Doing this right in Java (or any other environment) requires some knowledge about how things work
As I said, mistake might be on my part, but solving this problem in C++ was a piece of cake. However, I found this problem and Java better to be left alone. I simply do not wish Java so much to devote my life to study Java. If 3 months (that counts only the active testing for my needs, not learning from beginning, I already knew enough Java basics to write apps from long ago when I was using Borland JBuilder 1) was not enough then it is better for me to stick where I belong, pushing one technology where it doesn't fit naturaly is not something I want to do in my life.
There is no best hammer for every nail. That is why I stick with: - C, C++ and personaly modified freepascal for hardcore (whichever fits best to the job) - Lately C# for simple GUI things (I find C# aesthetics more pleasing than Java, while during my testing of both I only discovered that each has its own flaws (again read as: according to my expectations), but aesthetics were the only thing where C# won) - bash for system needs - a little of php and python for external scripts in my software
Did you run a profiler to find out where the bottlenecks were?
No need for that. Since I can't specify max mem (well as I learned today in one of the responses I can (at least in J5, link of documentation reffers to J5, but my Java testing ended 2-3 months before J5 was out) do that in runtime with Java, so solving my problem would probably need almost the same solution as C++) or cpu resources (I explained that in one of my replies, in c++ I solved this with simple IPC and all apps work as one) GC took everything for him self, everything started swapping which resulted in system crawl.
Optimization in Java is possible, just like in C
I know that;) Believe it or not sometimes I profile (:if you can call that profiling:) my bash scripts.
did you use the nio classes
Reason why I tried default Java was to avoid complex sources:) Using NIO is far less readable than C++ (that would be my personal opinion and my personal aesthetics)
p.s. Both.Net and Java have very bad coverage (documents) of my problem, they try to redirect your coding to managed memory and managed resources. And if documentation is hard to find then documentation is just as good as non existing. Maybe that is why more people try'n'leave Java than try'n'stick.
The example you give where allocation/deallocation happens in a loop over 2^32 iteration can be made to run by knowing something about how/whats being allocated. With smaller objects you could decrease the size of the young generation to make incremental collections faster. With variable size object you could set and adaptive resize policy for the tenured/young generation size-- this would allow the heap to grow and shrink as needed (this can make GC faster since it may have to collect on a smaller space). You can also increase the amount of GC can consume; sometimes gc can be too timid and allow memory allocation to get out of control, resulting in a heap that grows too big for GC to collect in a reasonable amount of time.
Same thing would be to alloc once and just reuse that block everytime wouldn't it? That was how I solved my problems. Having for example loop
allocation
work
deallocation end loop does take a pretty stupid coder to write it like that (if data type was known). I think we both agree on that;)
That said there are still times when it would be nice to have direct memory access
And that is why I went back. It is much easier to write specialized GC for your needs, than hoping somebody predicted what you need.
If you read my answers below you can see that I can't use that. But hey, thanks anyway. Unfortunatelly for me it was a special case where any auto-solution failed.
Sure this is a problem, but isn't it also a problem with traditional 'malloc'? From what I've seen, malloc gets pretty pathological under high load also, does it not? Eagerly reclaiming memory may not actually be as good as it sounds.
Agreed, but at the same time optimizing memory (de)/allocation is possible. Had to do that in C++ too. Had to set up completely controlled environment. Problem was I couldn't do that in Java.
Even with traditional explicit memory reclamation, you aren't ever *really* reclaiming the memory yourself. It has to first jump through whatever the malloc strategy is, and then the operating systems virtual memory and cache subsystems, all of which may choose to do arbitrary things. In/general/ I'd say that amortized over time, allocating and freeing memory in bulk, like say, a GC does, is optimal for throughput if not latency.
Same opinion. But automatic controlled environments can act correctly (and maybe even better) only in predicted occasions (and that would 98% probably) same as any auto-thing, as long as designer predicted those parameters system mostly works. Trouble is that I need those 2%.
If anything, it is not GC's fault, but perhaps the fault of whatever is causing such large memory allocations to begin with (e.g. the language or libraries or VM design).
