Return Styles: Pseud0ch, Terminal, Valhalla, NES, Geocities, Blue Moon. Entire thread

GC is shit

Name: Anonymous 2011-09-07 9:18

discuss

Name: Anonymous 2011-09-10 7:24

So is there any speed benchmarks about garbage collection vs manual memory managment? Is there really so big penalty in using GC?

Or is this thread just trolling?

Name: Anonymous 2011-09-10 7:40

>>81
Garbage collection is on par with or better than manual memory management when there are lots of cycles. Otherwise it's much slower.

Name: Anonymous 2011-09-10 8:33

>>81

If there weren't significant performance difference, everyone would use garbage collected languages.

However, the difference is so huge that anything that is important (drivers, kernels, etc) is written with the consideration of the lowest level, so using c,c++ and even assembly.

Name: Anonymous 2011-09-10 9:11

>>83
implying you can't use garbage collectors in assembly.

Name: Anonymous 2011-09-10 9:21

>>84
A language is more than just a tool, it is a mindset. And a programmer with an assembly language mindset would use a garbage collector only in really exceptional circonstances.
Also, some languages implicitly force the implementations to do garbage collection.

P.S. Go back to /g/, ``please''!!

Name: Anonymous 2011-09-10 10:45

check

Name: Anonymous 2011-09-10 10:45

my

Name: Anonymous 2011-09-10 10:45

<---- doubles

Name: Anonymous 2011-09-10 11:20

>>85
A language is more than just a tool, it is a mindset.
For people who have no mindset of their own -- maybe.

Name: Anonymous 2011-09-10 11:30

>>89
Like a wise man once said, a Real Programmer can write Fortran programs in any language.

Name: Anonymous 2011-09-10 11:40

>>90
I just puked a lot in my mouth.

Name: Anonymous 2011-09-10 12:31

>>91
You need to open your mouth, idiot.

Anyways, GC is shit

Name: Anonymous 2011-09-10 12:52

>>92
GC is great. You are shit.

Name: Anonymous 2011-09-10 13:00

>>93
GC is jewish. You are mom.

Name: Anonymous 2011-09-11 3:03

the stack/heap distinction is useless and distracting.

Name: Anonymous 2011-09-12 17:58

GC is shit.

Name: Anonymous 2011-09-12 18:06

GC is clit.

Name: Anonymous 2011-09-12 19:36

>>94
Your mom is jewish. You are fat.

Name: Anonymous 2011-09-13 9:38

one word: Azul's C4 pauseless garbage collector

Name: Anonymous 2011-09-13 13:02

OP here

I just wanted to say I'm glad this managed to get 100 post thread.

Also, GC is shit.

Name: Anonymous 2011-09-13 14:05

GC is great. fuck off and die, faggots

Name: Anonymous 2011-09-13 14:27

http://blog.buschnick.net/2010/02/garbage-collection-considered-harmful.shtml

http://john.freml.in/static-blog/tsac-harmful-garbage/talk.pdf

>>101
Now, how is GC great? It might not be shit, but certainly not great. I think GC is ok for scripting and prototyping, when you want fast results and quality doesn't matter.

I hate it when I have to run slow real world programs written in Java or C#. Damn.

Name: Anonymous 2011-09-13 14:36

>>102
What I like about GC and managed code in general (I hate JVM/.NET, though) is that the worst thing retards can do is write inefficient code. Make the exact same retards write C code and you're on for a load of buffer (under|over)flows, security holes and segmentation faults.

Name: Anonymous 2011-09-13 14:49

>>103
That is true. I hate those retard programmers.

Name: Anonymous 2011-09-13 15:03

>>102
you clearly don't actually understand those articles.

manual memory management is a distraction. There's no reason to manually manage memory and not manually manage CPU registers. It's a hardware detail that C/C++ actually already put a layer of abstraction over (the stack/1D heap abstraction) that isn't even that good.

Everybody who cares about shipping feature-rich good products and doesn't care about hardware-level performance wants automatic memory management.

Whether various garbage collection algorithms (or the idea of GC, as opposed to, say, reference counting) is good is a matter for the guys who write them, not the guy who use them.

Name: Anonymous 2011-09-13 15:11

>>103
That should be exactly what you dislike about managed code in general.  You're essentially saying that you like managed code because it encourages bad programming by softening the consequences.

The only logical conclusion is that you are a 66retard programmer.99  But hey, at least you can stick with managed code and be safe.

Name: Anonymous 2011-09-13 15:24

>>106
Your last paragraph gives me the urge to decapitate you then proceeding to fucking your skull. Leaving that aside, I would much rather have inefficient than insecure software. Sure, in an ideal world where all programmers are superheros and never make any mistakes, managed code is probably pointless. But this is the real world and if managed code is what it will take to prevent fucktards from making my system insecure, then managed code it will be.

Name: Anonymous 2011-09-13 16:23

>>107
So, which is it?  Are you talking about the real world, or fantasy world?  Because this statement:

if managed code is what it will take to prevent fucktards from making my system insecure
is only true in fantasy world.  The only logical conclusion is that you are a 66retard programmer.99  But hey, at least you can stick with managed code and be safe.

Name: Anonymous 2011-09-13 16:57

>>108
Your last paragraph gives me the urge to decapitate you then proceeding to fucking your skull. Leaving that aside, you gave no arguments to your statement.

Name: Anonymous 2011-09-13 17:45

>>108
Dumbass, nobody ever made a security hole in PHP.  Can you say the same about C?

Name: Anonymous 2011-09-13 17:53

>>109
Your last paragraph gives me the urge to decapitate you then proceeding to fucking your skull.
Then I'd say it was equally as successful this time as it was the first time it made you butt ranged.

>you gave no arguments to your statement
I don't think any arguments are really needed to justify the statement that managed code does not make a system secure.  It would save everyone time if you would at least read a Wikipedia article before pretending to be an authority on a subject when you are clearly just a 66retard programmer.99.

>>110
Did you forget [sarcasm] tags or do you really not no anything at all about programming?

Name: Anonymous 2011-09-13 18:35

>>111
Managed code does prevent a lot of security flaws that even experienced programmers may create.

Name: Anonymous 2011-09-13 23:24

the essence of programming is selective detail ignorance.

If you aren't writing a memory manager, you should be ignorant of memory.

Simple.

Name: Anonymous 2011-09-13 23:29

>>113
Now >>111's going to go Postal on your anus.

Name: Anonymous 2011-09-13 23:38

>>113

+1

Name: Anonymous 2011-09-14 0:42

>>38
I like youSegmentation fault

Name: Anonymous 2011-09-14 2:48

>>113
Your example is valid only if you're willing to allow it to be extended to:
If you aren't writing a memory manager, you should be ignorant of memory.  However, if your program manipulates large amounts of data, then you need to write a memory manager.

All you've done is rename the job.  There's some merit to that.

Name: Anonymous 2011-09-14 3:50

However, if your program manipulates large amounts of data, then you need to write a memory manager.

No it doesn't. Lots of general-purpose memory managers have already been written. You need a good reason to reinvent the wheel.

Name: Anonymous 2011-09-14 3:52

>>118
writing your own atoi function and writing your own memory management have a lot in common, actually. Yours will suck, and it's a waste of time anyway.

Name: Anonymous 2011-09-14 4:03

Did you hear?! GC is so shit, that Microsoft got rid of it from the .NET virtual machine for Windows 8 and made it optional.
Newer Posts
Don't change these.
Name: Email:
Entire Thread Thread List
All trademarks and copyrights on this site are owned by their respective parties. All comments and posts are the responsibility of their own posters.
- Tinychan + 2ch Mode -