State of the CF Union 2021 Survey Results

Question about conference should have been altered due to the pandemic nobody can move at the moment.

Where you had operating systems you did not have Unix we use Oracle Solaris

Love CFML. My personal stuff is all done in Lucee.

Adobe has to fix the licensing per core issue. Last time we ordered new servers I had to custom order Dell to get 4 core CPUs so I could get by with a mere 2x Standard Licenses per server. Most standard CPUs in servers are 8-16 cores. No one is going to pay $5,000 for a server + Windows and then $25,000 for CF. You want to force people to Lucee? THIS is how you force people to Lucee.

Thanks for putting these together each year. 🙂

Share these results with Mark Takata at Adobe! And anyone else that has responsibility for promoting and improving CF.

Community is good and helpful, like most developer communities. It's also small (compared to communities for languages/environments with much wider adoption) and the Adobe team is unhelpful in most cases, especially with support for ACF performance and ColdFusion API Manager.

We are migrating away from CF. Both in my last company we migrated from CF. It is hard to justify cost in open source world. It is very very difficult to get good CF developers who has experience in OOPS. It is definitely not easy to find CF jobs to advance the career. There is not single job in California in my salary range.

Adobe killed CF. Lucee is a great retirement home for legacy CFML apps to live out the remainder of their lives, but it will never know the glory days of old again.

I love CFML. I've championed it at every employer I've worked for in the past 20+ years. I wrote and maintain trycf.com. I'd love to see it catch up to the modern stack, but that will not happen. ACF is too slow to introduce language features because fluff features pay the bills. (API manager, CFClient, etc...). They are also too slow to upkeep their fluff (hibernate, extjs, etc...). Likewise, Lucee is too afraid to move away from ACF compatibility and therefore are stuck in the same decade's old mindset. It sucks, but that's the way it is.

Far too long have we pretended that CFML is easy. I work with a lot of OO CFML that is by far the most complicated (unnecessarily) and hard to debug code I've ever seen. That's not all (much of it, but not all) CFML's fault, but OO is hard, with very little benefit. Too much mysterious mutations happening and no telling where functions, data, variables come from or why they change. ColdBox, sorry, but it sucks by association. It tries to help the not really-OO-language-but-tries-to-look-like-it, but it falls short (don't tell the Ortus folks, they're awesome and do an awesome job, but ColdBox is too OO for not enough gain -- too much black box!). FW/1 is much simpler, but not as many resources (human, oss, support, etc...). Functional programming is also hard, but has so many benefits that overcome all the headache we've found in OO programming. It is impossible to achieve anything close to pure FP in CFML (not that pure would be good either), which is one major reason we're ditching it. Moving to a much flatter more manageable and readable code platform that is much leaner and can spin up and be deployed much faster (nodeJS servers & serverless AWS Lambda).

CommandBox and especially the ComamndBox docker image and CFConfig are awesome and are far ahead of many web tech stacks out there, but developers like to use modern language features that make their jobs easier and their code better. CFML just doesn't see that, never have, never will.

I'll likely continue the CFML ride until the end, but as a manager and one responsible for hiring and tech decisions, I will not be doing any new development in CFML.

1. I don't understand why you are not doing a scientific survey. It would be much more accurate.
2. I don't use any frameworks because I don't know how to use them. Training classes would be helpful.
3. It would be good to conduct qualitative research after you get the results from this survey. I would recommend a focus group on a particular survey question.

I love ColdFusion but it is being destroyed.

Please put ColdFusion back in the USA with better marketing and sales and better communicators and better support and better training. Hire professionals. Hire back in the USA for better coders and better coding. Testing. Test this stuff please.

Get out and explain that this is a JVM or Java fronting language.

REST in Peace!

There's a few sets radio/checkbox answers here that could benefit from "dunno" / "N/A"

stop using AWS products as categories, use 'serverless' for instance as each cloud has it's own name and AWS is a dogs breakfast atm which is why we're never intending to use them again.

Nothing about using message queues and other tech here. Nothing about ML either.

Also no talk of technology stacks really around cf, what WAF other than cf-based ones, etc

No

None.

I really love CF and worked with it for many years ...
BUT:
I think the licensing model of CF is killing this language. the kind of thinking (8 cores for one enterprise-license) is out of todays needs and reality - even hard to find a server with 8 cores only :-))) i switched to VMs and there the costs of the license it is even more strange to me. it should be at least countable for 32 VM-cores or more ...
young developers will choose other languages and technologies.

Thanks for doing this! I look forward to seeing the results.

We would never - under any circumstances - pay for a web programming language/platform. If not for Lucee, CFML would have been in the rear view mirror ages ago.

It would be interesting to have a question about which devs/blogs people read or follow on Twitter etc...

Need an IDE, like visual studio and need a debugger, breakpoints like visual studio

Might be good to ask what modules people are using from ForgeBox.

documentation can be better (can't it always?). The examples provided are generally not useful nor realistic nor do they show some of the more complex aspects of various functions. Better cfscript documentation needed.

The biggest issue with CFML dev was and continues to be, Adobe itself. They just don't care enough about the product, and if they don't take it seriously how can anybody else do that? Having a hidden away team in India do all the development, with no community involvement whatsoever, results in features and updates that just serve to maintain the status quo for their existing clients, but are not bringing in any new devs or big clients.

CF Makes me sad because Adobe seems stuck: they've reduced the user base so less revenue, but they need revenue to justify supporting the product. I don't know the exit or the right move on this. It's up to them. But yeah, it's not showing growth because it's not their focus (their fault) but they can not play with pricing because they need that to sustain the specific Adobe division.

Question 49: 100% of my server side code time is in CF. The rest is in React for our front end development.