Live data from Hacker News

RethinkDB: why we failed (2017)

defmacro.org

231–237 of 237 posts

Re: RethinkDB: why we failed (2017)

#231

I was under the impression that rethink db was not open source (until after it “failed”). That’s not a reason for failure, of course, but I have mentioned before that we (as in developers) often only choose things with permissive licenses, often to the detriment of the products we “support” (see the recent elastic drama, where we blamed elastic being forked by Amazon for being too permissive!).

It was open source from the start, but under AGPL which is a fairly restrictive open source license. I remember diving through the code trying to debug bugs in RethinkDB when using it in production long before they went out of business. See my blog post http://sagemath.blogspot.com/2016/10/rethinkdb-must-relicens...

Yeah, AGPL seems to be one of those licenses that aren’t popular here. I’ve never understood it myself - especially for something like a db which folks are generally less likely to modify.

What would you say is the problem with AGPL? I was thinking of releasing a product under it myself.

Re: RethinkDB: why we failed (2017)

#232
post #153

Earlier quoted context omitted.

This is exactly why Oracle in the 90s and 00s starting buying up software Application companies (Peoplesoft, JD Edwards, Hyperion, etc). And then in the old on-premise world, you sell them both the App and require the use of your database which is an additional sale. This strategy doesn’t work in a Cloud world though. Since customers are no longer buying the individual components (like the database) but are instead j…

IIRC, Oracle managed to get a huge, well paying customer (DoD) behind their "we build a better database" pitch. That provided enough funding, users and support to move into applications. In fact even if they didn't expand into other areas, that support could have kept them as a profitable DB company for decades.

Ten years ago, when I still worked in consulting, all my clients (large institutional companies) used Oracle and they'd say they were willing to pay the premium because Oracle was 20% faster than everyone else.

Re: RethinkDB: why we failed (2017)

#233

Earlier quoted context omitted.

Did RethinkDB ever actually attempt to get people to pay for it? The massive, gaping hole at the heart of this analysis is "maybe we failed because we gave our product away for free as open source"?

Yes, my company paid for RethinkDB over a year before they shut down. They sold support to us for $100, and of course their engineers did far more than $100 of bug fixing for us. The reality was that RethinkDB was a very difficult product to create and when you pushed it hard in production it had some serious bugs. The other core issue is that pretty much the same unique problem RethinkDB solved for us (data storage…

A support contract for $100 !!! Wow, I thought I'd heard it all but that one is new. Even cheap consultants I know won't sell you only $100 worth of time.

Re: RethinkDB: why we failed (2017)

#234
post #109

Earlier quoted context omitted.

Andy Gocke put it on twitter better than I've ever seen it anywhere else: 'Developer tools seemed like a good industry to be in, "sell shovels in the gold rush" and all, but it turns out developers prefer to dig for gold with their teeth.' https://twitter.com/andygocke/status/1017509689695715328?lan...

I chuckled a bit, but you and I know that devs are not savages :) I'd probably word it differently: Developer tools seemed like a good industry to be in, "sell shovels in the gold rush" and all, but it turns out developers prefer to make their own shovels to dig for gold (because it is often cheaper than buying).

Based on the tools that I've often seen developers build in-house, "dig for gold with their teeth" is fair, much of the time.

Re: RethinkDB: why we failed (2017)

#235

Earlier quoted context omitted.

https://www.prnewswire.com/news-releases/mongodb-inc-announc... - Revenue: Total revenue was $150.8 million in the third quarter fiscal 2021, an increase of 38% year-over-year. Subscription revenue was $144.1 million, an increase of 39% year-over-year, and services revenue was $6.7 million, an increase of 19% year-over-year. - Gross Profit: Gross profit was $104.7 million in the third quarter fiscal 2021, representin…

MongoDB Atlas's business model seems very, very odd to me. The prices for hosted Mongo are so high that I really find myself asking who is paying for this and what the target market is. In particular I wonder how they can possibly be losing money when charging such insane markups unless they have virtually no customers. For a dedicated database that has 32 GB of RAM, 8 virtualised CPUs, and only 160GB of storage you…

Don’t you get a replica set for that $? That's three of your laptops. And if any of them fail, a free replacement. And that FT sysadmin of yours is willing to work 24x7? Cool!

Re: RethinkDB: why we failed (2017)

#236

Earlier quoted context omitted.

But most companies could do things their suppliers do, because companies always have the option of hiring people with those skills and developing those competencies in house. E.g. Google moved from running their software in commercial datacenters to building their own, or how Amazon has progressively taken over more and more of the supply chain. Android is sold to firms that can build their own mobile operating syste…

That's an option in the same way that anyone can go back to college, specialize in pre-med, go to med school, complete residency, and then become a doctor. It's possible in theory. In practice, you're looking at a decade+ of your life, with significant risk that each step fails. So it is with companies bringing core competencies in-house. Each step takes time, and is fraught with risk. It's not a matter of simply hir…

I think you may be over-exaggerating the difficulty of doing this. Companies both insource and outsource things pretty frequently. For example it's common for tech firms to insource catering. Building restaurants is not a core competency, yet they are perfectly able to do it and haven't suffered any great distraction from doing so.

In practice one company acquiring another, or even just poaching some employees, is a common thing that happens so frequently we take it for granted. When Apple acquired PA Semi nobody batted an eyelid even though they were in-sourcing something as complex and difficult as CPU design. And there are many examples outside of the computing industry too.

With developer tools specifically, the alternative to buying the product is often "We'll have a team of engineers spend a quarter or two building something that is specific to our needs"

If it takes a small team of engineers only six months to roll their own version of your product, that's what I meant by the value proposition being quite thin (or your price being too high). Also, most companies outside of very rich tech firms will not actually staff up an entire project just to clone a tool they could acquire off the shelf, unless that really is the only reasonable path forward e.g. their needs genuinely are unique.

Now, in computing we do have the issue that the prevalence of VC money, and indifference of VCs to whether it's being well spent, has created a large population of "very rich tech firms" with lots of apparently bored engineers working at them. So they spend a lot of time churning out open source frameworks that they maintain for a few years and then abandon, even when they could have used something else. That's rather unique to software but I don't think it's something inherent to developers as a type of person. Rather, it's inherent to a market where money is free and firms compare themselves to each other by the sheer size of their engineering teams. Outside of the Valley that problem mostly goes away.

Re: RethinkDB: why we failed (2017)

#237

> To see how this plays out for other companies consider MongoDB (valued at roughly $1.6B with ~700 employees), Maybe another, important, lesson the RethinkDB team left out was: raise more capital than you need, and go public fast. Two years later, MongoDB the company is now valued at $19.491B. Like many other public tech companies, it is yet to turn a single profitable quarter, and is not expected to in the next few…

>Annual YoY revenue growth is flat too at around $500m. 2018: $154M 2019: $267M 2020: $421M 2021: $590M Source: https://finance.yahoo.com/quote/MDB/financials?p=MDB

You're right, thanks. 40% YoY revenue growth at size is considerable.
Post reply on HN