I'm so tired of these shitty "industry standards". Managers that shouldn't be there because they don't have a fucking idea of it read in their monthly EasyPC magazine featuring step-by-step guide to turn on your PC that OO/Java/.NET/XML/bells and whistles are kewl. Then they demand it, pretending to be experts. Hey, everybody else seems to be doing it, so it has to be good.
My boss ditched Java the second he saw what I could do in 2 hours of spare time with PHP.
Name:
Anonymous2005-08-10 17:50
Damn your boss is awesome.
It's the "3-5 years commercial experience" that gets me. Like, where exactly do I get all these years of experience with this huge list of "technologies" if literally every job requires experience? Hell even so-called entry level positions require 1 years experience. And then they bitch about the ongoing software crisis. Idiots.
Name:
Anonymous2005-08-10 18:06
>>5 People ask me why I don't work with computers for a living. This is a great example why. I don't want get burned out on the one thing I enjoy in life.
Name:
Anonymous2005-08-11 7:29
Actually, after working with C# in a professional environment for the last 8 months or so, I must admit that it isn't as bad as all that. It has weaknesses (as all languages do), but I can't deny it's general technical merit for its target.
My primary concern is how tightly it's coupled with MS. All this C# code is going to bite us in the ass one of these days when MS decides to abandon the technology.
Name:
Anonymous2005-08-11 10:26
True, although Mono should soften the blow.
Name:
Anonymous2005-08-11 13:19 (sage)
lol mono
Name:
Anonymous2005-08-11 18:43
>>7
That was why I thought C# would be dropped because of its close tie with M$. C++ was independent and not-commercially developed but rather developed from an academic standpoint. C# was like Microsoft's way of controlling the language.
Anyone knos the major difference between C# and C++? IDE-wise, C# IDEs not just M$ products are they?
Name:
Anonymous2005-08-11 21:12
>>10
C# is garbage-collected and bounds-checked. The syntax is more Java-like.
C# is much more closely related to Java than to C++. I flippantly refer to it as "Java, but with less stupid." Microsoft has added managed code/IL compilation to C++, but that's independent of C#.
Name:
Anonymous2005-08-12 5:36
It's Microsoft's way of telling Sun "yeah we've got that too"
Name:
Anonymous2005-08-12 10:46
How do I pronounce C#? See-Hash?
Name:
Anonymous2005-08-12 10:50
I agree with the "Java, but less stupid" standpoint, but I still prefer a mix of C and C++ armed with a garbage collector.
Also, managed code is bullshit. Why wasting CPU and RAM like that when you can just compile to fucking native code? It's not like Microsoft's gonna support other platforms with it. Managed code is just a marketing produced, hype mongering, magazine fagging, manager fooling, resource wasting, system bloating piece of shit.
>>15
It's because managers demand 100% neatness and organisation in your coding I am guessing. Thats why C# appeals to them because of the managed code. Now THEY can dictate how it should be done.
Name:
Anonymous2005-08-12 16:17
>>17
Response to imaginary boss:
When you learn how to read code, then you stand a chance of telling me how to write it.
Name:
Anonymous2005-08-12 17:36
>>18
So true it hurts. I wish they printed that on the Faggy PC magazines they read.
"Managed code is...shit." (I think I caught the gist, if not the entertaining vocabulary.)
It may well be. I haven't been convinced one way or the other, yet. I think I'll make up my mind regarding managed code when I see how my current product turns out.
I don't deny that the entire brand-awareness marketing push was extremely offputting, though. I tend to agree with you just because they sold it so danged hard. If you've got a technology like this, I prefer the approach of "advertise lightly and let it stand on its merits" to the .NET marketing bonanza.
Name:
Anonymous2005-08-17 19:07
>>20
Well, just name one reason, just one, why you would want managed code, that can't be done with real code faster.
Name:
Anonymous2005-08-17 19:27
>>21
Because the boss that signs your payroll ask you to?
Name:
Anonymous2005-08-18 9:36
>>22
Exactly. That's the reason why people use it.
Name:
Anonymous2005-08-18 11:14
Someone define "managed code" for me so I know what we're arguing about.
I DUNNO ABOUT YUO GUYS BUT I THINK BUFFER OVERRUNS ARE THE MARK OF A REEL PROGRAMMER SO IM GONNA KEEP USING C BECAUSE IT IS THE ONE TRUE LANGUAGE
Name:
Anonymous2005-08-19 8:09
>>26 yeah 'cos it's sooooo hard to write decent code and there's no alternative to routines that are vulnerable to buffer overruns. :gonk:
Name:
Anonymous2005-08-19 11:10
>>27 yeah which is why none of the security bulletins ever have anything to do with buffer overruns!
Name:
Anonymous2005-08-20 8:16
>>24
Virtualizing your non-portable code for the absolute heck of it (and having shit said on magazines so stupid bosses want it). Think the Java VM, only better, but not portable, so WTF would you want to compile for that instead of >>25 real code: code that can be run directly by your processor.
>>26
Exactly. Actually, buffer underruns are not a problem in C, but the C standard library. It's your problem if you use it. Real programmers don't need standard libraries :P .
Name:
Anonymous2005-08-20 10:42
A language that considers a[i] to be equivalent to *(a + i) has nothing to do with the standard library.
If I wanted to be safe I would have to wrap all arrays in a struct that also contains the array length. Then I would have to access the struct through specific functions which automatically check the length before trying to access, and return error codes if the index is out of bounds. If I were using someone else's code in my program I would have to either edit their code to use my structs instead of raw arrays, or just trust that they are awesome programmers who never allow a situation where an overrun could occur.
I'm sure there are libraries that do what I described; it's trivial, I could write a simple one in an hour. The point remains that I'd still have to check other people's code if I wanted to use it.
