C#/ / .NET Observation
-
It does not surprise me that people did not use std::map, most people simply used CArray and didn't move beyond that. Personally, I used std::map all the time, although not for collections of numbers or collections of 5 objects :-) The syndrome is the same in both cases, only a subset of the people using a language are interested in learning to use it well.
Christian Graus - Microsoft MVP - C++ "also I don't think "TranslateOneToTwoBillion OneHundredAndFortySevenMillion FourHundredAndEightyThreeThousand SixHundredAndFortySeven()" is a very good choice for a function name" - SpacixOne ( offering help to someone who really needed it ) ( spaces added for the benefit of people running at < 1280x1024 )
It's not just the abuse of reflection which is abundant but as mentioned blind 'hey it'll handle it for me' CLR collection following and not understanding how expensive those, default and often dev induced, lousy Equality tests are. There is much hacking there and MS attempting to fix by-large an inefficient and at times broken model. It isn't funny, IEquatable, IComparable, casts, boxing even when many think they are avoiding it, virtual dispatch, devs not understanding good hash impls, you name it. I mean it is obvious when a C# guy asks for VB.NET to be converted and vice versa. And I agree whenever a return to C++ is due, it ends up always refreshing. Even if you have no System.Object, KeyValueDictionary generic is involved in hash as well, it makes most of the code involved, or better put modeling/naming with it suck too ( Heck the hash in object is the Java idea, but I'll agree hash has its purposes and at least one dynamic language implementation excels on it). Anyway, after 7 years of working with collections in .NET I think it just shows how badly the entire show is for Key/Value, value vs reference, and runtime binding.. Don't even get me on multiple inheritance, something no matter what runtime zealots say is bad, it just blasts away all key/value attempts to 100% .NET runtime blindess. It is all meant to be about building types, but when you look at it CLR/Java mentality prefers to get a guy thinking he is up and running quick and cool. And after a while you just see the effect it has on software quality: Incorectness/incompleteness, RAM and CPU over-utilisation.
-
It's not just the abuse of reflection which is abundant but as mentioned blind 'hey it'll handle it for me' CLR collection following and not understanding how expensive those, default and often dev induced, lousy Equality tests are. There is much hacking there and MS attempting to fix by-large an inefficient and at times broken model. It isn't funny, IEquatable, IComparable, casts, boxing even when many think they are avoiding it, virtual dispatch, devs not understanding good hash impls, you name it. I mean it is obvious when a C# guy asks for VB.NET to be converted and vice versa. And I agree whenever a return to C++ is due, it ends up always refreshing. Even if you have no System.Object, KeyValueDictionary generic is involved in hash as well, it makes most of the code involved, or better put modeling/naming with it suck too ( Heck the hash in object is the Java idea, but I'll agree hash has its purposes and at least one dynamic language implementation excels on it). Anyway, after 7 years of working with collections in .NET I think it just shows how badly the entire show is for Key/Value, value vs reference, and runtime binding.. Don't even get me on multiple inheritance, something no matter what runtime zealots say is bad, it just blasts away all key/value attempts to 100% .NET runtime blindess. It is all meant to be about building types, but when you look at it CLR/Java mentality prefers to get a guy thinking he is up and running quick and cool. And after a while you just see the effect it has on software quality: Incorectness/incompleteness, RAM and CPU over-utilisation.
-
User of Users Group wrote:
and MS attempting to fix by-large an inefficient and at times broken model.
Mostly the model is broken by the loose nut between the chair and the keyboard.
led mike
Yeah right, I can see your post below and you are probably not aware (or the guy voting you 5, plenty of those runtime lovers around) most of those Generics still use the hash (and some broken/unchecked runtime comparisons if you prefer), it has nothing to do with 1.1 :) Besides when they initially did that Rotor hack in MSR, many people in languages field (and less broken models) just went: *sigh*. Generics (List and some other bits aside) are so weak in comparison to alternatives by not just C++, it doesn't surprise me to see people believe they are sooo, oh soo 'powerful', or thinking that by default they really eliminate boxing or virtual dispatch (that is the type responsibility and implementing one as complete as MS does for its own, shows you how much bloat is involved to express a simple primitive).
modified on Thursday, January 31, 2008 3:07:44 PM
-
When I was working with C# before happily returning back to C++ one thing I noticed being *really* abused was the reflection. I understand there are valid uses of reflection/introspection - mostly for the dev tools and code analysis - but creating the complete UI out of some crazy XML files with reflection was simply wrong - it made the application less robust and it would take forever to start.
