The FM developer is no more. Long live the license-buying customer!
1 Like
Yes. That's madness.
1 Like
I know someone who does the same. Insane or simply methodical and debugging champ?
1 Like
We need to. We log like crazy and have multiple developers. I barely ever have to open the debugger because our log reports everything and errors are gracefully handled. In effect it IS try/catch, but because of FM’s lack of that flow structure, it’s more verbose.
We’re not doing anything unique relative to the programming world, it’s just very verbose in FM script.
And frankly, that’s FM’s shortcoming.
3 Likes
Makes debugging and spotting errors faster so it saves time. I’ve automated creating the error steps xml.
It’s akin to writing unit tests. More work up front = less work overall.
Code you don’t have to remember anything about, which will tell you about errors automatically is better code IMO.
5 Likes
This is a good list and of course the list of missing functionality we all would want is probably 10 times longer. The truth is I think most FM full time devs like myself want a truly professional development platform that gives us much more control over things and much better built in analysis and debugger tools. (think 4D with FM's ease of use). I spend over twice as much on third party tools to complete my home grown FM "IDE" (DA, FM perception, MBS Plugin etc..) then I spend on FM itself which I don't think is the way it should be.
The problem has always been that what I/We want the product to be has always been at odds with want Claris wants it to be. They don't want to create the best database development tool in the world for developers they want to build it for "problem solvers" and "knowledge workers" and hope they become developers like most of us have.
That is their right as a strategy but in my opinion it's at the core of all their problems and limitations they have placed on the platform. 19 will most likely be the last version of would I would call "classic" FileMaker what it turns into next is anyone's guess so we just have to pray Claris get's it right and the new generation of the platform will finally deliver on it's untapped potential for professional developers.
3 Likes
When you talk to developers in other languages, they often talk about how much they spend on tools to do their job. The IDE itself, free or not, is only a piece of it. There are a few developers that I talk with that claim to have nothing but free tools. That sounds great. In my experience, the free tools help some, but no where near the level that professional products/tools have. I've watched both types of devs work. The free tools...still have a cost. Watching a professional dev with professional tools is a little like watching a magician, even though I know how the trick is done, I'm still in awe.
It may just be the way I develop, or just don't recognize the pain, but FMPerception ( and hopefully soon FMComparison ), are the only tools I really need. Though, a few tweaks in UX, and some added features ( some of the above included ), would be fully welcomed.
Exactly.
Can you imagine a professional development tool not having global search and replace or regex or professional level SQL support (or the bazillion other missing features)?
Oh, wait a minute...that's FMP!
(see my list above -- the purpose of this posting).
Thanks, J.
2 Likes
Excellent posting!
1 Like
(Duplicated reply. Removed.)
The truth is Josh that Claris has to do one or the other but currently they do neither.
Meaning if they don't want to be a professional developer tool and instead be for "problem solvers" then give these people the tools they need to solve their problems like search and replace in your code, and in others areas (basically what MBS and DA provide), plus some lightweight built in analysis tool (say 25% of what FM Perception or BE does) OR
Don't do that and give us better professional features inside the platform and since most of us make our living from it we'll keep buying third party tools like we do now. Instead after 30 plus years we still don't have basic search, replace and analysis tools that most newbies would expect or professional level controls that most full time devs would expect.
Claris has always struggled trying to figure out how to make the product not seem difficult or intimidating to new users (You can be a developer of "apps" in 5 minutes with FileMaker!!) while still trying to convince traditional developers to adopt the platform.
This has resulted in them making many compromises that has left all the stake holders in their platform frustrated on many fronts. Let's hope their new direction as company will finally address these issues.
9 Likes
Wow. Great!! So well put.
I'm of the opinion that this is a matter of Developers using the tool because it's powerful vs Claris making a developer tool.
The full reality for me is that, while there are limitations and quirks I don't like, there are ton more quirks in other stuff.
The find and replace stuff just doesn't work with the way directly in the way FileMaker is set up. Setting up a dev environment the other way, so you can find/replace would cost us some of the native functionality that we assume is handled for us. Again, I'm not saying I don't want it. I would love to see that. But it's not a deal breaker, it very little affects me getting the project done and providing a great user experience.
Claris has not struggle with who their target is. It has always been us, as developers, wanting them to expand the circle to include us. And they have done that some. So I can't say I agree with that sentiment.