Would it really have been so bad to make this a core language feature? It would have saved us a lifetime of hacked webservers.
Name:
Anonymous2005-08-20 17:35
Would it really have been so bad to make this a core language feature?
C is an older language that sits right next to assembly. Back then they counted every memory access and cycle. Even now, in frequently-called inner loops, such checks can slow things down dramatically.
C is a pleasant and portable assembler language, nothing more; if you want the safety you'll have to add it yourself. The problem isn't C, but rather people using it in inappropriate domains.
The standard library could desperately use an update though.
Name:
Anonymous2005-08-20 18:03
>> The problem isn't C, but rather people using it in inappropriate domains.
This is precisely what I've always said. It's great at writing device drivers, but keep it away from my web server.
Name:
Anonymous2005-08-21 7:05
>>30
What's the problem with simple pointer arithmetics? That's one of the advantages of C. Why do you want bound checking? That's for Basic kids. Proper programs don't require to waste time bound checking; however, if you want that, you can overload operator[] , something you can't dream of doing in Java.
Oh, I see what you mean about others' code. But then again, if they are bad programmers who write buggy code, you shouldn't be so interested on using it. You should use more trusted libraries, or, at least, have a look at 5% of the code so you get a glimpse of how good it looks.
Perhaps bound checking is not so bad for Internet software, although you don't really want that done when you're coding low-level software. Remember C is the fastest thing ever (now that compilers will do a better job of pairing and optimizing than an assembly guru). You gotta keep it the fastest thing ever; no point in having a bunch of slower languages and no lower level better performing language.
>>31
What's killing C is the lack of a decent standard library. They went for stdin/stdout only, with a couple very crappy date and string functions. If it had a strong, comprehensive standard library, with proper strings and Unicode support, etc., it would be even more popular than it is. Think a standard library as powerful as Java's, only not a piece of OO shit like Java's.
Name:
Anonymous2005-08-21 7:24
>> Proper programs don't require to waste time bound checking
Proper programs show up on the security alerts every week.
>> you can overload operator[]
Surely C doesn't support overloading? I thought that was only for C++.
Name:
Anonymous2005-08-21 9:55
>>34
Not strict C99, but C++-- which everybody uses does.
Name:
Anonymous2005-08-21 9:59
>>33
But you don't always have the option of choosing who's code to work with. Sometimes, you're forced to use a specific library or you have to pick up where your sucessor left off.
For what it's worth though, I don't think bounds checking should be built into C. But implementing it via a library would lead to crumbleson syntax I guess. getArrayItem( a, i )
Name:
Anonymous2005-08-21 10:36
>>36
You can try overloading in C++--, or have function pointers in the structure for a get() and put() functions (zomg is this OO!? Magic!).
Name:
Anonymous2005-08-21 11:43
>>37
Not technically OO, but halfway there. Add in a mechanism for inheritance and polymorphism (encapsulation would be a bonus too) and you have a winrar.
>>35
Well in the past when I tried to overload a function the compiler (gcc) complained that the function was already defined.
Name:
Anonymous2005-08-29 14:05
Is it true that C# and ASP has much better OO capabilities? Some even dare say that C# is the only language that can do true OO.
>>44
Yes but it has so-called "generics" which attempt to bring some of the usefulness of dynamic typing to the language without actually allowing the programmer any extra freedom.
Name:
Anonymous2005-08-31 17:29
Generics are a useful construct for containers. They lessen the pathological casting common in languages that pretend to be statically typed while deriving everything from the ubiquitous "Object".
Generics are good at what they do. That's all they're good at. Unlike C++ generics/templates, of course, which are really *really* good at not only parameterizing compilation of containers, but *also* making extensive and complex code even *harder* to read.
Name:
Anonymous2005-09-01 6:21
Yes, like I said, they're a poor attempt to compensate for the inadequacies of static typing.
Name:
Anonymous2005-09-01 10:34
>>47 has obviously never used a real statically typed language
>>50
I've used Ada, it does OO, it has generics, a very nice package system and it merges its type system with its class system way more effectively than Java.
>>52
ladida. OCaml/haskell has polymorphism(generics,sort of) that doesn't hurt the eyes. and Haskell has type classes. complain about static typeing with that in the equation was what i wanted to say
>>57
Given that >>54 was in response to "No True Scotsman", please explain the meaning that I am so obviously failing to comprehend due to my poor reading skills.
Also, lol @ the bumping from 2005. You get that flux capacitor working again, Doc?
Name:
HAX MY ANUS MEME FAN2008-05-01 11:30
Yet another good thread has been tainted by the hax my anus meme.
Also, managed code is still shit.
Name:
Anonymous2009-11-02 23:12
>>26,29,31 Actually, buffer underruns are not a problem in C, but the C standard library. It's your problem if you use it. Real programmers don't need standard libraries
>>33
Vouch for this. I hope they improve and expand the standard library when C1x comes out (whenever that will be).
>>66
Oops, was about to say that the comment I quoted is fucking retarded. Are you really going to try to use C with no standard library because you're too fucking stupid to use it safely? It's really not that difficult.
Name:
Anonymous2009-11-03 3:57
>>15
This. The CLR even compiles the code right before it's run, which is bad because you can't spend too much time doing that or the startup time will be slow (as if it isn't already), and as anyone should know, the faster the compiler the less optimized the code it generates.
>>67
I've written a program that didn't use the C standard library, since I didn't need the bloat
Name:
Anonymous2009-11-03 9:11
>>4
doesnt that just mean you're more skilled with PHP? i could probably code whatever the fuck it is you coded in 2 days of PHP simply because i dont have much experience in it.
Name:
Anonymous2009-11-03 13:41
I can safely say that I believe Sharpples is way better than Sepples for simple tools and speed of development.