostly on how much it can pay to create or buy that information from some news service or equivalent. The subsidized model which supports knowledge creation is more like the net at large with hyperlinks forming the votes for who's cool and who's not. Even without a framework specifically designed to support it, google seems capable to extract useful information from those votes:-).
Slashdot Mirror
If I remember correctly, the speedup of turning on this numa-awareness on an opteron multiprocessor system is negligible even on benchmarks which should stress it. It's a neat feature but doesn't appearantly really help the opteron (partially because the HT bus is very fast and the number of CPU's involved isn't so great).
--Eamon
This isn't news about a point release - I've never heard of VCF before this so to me this is actually new. And it's actually vaguely interesting - as in something to consider using as opposed to many other announcements on slashdot. Obviously a disclaimer would be nice, but it's interesting enough to warrant some attention, and for me therefore leniency :-).
And - nobody's reading slashdot with the expectation of well-filtered news anyhow.
--Eamon
A hypocrite is a person who professes beliefs and opinions that he does not hold. I don't believe that's the case here... rather; it's a bit rich to here about media consilidation evils coming from such a man - But the more Ted Turner represents the very media consolidation he now says are so detrimental, the more you should take his arguments seriously. I can hardly image a more appropriate preacher for this gospel :-).
--Eamon
The article you're referring to is probably one of the worst hardware-review/benchmark articles in a long long time. Taking a game engine which was so obviously immature and pretending that the benchmark results it provides have any real meaning whatsoever isn't very honest - especially considering that ATI was partnered with valve at the time. HL2 was at the time the first game to really use such new features as well meaning that it's very very likely that future driver releases would have changed the picture dramatically.
If you want to by ATI that's fine. Frankly that's not a good idea: not because nVidia's cards are so much better but because you should look at the cards individually. For example, the high-end NV30 cards (5800 and related) performed badly. Don't buy those then! However it's also true that these new nVidia cards (esp. the 6800 GT) price/perform very nicely all around. Slightly lower end, the 9800Pro still looks very good for all games currently on the market; including doom 3 and far cry...
Anyways - that article suggests conclusions that, based on the facts it presents, are completely off base.
Not that it's in any livable price-range; but two 30" displays connected to a card like that running MacOS-X, can you image that? ;-)
The beauty, the beauty...
Ooh and then you be like... ...all creative!
Or something...
I don't believe that source code theft is really such a problems for such companies - I really really doubt microsoft would use much of altavista's code even if they legally could! (It's so unbelievably much work to figure out someone elses mature code....)
However, employee education leakage is far more important. The raison d'etre for some of those architectural choices, or experiences with certain emergent pattern in large scale systems, and similarly complex issues are very, very valuable.
So really - feel sorry for microsoft... this just gives them bad PR, potentially opens them up for lawsuits (however unfounded), and generally doesn't do them any good..
I wonder where the claim of 5000 dollars damage comes from? The article says he claims he was curious about the progression of the product (which honestly, however illegal, I sympathize with - you put so much of yourself in these systems and then all of a sudden you're not allowed to know anything about them... arg!), so maybe it's all just much ado about nothing.
Why would social and financial systems be safe even if they took "default user behaviour" into account? And safe for whom?
:-)
I don't think things are quite that simple
Back on topic; The ethics of microsoft's (comprehensibly lacking) user education are further muddied by their being a monopoly...
Requiring a goody-goody warm and fuzzy company is obviously neither realistic nor particularly useful. However, with power comes responsibility; and microsoft, being a monopoly, certainly has power in the market, whereby some responsibility for that market is assumed.
Having said that, at a more emotional level I reject this deplorable reduction of such issues to mere market mechanics and capitalism (oh how we pray to the yadda yadda yadda). I just want to live in a nice society - and having big bullies around which rationalize their actions (i.e. distributing IE with windows) by essentially saying "just because we can", is pretty damn lame.
Maybe it's fair that the schoolyard hulk that has trained and learned how to fight well always beats other people up. Fair in some darwinian sense... certainly it's not ideal for me!
--Eamon
I don't exactly know which spikes the parent poster is referring too; but I'm quite sure they're not the spread-spectrum spikes. Those are frequency (not temporal) spikes and can increase EMI. Spread spectrum reduces the sharpness of the frequency spike by modulating the FSB - reducing, not increasing the stability.
