Josh the Outspoken

Posted by josh at July 7th, 2008

Today Google open-sourced a component we’ve used in-house for a long time called Protocol Buffers. It’s a binary format that we use for almost all of our on-the-wire messages and lots of disk-based long-term storage as well. For many (maybe most, though not all) uses, Protocol Buffers kick XML’s ass. In a big way. Seriously, if you’re XML this is the part where you sulk home with your tail between your legs.

Why not just use XML?

Protocol buffers have many advantages over XML for serializing structured data. Protocol buffers:

• are simpler
• are 3 to 10 times smaller
• are 20 to 100 times faster
• are less ambiguous
• generate data access classes that are easier to use programmatically

–Protobuf Developer Guide

If you or your company needs a very compact, very fast, extensible format for structured data, you should give Protobufs a good look!

(P.S. Of course I don’t speak for Google. The attitude is all me talking and my personal disdain for XML. Google’s official attitude is, of course, much more diplomatic).

Posted by josh at May 23rd, 2008

Steve Vinoski has come out very vocally against RPC in the last few days: see this blog entry and this mailing list post. The blog entry (which I read first) made him sound like someone who just hasn’t been around large systems much, but then I was surprised to see that he’s a senior fellow or architect or something along those lines at a company that does distributed systems.

His blog entry basically makes fun of Cisco for inventing/releasing another RPC system. It’s not clear exactly what he thinks they should have done instead. What is strange about this criticism is that tons of technology companies have developed their own RPC system — Facebook and Cisco publicly, and other technology companies I am familiar with in a not-so-public way. Guess what: large commercial distributed systems are built largely on RPC. Is he arguing that all of the engineers at these companies simultaneously got the bad idea of investing in something they don’t need? If RPC is such a bad idea, then why is everybody doing it?

“Everybody’s doing it” obviously isn’t a justification alone, but it definitely puts the onus on the person making the critique to show why it’s a bad idea. I got a better idea where he was coming from when I read the mailing list post. Here’s the heart of his argument:

the fundamental problem is that RPC tries to make a distributed invocation look like a local one.This can’t work because the failure modes in distributed systems are quite different from those in local systems, so you find yourself having to introduce more and more infrastructure that tries to hide all the hard details and problems that lurk beneath. That’s how we got Apollo NCS and Sun RPC and DCE and CORBA and DSOM and DCOM and EJB and SOAP and JAX-RPC, to name a few off the top of my head, each better than what came before in some ways but worse in other ways, especially footprint and complexity. But it’s all for naught because no amount of infrastructure can ever hide those problems of distribution. Network partitions are real, timeouts are real, remote host and service crashes are real, the need for piecemeal system upgrade and handling version differences between systems is real, etc. The distributed systems programmer *must* deal with these and other issues because they affect different applications very differently; no amount of hiding or abstraction can make these problems disappear.

Finally something we can agree on! Yes, on a network shit happens, and no sane RPC system will try to hide this from you.

But then again, I don’t know of any RPC system that tries to hide this from you except possibly CORBA. Maybe there’s a horrible history here I don’t know about, but no RPC system I have ever encountered tries to hide from you the fact that on a network, shit happens.

So what are his other criticisms?

RPC systems in C++, Java, etc. also tend to introduce higher degrees of coupling than one would like in a distributed system. Typically you have some sort of IDL that’s used to generate stubs/proxies/skeletons — code that turns the local calls into remote ones, which nobody wants to write or maintain by hand. The IDL is often simple, but the generated code is usually not. That code is normally compiled into each app in the system. Change the IDL and you have to regenerate the code, recompile it, and then retest and redeploy your apps, and you typically have to do that atomically, either all apps or none, because versioning is not accounted for.

Yay, we can agree again. RPC systems that make you do an “all at once” upgrade are a bad idea. But again, no RPC system I have encountered makes you do this. Does this mean that the RPC system guarantees for you that the old and new protocols are compatible? Of course not — you don’t want your framework to be some big “I know what’s best for you” mommy that does really expensive things to solve this problem, like loading both versions of your code at the same time. But any RPC framework worth its salt makes it possible to have different interface versions interoperate. Adding a new parameter? No problem, old servers simply won’t see it. Completely changing the semantics of your call? No problem — just give the new call a new name.

Steve’s criticism amounts to “sucky RPC systems suck.” Yes Steve, yes they do. But a lot of the technology world is running on non-sucky RPC systems, and from time to time you get a glimpse of that when a company like Facebook or Cisco releases their internal RPC system to the outside world. Did Steve check to see if Cisco’s new RPC system is subject to any of his critiques? I haven’t, but I would suspect it isn’t.

