Been checking these other Frameworks -- nothing as good as CI (so far)

[eluser]jedd[/eluser]
[quote author="jayrulez" date="1245870219"]
Yes i usually update all my apps as soon as changes are made in the repository as long as running with that update wont affect the stability of the program.[/quote]

The trouble with that approach is the hidden workload embedded in that little caveat 'so long as running with that update won't affect the stability' - as that's a fairly serious amount of testing. I think CI typically suggests that new versions of the codebase aren't reason enough to update extant applications. Security releases (few and far between I believe) are obviously a different matter.

Its not as simple as "simple things". its granted that php is somewhat simpler than the other languages but alot of people have trouble with php any way since there are varying degrees of difficulty.

This is true. It's always surprised me the number of quite competent programmers, that can get their head around pointer handling in C, or 600 page books detailing standard libraries ... but can't work out how to spell 'its' properly. This is not specifically a dig at you, but in the context of simple things, it's always worried me that the necessary care, caution, attention to detail, ability to understand rules and exceptions - some might even say OCD or plain old anal retentiveness that is typical in this demographic - and that occurs with geeks when they programme, doesn't extend to the language they use to verbalise their thinking in. Disclaimer - I'm a big proponent of the Sapir-Whorf Hypothesis, particularly as you scale it down to the individual, so this phenomenon is especially fascinating for me.

I'm not bashing codeignitor, however i think for such a "mature/established" framework, there are alot of things CI could have done differently.

Go on .. name names! I have little exposure to other frameworks, so I really am quite naive on this one. Unless you're just talking about bundling in more and more libraries and helpers (?) in which case I can see your point.

The thing with the vanilla forum / authentication system - I like the idea, but I suspect that the same problems will crop up as CI-users hit when they look around for an authentication system. The available options, whilst quite superb in many ways, are rarely an accurate fit to the requirements du jour, and/or require a huge effort to master and modify (and maintain), and/or will have performance problems because they are necessarily heavier than the actual requirements.

I believe that with the appearance of so many new eye-catching frameworks, codeignitor's creator/s need/s to strenghten its weaknesses or it will be dethroned relatively soon.

I think that like so many other things, we'll stick with the framework that we first properly engage with - so long as it's mostly competent - as the cost of switching is quite high compared to the benefits of that flattening range of the learning curve. I'm not saying this is a spectacularly good thing, I think it's just the way that we work - triage, if you will, for the learning centre of the brain. Otherwise you'd flip between frameworks (and cars and music styles and jobs and friends and banks and mobile phone providers and homes and so on) every few minutes, as a 'slightly improved' version of each becomes apparent.

Combine that with the natural preference to follow the herd, along with CI's huge popularity - means that most people will assess a framework based, at least in part (and probably a large part) on the size of its userbase - this in turn means that CI (and the other incumbents at the top of the 'PHP framework' hit list on Google - will retain their lead for quite a while I reckon.

[eluser]garymardell[/eluser]
I think taking the time to learn how to use other frameworks will always be a good thing, despite the learning curve. Using different frameworks will show you a large number of approaches to a problem and then you have the knowledge to judge what you think is best. It will also make you a more versatile programmer.

[eluser]jayrulez[/eluser]

This is true. It's always surprised me the number of quite competent programmers, that can get their head around pointer handling in C, or 600 page books detailing standard libraries ... but can't work out how to spell 'its' properly. This is not specifically a dig at you, but in the context of simple things, it's always worried me that the necessary care, caution, attention to detail, ability to understand rules and exceptions - some might even say OCD or plain old anal retentiveness that is typical in this demographic - and that occurs with geeks when they programme, doesn't extend to the language they use to verbalise their thinking in. Disclaimer - I'm a big proponent of the Sapir-Whorf Hypothesis, particularly as you scale it down to the individual, so this phenomenon is especially fascinating for me.

Even with error in my grammatical syntax you were able to understand what i said, hoevwer, in the case of programming a syntax error will cause the program to break. On the other hand i must agree with you to some extent (I must also thank you for pointing this out. It's become a bad habit of mine to use "it's" and "its" interchangebly.).

[quote]
The thing with the vanilla forum / authentication system - I like the idea, but I suspect that the same problems will crop up as CI-users hit when they look around for an authentication system. The available options, whilst quite superb in many ways, are rarely an accurate fit to the requirements du jour, and/or require a huge effort to master and modify (and maintain), and/or will have performance problems because they are necessarily heavier than the actual requirements.

That i only true for some systems.


[quote author="garymardell" date="1245880000"]I think taking the time to learn how to use other frameworks will always be a good thing, despite the learning curve. Using different frameworks will show you a large number of approaches to a problem and then you have the knowledge to judge what you think is best. It will also make you a more versatile programmer.