Actualy it is nature of the job you're processing that controls this factor;)
Thanks, if you read my post again you can see that I know that there is a case where I could be wrong. That is why countering might be bad word to use. More like explanation;) So, here it is again: "Yes, I know I (or my mistakes) could be the problem here"
People like you run into trouble when they manually call the garbage collector (bad idea, can cause severe performance issues) or when they misconfigure the garbage collector (e.g. if you are running a server with multiple cpus and multiple gigabytes of memory). Also a big issue is C programmers trying to program Java like it is C and making wrong assumptions about the cost of certain code and the impact of certain optimizations (e.g. pooling is actually bad for performance in some cases).
Might be some truth in that, although I tried my best to go by the books and articles. But I can say this, that if I leave it out to GC then performace is really bad (same goes for C++ GC).
The only real problem that Java has with respect to memory management is that java programmers tend to use lots of memory when writing their software. You don't have to do that but using the default APIs and recommended practices in e.g. the java tutorial and many books on this matter you will probably end up using substantial amounts of memory. Many Java programs continually create and destroy large amounts of small objects. If you'd do that in C/C++ you would really notice this. In Java you hardly notice this at all because it allocates and deallocates in a much smarter way.
As I already said I tried by the books. And using base Java only, no additional packages.
I don't know what 'personal tests' you have been performing. Up front I'd say you may have been measuring the wrong things the wrong way and have been drawing the wrong conclusions based on what your results. The JVM is a highly complex piece of software and very difficult to benchmark properly. Doing so requires that you know
Actualy exactly the same job I had already done in C++. I tried to rewrite (and not line by line, but completely new one) the same job in Java (and then in.Net too), because I would benefit if it would work out and I had some time to spare.
Job consisted out of few separate apps (can't afford to make one app, have to separate them, because sometimes they do not reside on the same computer). All apps work multithreaded. A lot of memory blocks of various size, but one block (where various data builds and as long as it builds it can't be freed, since job must complete first then being processed or dropped, somewhat like transactions) could go up to 150MB. Trouble is that I can't predict how many blocks and how big will reside in memory at once and swapping data to disk is not possible.
Anyway, thanks again. But, to speak the truth, I won't consider Java now for a long time (which means I won't be testing it), not that I have something against Java, I simply have no time to waste anymore (I'm in a middle of the project right now) there was a time but Java failed and if that is because of Java or me, result is the same.
While I agree with you on all accounts I can't help but comment you.
Both Java and.Net have the same problem. Sloppy memory. As long as you don't use a lot of atomic memory blocks with higher load on machine where it runs, everything is ok and just as you said it. I tested both on the same tests and always fallen in the same problem. No direct memory control, GC waited until it was too late, and everything started to crawl. GC somehow avoids doing work if software is taking most of the CPU, it is also the same reason why Java beats C or C++ on speed tests (Every speed test where they try to proove that Java or.Net is faster than C or C++) uses allocation and freeing of memory in some loop. And while both C and C++ actualy do free memory, Java and.Net just mark that as garbage and wait for GC to clean up the mess which doesn't happen if load is too high. Just take any test where Java was faster and test loop to make 2^32 instead of 1000 or 10000 calls. This way you will actualy use more memory than you actualy have with allocations.
p.s. Not bashing, just saying results of my testing. If you can suggest some approach, do that, but so far not even one person suggested something I haven't tried yet. Both Java or.Net would be a real gift for me if only I could use them for my needs.
and only a tiny minority don't dual-boot with Windows.
Yeah, I admit it is (or it was) true. But as soon as person buys PS2 or other game console, the only reason for windows to exist on desktop dissapears. 99% of dual booting is just to play games.
I really don't see the use for personal desktop in Windows. What can it do? (and no, viruses I don't need. If I get virus better that I got it by sex, while spyware position is already taken by my annoying neighboors. All other jobs are much better performed with linux)
p.s. What would be Unixy app? Some Gay-MS-Word spellchecker result?
In my case, Outlook (2000 and 2003, not Express) problems account for 93% of my computer problems I have to solve. Other is accounted mostly to Explorer. Outlook Express is actualy having far less problems. Every folder in OE is separate file, while the pro version has one pst file for everything. Which means 2GB file limit is reached far sooner. Maybe Outlook solved that problem, maybe not. But 2000 never will. So considering the expense of moving to newer version, which would demand puting XP on every machine, it was far cheaper to go with 2000 which doesn't require buying anything, abandon Outlook and move to Lotus Notes (damn you IBM, for not putting Linux version) with some proprietary add-ons.
Every machine (for now) where I succed to change user to Thunderbird and Firefox means one more machine that requires regular checks only.
after few active (very active) minutes as any Java app consumes your memory, swap and starts working 0.5fps. And from that point in 10 minutes it didn't go anything better. Actualy first wrong response was when LG tripled menu.
