Ironically, I'd argue that FLOSS has simply served to enable proprietary software vendors.

In embedded/IoT, devices might use FLOSS but they are usually violating the licenses and the parts that make the device special are usually proprietary (think robot vacuums running Linux with a proprietary daemon doing house mapping etc). Often there are proprietary drivers or firmware too.

In servers FLOSS operating systems have indeed won, but lower level hardware-facing software like UEFI has not and indeed most proprietary UEFI implementations are based on FLOSS implementations, but with proprietary features or hardware support added that means users can't use the FLOSS version.

On the web FLOSS has indeed won, it is fairly rare to use proprietary JS/backend frameworks, but the things that are built using FLOSS web tools are mostly proprietary.

On the cloud FLOSS may have won the VM side, but the hypervisor side is trade secret land and management interfaces are mostly proprietary, very few cloud vendors use OpenStack.

Gaming consoles may use some bits of FLOSS internally, but they are locked down and proprietarised.

Mobile OSes may use some bits of FLOSS internally, but they are locked down and proprietarised to the extent that it takes a lot of work to get Linux running on a newly released device, sometimes even requiring exploiting security vulnerabilities.

Enterprise Software might be based on FLOSS but it is all proprietary at the end of the day.

On the desktop, the trend continues, macOS is FLOSS+proprietary, Chrome OS is FLOSS+proprietary.

Precisely, this is why I think open data should be the next battle. All code takes input and gives output. I don't mind closed source as long as the interfaces are open ... but data being bought and sold is really troubling. Apart from public data sets, most data sets are closed. I am not sure if there are companies that publish all the data they collected openly.

The data issue is that it most of it shouldn't be collected in the first place, not that it should be public.

Well thats a chicken and egg problem in a way. Without data you can't build some types of software. I think you have three types of data.

1. Private data / Linked Data

This is the data contributed by the user. Shopping data for example. I think the EU data laws are meant to guard this. I think it is both unethical and illegal to sell / publish primary data without the user's consent.

2. Public data

In social networks you also get interaction data, which the user had made public. I don't see any reason for companies not to publish this. Right now people use crawling to get this data.

3. Insights

This is the data generated by ML and statistical analytics.

By open data I mean user should be able to get a dump of all three and the organisation should? publish 2 openly and 3 if it wants to openly as well.

Also see: http://www.veen.com/jeff/archives/000810.html

I don't think any of those things should be collected nor published.

How would you solve a basic e-commerce use case then ?

OSS has won the cloud. This makes sense as companies gain a lot by collaborating on Server Technology, and don't lose much.

OSS has won mobile (kind of). Android makes sense as multiple companies can collaborate on the same OS and put their own layers on top.

Basically FOSS has excelled in every field except the desktop. This is because:

- Hardware is tightly controlled by a lot of PC manufacturers, which aren't all that interested in supporting a different OS.

- Apple and Microsoft which have developed their OS's for decades, gaining private experience in GUI design and user experience, which doesn't seem to be as popular in Open-Source desktop OS's.

- Majority of people don't even know you can change the OS.

I think until manufacturers start releasing desktop computers with FOSS preinstalled, like Android phones... FOSS has not won

Depends on the market. The following ones have definitely seen open source win:

Servers (as mentioned, various Linux distributions with things like Apache also being open source on top of that)

CMS systems (WordPress, Drupal, Magento, MediaWiki et al)

Programming Languages (virtually no one pays for one anymore)

Source control (Git)

Web browsers

These are dubious though:

Mobile (Android is open source, but iOS still holds a decent percentage of the market)

Enterprise software

And I don't think the gaming world has been won by open source software just yet. I don't see Nintendo/Sony/Microsoft/Valve open sourcing their dev kits/OS software/firmware/whatever, and those companies dominate most of the market.

Most Enterprise software is not FOSS - it's either written in house and owned by the corporation which wrote it, or it's SAP etc. You probably meant the stack that enterprise software is running on and not enterprise software itself?