Would you agree this is the definition of a PHP framework?

[TomB]

@tony_marston404

I suggest you learn to read.

Please link me to a post where you’ve made an actual apples to apples comparison. E.g. You have shown more than one way a problem can be solved and then identified the strengths/weaknesses of each approach.

[tony_marston404]

@TomB

tony_marston404:
If it does not fulfil the conditions of a framework then it is not a framework.

Tony Marson logic. Redefine common terms on a whim and then tell everyone else they’re wrong for using the well known definition.

Hark at the pot calling the kettle black! You are the one who redefines the circumstances under which Dependency Injection provides benefits, then tells me that I am wrong for not agreeing with your definition.

[TechnoBear]

OK - enough. Let’s get back to discussing issues, rather than each other.

[mAAdhaTTah]

@tony_marston404

This tells me that the framework is in control, not the developer, that the framework should provide default behaviour, and the developer should only need to write code when this default behaviour needs to be changed. How else can this be accomplished?

The framework is in control. Writing:

$app = new Application();
$app->boot();

puts the framework in control. The default behavior is to run your application. All you’re in control of is starting the Application; you are not in control of how the application runs–the framework is.

[s_molinari]

@tony_marston404

certain people don’t like their rules being ignored or questioned

This isn’t quite correct. It isn’t a problem questioning. It’s just that you don’t just question, you also say your code is just as good (or at times one would think you think it is better). It isn’t better. We’ve proven that without a doubt.

@tony_marston404

There are some basic programming practices which are universally accepted as being “good” and have been around for decades

And one of them has NEVER been a 9000 loc monster class. Yet you insist, it is ok. You don’t even realize, your argumenting over it being “ok” just takes the “competency” legs right from underneath you.

@tony_marston404

Why not? Unlike you I do not write “use once then throw away” applications,

Nobody, who programs, goes through the effort to throw away the code after a single use. That is a silly supposition to make and rather lame attempt to make yourself out to be a better programmer, which you are not.

And you are married to your code badly. You can’t even look at other developer’s code without having to compare it to yours and to find fault with theirs, so you can push yours up as “the better framework”. You can’t learn from others, without looking at your code and basically cherry picking the written articles on coding theory to try and show how you’ve done things differently and in places incorrectly, yet try and show it is right. “The experts have said so.”

Scott

[tony_marston404]

@TomB

tony_marston404:
default behavior: A framework has a default behavior. This default behavior must be some useful behavior and not a series of no-ops.[citation needed]

You put [citation needed] in your own post rather than providing a citation?

That text was copied directly from https://en.wikipedia.org/wiki/Software_framework

@TomB

Let’s find some actual primary sources for software frameworks.

Just because you find articles which offer a different definition than the one found in https://en.wikipedia.org/wiki/Software_framework does not means that those articles are correct. As far as I am concerned they are merely offering different opinions, and I do not agree with those opinions.

[TomB]

@tony_marston404

Just because you find articles which offer a different definition than the one found in https://en.wikipedia.org/wiki/Software_framework does not means that those articles are correct. As far as I am concerned they are merely offering different opinions, and I do not agree with those opinions.

So you’re seriously suggesting we should take a line on a wikipedia page which has [citation needed] as more authoritative than a peer reviewed paper from CERN or Manchester University.

[TechnoBear]

This topic is now closed. I will reopen it in ten minutes time, which I hope will have given everyone the chance to calm down and think of less personal responses.

[TechnoBear]

This topic is now opened. New replies are allowed - but please keep them civil.

[tony_marston404]

@s_molinari

Tony, why don’t you quote the fourth description?

I deliberately missed out the fourth line as it was not relevant to the particular question I was answering.

@s_molinari

Does the dev have to modify your code, to get their own customizations to work?

No they do not. All the Controllers, Views, Data Access Objects and the abstract class which is inherited by all Models are built into the framework and do not have to be modified by any developer at any time.

[TomB]

@tony_marston404

I deliberately missed out the fourth line as it was not relevant to the particular question I was answering.

Cherry picking

[tony_marston404]

@TomB

tony_marston404:
so at this stage in my career I see very little opportunity for anything other than very minor improvements.

Unit testing? Separation of concerns? Better modularity?

