>>41 >>42
Now that we are talking about the performance of Common Lisp implementations, how fast are the commercial Lisps such as Allegro and LispWorks?
>>45-46
If something accelerates uniformly instead of maintaining a constant speed, it makes sense to answer your question with the acceleration, assuming your not dumb enough to figure it out by yourself, ``fucking lithp faggot''.
Name:
Anonymous2011-07-12 15:00
What modern languages there are nowdays that do not have garbage collector and is not C or C++?
You think it's not possible to code withouut GC. Just fucking release your memory in destructor and all is fine. WTF is wrong with you. Fucking Java faggots.
I mean Java does not suck that much, and GC is almost OK but it kills the performance and I just like to RAII more. If you can't code without GC, then there are something wrong with your programming skills.
And I think manual memory management is about as silly as manual CPU register management. (Seriously, demonstrate that there is an essential difference, please.)
Shared references, as a general technique, is also insanely useful. It's not just that having to keep track of "owners" is tedious, it also limits your design considerably.
I tend to write ad-hoc reference counting and GC pretty regularly while writing C++.
Name:
Anonymous2011-07-12 17:43
>>58
Manual register allocation is silly, because the compiler will do it better.
Manual memory managment is not silly, because it is very easy, more predictable and more effiecient.
>>59 more effiecient.
Allocation on copying GCs is done in constant time, malloc is not.
A minor garbage collection (assuming a generational GC) is proportional to the amount of live data. Most object die shortly after being allocated. Free is not smart. The GC can do major garbage collections in a separate thread.
>>60
１．Time complexity != execution time
２．Use system calls instead of glibbloat.
３．What do you plan to do on single-processor systems?
４．You are a ``faggot'' for reliying on another things than your cock and your fingers.
Name:
Anonymous2011-07-12 18:35
Manual memory management is stupid unless you're working on embedded systems in C. If you're working in anything higher level than C you have no right to complain about GC and doing so makes you an idiot.
>>61 １．Time complexity != execution time
You just bump a pointer. Like with stack allocation. ２．Use system calls instead of glibbloat.
You still have to search for a free block of memory. ３．What do you plan to do on single-processor systems?
Use an incremental GC. ４．You are a ``faggot'' for reliying on another things than your cock and your fingers.
Make your own processor, with your own instruction set, then write your own assembler for it, then your own programming language and compiler for it, then the library, then the OS.
Name:
Anonymous2011-07-12 20:47
>>59 very easy
defect rates disagree. Garbage collection is easily correlated with fewer and easier to correct bugs. Enjoy ur heap corruptions.
more predictable
What are you trying to predict? If you know your machine has 4gb available, predictability isn't important. Regardless, there are garbage collected languages with EXTREMELY predictable memory usage, such as PicoLisp.
more efficient
not unless you write your own manager. malloc's performance is not that predictable and C doesn't really give you the tools to reason well about cache coherency anyway. If you're going to abstract SOME aspects of memory, you might as well abstract all of it. Why aren't you arguing for assembly? The C stack is a very rudimentary form of automatic memory management. It just sucks. Why does it get a free pass?
If you care about efficiency you ARE going to write your own memory manager. It may or may not be faster and less buggy than one used by a language implementation. Probably not. It may or may not include reference counting or garbage collection, and therefore make everyone's life easier. Probably not.
Name:
Anonymous2011-07-12 20:51
>>59 Manual register allocation is silly, because the compiler will do it better.
Nope. It's silly because it can easily be made automatic. Programmers shouldn't do things that can be made automatic. Optimization is almost never a good reason. For anything. And at almost never conflicts with the principle of automation. Especially with time. Soon garbage collectors will be faster than all but the most skilled* C programmers, just like compilers are now faster than all but the most skilled* assembly programmers.
Manual memory managment is not silly, because the compiler can't do it at all.
FTFY. Seriously guys, static analysis.
Static analysis: it doesn't always make it faster but it never makes it slower at run time.
I don't mind GC usually. I do however avoid java and python for anything with a UI bec ause of it. If you're going to use GC ffs choose a responsive one.
The US Government is due to establish a Federal Programming Administration (FPA) over the next few weeks. During this time, government officials will be screening individuals for administrative positions and will be laying the groundwork for guidelines which all software companies will soon be required to follow.
Under these guidelines, all computer code will be required to meet a level of quality. "Obviously, there is a lot of code in the market already," explains government technical expert Ryan Jennings. "Thus, the FPA will only screen software by a complaint-by-complaint basis." Jennings goes on to describe a couple of the guidelines. "In the future, programs will have to start up in less than a couple of seconds. The programs themselves have to be programmed in one of the many enterprise languages which are widespread in use, such as Java, VB.NET, and PHP. Remember, a positive user experience is everything here and the FPA will ensure that it happens."
US Congress is due to pass a set of laws in the future which will halt the distribution of computer software deemed "low-quality" by the FPA. According to Jennings, "The goal here is to lower and eventually eliminate the already high amount of poor quality software. Top issues in the agenda include the elimination of functional programming, a sub-par programming practice, and particularly open-source software. Open-source software is seen as destructive to the economy so we already have plans to ban its production." The screening process is estimated to be completed in a few days. "It's only a matter of time, but I'm certain this is going to work out. I really see a bright future in the computer industry."
Name:
Anonymous2011-07-14 0:53
>>55 I've never written a concurrent, multithreaded program or anything over 100 LOC and have no idea what the fuck I'm talking about.
A good language should allow you to trivially express common OO abstractions with higher level, simpler constructs, without needing OO explicitly wired into the language syntax and semantics. Then you will realize while using these higher level constructs that you don't fucking want to do OO in the first place.