Bryan Cantrill, co-inventor of DTrace, leaving Sun | Hacker News Reader

Bryan Cantrill, co-inventor of DTrace, leaving Sun

dtrace.org

42 pointskmavm19 comments

Interestingly, this is the same guy that in a discussion of open source economics coined a term for the point when your closed source vendor pushes you too far and you switch to open source and called it the Fuck You, Oracle Point

(though there is historical reasons for that name, not just random swearing)

http://blogs.sun.com/bmc/entry/the_economics_of_software

random quote: "doesn't it strike you as odd that your operating system is essentially free, but your database is still costing you forty grand per CPU?"

i find this [http://cryptnet.net/mirrors/texts/kissedagirl.html]to be far more interesting :o)

Correct link: http://cryptnet.net/mirrors/texts/kissedagirl.html

A classic, although if I remember correctly the comment makes more sense in the context of the entire thread.

I don't understand why binary Oracle compatibly isn't the number one effort in the OS DB market. That would significantly move the FYO-point.

Purchasing decisions are made by managers, not the ones doing the actual work.

Hence Oracle fits the bill if your company can afford it: It's number one, expensive (which implies good quality), and everyone else uses it. Ergo, no one ever got fired for choosing Oracle, licensing costs be damned.

Did you read the linked article? As long as there is a FYO point, there's a market for moving that point downwards.

Also, I'm not aiming to kill Oracle overnight, I'm aiming at slowly chipping away at Oracles dominance. Right now there's a huge market for serving apps that for some reason are locked into Oracle, and are too expensive to re-write.

That may be the case in some places, but certainly not everywhere. One of the public companies I've worked with uses Oracle to the tune of several million per year in licensing fees. The CEO is responsible for maximizing shareholder value and he's not stupid - he approves the checks to vendors and knows very well that a single company accounts for an overwhelming percentage of their annual software costs. Therefore they are actively looking for alternatives where the cost-benefit ratio of switching is favorable.

Heck, they would be negligent not to investigate ways to reduce their costs on behalf of the shareholders.

It's already being done, based on Postgres.

http://www.enterprisedb.com/products/postgres_plus_as/overvi...

Why would anyone consciously cripple a fine database with a "varchar2" type is beyond my understanding.

Diversity is more important than Oracle compatibility.

Also, migrating a big Oracle install is not something businesses do. I see lots of big databases that are two or three releases back. And, lets be sincere, Oracle's RDBMS has a lot of clustering/availability features that no free alternative can touch. Its fate will be the one of the mainframe - Oracle (along with Microsoft) will keep losing the lower end at the same time Oracle pushes the higher end forward with new features that don't exist elsewhere.

Solaris SPARC was never "free", you got the license with the hardware, which you paid for.

Tangent: The irony of a blog post discussing FYO point with the branded Oracle O is quite something.

"One of Sun’s greatest strengths was that we technologists were never discouraged from interacting directly and candidly with our customers and users, and many of our most important innovations came from these relationships."

Oracle discourages technologists from doing this. Bryan has a nice way of phrasing his reasons for leaving. I wish he had left a clue about his future venture too.

There is almost a sense of inevitability to all of this. The best and the brightest leave , the foot-soldiers get cut and the company culture/technology dies a slow,quiet death.

Sad but inevitable.

What is Oracle doing that is enraging the top engineers so much? Bray, Gosling and now him are all saying that they don't want to work for Oracle.

I thought if you're that far up the ladder you have a lot of freedom.

I seem to recall from Brays blog that he too shared Cantrills lack of enthusiasm for working for Sun, but was drawn in by the culture. It seems that Sun was an outlier in the huge-company-enterprise space in having a great hacker culture, and that brought in some people that would otherwise be more at ease in smaller companies.

Oracle can't/won't/isn't interested in keeping that culture around, and it is questionable what that culture actually brought about in value at Sun. After all, they weren't very good at making money. Oracle, on the other hand, is very good at making money, and perfectly happy letting the cool people work other places, acquiring their work if they decide they need it.

It might not even be that Oracle isn't interested in that culture so much as Oracle simply isn't Sun and that was Sun's culture. Sun believed Sun was the best and everything else wasn't. Sun believed in Sun, quite a bit, enough so that Sun wouldn't wake up and smell the coffee when the CEO was blogging about a filesystem in the 21st century.. (You know, I couldn't tell you what filesystem gmail is on or amazon uses or salesforce.com uses, since we dumped FAT nearly 20 years ago, I haven't really had much reason to care other than because I'm geek. It's not impacted performance in anyway that would hurt business, just as one example.)

If there was some sort of news from Oracle, how they weren't interested in a class of products or they weren't interested in technology or anything like that, or maybe all all the senior sun guys were being demoted a notch or seomthing then folks would be screaming bloody murder and we'd all know about it. Oracle wants to continue to make money and Sun was more interested in making their team think they were the best, regardless of what the market showed. That's the difference.

Some of these guys simply might be better off elsewhere, their time might be past, but if they really want to show their greatness then they have an opportunity to do so now with real marketing might. I went through this with a company a few years back, we boot strapped it, got it going, came up with some good ideas and some products but didn't have the money to make them robust and didn't know how to sell; we got bought and maybe a quarter of the team saw that as the ending (nobody made any money at that point) and they put that kind of effort in, the others saw it as time to get real and do the hard work, take things to the next level.. It's just a natural sort of thing when companies get bought.

On the up side, I give the guys that are leaving some credit. They could just hang out and be dead weight, draw probably giant pay checks and not do a whole lot.

I find your post extremely pessimistic about the people involved. People of Cantrill, Gosling's caliber do great stuff no matter where they go.

Based on Gosling's blog, "enrage" might be a correct characterization ("an awful lot [ of my time ] has been taken up talking to people about life at Oracle. They need a place to vent, and I try to be a good listener.", that's from about a month ago and the whole short item is worth reading: http://nighthacks.com/roller/jag/entry/escaped_from_reality)

One very simple thing we do know about is that many/most?/nearly all??? of Sun's software people have seen their salaries significantly reduced. Even if this didn't hit directly people at Gosling's level it would substantially change a whole lot of things and certainly sends a number of negative messages.

It's not completely grim, the Maxine (JVM in Java) research group seems to be very productive and not losing principles, but they're working on an area that Oracle considers to be critical (JVMs in general and JVMs on top of VMs in particular).

Congrats to Bryan. I hope he moves on to something interesting.

He's one of the most technically intimidating people I've ever met.

Reply on news.ycombinator.com