News for hackers. Stuff that matters.

A comment on Malleable Systems: Software must be as easy to change as it is to use it.

Re:Malleable Systems: Software must be as easy to change as ... by shalabhc on Monday April 13, 2020 @06:45PM (#22861015)
This isn't about building software similar to what we currently have and just making it more end user configurable. This is about rethinking software so that end user malleability makes sense.

What does such software even look like? We want some kind of end user modifiable artifacts, still contained within some constrained substrate. Think Minecraft or Excel - I can give you a spreadsheet to calculate some taxes, but you can then dig in, change the formulae, remix with other sheets, and so on. Most software is not like this - typically you get some pre canned views and switches, not a canvas of composition.

[ Parent | on HN ]
Re:Malleable Systems: Software must be as easy to change as ... by denster on Monday April 13, 2020 @06:54PM (#22861055)
This is exactly what we built at MintData. [1]

A new way to construct software by using a spreadsheet to define application logic, instead of today’s spreadsheets which breed chaos.

“A canvas of composition” - I really like that phrase to describe this approach of re-mixable software creation and usage.

—-

[1] I’d add a link here, but I think folks below might be sharpening their pitchforks by now :D

[ Parent | on HN ]
Re:Malleable Systems: Software must be as easy to change as ... by qnsi on Tuesday April 14, 2020 @10:40AM (#22866551)
Dude pls with the spamming
[ Parent | on HN ]
Re:Malleable Systems: Software must be as easy to change as ... by jcelerier on Monday April 13, 2020 @06:49PM (#22861033)
plenty of people use software such as Max/MSP, PureData to do exactly that
[ Parent | on HN ]
Re:Malleable Systems: Software must be as easy to change as ... by denster on Monday April 13, 2020 @07:01PM (#22861114)
The issue is that FBP (flow based programming) scales to a certain point of complexity.

After that, it stops being more effective than a textual representation or (gasp!) even old fashioned imperative programming.

As a result, we need new approaches for more powerful tooling that scales better along the axis of complexity (for the idea / thing we want to create)

[ Parent | on HN ]