I needed to get the root item of a TreeView. The obvious way to get it is to use the getRoot() on the TreeView. Which I use.
I like to experiment, and was wondering if I can get same root, buy climbing up the tree from a leaf item (a TreeItem), using recursively getParent() until the result is NULL.
It is working as well, and, in my custom TreeItem, I added a public method 'getRoot()' to play around with it. Thus finding out this method does already exist in parent TreeItem, but is not exposed.
My question : Why would it not be exposed ? Is is a bad practice regarding OOP / MVC architecture ?
The reason for the design is summed up by kleopatra's comment:
Why would it not be exposed I would pose it the other way round: why should it? It's convenience api at best, easy to implement by clients, not really needed - adding such to a framework/toolkit tends to exploding api/implementation to maintain.
JavaFX is filled with decisions like this on purpose. A lot of the reasoning is based on experience (good and bad) from AWT/Spring. Just some examples:
For specifying execution on the UI thread, there is a runLater API, but no invokeAndWait API like Swing, even though it would be easy for the framework to provide such an API and it has been requested.
Providing an invokeAndWait API means that naive (and experienced :-) developers could use it incorrectly to accidentally deadlock threads.
Lots of classes are final and not extensible.
Sometimes developers want to extend classes, but can't because they are final. This means that they can't over-ride a lot of the built-in tested functionality of the framework and accidentally break it that way. Instead they can usually use aggregation over inheritance to do what they need. The framework forces them to do so in order to protect itself and them.
Color objects are immutable.
Immutable objects in general make stuff easier to maintain.
Native look and feels aren't part of the framework.
You can still create them if you want, and there are 3rd party libraries that do that, but it doesn't need to be in the core framework.
The application programming interface is single threaded not multi-threaded.
Because the developers of the framework realized that multi-threaded UI frameworks are a failed dream.
The philosophy was to code to make the 80% use case easier and the the 20% use case (usually) possible, using additional user or 3rd party code, while making it difficult for the user code to accidentally (or intentionally) break the framework. You just stumbled upon one instance of an application of this philosophy.
There are a whole host of catch-phrases that you could use to describe the reason for this design approach. None of them are OOP or MVC specific. The underlying principles have been around far longer than software engineering, they are just approaches towards work and engineering in general. Here are some links if interested:
You ain't going to need it YAGNI
Minimal viable product MVP
Worse-is-better
Muntzing
Feature creep prevention
Keep it simple stupid KISS
Occam's razor
Why Neto/Shopify use their own templating language instead using any popular popular language ?
When you have your own templating language you have full freedom to implement or limit the logic of the language to meet your needs.
You don't need to wait for a third party update in order to implement new stuff or objects.
Why do you think Samsung creates clones of google apps on their phones? To create an experience that they can control in some way and if they like to change or add something to do so and not to wait for google to do so. ( and some other things but that is outside the scope of this question )
Since liquid was a Tobias Lütke creation ( co-founder and CEO at Shopify ) and now is an Open Source project it was written in specific fashion in order to meet specific needs and those needs seems to be fitting for Shopify and other platforms as well.
Being popular doesn't mean being better! This is the wrong way to go.
WordPress is the most popular platform, but is it the best one - NO! There are a lot of hole provided by plugins, poorly written themes and some times core issues. While it's easy to use and extendable it opens a lot of doors for issues if you don't manage it properly.
Beats by Dre is the most popular headphones and quite expensive, does they sound as good as the price tag put on them - NO! You can buy the same quality headphones ( even better ) for less, but you are paying for the brand.
Creating new languages in the coding world is ALWAYS a good thing It might be a chore to learn it if it becomes a standard but that means that it provided something that the other popular languages didn't and this pushes the coding world forward. It's a much better alternative than to be in a standstill like when we had only jQuery and there were no new stuffs to excite the developers.
Now we have so much different things that you can choose the direction you like to go and you won't be able to learn all of them even if you try, which is a great thing to a developer who likes to grow.
Conclusion:
Being different is OK as long as that fits your needs and you are not doing it just because it's popular to be different. ( so true IRL now too :D )
This reasoning from their Github Wiki lists some of the reasons. Why Liquid Templating Engine ?
Liquid is a template engine which was crafted for very specific
requirements
It has to have simple markup and beautiful results. Template engines which don't produce good looking results are no fun to use.
It needs to be non-evaling and secure. Liquid templates are made so that users can edit them. You don't want your server running code
that your users wrote.