That is true but for some accomplishing it would be time consuming.

[eluser]garymardell[/eluser]
Yeah, would definitely have to be a spare time thing. or a personal project which wouldn't matter if it was on going. A good way to get used to the framework is making a project that interests you.

[eluser]jayrulez[/eluser]
Yes that is true. I'm on a break from school now and I have taken the initiative of doing this. Currently i'm building a single app on 4 different frameworks (CI,Yii,TP&IF, with the fundamental purpose being to choose the one that i will use in official projects.

[eluser]jedd[/eluser]
jayrulez - yes, I can understand precisely what you meant. I was just observing the interesting paradox (as I see it) with your average programmer. I was at a tourist site recently - a castle in Spain, so some forgiveness there - and a sign blocked off entry to one of the corridors. It read "Access restricted to unauthorised personnel". Trez amusement. My Spanish friend read the original (Spanish) that accompanied that, and said that it was equally ambiguous in their native tongue, so I retracted my forgiveness.

And yes, the problems I mentioned with the lack of common alignment between bundled bits, and what designers want on the day - are only true for some systems. But consider the huge number of auth systems that CI has engendered. I've only been loitering with intent around these forums for several months, but in that time I've seen plenty of people pop up, assess the extant offerings, and then proclaim that they're going to write their own instead. Usually the primary reason is this lack of alignment - not any particular security concern, or documentation problem, or even methodology. They've all basically assessed that it's easier to write one from scratch, than to try to adapt a frequently very impressive lump of code into their app.

I'm not saying that you can't get a 'one size fits most' solution - just that it hasn't happened here, and I'm doubtful that it'll happen with other contenders in the 'fast, lightweight, versatile framework' space.

Thinking about it now, I suspect that the line between framework and CMS blurs even further with products like this one. I often encourage people to head down the path of a CMS, and put the effort into writing modules to plug into it, rather than writing their system from scratch - simply because it'll be cheaper/easier. My gut feel, though, is that most people will be split into one of two broad categories - wanting a very light weight and very fast platform, or a very busy and complete platform where performance problems are solved by throwing more hardware at them.

garymardell - the same could be said of each and every programming language, too - it'll give you great insight, but at what cost to your proficiency in any given language (or framework).

[eluser]jedd[/eluser]
Oh, I meant to ask - if you (jayrulez - or indeed anyone else) takes a good enough look at this framework - I'd love to know how useful they find the database structure class. I think it's strongly tied to their one-model-per-table approach (which I find uncomfortable) so it might not be that useful for me, but, yeah, that one thing really piqued my interest in their doco.

[eluser]ntheorist[/eluser]
I'd agree that CI doesn't seem as mature as it could be. It has some oddities in it and lacks some depth but i suppose that was a conscious decision on the EL's part to keep it small and the learning curve low compared to other frameworks. This makes it easy to pick up and master its libraries but offers little on it's own beyond this.

The rest relies on communal development of libraries and bits of code. It may be that it relies TOO heavily on it though and thus you end up with 12 contributed authentication libraries to choose from. Part of framework development is creating new 'standards' for doing/naming things etc., and then the community builds off that. It's understandable that people would look to the developers for some form of leadership when they sort of reach 'the end' of the docs n such. I do believe tho there will be more upon release of EE 2.0 because its being built on CI directly, and i'm very interested in seeing what they come up with.

The fact that it is relatively basic can work for you very well even for advanced projects, as it simply provides a core to work on. Not to say you should have to build everything from scratch though. Truthfully i don't think it matters what framework/language/approach you use as long as you're creative and productive with your ideas, The rest is performance tuning.

btw, has anyone tried using Kohana yet? It's php 5, has an ORM, has active forums etc and would probably be a snap to jump into based on its similarity to CI, yet does seem more evolved.

[eluser]jayrulez[/eluser]
[quote author="jedd" date="1245884814"]Oh, I meant to ask - if you (jayrulez - or indeed anyone else) takes a good enough look at this framework - I'd love to know how useful they find the database structure class. I think it's strongly tied to their one-model-per-table approach (which I find uncomfortable) so it might not be that useful for me, but, yeah, that one thing really piqued my interest in their doco.[/quote]

which framework?

[eluser]Neovive[/eluser]
[quote author="ntheorist" date="1247421394"]
btw, has anyone tried using Kohana yet? It's php 5, has an ORM, has active forums etc and would probably be a snap to jump into based on its similarity to CI, yet does seem more evolved.[/quote]

I switch between CI and Kohana for various projects, but typically choose Kohana for new projects due to it's cascading filesystem and built-in support for ORM, Auth and Events. In the end, you can accomplish most of the same things in both frameworks ... it just comes down to coding style.