Reading this I almost gave into the urge to make a bet
Of course there are things that you can do on Excel that you cannot do on Coda, such as:
connect to Windows API (COM objects) through VBA
play MIDI files
interact with your filesystem, devices, drivers
make a 3D shooter game on it (although I tried lol)
manipulate doc schema (thatās a real bummer yeah; one of the reasons why I didnāt gave in. Although in many cases one could work around that with some smart schema design)
You have to do these with VBA though, which is an un-nice language to learn. Or hire $200+/hr consultants to code those for you without a slightest chance of you being able to maintain it yourself ĀÆ_(ć)_/ĀÆ
(it would be fair to say here that Iām myself a $200+/hr consultant on Coda, albeit I share a lot openly, and some more with patrons.)
Ok yes, those are some things Coda cannot do. Also, Coda is terrible for training neural networks. Though a fun exercise Maybe one day Iāll build a neural network doc to share.
Iām actually very pragmatic and that means Coda as well. I would dissuade a customer to use it where itās not a good fit. I actually donāt care about losing profits there, I think about providing care first.
I donāt support that whole āno codeā hype. Tools are tools, each fit for its purpose. Platforms like Coda solve some pains really well, whereas some things are much better done traditionally. Truth is, serious customers donāt care if itās code or no code ā they want their pains solved as fast as possible with as few pains emerging eventually as possible. Mindlessly pushing airtable/coda/zaps/webflow/bubble/name-your-nocode-tool is blatant fanaticism and short-sightedness
I think even that the terms āNo Codeā and āLow Codeā are a little bit disingenuous.
The reality is that if you learn Coda (or some other tools like Bubble, FileMakerPro, perhaps others) you really have learned how to code!
Lots of companies seem to want to shy away from that. Iād guess itās because they want to make it seem easier. The reality is that:
Sometimes building what you want is hard. Thatās ok.
Learning a programming language | new paradigm is frustrating and sometimes discouraging. Thatās ok, too.
I see it as a different sort of opportunity. Thanks to Codaās amazing language design and your great videos, @Paul_Danyliuk, @Brandt_Smith1 can learn to code! Itās a skill eventually everyone will learn, just like reading and writing.
And perhaps like reading and writing, there was a time when it had to be done painstakingly and couldnāt easily be shared, and as technology has progressed itās now much easier for anyone to create and share.
This is why I like Codaās Maker framing (not Low-Coders). Because the point isnāt that itās simple (as youāve noted @Paul_Danyliuk, sometimes itās not, and @Peter_Wills youāve pointed out that some things are impossible in Coda) but itās incredibly fast to develop.
That part is cool. Perhaps after youāve built a proof of concept in Coda, youāll be forced to build the final thing in C++ for performance. But Coda is a wonderful tool to think and explore with, and when thatās the case many things end up being completely built there (see python).
Exactly! I couldnāt agree more. I could compile many examples of things I do in a spreadsheet that may or may not be possible in Coda, Notion or Airtable but it would take hours/days and would likely not convince the āanything a spreadsheet can do Coda can doā fraternity so I choose not to waste time on that exercise.
I repeat, I have used spreadsheets and databases since the late 70s, read VisiCalc and dBase, and Coda and Notion over the past 2 years or so and I do have an inkling of what I am talking about.
Now that I have some Coda brains, the balance between what I would do in Coda and in Spreadsheets has changed quite a bit, towards the Coda side.
It requires a bit more thinking and planning, but I am at the moment busy doing a financial planning forecast, and doing so in Coda.
I want it it to be able to do scenario planning:
what happens when I retire (So an arbitrary number of years into the future based on the userās age)
what happens if I die,
get retrenched,
get ill tomorrow
etc
I started doing that in Excel, a while ago, but when I now restarted the project, I did so in Coda. I think I have the solution for a MVP in my head, and is now busy making it.
There are some bits that I hope is still either on the Coda development list, or I can implement a workaround - e.g. how to get several people getting to use the functionality with confidentiality. There are a couple of existing ways I can think of, but they all have drawbacks. Another topic is more formulasā¦
Maybe one day Iāll build a neural network doc to share.