It has to be stateless. The compile and render steps have to be separate, so that the expensive parsing and compiling can be done
once; later on, you can just render it by passing in a hash with local
variables and objects.
It needs to be able to style emails as well as HTML.
CakePHP Applications being made in our company tends to become unmaintainable as it becomes more complex. I figured that one specific reason is inheritance which makes the functions in child classes depends a lot on it's parent classes and vice-versa (implementing template method pattern). Why is CakePHP designed this way and not friendly in using Dependency Injection, Strategies, or Factory patterns?
There is not such a bad design as you claim in the framework. Sure, there are probably things that could be done better but I would like to see a more substantial critic including solid arguments and examples. I assume you're not using the framework as it was intended.
Let me quote the first paragraph from this page.
According to Eric Evans, Domain-driven design (DDD) is not a technology or a methodology. It’s a different way of thinking about how to organize your applications and structure your code. This way of thinking complements very well the popular MVC architecture. The domain model provides a structural view of the system. Most of the time, applications don’t change, what changes is the domain. MVC, however, doesn’t really tell you how your model should be structured. That’s why some frameworks don’t force you to use a specific model structure, instead, they let your model evolve as your knowledge and expertise grows.
You're not showing code (for a reason?) so I guess your problem comes from stuffing everything into the table objects in src/Model/Table/ or doing something similar.
But you're totally free to create a folder structure like
/src/Service
/src/Model/Domain
and then simply instantiate services as you need them in your controller actions. A service could be for example \App\Service\User\Registration and using objects from App\Model\Domain\User.
I agree that the framework in fact doesn't provide any recommendation or template structure for how this could look like. For exactly this topic there is a discussion going on here. Because of a lack of such a structure I've started working on a plugin that provides this. The plugin doesn't require but suggest the usage of DI containers for the people who want them.
Given the whole fancy topic around DI and DDD so far I would say there is not the one way to get things right but different paths as long as the code is easy to maintain. And honestly, as long as this goal is archived I really don't care about how you call it. :) I think many people tend do make this topic to academic instead of simply trying to be practical.
Not everybody is even needing that structure. It depends on if you're building a RAD CRUD application or a more complex app. Not every application needs a DDD approach. There are so many shades of gray when it comes to design the business layer, no matter how the framework would do it, somebody would always complain about it.
I personally almost never missed a DI container in CakePHP, not even in the biggest project having more than ~560 database tables which was a hospital management solution and it just worked well.
I would suggest you to ask a more specific question about your approach how you structured your code and showing your structure and code and then asking for advice on how to improve it instead of blaming the tool you're using in the first place without providing context.
Unfortunately CakePHP v3 can not compare to the Zend3/Laminas, Symfony or Laravel.It is 7-8 years behind the other frameworks.If you are using cake for years or it is your 1st and last framework it is normal to not realise that.But if you have to use it after Zend 3... cake seems like really bad ecosystem.
Bad documentation
Bad ORM
Poor Routing system
Bad Templating engine
Bad idea to mix Data Mapper and Active Record
DIC is totally missing
Components - not good but not terrible
...
And many more thinks that should not be underestimated like - lack of GOOD tutorials, pluigns/addons/packages
The above thinks make developers to follow bad practices that adds a lot of technical depth.
If you care just for - it works! But not how it works and why it is bad, cake will fit ok for you.
Cake can not scale as good as Symfony/Laminas if you are doing big project.(yea AWS/GC can help for scaling a lot of thinks but not for scaling source code)
Cake doesn't allow you rapid development like Laravel/Symfony for decent project.
I'm wondering who and WHY would start a new project today using Cake as it has zero benefits over the other frameworks.
Probably only devs who used only Cake for last decade and do not want to start learning new technologies or devs that thinks SOLID is just a fancy hype with zero benefits like design patterns, DRY and KISS
CakePHP framework supplies user interaction with databases using Active record, it means that exist a high coupling between business layer and database layer which has negative effects in unit testing and because of that the framework is not friendly with Dependency Injection. The same issue happens with Factory pattern, high coupling mentioned before makes more difficult use simulated objects in unit testing.
Hope it helps!
Alberto
I am new to Magento and impressed by the MVC framework that powers it, making module development a well thought out solution. I am strong CakePHP developer.
I am working on a project that uses a dropshipper for the physical products. As a result, every day at 4am a feed needs to be parsed and the products / categories modified, plus stock information. A CRON will be setup to do this.
Additional requirements are:
Upon a sucessful order, the system must upload a CSV feed to the Dropshipper via FTP with the order details for distribution.