Would be interesting if it wouldn't be in Java. Don't get me wrong, I don't mind Java as long as it is used for standalone desktop application, but writing desktop or server service in any language that has sloppy memory is plain stupid. These work non-stop.
Yeah, Rob Enderle (aka. MS payed public (dis)figure) is a new viral plague.
It sounds like any MS battle plan: 1. Prepare for the holy war 2. Kill and screw anything and everything by disregarding any moral issue or humanity 3. In case you find your self on a loosing ground start runing and franticaly screaming "NOT FAIR, NOT FAIR..."
After RTFA
My thesis completely fails here because of "Lu Shouqun, president of the China Open Source Software Promotion Union."
?? If I would only know where this world is going:? ??? And next week, pope preaching about safe usage of nuclear bombs in terrorism and how baptizing people before killing them is a thing to do, a show entitled "who cares for Geneva convention if you have to kill, at least do it properly! says your favorite pope"
I think MS Rep comment on this topic would be something like that:
MS Rep: What is this about Windows having no coverage? Microsoft has Rob Enderle to speak for us. ? . . . [ thinks ] ?? . . . [ thinks and taps his fingers ] ??? . . . (enderle = moron) + (enderle->public_opinion = MORON) . . . (speaking -> us) . . . = suicide ??? ??? . . . UAAAAAHHH? [screaming, but without Ballmers patented friendly gesture like throwing chairs] [runs into the sunset waiting for his XBox 360 to get HD-DVD support]
It is quite simple in fact. Intel was always used to be followed so there was no rush on their R&D department. AMD and others always only tried to achieve what Intel achieved. So AMD putting out 64-bit before them was probably quite a surprise. Now it is obvious that Intel is following AMDs R&D footsteps.
Intel somehow just can't achieve former glory. And it won't. Intel was always just boosting and never stand behind the words. Anyone remembering 64-bit RISC based Merced. Well, Merced was out (with delay) but for sure it wasn't 64-bit or based on RISC core. Same goes for P4 fiasco. P4 was nothing but a nice commercial. Itanium (or better Itanic was not better either). The only good thing Intel put out in last imes was P4M (which has less to do with P4 than with P3), but even here I can't wait nothing but to get rid of my notebook as soon as possible and buy either Turion or hopefully some Cell based notebook.
Basicaly, Intels R&D department sucks under pressure. It was nice until others were just trying to achive what they did, but now that others put out things before Intel? Well, result is here.
Unusual? Nothing. Moronic, maybe stupid and biased view on what is overpriced and how he expressed that would be better expression.
I never liked BS people belonging to the "moron" tribe. This is the reason for my response as it was.
p.s. Not that I own PSP, I just saw one belonging to a friend of mine. And I was favorably impressed too. I would maybe even buy that thing if phone would be integrated somehow. But now that just means one not really necessary gadget too much, and like there wouldn't be few of them already in "too much group".
??? And you're the one to talk about overpriced? What now? Next week, birds hosting egg cooking show?
PB 17"? 30" ACD?
These are overpriced items (at least price-performance, when I was selecting 24" (30" is too big for coding, so I decided on 24) LCD display ACD has fallen out the first. Quality is poor and price is enourmous) and the fact that I hate Apple has nothing to do with that.
btw. You're not the only one with 50" HDTV amd two notebooks, so you can stop showing your wang and step in line as the rest of us that own them.
I was having the same opinion until I read mailing list on Inkscape about this. Talk about one sided canibailzing without prejudice (would take theirs, but hell no, they won't resell mine).
Does natural selection not lead to intelligent design?
Natural selection in cases where people are involved? NO!!!
Have you ever seen 600lb geek (and even if you have, he still has to know how not to punch like a girl!! Lazy, overfat and living in moms basement geeks do not count here)???
New race of people would be mostly consisting people that have their body and brain constitution similiar to gorillas (mostly from boxer, kungfu or club bouncer tribes or warrior casts with machine guns, with common fact: "punch|shoot first, ask later").
Now how could this lead to intelligent design? By a miracle?
Nope. "Basic security" comes from a decent security model.
Agreed. Now think a little. Most of spy and virusware is users mistake or downloading software on the net. What does for example spyware differently than usual program. How could system detect spyware if user installed spyware along with some software? By guessing? By counting executables? Spyware has to be corrected by approach that they inteded to go.
