Sounds like Google's position is that they did do what was alleged, but that Attia had signed terms that made it legal.
What you described above is similar. Legal, but shitty. Some amount of information obtained under false pretenses is surely passed to the "clean room" team. Otherwise, why meet with the target at all?
Edit: This Verge story has pretty good detail on the backstory: https://www.theverge.com/2015/2/17/8048779/google-x-eli-atti...
I guess it gives the employees plausible deniability, but the company knows what it's doing.
The room can be clean, but what about everything else? There's more to software than the actual written code.
Part of the contract that both sides had to sign included a clause that everyone that was involved in the vetting process could not work on anything related technology wise for 18 months after if the deal fell through. (As Cisco had a competing product already that our company made).
This was more in cisco's favor as they have a full team that does M&A and handles this vetting. So the engineers involved on cisco's side in the vetting would likely not be key engineers on the product.
Large companies are accustomed to isolating people in M&A deals from the rest of the company to be able to prove clean room easily.
To over-simplify, Google probably just tunes the Hotspot JVM.
And now, of course, dart and flutter, on a new OS to replace the Linux underpinnings.
Although I’m sure Zuck would have turned them down regardless.
Would they pass up booking.com solely because it's Perl? The value there hasn't much to do with the plumbing.
I’d wager they’d take a closer look these days.
That said, there is a lot of poorly written legacy php code out there before frameworks really took off for that language.
https://np.reddit.com/r/Windows10/comments/74xc2z/windows_an...
It used to be a deal-killer. It's less of one now, but that's more because they'll acquire a company and have the staff re-invent the project in one of the "blessed" languages (Java, Python, C++, Go) while they run the old code on isolated VMs.
Actually, they do that even if the project was in one of the blessed languages. There's a tech stack my internal project uses that was acquired. We can't use it for certain sensitive data because it's considered "untrusted", which is kind of a pain. We're waiting for the re-write to stabilize so we can move over to that.
http://www.cvedetails.com/product/128/PHP-PHP.html?vendor_id...
I have gotten out of the consulting world a bit, but for a while a php codebase meant I would double my estimated time on any quote I sent out. There is terrible php out there and most of it is vulnerable to every one of the OWASP top 10.
The only other thing I like other than PHP is something that's even dumber, and that's rewriting it in C++. That's the language I love, using it like C with classes instead of structs, and overloaded operators. That's all I ever wanted in life, really. Anything else? Let it mature 10 years, and not be pushed by one of these companies so super keen on being middlemen. Until then I'd rather derp around than buy into any of it.
I know I'm full of it, but here's how I write super secure code: I simply start with nothing, and add things slowly and deliberately. The program doesn't do anything I don't whitelist it to do. Orwell said if people can't write properly, they can't think properly... well, if you need a linter to learn to put a semicolon every single time, or a safe language [sic] to realize you're dealing with user input, can you code properly? I say this as the worst coder in the world, but I still wonder. Orwell probably had people who proofread for him which is where it all falls down, and I honestly don't consider myself a good programmer. I'm not above compiler warnings. But I don't believe in that to the extreme, I don't believe in looking at the GPS instead of the road.
Granted, that's all for things I intend to use for myself, with me being the only user, and I'm also fond of having a local authoritative copy and just syncing the changes one way, other than simple stats that get generated on the server. That way, not having critical data that could be stolen, I only have to worry about maybe rogue files getting dropped on the server. Didn't happen yet, I'm sure if I was paranoid I'd find a way to automate that. Bam, done. Not useful for a lot, but if everybody did that instead of flocking to yet another silo, just make up their own crummy stuff but be as careful as they can while they do it, there'd be a lot less tears over data breaches. Security through individuality.
Okay, I guess I am being argumentative, sorry. But it's not because I feel left behind, it's because I'm bored. Not at the moment, but with web dev trends, and Google. I just have to roll my eyes so much when I see screens of requirements to tinker with the same 3 tutorial type projects. If you need more than PHP to make something like, say, HN or blogspot.com, what are you even doing? And then look at the HTML source of just about any Google page. I don't care what they use to spit that out, I'm sure it's a million best practices used perfectly, but the result turns me off so badly. If that's super useful and fun, good for the people enjoying it, then they won't miss me doing it, too.
* Google employee comes up with an idea.
* They go and research the idea to check if there are any already existing companies which do it.
* If any are found, they meet and decide if they should buy the company, reinvent the idea, or that it isn't relevant.
* If, after investigation it is determined the company's tech isn't good enough, they will re-invent. Google has fairly strict tech requirements (no php, no shady licenses/ownership, no pirated stuff, etc.), so many companies don't pass.
* When they reinvent, they will do it "clean room" - ie. none of the people who reviewed the original company will be involved in the re-invention.