Realtime stock checks, either every hour by CRON or a lookup on the product page
I can think of 2 approaches:
Write everything natively into Magento. As a newbie, this is going to be a big learning curve, but it is the right solution?
Write a simple CakePHP app that runs as a shell. This will use the Magento API to manage all dropshipper processes. This solution will be easier to rollout but introduces an additional system to support.
Does anyone have an advice relating to dropshipping in Magento?
First, with respect to the product import (product, stock data), make sure to do the real data saving inside of Magento. There have been changes to the catalog implementation in the past, and it's likely with a framework like Magento that there will be more. Keeping it inside the framework will reduce the likelihood of it simply no longer operating and you getting a very unpleasant phone call.
Another advantage to this approach is that, in contrast to the API approach, the native code will not try to spin up the entire framework for every request. This is expensive and to be avoided. Depending on how many products there are, you may need to break the script into multiple executions due to memory leaks when saving catalog products.
Don't tie the stock checks to a catalog page view. Some web crawler will come eat your lunch.
Finally, there's no easy FTP library built into Magento, but throwing that on another cronjob and using system calls to perform the actual (S)FTP call is possibly your easiest option.
Hope that helps!
Thanks,
Joe
I think the answer to this question is simple. Write it in what you know. The biggest reason is "UPGRADES"... with Magento being as high profile, the possibility of being hacked with older versions increase every day. Therefore, when they release new versions, you are going to want to upgrade. With that in mind, are you going to want to add all of your changes into each new version as it is released? Probably not. If there is a solution to write this as a separate tool, that is what you should do.
PROS TO BUILDING OUTSIDE OF MAGENTO
No need to re-integrate the upgrades
every time a new version of Magento
is released.
Code is easier to maintain.
Tool is easier to write in something
you are familiar with.
No learning curve.
Integration speed will be much
quicker.
More flexibility since you do not
have to fit inside Magento code
limitations.
I spend a lot of time (actually too much time) developping back-office applications whose main purpose is content management and web application configurations. Here is how I can describe these apps :
- Made with PHP
- Using a MySQL or Postgres or SQLite database
- Made of a lot of pages and features
- Very simple features, mostly data CRUD (create+read+update+delete into the database)
- Mostly made of forms
- UIs are usually quite simple (html + css + very basic javascript)
All of the data access code in these apps relies on a library I developped years ago and re-use every time I can. This part is not time-consuming.
What's time-consuming is the UI part, and mostly designing data-lists and forms. Using a WYSIWYG editor would make a lot of sense here, except those I tried (Dreamweaver, Frontpage, Expression, Eclipse, ...) don't really make it much faster, because the generated code is often bloated, and these tools can't rely on custom libraries such as the one I made and use.
I figured using a Web Tookit could be another way to spend less time developping these tools. So before I spend too much time looking for the perfect toolkit, I would appreciate your opinions and experiences on that kind of matter.
Disclaimer : I'm not looking for advices on how MVC is the way to go and how CodeIgniter/Zend/WhatEver is the framework I should use. My question is not about the frameworks or the design patterns I should build my applications upon. My question is about using the right tool to make simple web-applications development faster, and their code even more re-usable.
Is there an awesome web-application RAD tool I don't know about ?
Which toolkit do you use for simple but form-heavy web applications ?
Are there good, light, non-bloated, reliable toolkits written in PHP ?
Thanks in advance !
Edit : Not getting much feedback so far :/ I'm aware that my question is very broad, but I'm sure lots of people work on the same kind of projects I'm talking about, and have improved their productivity by using toolkits such as GWT, Wicket and such. Tell me about it, please :)
September 28 edit : Thanks everyone for the interesting answers. What I'm looking for is not covered by any framework I could try in the past months. PHP is probably not the best language to use for my vision of RAD, but since it's a language I know very well, and since I don't want to spend too much time learning Python as well as I know PHP (for the moment), I decided to do it by myself. Everytime I have a specific need for a widget, I code it in the most re-usable way...so far so good :)
I might open-source that toolkit at some point, and will let you guys know.
The PHP project I've been working on the past few years is a lot like that. Heavy on forms, heavy on server-side logic, but lots of redundant form coding. Too make matters worse, it wasn't all forms, sometimes we actually need to do fancy layout (even just doing a tree control is a pain without a library), and the home-grown nature of the UI meant that I would be battling browser quirks from start to finish.
