How many open source projects are 'open' in the sense of being interoperable with other software systems? Coming up with a conforming common standard is a non-starter IMO.

They tend to be 'silos of automation'.

Interoperability is a characteristic of a product or system, whose interfaces are completely understood, to work with other products or systems, at present or future, in either implementation or access, without any restrictions.

e.g. Quantlib, Boost etc.?

That's not what "open" means although having unrestricted access to the code certainly means the interfaces are completely understood (unlike proprietary software).

It would seem that Microsoft is a far worse silo with far less interoperability than most open source projects.

I think that was true in the 90s. But you probably are not aware of developments in MS the last few years. You might be surprised.
Personally, I have no interest in smart phone interop, so no need starting a brush fire,

Back to topic: how many open-source codes are interoperable? (btw there is a precise definition, so I won't give a link)?

I assume when you say "there is a precise definition, so I won't give a link" you mean that you've already given the precise definition in italics above (otherwise you are providing poor interoperability yourself by hiding the definition).

In any case, I assume we need to construct an (M+N)x(M+N) matrix of the M proprietary & N open-source code in the world and test if software i interoperates with software j. Of course, if i doesn't interoperate with j, then there's the question of who's to blame.

That definition also alludes to "the future" as an element of interoperability. Are the file formats of Microsoft products open and controlled by an independent body or are they closed and subject to the whim's of Microsoft product development? Maybe that's something that's changed since the 90s?

But why should vendors should support BLAS 3 (Fortran) versus expecting BLAS to support them? Is the first system to reach 90% marketshare declared "the standard" thus placing the burden of interoperability on all others?

P.S. I looked at the 6-levels (https://en.wikipedia.org/wiki/Conceptua ... perability) and what struck me is that the lower levels are generally "good" but the upper levels may be "bad."

A command line interface, mouse interface, and touchscreen interface are not interoperable beyond some minimal level nor should they be.

To the extent that two pieces of software are actually trying to do different tasks in different ways, interoperability may be impossible and undesirable because different users have different requirements. In that regard, I would expect open source projects to have lower interoperability than consumer-grade software because so many open source projects are designed by specific developers trying do specific tasks in their preferred specific way.

The medical device software arena is probably one area where companies cannot afford to play the maverick. And FDA breathing down your neck all day.

In QFopen-source  libraries, interop << 1 in general.

I seriously doubt Medtronic and J&J pacemakers (or their control devices) are interoperable. From what I've seen doctors are trained by the manufacturer to use a specific device and it's associated support equipment.

The only required interoperability might be in the need for such devices to transmit medical records in some standard form acceptable in EHR (Electronic Health Records) system.

My 2 cents.

I like Open Soure. Like Alan I have used all sorts of Open Source software. However like evolution it takes time to come up with the optimal solution.
 
Strata looks basic in terms of its offering.

To develop anything it should be Python as more and more people will know it. But I don't really like the language (and I don't develop professionally anymore).

Cuch: "Speaking of costs,  I remember NEXT (but never bought one)

Steve Jobs pooled the finest of over-specified hardware and software (from PC standards) into NeXT and the company added its own innovations. NeXT was the machine-of-choice for well-funded Unix-friendly science departments.to get excel add-ins, then yes used it for trading on my PC. "

I think it was UBS that where running NeXt computers on their trading floor in early 1990's or so, at least in London trading floor, remember I visited them and liked it. This was when I always carried around my Apple SE, it came before the apple portable, and was actually more portable than the portable. I carried it every day back and forth to the trading floor, and were running Wingz on it. Wingz were light years ahead of Excel back then. 

We live in a Democracy and ended up with what the majority voted for, average computers and average spreadsheets. 

What would we have done without Steve and NeXt: ``The NeXT Computer was used by Tim Berners-Lee at CERN, and became the world's first Web server and ran the world's first Web browser in 1990"

when 3D printers get a bit better, we can print our own computers, more variety? And yes they already talk about open source computers too.

Wingz was nice but my favorite was Trapeze which was like a visual version of APL complete with named variables (instead of row-column index nonsense) and vector & array functions that auto-sized according to the inputs (instead of Excel's silly manual approach of copy-pasting formulas to fill out vector and arrays). But Trapeze got bad reviews because it was not like Excel! It's the triumph of mediocrity!

Job sure had a thing about cubes: Pixar Image Computer, NeXT, and the Mac Cube.

Cuch: "Speaking of costs,  I remember NEXT (but never bought one)

Steve Jobs pooled the finest of over-specified hardware and software (from PC standards) into NeXT and the company added its own innovations. NeXT was the machine-of-choice for well-funded Unix-friendly science departments.to get excel add-ins, then yes used it for trading on my PC. "

I think it was UBS that where running NeXt computers on their trading floor in early 1990's or so, at least in London trading floor, remember I visited them and liked it. This was when I always carried around my Apple SE, it came before the apple portable, and was actually more portable than the portable. I carried it every day back and forth to the trading floor, and were running Wingz on it. Wingz were light years ahead of Excel back then. 

We live in a Democracy and ended up with what the majority voted for, average computers and average spreadsheets. 

What would we have done without Steve and NeXt: ``The NeXT Computer was used by Tim Berners-Lee at CERN, and became the world's first Web server and ran the world's first Web browser in 1990"

when 3D printers get a bit better, we can print our own computers, more variety? And yes they already talk about open source computers too.

I can verify the UBS story, although the Next usage was limited to the global equity derivatives group. Exotics were managed based on models implemented in Mathematica notebooks. Simultaneously, UBS IT was agitating for quants to switch to programming in K, while some of said quants were busy migrating USD interest rate derivatives from a system written in APL into an Excel based system relying heavily on the Excel 4 (pre VBA) Macro language. Ah - the good old days!