[benjaminwootton]

And people wonder why enterprises are risk averse and nervous of open source and new innovations from startups.

Are there any circumstances where a company can reasonably go dark for a week?

If it is an acquisition it's pretty disrepectful of people investing in their solution.

[wingless]

On the contrary, I would say that enterprises should take the risk and use such projects. They likely have enough capital to hire the correct people (e.g. one of the core devs) when a problem shows up.

Startups can't afford that. Therefore they should use older technology until they grow big enough to tank the damage.

[bnastic]

That's not how enterprises work.

[moreentropy]

Ack. Sadly, enterprise procurement is all about avoiding blame in case something goes wrong. Another issue is contributing to open source projects. I know a case where a customer's in house developer had to wait two months to contribute a few lines patch to a open source project because legal objected. They never dealt with that situation before.

[nikanj]

Enterprise procurement is all about avoiding things going wrong, avoiding blame is a distant second. When every hour of downtime costs hundreds of thousands, the ridiculously expensive, 24/7 SLA supported enterprise solutions don't seem so overpriced anymore.

[dredmorbius]

That's how at least one of my start-up experiences wound up. The large enterprise entity who'd committed strongly to the product ended up bringing much of the development effort in-house, for several years.

You've got to realise that for enterprise software, there's $10 sales, $100 sales, $1,000 sales ... and then $1,000,000 sales. At the top level, you're looking at entities who have the resources, and occasionally the motivation, to bring product entirely in-house. It might not be their first choice, but it's an option, and quite likely something they've already considered when negotiating for services.

[rubber_duck]

Can't say I would want to depend on being able to hire one of the few devs from the core team as my contingency plan in a high risk situation. Precisely because they are fat and have a lot of money to throw around - that leaves room for inefficiency and getting squeezed by Oracle and the likes in exchange for covering their ass.

Didn't the same problem happen to the CloudFlare with LuaJIT where they would like to hire guys that worked on it but couldn't ?

[chx]

Nope, it didn't. CloudFlare is listed https://luajit.org/sponsors.html here.

[rubber_duck]

Sure but they are still trying to hire someone with direct experience working on LuaJIT and failed to recruit prominent community members from what I read around here.

[rtpg]

Well the great thing is that since it's open source (esp. if the OSS community itself is vibrant), the software can outlive the company.

Things can get difficult in some situations, we're all human. But Joyent going bankrupt wouldn't mean the end of Node. But where is Datomic without Cognitent?