Their security sucks, true. Their patching speed is likespeed of hardcore snail race, true. Overexpensed, true. But at least give them a chance to make something (or not, somehow reminds me on yesterdays story where Gates demanded from Sony that they don't ship with Blue-Ray).
As I said, I somehow feel sorry and somehow laugh (and enjoy the fact) because they are getting beaten with their own stick more and more often.
Which is why you need to continuously update the virus signatures.
No I don't, I don't use Windows. It is about 8 years now that I'm running Linux 99% of time. In reality I hate Apple and MS.
The correct approach is to fix the real problem.
In basics, if they integrated those two softwares without GUI and say that they don't exist one could call that correct approach.
The PROBLEM here is that when Microsoft starts shipping its own anti-virus/spyware app, the other companies will all die.
Companies producing this type of software are all more or less producing crapware. So probably you could call this biased. And since "king of the crapware" Symantec is making these claims I'm even more biased.
Which means that within a couple years, the ONLY commercial option you will have for anti-virus on Windows will be... Microsoft.
No, they will (not sure if it isn't part of MSDN only) bundle Acrillic, do you think that Photoshop will die? They bundle movie editor, did any movie application die? You can record CDs from Explorer, did Nero die? Frontpage, Dreamweaver? Want more?
Now, to see how much effort Microsoft will be putting into that... look how long the problem has already existed and look how long it took Microsoft to improve IE.
And here is the reason why those companies could stay alive. They should just be better.
Slashdot Mirror
How about their free software for Linux first?
Doing this right in Java (or any other environment) requires some knowledge about how things work
.Net and Java have very bad coverage (documents) of my problem, they try to redirect your coding to managed memory and managed resources. And if documentation is hard to find then documentation is just as good as non existing. Maybe that is why more people try'n'leave Java than try'n'stick.
As I said, mistake might be on my part, but solving this problem in C++ was a piece of cake.
However, I found this problem and Java better to be left alone. I simply do not wish Java so much to devote my life to study Java. If 3 months (that counts only the active testing for my needs, not learning from beginning, I already knew enough Java basics to write apps from long ago when I was using Borland JBuilder 1) was not enough then it is better for me to stick where I belong, pushing one technology where it doesn't fit naturaly is not something I want to do in my life.
There is no best hammer for every nail.
That is why I stick with:
- C, C++ and personaly modified freepascal for hardcore (whichever fits best to the job)
- Lately C# for simple GUI things (I find C# aesthetics more pleasing than Java, while during my testing of both I only discovered that each has its own flaws (again read as: according to my expectations), but aesthetics were the only thing where C# won)
- bash for system needs
- a little of php and python for external scripts in my software
Did you run a profiler to find out where the bottlenecks were?
No need for that. Since I can't specify max mem (well as I learned today in one of the responses I can (at least in J5, link of documentation reffers to J5, but my Java testing ended 2-3 months before J5 was out) do that in runtime with Java, so solving my problem would probably need almost the same solution as C++) or cpu resources (I explained that in one of my replies, in c++ I solved this with simple IPC and all apps work as one) GC took everything for him self, everything started swapping which resulted in system crawl.
Optimization in Java is possible, just like in C
I know that;) Believe it or not sometimes I profile (:if you can call that profiling:) my bash scripts.
did you use the nio classes
Reason why I tried default Java was to avoid complex sources:) Using NIO is far less readable than C++ (that would be my personal opinion and my personal aesthetics)
p.s. Both
If this is what you suggestp ackage-summary.html
http://java.sun.com/j2se/1.4.2/docs/api/java/nio/
Yes, I tried it and I found no more Java advantage. Source became even more complex and harder to read than C++
Yep, that answered few questions of mine. Thanks.
The example you give where allocation/deallocation happens in a loop over 2^32 iteration can be made to run by knowing something about how/whats being allocated. With smaller objects you could decrease the size of the young generation to make incremental collections faster. With variable size object you could set and adaptive resize policy for the tenured/young generation size-- this would allow the heap to grow and shrink as needed (this can make GC faster since it may have to collect on a smaller space). You can also increase the amount of GC can consume; sometimes gc can be too timid and allow memory allocation to get out of control, resulting in a heap that grows too big for GC to collect in a reasonable amount of time.
