The secret history of ColdFusion MX
It was the year 2001. Macromedia had only recently-acquired Allaire and were moving forward with their MX project when I got a call. It was ***** from Macromedia asking if I'd like to come up and be part of an experts panel talking about the next release from a community standpoint. What else could I do but say yes?
We talked about a number of things that were happening in the community at the time and what Macromedia planned for the future. We all had something to say and some tweak to give to the language that was coming. ***** went into great detail on what he wanted for the creation of PDFs.
When it came to their proposal for CFCs, I actually took offense. This was not ColdFusion. This was Java encroaching on ColdFusion. At the time, CFMX was in alpha and CFCs looked like Java tag libraries - ugly as sin and something that would not make the average ColdFusion programmer happy. As they detailed each part, I was very vehement about what was right and wrong about it. "Was this the right way of doing it in ColdFusion? Is there something that can be leveraged to do it instead?" My opinion was that CFCs as they stood would be harder and scarier than they needed to be and added new tags where they were not needed. Need to validate variables? What about CFPARAM? Can we use the same approach there? The same thing was said for just about every part of the CFC spec and just about everyone had the same opinion. It went on and on but the bottom line was that Macromedia listened to us all and made what looked to me like major changes in the new feature spec before the next beta.
What is the point of all this? The point is a solid question to all those who are pushing for Java language constructs in ColdFusion: Why? Will Interfaces really help anyone or will it be only used by a top tier and make CFCs just that much scarier to people new to ColdFusion? Will it be something that slows down code execution just to 'look like other OOP languages'? Is there a point?
The same question is asked about new feature requests. Are they useful and can they be made to fit into the spirit of ColdFusion. When **** wanted an entire 'outside' language for PDF creation, it 'felt' wrong because it meant another language would be needed to make the tag work. This isn't in the spirit of ColdFusion. ColdFusion is about smooth, easy-to-use and human-readable tags that are used to do the work that would normally need much longer chunks of code.
Don't get me wrong; I'm all for advanced features. I just want people to stop and think about whether the features are actually needed or are just being thrown in because OOP, Java, Ruby or some other language/concept has it.
ColdFusion does already so much and there are already many frameworks for it available as well. So let the people who want to use a framework, use one of there. Let Adobe concentrate on the core. I bet there will be some cool stuff in CF 8.
If someone asks me what I want in CF8, I'm going to tell them what *I* want, and that will be based on what would be most beneficial to me, and my team. I am not going to second guess what's better for the product. I'm also not going to balance up whether it might or might not not suit [insert name of someone else here]. That might sound mercenary, but there's good rationale behind it.
I read with interest everyone's ideas, and the explanation of their ideas. I agree with some; I disagree with others. However I think it's a bit stupid for people to go "oh, you shouldn't want that, you should want [something else] instead, because it's better for the CF community and CF's future".
Let's leave that analysis to Adobe. I'll tell them (and you lot out there in the community, in case you have an opinion) what I want. You tell them what you want. Everyone else tell everyone else what they want. This will result in a list of possible features which possibly could be considerd for CF8.
If all of us want one particular feature, it would be a good one to add; if 50% of us want a feature: still probably a good idea; if someone suggests some feature and only they want it: fair enough for it to not be considered.
I'm growing a bit tired of people suggesting "oh, you don't want CFINTERFACE, you should want CFMAKEITPRETTY, because that's what CF is all about"; or "ah, you can already do CFIMAGE with [somethingelse], what we need is CFMAKEITHARDER, because that's what developers like me want". My considered response to that is: "screw you". I know what *I* want, and I think other people making their suggestions aren't just making them to sound clever: it's what they actually want too.
Everyone should list what they want to see. Everyone should discuss those requests, so they can be improved on, explained, and elucidated. No-one should say "no, that shouldn't happen, because [whatever]".
But [some people] should quit telling [other people] what they should be thinking.
Other than me, telling you you should automatically agree with all of the above, natch ;-)
--
Adam
This is not the blog entry I actually added the comment to...
(it was supposed to attach to the "What I want in CF 8 - part 1" one.
Not sure what happened there.
Michael, is it possible to move it to the place I intended it to be? Cheers.
--
Adam
- more speed
- fix the bugs that already exists
- more reliable JDBC drivers
- lower the damn price.
I think those 4 points are worth a lot more to people then interfaces and constructors.
"- more speed
- fix the bugs that already exists
- more reliable JDBC drivers
- lower the damn price."
Finally, someone that makes sense!
What's worse is that if Adobe's discussions with a much larger customer base indicate feature X is not wanted and they don't implement it (rightly) but the blogs all cry for feature X and then lambast Adobe for not implementing it. Do you see what I'm getting at here?
Yes, I see what you're getting at, but it kinda misses my point (I think). I don't expect Adobe to take any of these blog entries as some sort of grail. I do not expect them to put any more weight behind them than their own research. I give Adobe more credit than that.
My point was that people on these blogs should quit with trying to explain why someone else's ideas for enhancements are poor ones, and not good for CF, the community, and [other ballocks here]. Plain and simple. We are just discussing what WE (each individual) would like to see in CF. I don't think anyone here actually thinks we're compiling a list of the features that Adobe MUST put into CF.
As for your second paragraph, I personally have no problem with what Allaire / Macromedia / Adobe have thusfar released in any versions of CF, on the whole (bad implementation issues like <cfgraph> and CFMX6.0's implementation of CFCs aside). This might seem strange coming from me - as you well know I'm not exactly a CF evangelist - but it's true. And I have seen no evidence of other people thinking that way after each release of CF. I guess we are all disappointed that our pet feature request doesn't make it, but I've not seen anyone getting antsy about it. Have you?
Anyway... this is all a bit boring cf actually discussing the ideas, really, innit? Shall we move on?
--
Adam