Nemanja Trifunovic wrote:
but creating the complete UI out of some crazy XML files with reflection was simply wrong
(cough)XAML(cough)(gag)(spittle)
¡El diablo está en mis pantalones! ¡Mire, mire! Real Mentats use only 100% pure, unfooled around with Sapho Juice(tm)! SELECT * FROM User WHERE Clue > 0 0 rows returned Save an Orange - Use the VCF! VCF Blog
-
I have never knowingly used a hashtable in C#, and I stayed as far away as possible from STL when I was doing C++.
"Why don't you tie a kerosene-soaked rag around your ankles so the ants won't climb up and eat your candy ass..." - Dale Earnhardt, 1997
-----
"...the staggering layers of obscenity in your statement make it a work of art on so many levels." - Jason Jystad, 10/26/2001Why? Associative arrays have all sort of legitimate uses. Why wouldn't you use them?
¡El diablo está en mis pantalones! ¡Mire, mire! Real Mentats use only 100% pure, unfooled around with Sapho Juice(tm)! SELECT * FROM User WHERE Clue > 0 0 rows returned Save an Orange - Use the VCF! VCF Blog
-
Yeah right, I can see your post below and you are probably not aware (or the guy voting you 5, plenty of those runtime lovers around) most of those Generics still use the hash (and some broken/unchecked runtime comparisons if you prefer), it has nothing to do with 1.1 :) Besides when they initially did that Rotor hack in MSR, many people in languages field (and less broken models) just went: *sigh*. Generics (List and some other bits aside) are so weak in comparison to alternatives by not just C++, it doesn't surprise me to see people believe they are sooo, oh soo 'powerful', or thinking that by default they really eliminate boxing or virtual dispatch (that is the type responsibility and implementing one as complete as MS does for its own, shows you how much bloat is involved to express a simple primitive).
modified on Thursday, January 31, 2008 3:07:44 PM
User of Users Group wrote:
Generics still use the hash
Yes i reacted to
Hashtableand apparently missed the overall point of the thread, oops :doh:User of Users Group wrote:
it has nothing to do with 1.1
Of course since i missed the point I was talking about this.[^] Generics are the most powerful feature of C# 2.0. Generics allow you to define type-safe data structures, without committing to actual data types. This results in a significant performance boost and higher quality code, because you get to reuse data processing algorithms without duplicating type-specific code. and[^] Collision Resolution in the Dictionary Class
The Dictionary class differs from the Hashtable class in more ways than one. In addition to being strongly-typed, the Dictionary also employs a different collision resolution strategy than the Hashtable class, using a technique referred to as chaining. Recall that with probing, in the event of a collision another slot in the list of buckets is tried. (With rehashing, the hash is recomputed, and that new slot is tried.) With chaining, however, a secondary data structure is utilized to hold any collisions. Specifically, each slot in the Dictionary has an array of elements that map to that bucket. In the event of a collision, the colliding element is prepended to the bucket's list.
User of Users Group wrote:
you are probably not aware
If and when that is true, I don't know nor will I ever know everything, I am the loose nut breaking things. You might have read my initial post as excluding myself as the nut causing the problems, not so.
Last modified: 31mins after originally posted --
-
Yeah right, I can see your post below and you are probably not aware (or the guy voting you 5, plenty of those runtime lovers around) most of those Generics still use the hash (and some broken/unchecked runtime comparisons if you prefer), it has nothing to do with 1.1 :) Besides when they initially did that Rotor hack in MSR, many people in languages field (and less broken models) just went: *sigh*. Generics (List and some other bits aside) are so weak in comparison to alternatives by not just C++, it doesn't surprise me to see people believe they are sooo, oh soo 'powerful', or thinking that by default they really eliminate boxing or virtual dispatch (that is the type responsibility and implementing one as complete as MS does for its own, shows you how much bloat is involved to express a simple primitive).