Same thing would be to alloc once and just reuse that block everytime wouldn't it? That was how I solved my problems. Having for example
loop
allocation
work
deallocation
end loop
does take a pretty stupid coder to write it like that (if data type was known). I think we both agree on that;)
That said there are still times when it would be nice to have direct memory access
And that is why I went back. It is much easier to write specialized GC for your needs, than hoping somebody predicted what you need.
If you read my answers below you can see that I can't use that. But hey, thanks anyway. Unfortunatelly for me it was a special case where any auto-solution failed.
Sure this is a problem, but isn't it also a problem with traditional 'malloc'? From what I've seen, malloc gets pretty pathological under high load also, does it not? Eagerly reclaiming memory may not actually be as good as it sounds.
/general/ I'd say that amortized over time, allocating and freeing memory in bulk, like say, a GC does, is optimal for throughput if not latency.
Agreed, but at the same time optimizing memory (de)/allocation is possible. Had to do that in C++ too. Had to set up completely controlled environment. Problem was I couldn't do that in Java.
Even with traditional explicit memory reclamation, you aren't ever *really* reclaiming the memory yourself. It has to first jump through whatever the malloc strategy is, and then the operating systems virtual memory and cache subsystems, all of which may choose to do arbitrary things. In
Same opinion. But automatic controlled environments can act correctly (and maybe even better) only in predicted occasions (and that would 98% probably) same as any auto-thing, as long as designer predicted those parameters system mostly works. Trouble is that I need those 2%.
If anything, it is not GC's fault, but perhaps the fault of whatever is causing such large memory allocations to begin with (e.g. the language or libraries or VM design).
Actualy it is nature of the job you're processing that controls this factor;)
Thanks, if you read my post again you can see that I know that there is a case where I could be wrong. That is why countering might be bad word to use. More like explanation;) So, here it is again: "Yes, I know I (or my mistakes) could be the problem here"
.Net too), because I would benefit if it would work out and I had some time to spare.
People like you run into trouble when they manually call the garbage collector (bad idea, can cause severe performance issues) or when they misconfigure the garbage collector (e.g. if you are running a server with multiple cpus and multiple gigabytes of memory). Also a big issue is C programmers trying to program Java like it is C and making wrong assumptions about the cost of certain code and the impact of certain optimizations (e.g. pooling is actually bad for performance in some cases).
Might be some truth in that, although I tried my best to go by the books and articles. But I can say this, that if I leave it out to GC then performace is really bad (same goes for C++ GC).
The only real problem that Java has with respect to memory management is that java programmers tend to use lots of memory when writing their software. You don't have to do that but using the default APIs and recommended practices in e.g. the java tutorial and many books on this matter you will probably end up using substantial amounts of memory. Many Java programs continually create and destroy large amounts of small objects. If you'd do that in C/C++ you would really notice this. In Java you hardly notice this at all because it allocates and deallocates in a much smarter way.
As I already said I tried by the books. And using base Java only, no additional packages.
I don't know what 'personal tests' you have been performing. Up front I'd say you may have been measuring the wrong things the wrong way and have been drawing the wrong conclusions based on what your results. The JVM is a highly complex piece of software and very difficult to benchmark properly. Doing so requires that you know
Actualy exactly the same job I had already done in C++. I tried to rewrite (and not line by line, but completely new one) the same job in Java (and then in
Job consisted out of few separate apps (can't afford to make one app, have to separate them, because sometimes they do not reside on the same computer). All apps work multithreaded.
A lot of memory blocks of various size, but one block (where various data builds and as long as it builds it can't be freed, since job must complete first then being processed or dropped, somewhat like transactions) could go up to 150MB.
Trouble is that I can't predict how many blocks and how big will reside in memory at once and swapping data to disk is not possible.
Anyway, thanks again. But, to speak the truth, I won't consider Java now for a long time (which means I won't be testing it), not that I have something against Java, I simply have no time to waste anymore (I'm in a middle of the project right now) there was a time but Java failed and if that is because of Java or me, result is the same.
While I agree with you on all accounts I can't help but comment you.
.Net have the same problem. Sloppy memory. As long as you don't use a lot of atomic memory blocks with higher load on machine where it runs, everything is ok and just as you said it. I tested both on the same tests and always fallen in the same problem. No direct memory control, GC waited until it was too late, and everything started to crawl. GC somehow avoids doing work if software is taking most of the CPU, it is also the same reason why Java beats C or C++ on speed tests (Every speed test where they try to proove that Java or .Net is faster than C or C++) uses allocation and freeing of memory in some loop. And while both C and C++ actualy do free memory, Java and .Net just mark that as garbage and wait for GC to clean up the mess which doesn't happen if load is too high. Just take any test where Java was faster and test loop to make 2^32 instead of 1000 or 10000 calls. This way you will actualy use more memory than you actualy have with allocations.