I do 99.9% of my dev without either of those tools. I have other ways of find and replace if I need it. And the way I develop, I almost never need to go make a change in more than one place. So I again can't agree that "all the stake holders" are frustrated. Sure there are some, there always is.
Go be a SwiftUI developer right now... Have fun. It will make you appreciate what you have in FileMaker, when you are sitting there writing code to update the screen and move the cursor one half-width character to the right and determining when it needs to reflow. lol ... We've had this discussion before, haven't we?! You and me.
The problem for a lot of us, is that we KNOW what it's like to have professional tools in a development product.
While I agree FileMaker has some excellent capabilities, I'm constantly frustrated by its long list of of missing professional capabilities and the constant ways I must work around those omissions (no search and replace in version 18? Please name another development tool that lacks basic search and replace.).
A lot of the features found in professional IDEs deal with object orientation. Intellij, for example, can not only do regex search and replace, but it can do a "Structured" search based on class structure. FileMaker gets a "pass" here since it's not OO at all, but it also gets a FAIL since it's not even object based and clings to an outdated 1990 programming model where you can't send messages to objects (Hello Visual Basic 3?)
Although from your postings, I realize you won't be moved, but please review my initial posting above.
If FMI made cars:
-
A steering wheel would be third party plug-in.
-
Brakes would only work going forward.
-
The roof would only cover half the car.
Etc..
1 Like
LOL 
4 Likes
Claris has not struggle with who their target is. It has always been us, as developers, wanting them to expand the circle to include us. And they have done that some. So I can't say I agree with that sentiment.
I'm not sure I can agree with that Josh. It always appeared to me that the technical and marketing teams had different messages. We've been through the 'it's easy, setup a database in 5-minutes (2 of our current projects have over the last 18 months had 723 and 745 development hours invested in them respectively and we're nowhere near finished yet) and moved on to the 'problem solver's problem solver/work place innovation platform' messages, none of which are aimed at developers, unless maybe they can talk whale song.
However, the most recent technical additions are definitely aimed at developers, I'd love to see the traditional list and forms developer (nothing against them, I was there myself once) dealing with JSON, APIs, SQL, data synchronisation, command line server maintenance, etc.
It is interesting that we've had 3 continuous years of significant growth, but most of this has come from traditional on-premises development. Our cloud business was stunted with the revised EULA introduced with v15 and we've now had to abandon any plans for SaaS in the traditional sense, again not due to technology, but due to licensing restrictions and inherent cost overheads. We're far from abandoning cloud, but are too busy developing to revamp our marketing at the moment.
The point of the above, is that when we received enquiries for our SaaS products, we could turn around enquiry to FileMaker license sales very quickly. With the near 1500 hours development (not charged by the hour) mentioned above we've received significant income over a lengthy period of time, but both clients are already FileMaker users, hence the only benefit to Claris is their existing annual maintenance payments. It isn't us who has lost out on this change of business that has been forced upon us.
So I guess my wish for v19 is (yet again) new licensing that allows developers to use FileMaker as a proper platform to be able to compete with the growing online semi-SaaS/semi customised products and to fix some of the things that have never worked that we are constantly having to work around. Oh, and of course 'OnRecordUnload' or 'OnRecordExit' script trigger, which I suspect will not happen.
4 Likes
Well put @AndyHibbs. The change of licensing conditions was against the interest of FMI’s hosting provider customers and squashed their SaaS business. Security concerns have been brought forward to justify the move. Not much later, Claris pulls the very same SaaS rabbit out of it’s hat, grabbing that part of the business. Honni soit qui mal y pense.
Coincidentally, the FM Cloud offer omits to say that each customer’s FMS deployment runs on its own VM...
1 Like
This is a tool developers use. It is not a developer product.
Talk to the CEO. While they do want to make things a little better for the developers that use it, the primary target audience is not developers. They don't exclude us, but all of their marketing efforts are for citizen developers and power users. Even in the testing program, the persona they use to describe features is an above-average user.
This is one piece I want to be clear on. I would love to see many of the features you listed added to the product. Full stop.
The piece I disagree with is being frustrated that many of them aren't there. I don't buy an family sedan and expecting to be able to race highly modified, custom built NASCAR machines. Going in public and complaining that my Saturn SL1 can't hit 205mph, and can't handle hard turns because the tires collapse at anything over 50mph when turning...just seems counter-productive.
None of this even touches on the work involved to set up a solid, professional dev environment. That is so different than just firing up FileMaker and clicking Create New... and you are up and running. The comparisons are Apples-to-Oranges.