;-)
All spikes are equal - but some are more equal than others
--Eamon
It looks fine to me too...
:-).
t ch .html ...gives an in depth overview. The standards mode generally works more consistantly and predicably especially with complicated effects, and you get a bit of cross-browser advantages for free :-).
But I think it would be even nicer if the time-stamp were less prominent (for example by placing it behind as opposed to in front of the title using a z-index).
I especially like the way the underline on hover breaks across lines
The rendering differences might be solved by using the "Standard Compliance mode" in IE and gecko, using a valid xhtml doctype is the simplest way of achieving that but...
http://gutfeldt.ch/matthias/articles/doctypeswi
--Eamon
I don't know anything about plain text SSL.
However... Just authenticating a site is really important! It's more likely that you'll be fooled by a hoaxed web site than that someone manages to sniff your packets or man-in-the-middle you. I don't know about most of you; but I'm on a reasonably trusted network (i.e. I know there aren't any sniffers on the LAN and once the connection's being routed you'ld have to sniff all kinds of gunk and manage to somehow, transparently get into that connection between the various routers - much harder than on a lan.)
So if I had to choose between authentication and encryption - well; I know what I would choose...
I run gentoo; my suggestion is just not to worry about it...
I've compiled with some rather experimental flags in my time and they often cause problems in things like xine or mplayer or whatnot; and the speed difference isn't that great... (not perceivably so anyhow)
Anyhow; just use stage3 if you feel like it - you can always recompile portions with new CFLAGS by
emerge --oneshot [my-package-here]
I run ~x86 and for me the real advantage of portage is the access to up-to-date packages.
Some people are really snobbish about this; pretending that updating other systems is even remotely as easy (it's not) or as up-to-date (witness debian)
With so many developments happening in linux-land is nice to be able to actually get those features.
For example kde 3.2 (in my opinion full of real world improvements over 3.1.x) was available for gentoo right away (the same day...) - not so for other distros. Also gentoo has ebuilds for pretty much anything under the sun... almost all programs you'll want to install are emerge-able.
Gentoo may not be so bug-free; it may not be as fast (at installing); but it sure as hell has a great forum and bug-support system. It's really a great way to get help and to post some help to others that suffer from the same problems.
So it has costs (more hassle to set up) but benefits too (more up-to-date, great support, good documentation).
I.e. don't worry about the compiling from source bit... that's not even so important.
This discussion obviously boils down to quasi-religious beliefs; so lets not get too carried away with things (not that you did, I just hate flamewars).
;-).
To take your example about the get-away car and the robbery-plans, I think it wouldn't harm to institute a rule allowing the government a lot of leeway in such investigations; however simultaneously guaranteeing that the government provides for replacement (or more likely: remuneration) for taken goods including lost business/opportunities.
Why?
To me it seems that the issue isn't so much guilt; rather such searches really do cause costs to the people involved. By making them explicit; they can better be taken account of in policymaking. Also, this kind of thing causes costs that aren't so transparent; namely risk costs: even people but especially corporations never effected by such a large and serious warrant might spend time and effort ( = money, money) to plan for such an event, a cost to society that the government really doesn't need to encourage!
Furthermore, I think if people weren't so frightened of these things (and if there wouldn't be any grevious losses I don't think people would be), perhaps more friendly cooperation would be encouraged. For instance; if I saw reasonable evidence of a computer crime spread over machines on a network I'm a stakeholder in (say as a user or administrator), then I'ld think twice before informing any law enforcement given the possibilty of such serious personal costs.
The above are just my personal beliefs... I'm sure you can find fault somewhere
--Eamon Nerbonne
Sure AGP is a sort of bottleneck - similarly to how Ultra-ATA133 can be a bottleneck and serial ATA can perform better - it's not just bandwidth that matters.
In PCI-Express's case (I haven't read up on this for months so it's not so fresh) some of the improved things are cache-control, which can now be driver controlled, error detection and handling (PCI doesn't have this, can be an annoying cause for hidden data-corruption), p2p in the sense that devices can send info directly to each other...
Another important issue is buffer management and relaxed ordering and just general smartness and better QoS / data-type-awareness (this should improve effective streaming bandwidth, and improve choices in case bandwidth is low, etc.)
