What happened to Microsoft's Midori operating system project? | Hacker News Reader

What happened to Microsoft's Midori operating system project?

zdnet.com

164 pointsrbanffy34 comments

Here's the blog of one of the developers:

http://joeduffyblog.com/2015/11/03/blogging-about-midori/

He is planning a whole series of entries about Midori.

Yes - a lot of this is on the edge of what I understand about low-level programming, but I find it fascinating!

He hasn't really gotten into low-level details yet, but rather design principles along with fault and security models. It should be interesting to see it unfold.

One man's design principles is another man's low-level :)

But this is one reason it helps to complete your Computer Science degree. A lot of "operating system" design principles seem low-level to me, as a simple high-level programmer.

Midori was Microsoft's version of DEC's "No output division". They basically shepherded all the super senior people there to fool around with shit and not impede progress elsewhere. No real product was ever expected.

That's a very interesting theory. It sounds plausible. However, do you have any sources to back this up?

that was my impression of it as well, a very senior employee retention scheme.

No sources per se, other than a few people I know who worked there. Of course this wasn't the official party line, but it wasn't that hard to deduce.

I'm curious why they keep the super senior people if all they can do is:

    - Produce no output in a no output division, or
    - Impede progress elsewhere.

Obviously I'm missing some part of the argument, can you elaborate?

It's so those people who can do research, the goal of which is not necessarily producing "output" in the form of products (what I assume GP was referring to), but to experiment with new ideas and learn lessons that can subsequently be applied to new products.

Microsoft's research division is one of the best in the industry: http://research.microsoft.com

Easy: - some of those senior employees are senior for a good reason, and they could do a ton of damage if you piss them off and they go to a competitor - some of the stuff they fool around with may end up being useful elsewhere, which it did, in some cases - and in the unlikely event that a usable product is produced, the company wins as well

It's a win-win-win, no matter how you slice it.

So they "may end up being useful" to degree "could do a ton of damage if they go to a competitor". That's not very similar to "produce no output" as you said before. And what is the reason why they could only "impede progress" here and "being useful" there?

Very senior people are best used sparingly. Put too many of them on a real project and pit them against each other (as the old MSFT performance review process tended to do), and they start beating their chest and trying to backstab the other people they're competing with. OTOH if you put them in research and tell them how super important their project is and how it's advancing the state of the art and stuff, and you might get a few years of reprieve.

DEC had an explicit "never fire anyone if you can possibly avoid it" culture.

There were no significant lay-offs until the last few years.

You might think that's insane, but it created a lot of loyalty. And I'm not sure it's any less sane than the current religion of cutting head counts at the first downturn in profitability, losing all the good people (because they'll be the first through the exits), and crushing the morale of the rest - never mind other popular stupidities like off-shoring to countries that don't speak English, or hiring back fired employees for vastly more money as contractors, because they're the only ones who can make stuff work.

Meanwhile I keep wondering if MS should reinvent itself as a straight R&D company. The research and blue sky side regularly does supercool stuff. The management, strategy, and customer satisfaction cadres - maybe not so great, actually.

That's not insane at all if the company have thoughtful hiring policy (no HR involved, reasonable trial period) and stable enough line of products.

They had a net income of $4.62 billion last quarter, how is that "not so great"?

http://www.reuters.com/article/2015/10/23/us-microsoft-resul...

He didn't say they don't make money, he said they don't do supercool stuff. Money is supercool stuff too, but not in the same way.

What other cool stuff can "management, strategy, and customer satisfaction cadres" even do? And if just because "management, strategy, and customer satisfaction cadres" is not doing "supercool stuff" (whatever it means), the company should focus on R&D, then the vast majority of companies should cease to exist. Like it or not, companies are here to make money.

Saying MS should reinvent itself to be a R&D company is like saying Toyota should stop making cars because their normal cars aren't running by water, auto driven, and charged by solar panels yet. Ya, what's the point? Use your money to research on energy, AI and what not. Never mind where that money comes from.

When I read these absurd statements I think the author is either trolling, has an agenda, or is living in a bubble.

Actually I'm just a typical user - albeit with some coding and product design experience, which most users don't have - but with opinions ahared by a significant percentage of other typical users.

Consider:

The Win 10 keylogger/spyware issue Win 10 forced upgrades. The nonsensical "one OS for all devices" strategy. The Nokia acquisition. Silverlight. Kin. Zune. IE6, and the disastrous effect it had on the web design industry. Win 8, and the disastrous effect it had on the PC hardware industry. (Never mind the fact that MS took a perfectly usable and popular OS - Win 7 - and broke it almost beyond all imagination.)

That list is nowhere close to being comprehensive. (How many people remember that MS tried to create a library services and publishing division?)

In what sense were these strategic successes?

Now, a lot of smart people work at MS, and every so often they produce something that could be a game changer - like the Courier project.

Somehow the game changers never ship. What does ship is - too rarely great.

Where does that disconnect come from, if not in strategic direction and management?

"The nonsensical "one OS for all devices" strategy."

Which will probably make their mobile devices rise from nothing to something relevant

"IE6"

IS6 was a amazing browser at the time, the most advanced for years. The main problem was it didn't have a auto updater.

"Win 8, and the disastrous effect it had on the PC hardware industry. (Never mind the fact that MS took a perfectly usable and popular OS - Win 7 - and broke it almost beyond all imagination.)"

That's just your opinion.

So true.

I honestly haven't seen better strategy people in the software world than those at Microsoft.

Iirc certain bits of it eventually found their way into shipping products

I think they actually did have a working os that could play media and do other things but it's unclear what the broader goal was other than pushing the boundaries on systems development

Some of the design documents were really a joy to read. Very smart and thoughtful people.

Except that Microsoft hired people out of college to work on Midori, which they surely wouldn't have done if Midori was simply a place to dump the people they couldn't/wouldn't fire, which was the point of DEC's "No output division"[1].

[1] http://archive.computerhistory.org/resources/text/DEC/dec.be...

Sure, it's not like super senior people would actually want to do the more mundane work. You need junior devs for that.

Except that this is not true. Almost everyone working on Midori were (very) high level engineers.

I work with a few of ex-Midori folks, all of whom are fantastic engineers, most of whom would not have been senior enough to have been shepherded off for being super senior.

R/programming discussion is very interesting, there are ever few Midori developers in it:

https://www.reddit.com/r/programming/comments/3t50xg/more_in...

It always seemed like MSFT was about to position M# right into same niche as Rust. Both languages had the mantra of safety and zero cost abstraction.

It certainty felt like the blog posts about M# struck a zeitgeist nerve. I doubt it was ever really intended for release, though.

Didn't they hire the guy from BitC? BitC and ATS looked like the only interesting safe high perf languages for a while eh? Rust now has all the interest though.

Both BitC and ATS are big on formal program verification; rust has the interest because the number of people outside of academia that are interested in formal program verification is approximately zero.

I thought their was some advancements in realtime GC attributed to the Midori project?

Reply on news.ycombinator.com