[FoRK] Who needs efficient code?

Eugen Leitl eugen at leitl.org
Fri Jul 30 08:13:24 PDT 2010

On Fri, Jul 30, 2010 at 10:40:40AM -0400, rst at ai.mit.edu wrote:

> Ummmm... it's not quite clear which guy you're talking about.

I did not realize above was a citation of a citation.
> The quoted paragraph was written by ex-Twitter API lead Alex Payne,
> who doesn't blog for The Atlantic; he was quoted by some other guy who
> does.  Payne's obviously talking about Rails, which his (former)

I wonder why they would consider Rails in the first place.
I'm just an egg, but I know I wouldn't be using it if I
expected the result to go places eventually.

> company had to abandon for its most high-volume requests, though they
> still use it elsewhere last I checked.
> The paragraph was an excerpt from a longer post on Alex's own blog
> about why Node.js isn't a scaling panacea either:
>   http://al3x.net/2010/07/27/node.html
> which, by the way, points to a lot of other interesting writing
> about more graceful ways of scaling large systems (in the paragraph
> starting "...taking a hybrid approach to concurrency").

> It's, of course, possible that Alex Payne doesn't know what he's
> talking about, but it's not obvious to me why you'd think so.  Was
> there something else in the Atlantic post that you meant to call
> bullshit on instead?

I definitely take offense at 

In a system of no significant scale, basically anything works.

The power of today’s hardware is such that, for example, you can 
build a web application that supports thousands of users using 
one of the slowest available programming languages, brutally 
inefficient datastore access and storage patterns, zero caching, 
no sensible distribution of work, no attention to locality, 
etc. etc. Basically, you can apply every available anti-pattern 
and still come out the other end with a workable system, simply 
because the hardware can move faster than your bad decision-making.

Stated so generically, this is not true.

He doesn't say how heavy-weight each user session is. Incidentally,
our stuff has no issues saturating your cluster of 12-core boxes 
for hours, by a single query. Making half a rack work for ~10
simultaneous users becomes a challenge, and suboptimal architecture
is *not* helping.

Moreover, perhaps using himself and his peers as a point of reference
he apparently does not realize how brain-dead so-called
developers can be. He. Has. No. Idea. I've seen things, you people
wouldn't believe. SANs on fire, etc.

Eugen* Leitl <a href="http://leitl.org">leitl</a> http://leitl.org
ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org
8B29F6BE: 099D 78BA 2FD3 B014 B08A  7779 75B0 2443 8B29 F6BE

More information about the FoRK mailing list