modified on Thursday, January 31, 2008 3:07:44 PM
User of Users Group wrote:
or the guy voting you 5, plenty of those runtime lovers around)
actually the guy voting him a 5 may support the statement but not the stance... ;) often cases, most models, in mosts languages, are broken by the nut behind the keyboard. The result is often something that works small but doesn't scale, or that needs rewriting consistently over years, or any number of other things that happen to it. supporting that statement neither supports nor refutes the use of generics, lambda operations or any number of newer language features over the years. Almost anything "can" be used for good, but usually the nut behind the keyboard is holding back either C++ or C# or even any other language you or I or the world can invent. :-D :-D
_________________________ Asu no koto o ieba, tenjo de nezumi ga warau. Talk about things of tomorrow and the mice in the ceiling laugh. (Japanese Proverb)
-
James R. Twine wrote:
I have seen LOTS of uses of the Hashtable type
In Pre .NET 2.0 code. 2.0 introduced Generics and everything collection wise changed drastically. All old collection code should be migrated to using Generics, well there are exceptions of course but as a rule of thumb, just do it. So stop looking at 1.X code. ;P
led mike
Plenty of loose nuts in ASP.NET source, WPF dragging itself, and default provision of method implementation by CLR in many, many types :-) Go on, look at their source, use shift operator on the bits, all in an idea of runtime that is supposed to abstract blitting and bits.. You'll end up with Java or Java IDE feeling when users stress your app, reasons are outlined in that generic, one that cannot even properly compare value types, and frankly that is a disaster, mass-market, language design, Anders Legend-Lover or not. But you knew that anyway..
-
As I now need to start getting into C# after a very happy and fruitful live with C++, I have noticed something and wanted to know if any other pre-C#/.NET developers have noticed the same thing. I have seen LOTS of uses of the
Hashtabletype - far more than I would expect. And in places where I would not normally expect. For example, I have seen keys like "1", "2", "3", etc. (i.e. simple indexers), or seen them used for collections of less than 5 items or so. It just seems a bit weird to me - I mean,std::mapandstd::multimaphave been around for a number of years, and I have seen plenty of complex systems that rarely used them. But now I see the simplest systems _(ab)_useHashtable. Is it just me, or does anyone else notice that as well? I am not sure if it is the child with a hammer syndrome or not... What do you think? Peace!-=- James
Please rate this message - let me know if I helped or not! * * *If you think it costs a lot to do it right, just wait until you find out how much it costs to do it wrong!
Remember that Professional Driver on Closed Course does not mean your Dumb Ass on a Public Road!
See DeleteFXPFilesGiven that I do C++,not C#...when I use STL, I am pretty aware of the big-O characteristics of the different containers - they're specified in the standard, for crying out loud! C#...dunno - what are the big-O characteristics? I'd presume they are similar to the STL containers...but I really don't know. Also - C#..is it VB6 for the 00's? Not entirely, I'm sure, but I think the VB6 programmers have moved to C# :-)
-
Nemanja Trifunovic wrote:
but creating the complete UI out of some crazy XML files with reflection was simply wrong
(cough)XAML(cough)(gag)(spittle)
¡El diablo está en mis pantalones! ¡Mire, mire! Real Mentats use only 100% pure, unfooled around with Sapho Juice(tm)! SELECT * FROM User WHERE Clue > 0 0 rows returned Save an Orange - Use the VCF! VCF Blog
Jim Crafton wrote:
(cough)XAML(cough)(gag)(spittle)
Nay. MyXAML - that's the way. ;)
Deja View - the feeling that you've seen this post before.
-
protected KeyedList<ViewListenerKey, List<SessionContainerListener>> viewListeners;
protected KeyedList<ViewListenerKey, ViewKeyData> viewKeyData;
protected KeyedList<SessionContainerViewListenerKey, List<DataTable>> sessionTransactions;Run. Don't walk, run. :-D Marc
Marc Clifton wrote:
protected KeyedList> viewListeners;protected KeyedList viewKeyData;protected KeyedList> sessionTransactions;
Be scared. Be very scared. Generic List Guy's about.
Deja View - the feeling that you've seen this post before.