.Net would be a real gift for me if only I could use them for my needs.
Both Java and
p.s. Not bashing, just saying results of my testing. If you can suggest some approach, do that, but so far not even one person suggested something I haven't tried yet. Both Java or
and only a tiny minority don't dual-boot with Windows.
Yeah, I admit it is (or it was) true. But as soon as person buys PS2 or other game console, the only reason for windows to exist on desktop dissapears. 99% of dual booting is just to play games.
I really don't see the use for personal desktop in Windows. What can it do? (and no, viruses I don't need. If I get virus better that I got it by sex, while spyware position is already taken by my annoying neighboors. All other jobs are much better performed with linux)
p.s. What would be Unixy app? Some Gay-MS-Word spellchecker result?
Really? Where?
In my case, Outlook (2000 and 2003, not Express) problems account for 93% of my computer problems I have to solve. Other is accounted mostly to Explorer. Outlook Express is actualy having far less problems. Every folder in OE is separate file, while the pro version has one pst file for everything. Which means 2GB file limit is reached far sooner. Maybe Outlook solved that problem, maybe not. But 2000 never will. So considering the expense of moving to newer version, which would demand puting XP on every machine, it was far cheaper to go with 2000 which doesn't require buying anything, abandon Outlook and move to Lotus Notes (damn you IBM, for not putting Linux version) with some proprietary add-ons.
Every machine (for now) where I succed to change user to Thunderbird and Firefox means one more machine that requires regular checks only.
Having more problems != more decent
Yes, well and responsive, but....
after few active (very active) minutes as any Java app consumes your memory, swap and starts working 0.5fps. And from that point in 10 minutes it didn't go anything better. Actualy first wrong response was when LG tripled menu.
Would be interesting if it wouldn't be in Java. Don't get me wrong, I don't mind Java as long as it is used for standalone desktop application, but writing desktop or server service in any language that has sloppy memory is plain stupid. These work non-stop.
Even more...
:?
My first assumtion before RTFA
Yeah, Rob Enderle (aka. MS payed public (dis)figure) is a new viral plague.
It sounds like any MS battle plan:
1. Prepare for the holy war
2. Kill and screw anything and everything by disregarding any moral issue or humanity
3. In case you find your self on a loosing ground start runing and franticaly screaming "NOT FAIR, NOT FAIR..."
After RTFA
My thesis completely fails here because of "Lu Shouqun, president of the China Open Source Software Promotion Union."
?? If I would only know where this world is going
??? And next week, pope preaching about safe usage of nuclear bombs in terrorism and how baptizing people before killing them is a thing to do, a show entitled "who cares for Geneva convention if you have to kill, at least do it properly! says your favorite pope"
I think MS Rep comment on this topic would be something like that:
MS Rep: What is this about Windows having no coverage? Microsoft has Rob Enderle to speak for us.
? . . . [ thinks ]
?? . . . [ thinks and taps his fingers ]
??? . . . (enderle = moron) + (enderle->public_opinion = MORON) . . . (speaking -> us) . . . = suicide ???
??? . . . UAAAAAHHH?
[screaming, but without Ballmers patented friendly gesture like throwing chairs]
[runs into the sunset waiting for his XBox 360 to get HD-DVD support]
It is quite simple in fact. Intel was always used to be followed so there was no rush on their R&D department. AMD and others always only tried to achieve what Intel achieved. So AMD putting out 64-bit before them was probably quite a surprise. Now it is obvious that Intel is following AMDs R&D footsteps.
Intel somehow just can't achieve former glory. And it won't. Intel was always just boosting and never stand behind the words. Anyone remembering 64-bit RISC based Merced. Well, Merced was out (with delay) but for sure it wasn't 64-bit or based on RISC core. Same goes for P4 fiasco. P4 was nothing but a nice commercial. Itanium (or better Itanic was not better either). The only good thing Intel put out in last imes was P4M (which has less to do with P4 than with P3), but even here I can't wait nothing but to get rid of my notebook as soon as possible and buy either Turion or hopefully some Cell based notebook.
