What's wrong with OO?
>>54633094
It really doesn't enforce good design, the fucktards you hired just think they can't do anything wrong and you end up with an unholy, unmaintainable mess.
Also functional programming sounds like a great idea until you try and do anything useful with it.
>>54633094
There's nothing wrong with OOP.
That image is wrong though. The thing it mostly does is reduce development time for typical business applications as they are all the same fucking MVC based shit. However the quality of the code can be equally shitty as anything if it is being typed by some codemonkey that didn't get a proper degree.
>>54633094
>What's wrong with OO?
The entire premise of it.
OOP as a concept is beautiful and cool
what's wrong is people
>>54633148
Oh lord Im going to post on this and god help me.
I wouldn't really say its a problem with OO or any other type of programming methodology. In the end you have your fundamental data-types that are supported by the CPU. Then you have grouping of those fundamental data-types into a collection (class/struct) that more or less becomes its own type, a type that contains a collection of other types that itself is now a type. In the end you get to a point that all programming languages really fundamentally turn into a graph problem.
In the begging above the notion of Assembly you had C and to a later point in history C++. Though for the time being I will stick with C. It was all functional programming, and everyone was a rave writing program programs with it. Though soon programmers started thinking to themselves or doing it their own way. "Ha wouldn't it be cool instead of using all these function points in struct's we could write a struct with functions that only operated on that type?". In essence coupling the algorithms to that type. Iterate over that a couple of years and you really get the notion of classes that have member fields, constructors or de-constructors.
The notion of Inheritance, came along pretty fast soon after that. From memory it was added soon after with the C++ (originally named C with classes). Though I'm going to stop here and mostly start the rant.
The biggest problem isn't really the fact what language or methodology you choose. To a large part those decisions are made well before your time and the projects you had no control over and now have to maintain a huge mess of code that for what-ever reason has been migrated from flat files, to relational databases to java and then over to spring for java or guava.
The problem is the whole field of Computer Science is laughable at best. Down right fucken retarded stupid if you ask me. You've got a bunch of middle aged men, who think they......(continue)
>>54633658
They know best from reading opinions on message boards, reddit, joelonsoftware, coding horror, slashdot, stackoverflow. Without fully validating the opinons or hell even asking for citation. In other words you have programmers who think they're top shit because you know... you dont need evidence to backup your anecdotal evidence because you're a coding wizz.
You get into situation where I kid you fucken not they just decide to change the framework because its the newest thing on the block!
So have you ever wondered why some resumes require obscene requirements? It's mostly because they've had no control over what the fuck their programmers have been doing for the past 10 years and suddenly they have this giant turd that uses about 20 different technologies from different stacks that work, but are somehow slightly broken so they need a genius with python,c,c++,asm,perl,F#,C#,sharepoint expierence for $40K a year to fix their fucken shithole mess.
In the end, there isn't anything wrong with OOP, or any methodology. The problem is there is a complete lack of evidence based reasoning and evidence based research or consistency from one project to the next because the whole field is run by self appointed experts.
>>54633658
I think you need to work on your sentence construction, that was a difficult read.
C is an imperative programming language as opposed to functional languages such as lisp, haskel and erlang which are derived from the notion of lambda calculus.
Although some concepts within computer science abstract well to graphs, others do not and I'm not saying OO is not a good solution in such cases but it's too often treated as a golden hammer.
My original point was that bad programmers can break nearly anything and when you even begin to hint that your language protects code from bad practices a certain demographic withing the profession take that as a challenge. OO is often contorted into the most fantastic constructs to support absurd extensibility which actually makes maintaining and updating the code more difficult than it originally was to augment. Basically coding IS hard and it should be, we shouldn't be trying to make it easy because it never will be. We should be making it easier to learn how to do difficult and complicated things, too often these languages attempt to do the opposite and in trying to make the difficult things safe and easy transform basic and simple tasks into a nightmare.
>>54633750
You might want to add your name to the list of self appointed experts, but isn't that why we all come to argue on anonymous imageboards?
>>54634071
Been on a grind to get this project out the door and lack of sleep with a large consumption of coffee isn't letting me focus.
I think the goal shouldn't really be extensibility. You're not writing an API that will be used by thousand of developers around the world its you're own framework so no-one will know it better than you.
Instead of falling into the X/Y Trap you should be focusing on getting it out the door ASAP and only when its a commercial success you should focus on developing it further.
So my point? Well I think my point is instead of focusing on making it perfect first, focus on seeing if its commercially successful first. Get the first versions out there, get feedback. If you cannot sell a version move onto the next thing.