PCI Express is definitely a step forwards - however, that it will really matter too much for video cards initially, given that the drivers will need to be adapted make use of the new features, is unlikely.
--Eamon Nerbonne
There were some figures on power consumption that suggested Prescott runs quite a bit warmer than Northwood and the same clockspeed.
:-).
7
That's not really surprising either: A feature-size shrink (such as the move now from 130nm to 90nm) lowers the power per transistor, however, it increases the power per unit area (IIRC proportionally to the sqrt of the shrink).
Of course - a different processor layout could completely counteract these effects, and a change to 31 stages instead of 21 stages would qualify as such I suppose
http://www.aceshardware.com/read.jsp?id=6000031
does list some power data, and rates a 3.2 GHz prescott as having a 103W _typical_ power consumption...
Conclusion the prescott has a power consumption problem; hopefully later revisions will fix this...
I agree. It's simply to frequently to similar to previous episodes to remain interesting.
And perhaps it wouldn't damage the franchise to concentrate on books somewhat. Although I don't think they're that great usually (meaning the star trek books simply aren't that nice), books are much longer and more detailed than an episode. Perhaps a more vivid series can emerge from a period of pure book-ness.
This is probably slightly off-topic; but I'm curious about this too. On my machine Qt is much much faster than gtk in window-resizing; this is both in cpu-time user in X-server, and is cpu-time used in the program. This despite the fact that Qt actually shows many many more frames while resizing with content.
Of course I do have an athlon XP 2000+ as well. Using prelink means startup times are generally great; konq starts up and displays home-dir in about 1/2 a second; mozilla and firebird take around 1 second to start up and display a simple home page. I'm using an nvidia card with XRender acceleration on, and normal widget interactions are slightly less smooth (say about twice as slow) as win2k (which is completely smooth resizing simple windows at 85Hz) when using Qt, but when using Gtk+ is around a further 10 times slower than qt (really).
Mozilla (or even more clearly firebird) is a great example for this; it resizes with content quite smoothly in windows as long as the HTML page is simple enough, while even with a blank page it'll never get much beyond 5 fps with GTK+ under linux.
Am I missing some obviously GTK+ configuration step?
(I realize that window resizing isn't the only measure of worth; but it's a noticable, simple and irritating one)
So seriously - could you compare this to your experience with GTK+? Basically I just want to get GTK+ programs to run as smoothly as possible but I was under the impression that GTK simply is much worse at relayouting than Qt.
--Eamon
That's not so stupid at all. Although i have no experience with 2.3; I've heard it said that the general stability of 2.6-test is quite beyond that of the early 2.4-test - so much so that it's quite useable.
Furthermore; Before a kernel is "stable" it's going to have to be stable on most arches that it's to run on; support all the hardware correctly etc etc etc. For a distro specifically targetted at one arch, it can be much simpler to target a good stability because the problematic hardware interactions are far simpler.
Finally, it seems entirely normal, and indeed the opposite rather "questionable", for a _beta_ distribution to include that software that they intend to ship with. 2.6 Is definitly nearly at that point; as such it's the obvious choice to use. By the time AMD64 under linux is ready for prime time, i bet that 2.6 will be too.
--Eamon
Thanks for the update. I haven't heard of any of the problems you're referring to. As to those out-of-date coding practices, they sort of sound like my out of date c++ practice - I haven't (seriously) used it for years. I came across the mozilla guidelines recently and concluded I wasn't up for that anytime soon :-). Did you read the article? D, the language with templates and garbage collections sounds interesting.
Server side tasks are just great for java; your arguments don't add up.
Any platform I've heard of for server stuff will run java. C/C++ may have a wider distribution; but not amongst relevant platforms. Irrelevant to the discussion are portable/client side/ small / light platforms, and the computation heavy things. Servers are about "service" - the naming is not a coincidence.
Most server processes run forever, or as close as possibly; not very short lived at all. Starting a new process (aka application) for every client is generally not best practice - those are threads, and don't require constant re-JIT-ing or other JVM/CLR overhead.
Rather, the larger the system, and the more data, generally the less code in relation to that data, and even less JVM in relation to data. So for the quintessential server with tons of data gunk, the c/c++ advantage is much smaller than in the GUI.