-
User of Users Group wrote:
Generics still use the hash
Yes i reacted to
Hashtableand apparently missed the overall point of the thread, oops :doh:User of Users Group wrote:
it has nothing to do with 1.1
Of course since i missed the point I was talking about this.[^] Generics are the most powerful feature of C# 2.0. Generics allow you to define type-safe data structures, without committing to actual data types. This results in a significant performance boost and higher quality code, because you get to reuse data processing algorithms without duplicating type-specific code. and[^] Collision Resolution in the Dictionary Class
The Dictionary class differs from the Hashtable class in more ways than one. In addition to being strongly-typed, the Dictionary also employs a different collision resolution strategy than the Hashtable class, using a technique referred to as chaining. Recall that with probing, in the event of a collision another slot in the list of buckets is tried. (With rehashing, the hash is recomputed, and that new slot is tried.) With chaining, however, a secondary data structure is utilized to hold any collisions. Specifically, each slot in the Dictionary has an array of elements that map to that bucket. In the event of a collision, the colliding element is prepended to the bucket's list.
User of Users Group wrote:
you are probably not aware
If and when that is true, I don't know nor will I ever know everything, I am the loose nut breaking things. You might have read my initial post as excluding myself as the nut causing the problems, not so.
Last modified: 31mins after originally posted --
led mike wrote:
because you get to reuse data processing algorithms without duplicating type-specific code.
Although this is a good thing for some things, again as mentioned it is often used and abused in areas where use of generics do not save the algorithm because the algorithm cannot truly be generic. The case often used for examples of generics is a sort where identification of a number or string or other data item only requires identification of less than or greater than values, such capabilities existed with overloads, though generics allow for one typing, as do templates, etc. But the actual use in the field would probably be an algorithm for line of site for RF which is highly dependant upon frequency and the algorithm actually has to change at various frequencies along the way. Because the guy behind the keyboard thinks generics are a magical automatic process of saving his algorithm for reuse without any real knowledge of what it does, or how it works, or even if it applies to the process in question that he is coding. Such "magical saving graces" are more danger than savings because of that guy behind the keyboard. magic doesn't save anyone, if anything knowledge does, not lack therein. All of these things are more powerful when you put thought behind them, not when they are advertised as basis for less thought. That is where the danger lies.
_________________________ Asu no koto o ieba, tenjo de nezumi ga warau. Talk about things of tomorrow and the mice in the ceiling laugh. (Japanese Proverb)
-
As I now need to start getting into C# after a very happy and fruitful live with C++, I have noticed something and wanted to know if any other pre-C#/.NET developers have noticed the same thing. I have seen LOTS of uses of the
Hashtabletype - far more than I would expect. And in places where I would not normally expect. For example, I have seen keys like "1", "2", "3", etc. (i.e. simple indexers), or seen them used for collections of less than 5 items or so. It just seems a bit weird to me - I mean,std::mapandstd::multimaphave been around for a number of years, and I have seen plenty of complex systems that rarely used them. But now I see the simplest systems _(ab)_useHashtable. Is it just me, or does anyone else notice that as well? I am not sure if it is the child with a hammer syndrome or not... What do you think? Peace!-=- James
Please rate this message - let me know if I helped or not! * * *If you think it costs a lot to do it right, just wait until you find out how much it costs to do it wrong!
Remember that Professional Driver on Closed Course does not mean your Dumb Ass on a Public Road!
See DeleteFXPFilesJames R. Twine wrote:
I am not sure if it is the child with a hammer syndrome or not... What do you think?
My concern is more to do with the standard of developer let loose on the language. We keep seeing posts stating "How do I pass value from form 1 to form 2" and it's obvious that the developers aren't even bothering to think. Just because you can do something quickly in .NET it doesn't mean that you can ignore the basics of design. There's nothing inherently wrong with using a hashtable or a dictionary of some type, but there is something wrong with using it where a simple array would suffice.
Deja View - the feeling that you've seen this post before.
-
James R. Twine wrote:
I am not sure if it is the child with a hammer syndrome or not... What do you think?
My concern is more to do with the standard of developer let loose on the language. We keep seeing posts stating "How do I pass value from form 1 to form 2" and it's obvious that the developers aren't even bothering to think. Just because you can do something quickly in .NET it doesn't mean that you can ignore the basics of design. There's nothing inherently wrong with using a hashtable or a dictionary of some type, but there is something wrong with using it where a simple array would suffice.
