Microsoft is too big! Time to split it up in pieces.
2019-04-26 22:41 splitms [permalink]
The Microsoft Corporation reached a valuation of $1·1012, this could mean the US dollar is losing its value and thus its international relevance, but more probably that the company is positively humungous. Just like the behemoths is has recently joined, it's a worldwide employer, does worldwide business and is sitting on a big pile of cash it's not quite clear even how it could spend it. So when things like this happens, voices speak up about splitting up in pieces. But how? I'm listing just a possible set of fraction lines here through the little I know of the enterprise and what I understand about it. Please feel free to differ, or revel at the level of insight of this outsider, whichever feels most appropriate.
Products: Surface, XBox, HoloLens, other peripherals (and what's left of Zune and Nokia)
Revenue: from selling devices
Operating system and out-of-the-box software would no longer be available from a source in-house, so they'll have to shop around like the rest of us, or (try to) go with freeware.
Department: Developer Tools
Products: Visual Studio, compilers.
Do prices need to go up? Will this scare people away? Are the free community editions still possible? And what about Visual Studio Code? There's perhaps even a small chance of access to documentation turned into a payable service, but I hope it's clear that this most probably will scare away an important portion of the public. Anyway, this is where the true legacy of Microsoft lies, if you consider BASIC as one of early Microsoft's first successfull products, and the success of its compilers and platforms going hand in hand with the rise of Windows.
Department: Operating Systems
As I said before, Windows could do Coca-Cola and go open-source, and what people would be buying is the certified binaries to install, and — perhaps more importantly — a continuous stream of system updates refining (new) features and assuring better security. With an open-source Windows, one could go and try at it for oneself, carrying the full set of risks, and it's a big part of the equation if you're really sure that's what you want. Then again, it would open the playing field, and the few that could build up the substance to offer an alternative with a similar degree of dependability would at least get a fair chance. Don't forget, if this plays out and a few big ones compete like this, it keeps all of them on their tows, and typically will compete in innovative features and cooperate for collective security. Something that may regretfully be reverting in the field of web-browsers now that Microsoft too switched to a WebKit-derivative, and all we've got for technically true alternatives is Mozilla's offering.
Department: Productivity Software
Products: Office, Team System
Microsoft has been accused before of having an unfair advantage from the people building Office sitting next door to the people building Windows, being able to open hidden undocumented features for exclusive use. To their credit, they have been able to disprove any such claim, and have opened up the documentation on both sides about how it's all done, and how anyone could do the same. Still, as separate corporate identities, this would truly open the market, and offer a chance of other players to form a collection of stake-holders to jointly push for required features and revisions the operating system should provide.
Department: Business Software
Products: SQL Server, BizTalk
Much like with Office, the general public gains from having full transparency about how SQL Server embeds with the system. Unlike with Office, this hasn't been as much of a problem with SQL Server, perhaps because with advanced systems administration you naturally learn about the nooks and crannies of the system when it has to perform under load. Also with the port to Linux and other Unix-derivates complete, this is no longer an issue. What may be an issue in recent days — in my opinion — is the current pricing and how it may be employed to push people towards the cloud. I've heard from a number of instances in the field, that it's not working, and large projects are moving to PostgreSQL instead. Sometimes even also with a (partial) move to the cloud (and almost never Azure I noticed). But I could be wrong, and/or this could all be changing as cloud offerings are in continuous flux. But seperating this in to independent units would offer more operability and tranparency, I guess.
Products: Azure cloud
Revenue: cloud resource use billing
As with the hardware division, operation system and developer tools would no longer come from in-house, except perhaps the system modules and libraries used server-side specifically for running virtual payloads and managing the cloud services in use by the customers. But this also means that any services used and access to system internals provided by external providers, would be accessible to anyone else as well. Also as I've mentioned before, cloud offering is relatively new and in constant flux, and I expect an era of renewed interoperability between all kinds of cloud services of all providers. For now all we can do is frown upon vendor-lock-in very hard if we notice it, and only grudgingly accept if we really find no other viable alternative.
This is a hard one. If you split the research endeavour into a separate unit, it's not clear what you would expect of it. Research in hardware is important to advance the state of the art and develop new devices. Reasearch in development tools is important because of the constant influx of new developers learning to code and the quality of work we should be able to expect from them. Research in software is important because modern devices keep increasing performance and capabilities, and offers new options for human interaction that need to get developed. So each independent unit may need its specific research lab, but there's still a section of independent research done in the over-arching field of computer use, networking, man-machine psychology, how hardware and software could have less impact on the planet, etc. Where's the business model in this, I don't know, but if we have a peek over the fence, I guess the way ex-Google's X is managed by the Alphabet Corporation may serve as a nice example. Try new things here and there, and if anything sticks, spin it off, and house it with a different company in the group, or let it loose on the world on its own.