Furthermore, it is a mistake to compare the portability of java to that of c++ in the manner you do. c++ implementations aren't generally compatible. Take a look at the mozilla coding guidlines for portable c++: http://mozilla.org/hacking/portable-cpp.html
c++ isn't portable, normally. C++-- might be though. Then again, it might not be. The java language is very standardized; and in case you shouldn't have a compiler on that platform, the bytecode is too!
In conclusion, if you're using c++ for a server-side task you should consider using java instead. As a matter of fact, most scripting languages are probably better suited than c++, I can hardly image a worse fit.
No - hashish doesn't cost much more than beer here, I'm quite certain about that ;-)
--Eamon
Interesting - I live in the Netherlands and you should add two zero's there :-(
Seriously, notaries cost ridiculously much here.
The confession needs to be notarized. That makes it much harder to "fake out" and also makes it quite expensive otherwise - seeing as that notary will charge you :-). And if you're forging notarized documents, you're playing a new game...
Your counter argument isn't convincing.
First of all, the issue isn't what congress can or can't due; it's what should be done. I also don't believe that this really "can't" be done - generally, the interplay between te various government arms, the legal system, and the populace has been very sucessfull at doing this clearly beyond any initial boundaries - consider the closed military trials, and the various other crackdowns on civil liberties enacted and enforced after 9/11. The law isn't a religious text; it serves merely to support our values, morals (and religion depending on where you're from).
The federal government already does subsidize and tax at a large scale; doing is necessary for a functioning free market (without being great on the terminology: public goods are/cause market imperfections). Additionally, it's often a tool for a policy.
Given current subsidies in agriculture, social security, and education (which is almost knowledge), it's definitely possible to subsidize this.
I guess this is slightly offtopic; but with all this talk going on not only about the RIAA but also the software patents now in europe and DMCA etc etc etc, it's becomes hard not to notice the big pile of dung that copyrights et al seem to be causing. And for what? There's so many cool things one could do with a more relaxed information environment but instead, copyrights not only prevent this, but often, one of the original motivations behind copyright (namely that things get published at all) is rather side stepped. You can't learn anything from a compiled binary; yet nevertheless it enjoys copyright protection (effectively does in any case).
:-).
I don't think the right to exchange information is holy or somehow a human right which you're suggesting here. Consider slander, spam, or malicious information. Malicious information is for instance a virus, or even something as simple as telling a very gullible person that to cure his headache he merely needs to jump off that tower there...
Given the obvious advantages of free information flow (it is for instance the underpinning of a free market, and necessary also for a "democratic" society), I'ld say information should not be needlessly restricted unless there is a very good reason for it.
Supposedly, copyrights/patents are a required to encourage the production of new knowledge.
I would say it's clear that they do encourage some creation of knowledge. By their very nature, however, they also limit it's applicability and extension, therefore also discouraging the creation of such knowledge. Furthermore, I think a better system could be instituted.
Given that copyrights use market dynamics to encourage creation, whilst those dynamics work only in situations of scarcity, and that information itself (not the distribution thereof!) is not scarce, we can conclude that a system that tries to encourage new knowledge without enforcing scarcity would be optimal, as doing so would bring encouragement without destroying the actual point of the knowledge in the first place.
People regularly comment on the fact that communism (specifically in Russia) collapsed because it (it being the abstract administrative process that is communism) is a fundamentally bad match in the real world (in which resources are scarce). Generally it's not so widely noted that the same could be said of our current Intellectual Property mess.
Fortunately, we already have a mechanism to support non-scarce goods (aka social goods) in our society! Subsidizing knowledge production is a far superior solution... and we already do it to some extent with schools, art grants, universities, etc etc etc.
The question then becomes: how to divide such grants? I don't have an easy answer to that but a model ala de references by academic papers (or for that matter hyperlinks in the net) comes to mind.
To draw an analogy: in our current situation, knowledge is exclusively controlled by it's creator, which is comparable to how a completely "closed" internet portal would control its content and display information and news depending mostly on how much it can pay to create or buy that information from some news service or equivalent. The subsidized model which supports knowledge creation is more like the net at large with hyperlinks forming the votes for who's cool and who's not. Even without a framework specifically designed to support it, google seems capable to extract useful information from those votes