Basicaly, Intels R&D department sucks under pressure. It was nice until others were just trying to achive what they did, but now that others put out things before Intel? Well, result is here.
So what was so unusual about the response?
Unusual? Nothing. Moronic, maybe stupid and biased view on what is overpriced and how he expressed that would be better expression.
I never liked BS people belonging to the "moron" tribe. This is the reason for my response as it was.
p.s. Not that I own PSP, I just saw one belonging to a friend of mine. And I was favorably impressed too. I would maybe even buy that thing if phone would be integrated somehow. But now that just means one not really necessary gadget too much, and like there wouldn't be few of them already in "too much group".
??? And you're the one to talk about overpriced? What now? Next week, birds hosting egg cooking show?
PB 17"? 30" ACD?
These are overpriced items (at least price-performance, when I was selecting 24" (30" is too big for coding, so I decided on 24) LCD display ACD has fallen out the first. Quality is poor and price is enourmous) and the fact that I hate Apple has nothing to do with that.
btw. You're not the only one with 50" HDTV amd two notebooks, so you can stop showing your wang and step in line as the rest of us that own them.
Well, he was probably saying about "American Iraq problem". They solved their own and created few new general problems.
I was having the same opinion until I read mailing list on Inkscape about this. Talk about one sided canibailzing without prejudice (would take theirs, but hell no, they won't resell mine).
a d_id=8520852&forum_id=36054
http://sourceforge.net/mailarchive/forum.php?thre
p.s. Last comment (or it was last at the time I read was probably the only inteligent comment, about shared LGPL libraries)
Well, at least they show some inovation on their side. Even though it comes from from greed department:)
Either is that greed talking or they feel that people cheat with terminal servers to avoid buying OS licenses.
Does natural selection not lead to intelligent design?
Natural selection in cases where people are involved? NO!!!
Have you ever seen 600lb geek (and even if you have, he still has to know how not to punch like a girl!! Lazy, overfat and living in moms basement geeks do not count here)???
New race of people would be mostly consisting people that have their body and brain constitution similiar to gorillas (mostly from boxer, kungfu or club bouncer tribes or warrior casts with machine guns, with common fact: "punch|shoot first, ask later").
Now how could this lead to intelligent design? By a miracle?
Hexachrome is patentedc le=64&idPressRelease=38
http://www.pantone.com/aboutus/aboutus.asp?idArti
And there was a lot of talking about that on Inkscape mailing list
And it is no urban legend needed to tell me that Coffe wakes me up much faster than C++.
ogm? xvid?
You can play those two legaly.
And about DVD's, true, while HD-DVD does not exist yet.
Nope. "Basic security" comes from a decent security model.
... Microsoft.
... look how long the problem has already existed and look how long it took Microsoft to improve IE.
Agreed. Now think a little. Most of spy and virusware is users mistake or downloading software on the net. What does for example spyware differently than usual program. How could system detect spyware if user installed spyware along with some software? By guessing? By counting executables? Spyware has to be corrected by approach that they inteded to go.
Their security sucks, true. Their patching speed is likespeed of hardcore snail race, true. Overexpensed, true. But at least give them a chance to make something (or not, somehow reminds me on yesterdays story where Gates demanded from Sony that they don't ship with Blue-Ray).
As I said, I somehow feel sorry and somehow laugh (and enjoy the fact) because they are getting beaten with their own stick more and more often.
Which is why you need to continuously update the virus signatures.
No I don't, I don't use Windows. It is about 8 years now that I'm running Linux 99% of time. In reality I hate Apple and MS.
The correct approach is to fix the real problem.
In basics, if they integrated those two softwares without GUI and say that they don't exist one could call that correct approach.
The PROBLEM here is that when Microsoft starts shipping its own anti-virus/spyware app, the other companies will all die.
Companies producing this type of software are all more or less producing crapware. So probably you could call this biased. And since "king of the crapware" Symantec is making these claims I'm even more biased.
Which means that within a couple years, the ONLY commercial option you will have for anti-virus on Windows will be
No, they will (not sure if it isn't part of MSDN only) bundle Acrillic, do you think that Photoshop will die? They bundle movie editor, did any movie application die? You can record CDs from Explorer, did Nero die? Frontpage, Dreamweaver? Want more?
Now, to see how much effort Microsoft will be putting into that
And here is the reason why those companies could stay alive. They should just be better.