So, I got to thinking about what a better architecture would be. We needed very powerful form controls, rich grids, rich trees, advanced layout, and we needed to migrate to that gradually. None of the PHP frameworks seemed to fit. Then I took a step back and realized that it didn't have to be PHP, it could be javascript also. We already had a requirement on javascript, so it was fine to go the distance with it. First I looked at the smaller libraries, jquery, prototype, but it became obvious that they didn't do enough. So I looked at Dojo, ExtJS, YUI, all the really heavy javascript toolkits, and settled on ExtJS as having the best controls.
We had a UI structure that relied heavily on iframes, a navigation frame on the outside, application frames inside that, feature frames inside that, and so on. What we ended up with is we're migrating those from the outside in. It's all becoming ExtJS, and it's all living in the same page. The server-side code is kept the same, but it's migrated into web services. At the same time we've integrated zend framework, and are porting some of the stuff you really shouldn't do home-grown to it, like authentication and translation.
The end goal is being able to write just the business logic without having to mess with all the boilerplate. It's too early to know if my approach will pan out, but I think my message would be to be critical towards your code base and decide which parts you want to keep writing yourself, and which parts you want to outsource to a library.
Please try http://agiletoolkit.org/. I think it's what you need. Results with minimum time/code.
At the moment I'm using a solid forms class to render HTML forms with client- and server-side validation, and a database class to write the SQL. I can get a CRUD section of my admin console up in about 10 lines of code. I wrote those classes myself so I can re-use them in all my projects. Hopefully that gives you some ideas?
I would stay away from WYSIWYG tools personally.
I'm testing NuBuilder.com, I discovered it does
within days, at first looks promising. If you take a look
please send me your feedbacks!
Maybe adopt or create an "app-based" infrastructure like Django's? In Django's case, the community has created some powerful baselines like Pinax.
I think Symfony may be the way to go because, like your apps:
it's written in PHP
ORM based on Propel/doctrine (so you can use MySQL, Postgres or SQLite)
Architecture and patterns used will help you with complex applicatons
You'll find tools helping you to debug, document, and test your application
Forms creation, validation, l10n & i18n, testing, AJAX is easy (forms within symfony explained here, check it out)
prototyping you webpages while developing your application is easy
Other tools/practices implemented in the symfony framework that will make your life easy:
full configuration using YAML syntax (easy to read and understand)
the scaffolding feature generates for you a simple CRUD interface for editing your data.
you don't have to worry about coding form sanitization, security, caching, ACL; configuration is needed, but no heavy coding.
The only downside, you need to read some documentation to understand "the symfony way of doing things". But hey, a good framework is 20% code and 80% good practices.
My point is, even if you don't want to use Symfony for your project, you should check its features and built-in tools, because that's the kind of tools you want for your project.
I started using Django and it has very helpful features, esp. the built-in admin (for general CRUD stuff) and really great form-handling code & widget rendering. I'd suggest taking a look, even if you don't plan on using python, just to get an idea.
You mentioned that you don't want advice on "Use X framework", since this is more about RAD & UI/forms than system architecture. But I've found that a good framework helps just as much with the UI & forms side of things as it does the architecture. That means that while frameworks are great for big projects, they're also very helpful in reducing code redundancy. I started creating my own helper functions in PHP that I would copy from app to app that would automatically render an HTML form based on a few parameters. Even after a lot of work, this was very rudimentary compared to what Django offers, and basically I was writing my own framework.
I think you may be looking for a GUI-style tool to help, but you might find that a good PHP framework is more helpful in this case. At the very least, have you tried creating your own helper libraries? I know those helped me a lot.
Simple Example:
function renderInput($name, $value="") {
print "<input type=\"input\" name=\"" . htmlentities($name) . "\" value=\"" . htmlentities($value) . "\" >";
}
function renderRadios($name, $value="", $choices=array()) {
for ($choices as $cvalue) {
print "<input type=\"checkbox\" name=\"" . htmlentities($name) . "\" value=\"" . htmlentities($cvalue) . "\" " . ($cvalue == $value ? "checked" : "") . ">";
}
}
And build up from there. Stupid things like this tend to make form creation just that much faster. A good framework will blow this out of the water. And I'm sure the above has some typos, I haven't done PHP in a little bit.
If this isn't what you're looking for, could you add some more to the question? I'm curious.
Although you're asking for PHP + MySql, I would like to recommend you to give a try to the OutSystems Agile Platform.
You can create a simple CRUD app in less that 10 minutes and grow it as you go to a more complex system.
Download the Community Edition for free at www.outsystems.com.
Best,
Not sure but looks like Tibco General Interface (http://www.generalinterface.org ) is what your looking for.