The future of Fusebox

Question

The good old Fusebox was my first card, and I still really like it. Starting with a version of PHP, currently using the latest version of CFML.

But time passes, and I wonder: maybe I need to switch to another structure? Well, I do not want to start a holy war here. I just want to know the pros and cons of continuing to use FB.

Say, I believe that non-XML controllers are a very good idea and a step into the future. Or maybe I'm wrong, and this is not eogh, and I should focus on Mach-II or maybe Model-Glue or ... (enter your favorite)?

But what about PHP? It seems to be stuck in the past. Symfony, CakePHP, Zend, etc. They look much better and grow fast.

So, a rough list of aspects of the comparison is as follows:

• Time spent on development and maintenance. To me, FB seems good enough.
• ORM integration. I am currently using my own components (by the way, I was surprised to see very similar syntax in the cf9 preview), but with concern about their performance.
• Overall application performance. Caching Are parsed files still good enough?
• Integration with other products. For example, with unit testing tools - does anyone have experience?

Any thoughts and opinions are welcome. Thanks.

Answer 1

Fusebox is still under active development and recently changed hands, so lead developer is now Adam Haskell .

Should I switch to another structure?

This is a subjective question. The only good answer is that - given the endless time and possibilities - you should try all of them and see what you prefer. They all have their pros and cons, but most people agree that this is not a question of which structure, but also a question of structure. You have already decided that the tool you want on your belt is so good for you. Make it a tool that you understand and enjoy.

However, time and opportunities are not always available. In this case, you probably stick best to what you know and find out what's new with the latest changes to Fusebox. I don’t have time to study them myself, so I myself was a model glue. I see a few Fusebox in the near future, but again, this is subjective, and what matters is that you do what works best in your situation.

Php

I can’t talk about the status of PHP frameworks since I am a CFML developer. Again, if you have time, play with them and evaluate where they are and whether they are a tool that interests you.

ORM Integration

I know that Model-Glue has ORM integration - Reactor and Transfer are both very convenient. I suspect the same can be said of Mach-II and probably Fusebox, but I'm not sure either.

Hibernate's baked ColdFusion 9 will probably work well in any framework, but it's not yet visible.

Performance / caching; Designed files?

This is another ColdFusion vs .Net question, right? PHP is also a "parsed" language. Precompiled binary will always have at least a slight advantage at runtime, but consider that for most web applications, adding more convenient hardware is easier and cheaper than spending extra months (or more) on software development.

Are the “analyzed” files good enough? Yes! Hell!

Integration and Testing Structures

There are several testing schemes, including CFUnit, CFCUnit, and MXUnit, from the top of the head for unit testing (which work well for TDD ) and CFSpec for BDD . I am sure that there are many others.

CF8 provided integration with .Net and Exchange (and maybe a few other things that I forget), and we have had Java integration since version 6. It has never been easier to “smudge” some components written in these different languages, to get the best of all worlds.

Conclusion

The title of the question about the future of the Fusebox framework, and I can say that it will not go anywhere (except for continued growth and improvement, like the other CFML framework ...). If you are satisfied with Fusebox, there is no reason to leave it. This does not mean that you should not try everything, but there is no reason to abandon the ship.

Answer 2

Impossible to damage your horizons:

The comparison range is so vast that you cannot get a comprehensive and well-chosen answer to your four criteria in a single SO stream. This is a good question, but one for which no answer will be final.

Instead, I would like to ask what (if you will) will prevent you from trying a different structure and expanding your horizons (assuming that your exclusive or main experience was with FB).

Nothing will exceed your own assessment of your four criteria from direct experience, especially since you ask about factors that are either very subjective or reasonably eliminated by all the "high-profile" structures of web applications.

The key to FB in particular:

The foundation of Fusebox arose and gained strength before most people had heard of XML or web frameworks. This was one of the first “bursting” web development structures designed to actually make web application development more “interesting” (with the goal of eliminating some of the troubles and boredom from ColdFusion, which in itself was an exceptional foundation for its time),

Consequently, he has come a long way and has a relatively stable reputation (e.g. ColdFusion).

This, however, can be seen by some as significant damage against the FB (as for ColdFusion). There is a lot of “baggage” in the framework that, frankly, would not be there if they were the same age as many other MVC structures that are gaining confidence, like the “new children” on the block. There are many aspects that (in terms of language design) show some rough edges that can negatively affect your thinking about web application structures if you choose FB as your exclusive way to achieve results.

Without the names of the names (you've already heard them), I would advise you to keep FB well in your toolbox, but also move to a new framework, especially those based on programming languages ​​other than PHP and Cold Synthesis.

Thus, you will also expand your horizons and understand as a programmer as a whole.

Answer 3

At work, we use Fusebox (php) ... IT SUCKS !!

If at all possible, I would definitely suggest moving over the hips.

Although .. What I am doing, and it seems to facilitate most of my beef with the framework, is to write template files, include this from the switch, and also calculate any "runtime" parameters inside the same statement case. This contributes to good code reuse.

But I mean .. with one huge expression about switching? Isn't that the smell of code for "should it be an object?" This reminds me of the procedural version of the RoR controller class. (I'm not a guy from RoR .. just saying)

Answer 4

If you are looking for everything in one Rails type structure, check out cfwheels.

http://www.cfwheels.com

Answer 5

After some time, we can say that Fusebox is more dead than alive.

• FuseNG update from A.Haskell

• FuseNG status on Google Groups.

Answer 6

I used FB almost exclusively in my last work. Most of our code base was non-OO (cfc did not exist yet), so the difference in model / presentation really helped us. Designers knew they were heading straight to the view folder, rather than having to interpret too much elsewhere. Schemas have given us a better way to display areas of a site than just using directories. As a rule, he solved many problems with the page as a constructive way of working.

Over time, I found that most of the model / controller code ends in cfc and dao with view files only, and the indexes really remain in .cfm, so this separation happens almost naturally. This leads to a new problem that FB does not help with - the management of all cfc and resulting objects, as well as initialization and inheritance. Because of this, I started using coldspring , which focuses more on the problems that arise in modern OO CF, as indicated on the CF page that we wrote about many years ago.