Deja View - the feeling that you've seen this post before.
Sadly yes. This is the root of all evil; trying to tackle a problem without understanding what's needed first. Proper design comes after proper understanding and that comes from using your brain. These days, that's becoming as rare as hen's teeth.
"Every time Lotus Notes starts up, somewhere a puppy, a kitten, a lamb, and a baby seal are killed. Lotus Notes is a conspiracy by the forces of Satan to drive us over the brink into madness. The CRC-32 for each file in the installation includes the numbers 666." Gary Wheeler "The secret to a long and healthy life is simple. Don't get ill and don't die." Pete O'Hanlon, courtesy of Rama Krishna Vavilala
-
Given that I do C++,not C#...when I use STL, I am pretty aware of the big-O characteristics of the different containers - they're specified in the standard, for crying out loud! C#...dunno - what are the big-O characteristics? I'd presume they are similar to the STL containers...but I really don't know. Also - C#..is it VB6 for the 00's? Not entirely, I'm sure, but I think the VB6 programmers have moved to C# :-)
Stuart Dootson wrote:
I think the VB6 programmers have moved to C#
No, they either learned what OO is and moved to VB.NET or they got promoted into a position where they say, "back in the old days."
"Every time Lotus Notes starts up, somewhere a puppy, a kitten, a lamb, and a baby seal are killed. Lotus Notes is a conspiracy by the forces of Satan to drive us over the brink into madness. The CRC-32 for each file in the installation includes the numbers 666." Gary Wheeler "The secret to a long and healthy life is simple. Don't get ill and don't die." Pete O'Hanlon, courtesy of Rama Krishna Vavilala
-
As I now need to start getting into C# after a very happy and fruitful live with C++, I have noticed something and wanted to know if any other pre-C#/.NET developers have noticed the same thing. I have seen LOTS of uses of the
Hashtabletype - far more than I would expect. And in places where I would not normally expect. For example, I have seen keys like "1", "2", "3", etc. (i.e. simple indexers), or seen them used for collections of less than 5 items or so. It just seems a bit weird to me - I mean,std::mapandstd::multimaphave been around for a number of years, and I have seen plenty of complex systems that rarely used them. But now I see the simplest systems _(ab)_useHashtable. Is it just me, or does anyone else notice that as well? I am not sure if it is the child with a hammer syndrome or not... What do you think? Peace!-=- James
Please rate this message - let me know if I helped or not! * * *If you think it costs a lot to do it right, just wait until you find out how much it costs to do it wrong!
Remember that Professional Driver on Closed Course does not mean your Dumb Ass on a Public Road!
See DeleteFXPFilesNow that we have LINQ, I can see the List<> types getting some heavy abuse. Too easy with LINQ to sort, group, calcuate, etc all from a simple List<>. Of course, LINQ works with other objects, but I got a felling some developers will think list about the needs of an app and head straight for simple, quick to implement solutions.
Rocky <>< Blog Post: Handy utility app that is always on my machines! Tech Blog Post: Moving on up with Windows Live stuff and Plus!
-
Sadly yes. This is the root of all evil; trying to tackle a problem without understanding what's needed first. Proper design comes after proper understanding and that comes from using your brain. These days, that's becoming as rare as hen's teeth.
"Every time Lotus Notes starts up, somewhere a puppy, a kitten, a lamb, and a baby seal are killed. Lotus Notes is a conspiracy by the forces of Satan to drive us over the brink into madness. The CRC-32 for each file in the installation includes the numbers 666." Gary Wheeler "The secret to a long and healthy life is simple. Don't get ill and don't die." Pete O'Hanlon, courtesy of Rama Krishna Vavilala
Mustafa Ismail Mustafa wrote:
as rare as hen's teeth.
actually I think Hen's teeth are far more common.
_________________________ Asu no koto o ieba, tenjo de nezumi ga warau. Talk about things of tomorrow and the mice in the ceiling laugh. (Japanese Proverb)
-
Now that we have LINQ, I can see the List<> types getting some heavy abuse. Too easy with LINQ to sort, group, calcuate, etc all from a simple List<>. Of course, LINQ works with other objects, but I got a felling some developers will think list about the needs of an app and head straight for simple, quick to implement solutions.