Posted by josh at April 9th, 2008

Observation #1: stop-the-world garbage collection is a thorn in the side of latency-sensitive applications.

Observation #2: we will very soon have more cores than we know what to do with.

Prediction: fully concurrent garbage collection is the future of automatic memory management. I’m talking garbage collectors that run in other threads and clean up after me without ever stopping me in the middle of what I’m doing.

It will almost certainly be more expensive in terms of total CPU time, and probably can’t be as aggressive in terms of what it can reclaim at any point in time, but for most applications the latency guarantees will far outweigh.

Discuss.

Posted by josh at March 20th, 2008

Some Python fan please tell me that I’m missing something.

Is this really the boilerplate necessary for creating even the simplest thread in Python?

This is making me miss Ruby, for which the equivalent is:

P.S. Gazelle 0.2 is making a lot of progress, but unfortunately won’t hit the 1 month mark I hoped for. Surprise surprise. But when it does come, it’s going to be awesome.

Posted by josh at February 12th, 2008

I’m finally responding to Buffalo’s perspective on my last post about “Brilliant programmers.” I didn’t say anything at first because I couldn’t think of anything insightful to say. I still can’t, so I’m going to have to just make shit up.

Buffalo approaches the question from an educator’s perspective:

The real million dollar question in my mind, Josh, is what are these super-programmers doing that others are not? Even if there’s just something irreproducible in their genes - shouldn’t this be the sort of thing we could detect somehow? Or if there are strategies involved, could we use them to make the 90 percentile 5x better than everybody else?

[…]

Which of course brings me around to my research - say you had a group of the 50 cs students who enter an average program around the country. And say your goal was to get absolutely as many of them as possible into that 5% and screw everything else. What would you say to them? What would you do with them?

I guess it’s depressing to put this way, but my true belief is that programmers are born, not made. So I think the absolute number one thing a CS education can do for its students is help them understand whether they were born programmers or not.

This isn’t a binary thing, of course, and I don’t even think it’s single-dimensional. Everyone has areas of strength and weakness. And it’s a big world out there — succeeding at Amazon or Microsoft or Google is different than succeeding in the consulting world or at a startup or the technology department of an unrelated industry. I know at least one person who is a decent programmer but is doing quite well by combining good business understanding and fantastic people skills with his competent but not outstanding technical background. There are a lot of paths for people to take.

Then there is the question of what productivity is. In a corporate setting, productivity is making progress toward keeping your customer happy, whoever your customer may be and however bizarre their wants. Well I doubt that writing a self-compiling C compiler or figuring out how to calculate pi was at the top of any customer’s wishlist. Fabrice Bellard, for all his badassery, could end up being fairly unremarkable when you put him in a cube and tell him to write web applications. Or he might find it excruciating and spend the whole time inventing a framework that lets him write web apps the way he thinks. After all, that’s what Paul Graham did.

So disclaim, disclaim, disclaim. My point so far is just to highlight what is probably already obvious: that the landscape of talent is complex and multidimensional.

That said, I think it’s really key that students understand their strengths and weaknesses as early as possible. It will help them decide whether CS is right for them, and if it is what direction they should go within CS. And the best way to achieve this understanding is to expose students to lots of different things. As a bonus, this is also exactly what the overachievers need and want too; more problems to feast their minds on. So if I could sum it up in a word, I think the number one thing that university can do is expose young programmers to lots of topics and give them feedback about how they do in each one.

As to how we can help budding programmers be in the super-productive elite? I’m somewhat hesitant to answer that question, because I feel the answer reveals what the answerer thinks makes him/her a member of this elite class. But my biased answer is that the #1 most valuable trait a programmer can have is resourcefulness. It’s knowing what’s out there — what tools, strategies, file formats, blogs, or best practices — and knowing how they apply to the problem in front of you. If some programmers truly are outperforming others by 20x, it’s not because they’re typing 20x faster — it’s because they have such a solid understanding of the problem’s context that they have laser-like focus on the shortest possible path between them and their goal.

On the flip side, I think the worst thing that can happen for programmers is for them to get caught in their own little worlds where they only know one way of dealing with problems. Java-only programmers are the easiest to pick on example of this vice, but not the only one. You can probably get by only using/liking Java if you work at an all-Java shop where you are a cog in a machine, but you’ll never have any perspective of the big picture. I think we all fall prey to getting comfortable with our favorite languages/tools and thinking of everything through those lenses, but the more we can avoid that, the better we do IMO.

So that’s my answer: show students what’s out there, and encourage them to be resourceful. If there is any trace of programmer in them, it will come to life and demand that you feed it more. Anyone who doesn’t self-motivate at that point is not destined to be a programmer. They might not have the talent, or they might have the talent but not the motivation; either way, they probably should find something else instead.