Unit testing would be far too expensive to implement at this stage in a large enterprise application.

As for separation od concerns and better modularity, I have already stated several times that my framework is based around a combination of the 3-Tier architecture and the MVC design pattern, so that it has all the separation and modularity it needs. By taking my existing cohesive modules and breaking them down into artificially small units I would end up with nothing more than a collection of fragments in which all cohesion would be lost.

[TomB]

@tony_marston404

Unit testing would be far too expensive to implement at this stage in a large enterprise application.

That wasn’t the question though was it

@fretburner

You don’t believe it’s possible for you to improve as a developer (i.e. you’re at the peak of your profession)?

[tony_marston404]

@s_molinari

This isn’t quite correct. It isn’t a problem questioning. It’s just that you don’t just question, you also say your code is just as good (or at times one would think you think it is better).

Incorrect. I simply say that my results are just as good but that the code I use to achieve those results is different.

@s_molinari

It isn’t better. We’ve proven that without a doubt.

You may have proven it to yourself but not to me.

@s_molinari

And one of them has NEVER been a 9000 loc monster class. Yet you insist, it is ok. You don’t even realize, your argumenting over it being “ok” just takes the “competency” legs right from underneath you.

Every one of the 120 methods in that class needs t be there and nowhere else, so every one of the lines of code in each of those methods needs to be there. The fact that you cannot envisage a class ever needing 120 methods means that you have never written a large and complicated application.

If you read http://www.tonymarston.net/php-mysql/your-code-is-crap.html#table1 you will see that I use 42 of those methods in a simple UPDATE operation. If you look carefully you will see that each of those methods has a definite purpose, so to use fewer methods would do nothing but diminish the flexibility of the entire application.

@s_molinari

You can’t even look at other developer’s code without having to compare it to yours and to find fault with theirs

Just like you are criticising my code by comparing it to yours.

@s_molinari

You can’t learn from others

Yes I can, but only when they can teach something that has genuine value instead of being nothing more than a “personal preference”.

@s_molinari

without looking at your code and basically cherry picking the written articles on coding theory to try and show how you’ve done things differently and in places incorrectly, yet try and show it is right. “The experts have said so.”

Just like you cherry-pick articles which support YOUR point of view.

[TomB]

@tony_marston404

Incorrect. I simply say that my results are just as good but that the code I use to achieve those results is different.

What metric are you using to measure “good”?

[tony_marston404]

@TomB

tony_marston404:
Incorrect. I simply say that my results are just as good but that the code I use to achieve those results is different.

What metric are you using to measure “good”?

They produce the results that my customers expect and are willing to pay for.

[TomB]

@tony_marston404

They produce the results that my customers expect and are willing to pay for.

In that case the code itself is irrelevant. Why are you here arguing about specifics (DI, Autoloading, Separation of concerns)?

[tony_marston404]

@TomB

tony_marston404:
They produce the results that my customers expect and are willing to pay for.

In that case the code itself is irrelevant. Why are you here arguing about specifics (DI, Autoloading, Separation of concerns)?

I am defending my coding style for the simple reason that you, and many others like you, keep attacking it for no other reason than it does follow your preferred style. There is no “one size fits all” style, so stop trying to bend me to your will as it simply will not work.

[s_molinari]

@tony_marston404

The fact that you cannot envisage a class ever needing 120 methods means that you have never written a large and complicated application.

This is the epitome of it all. I am not the only person who cannot envisage a 120 method class. In fact, most programmers can’t. It is a clear break of SRP, which you deny to be breaking, which is clearly wrong.

Scott

[TomB]

@tony_marston404

I am defending my coding style for the simple reason that you, and many others like you, keep attacking it for no other reason than it does follow your preferred style. There is no “one size fits all” style, so stop trying to bend me to your will as it simply will not work.

For the hundredth or more time: You keep coming here and saying “DI IS EVIL” or “YERRR DOING IT WRONG LOOK AT MY MAGICAL AMAZING CODE, IT’S PERFECT!!!11!”. When we demonstrate this to be an incorrect assertion you have a hissy fit and do everything you can to argue that the code is perfect (It’s demonstrably not, as we’ve shown time and time again.)