Rocky <>< Blog Post: Handy utility app that is always on my machines! Tech Blog Post: Moving on up with Windows Live stuff and Plus!
I completely agree with everyone on this, but someone has put it very simply back in the day (forget who it was). "Professional programming isn't for the masses, MS is going against that. Same goes for similar Wrox titles." I disagree just a tad, because obviously .NET has pushed out the envelope and produced some good stuff, but it is producing big-heads far faster. In my mind, to consider oneself as a master of abstraction is very hard for me, and I wonder who would when they look at the number of times they refactor their code (if they can afford it), or say when they look Boost and go WTF (that is the leadership there, if you care about types. Sure it has its quirks in few areas , but if anything it demonstrates what kind of true abstract heads we are up against; Hibernate and Castle people, please stay clear as it is beyond ORM-loving childplay)
-
I completely agree with everyone on this, but someone has put it very simply back in the day (forget who it was). "Professional programming isn't for the masses, MS is going against that. Same goes for similar Wrox titles." I disagree just a tad, because obviously .NET has pushed out the envelope and produced some good stuff, but it is producing big-heads far faster. In my mind, to consider oneself as a master of abstraction is very hard for me, and I wonder who would when they look at the number of times they refactor their code (if they can afford it), or say when they look Boost and go WTF (that is the leadership there, if you care about types. Sure it has its quirks in few areas , but if anything it demonstrates what kind of true abstract heads we are up against; Hibernate and Castle people, please stay clear as it is beyond ORM-loving childplay)
User of Users Group wrote:
"Professional programming isn't for the masses, MS is going against that. Same goes for similar Wrox titles." I disagree just a tad, because obviously .NET has pushed out the envelope and produced some good stuff, but it is producing big-heads far faster.
I'm not sure it's quite that - what I think is true is that there is more of a take up by people who are inherently unsuited to developing. Good developers have a breadth of knowledge and an interest in expanding their knowledge in many different areas, by learning other languages or by trying different architectures/techniques. Then there are the others - the people who I would term The Google Cut and Paste Monkeys. They want answers spoon fed to them, and they look for the next quick coding hit/new technology without bothering to look into them at all. I blame MS for their statement in the early 90s that they wanted to create systems that would remove the need for developers.
Deja View - the feeling that you've seen this post before.
-
User of Users Group wrote:
"Professional programming isn't for the masses, MS is going against that. Same goes for similar Wrox titles." I disagree just a tad, because obviously .NET has pushed out the envelope and produced some good stuff, but it is producing big-heads far faster.
I'm not sure it's quite that - what I think is true is that there is more of a take up by people who are inherently unsuited to developing. Good developers have a breadth of knowledge and an interest in expanding their knowledge in many different areas, by learning other languages or by trying different architectures/techniques. Then there are the others - the people who I would term The Google Cut and Paste Monkeys. They want answers spoon fed to them, and they look for the next quick coding hit/new technology without bothering to look into them at all. I blame MS for their statement in the early 90s that they wanted to create systems that would remove the need for developers.
Deja View - the feeling that you've seen this post before.
> that would remove the need for developers. Rings some bells, and btw this is what I believe DB is talking about in his recent blog entry on 'overfactoring forewords', you know generics and that. After all, WCF is one of the pieces you see that is clearly showing less development effort and more configuration expertise( hey hate the word but that is what it is: IT!). WPF aims to take it into media and design space too. But I am supripsed at how repetative/similar it is to past tech, and especially as WS falters. So while it does look like a step forward in terms of factoring, at least in that ahm *research* attempt (bold to say but I don't think WinFX will really rock the world), apart from some awful looking code when dealing with communication endpoints. Hey if it produces a good idea where you pay slightly less in XML and runtime penalties down the road I am all up for it, but that is what is lacking out there.. implementation of environments that do not introduce any runtime penalty whatsoever. And that, ties into the topic, which is compiler work I believe (to 'ease'/force-to-think and warn very early; best of both worlds etc).. Pretty tough subject for me at the very least.
