jamwt

20 ms·

211.
by jamwt 16y ago
Can you say more about what you mean when you said they're "share-everything?" I'm not entirely sure what you mean by that, since, for example, we do diesel in what we consider to be a "shared nothing" approach at Bump. Redis is the transp
212.
by jamwt 16y ago
Fair enough; 100 was sort of arbitrary, but there is some metric when you're sort of at google scale and cutting down on power costs or achieving absolute minimum latencies between datacenters has a meaningful impact on the bottom line. Dr
213.
by jamwt 16y ago
I know it's not exactly in vogue these days to tout the merits of bare hardware, but.. after all the VPS hubbub over the last couple of years, the best progression for your website still seems to be: 1. No traction? Just put it anywhere, 'c
214.
by jamwt 16y ago
Good idea re: Arrows, and we can go all out as long as we're using them. :-) rle = group >>> map (length &&& head)
215.
by jamwt 16y ago
Haskell version that is fully lazy, (though not point-free or completely idiomatic). import Data.List (group) rle = map (\x-> (length x, head x)) . group main = print $ rle "aaaabccaadeeee"
216.
by jamwt 16y ago
Could you say more about why Bump's implementation of network-based queued logging is "hacked-up" while facebook's (by implication) isn't? To answer your question, simply put, no one here had heard about Scribe.
217.
by jamwt 16y ago
I've taken quite a deep look at it: https://github.com/jamwt/diesel/blame/e360313d3950a952110b1d...
218.
by jamwt 16y ago
Hi Antirez, thanks again for Redis. Despite our few problems with it, it rocks. A few comments: > With fsync everysec, there is the problem that form time to time we need to fsync. Guess what? Even if we fsync in a different thread, w
219.
by jamwt 16y ago
I'd chalk it up to the general benefits of eschewing needless complexity. I can say, empirically, none of the many, many challenges we've had building and scaling Bump, have been related to Redis's capabilities as a messaging bus. So "good
220.
by jamwt 16y ago
It is in 2.2. Check out: maxmemory-policy allkeys-lru How do I know? We ( http://bu.mp ) push 100s of GB though redis every day in LRU mode.
221.
by jamwt 16y ago
And, for very detailed play-by-play data for decades of games, check out retrosheet: http://www.retrosheet.org/game.htm
222.
by jamwt 16y ago
You're just jealous I worked it in so naturally.
223.
by jamwt 16y ago
Yeah, the whole pipeline is actually quite more faceted than can be deduced from this summary. This stage actually just persists the events into a consolidated transaction log. Then, there are secondary processes that scan these transacti
224.
by jamwt 16y ago
While I agree with you that Haskell (or, really, any GC'd language) is unsuitable for real-time systems, I disagree that my statement about its excellent suitability for systems programming in general is misleading. There are many, many do
225.
by jamwt 16y ago
Sure, I can say more--put bluntly, and despite anything you might hear to the contrary, you basically do need to grok monads and monad transformers to tackle any nontrivial project in Haskell. Even if only to understand the code and APIs
226.
by jamwt 16y ago
Nope, this is a process that BLPOP's logs from some Redis queue, does some processing on them, then writes them to disk.
227.
by jamwt 16y ago
Yeah, Haskell is roughly as fast as Java. I imagine, with the tuning I allude to in the blog post, we could restore about a 10x improvement, even on a single core. After all, 10x was about what we had with raw Redis ops/s before the seria
228.
by jamwt 16y ago
I would say the two real barriers to writing effective Haskell projects are a.) "getting" monads, and b.) understanding the implications of laziness, especially with regard to space leaks and unconsumed thunks. Everything else isn't that b
229.
by jamwt 16y ago
Hi, article author here. It's important to note that this particular job is largely bound on a.) I/O and b.) format serialization tasks. Both Python's BSON and JSON libraries are mature and have their critical sections written in C, so a s
230.
by jamwt 16y ago
I wouldn't try to shoehorn SSL/TLS in as a requirement of whatever language/framework you do network stuff in--it's unlikely to be a great implementation, anyway. Just put nginx or stunnel in front of your service and go about your busines
231.
by jamwt 16y ago
Others have mentioned coroutines; just to avoid repeating a good conversation that already happened on all this, here's the link: http://news.ycombinator.com/item?id=1549023
232.
by jamwt 16y ago
SoftLayer is a superb dedicated hosting provider--maybe the best. They have a competent staff, several datacenters all over the US, a very good, low-latency network (public and private), high-quality management tools, super high uptime and
233.
by jamwt 16y ago
Well, the guard knows Y, and the enemy knows Y... but the guard's authentication procedure involves taking an X ( not Y) supplied by the friendly spy, executing the calculation, and asserting the result is equal to Y. Since the enemy does
234.
by jamwt 16y ago
Just go into the Extensions window and then Vimium preferences; you can add this vi page to the list of "excluded URLS".
235.
by jamwt 16y ago
Yep, some Scala at Bump, too.
236.
by jamwt 16y ago
Do a lazy migration--aka, every document gets tagged with current version. The restoration/model routines have a transparent upgrade chain (1->2, 2->3) that move any loaded document up the chain until they reach the current version
237.
by jamwt 16y ago
I do this same thing (but with VMWare Fusion); the only significant drawback is the substantially slower disk I/O within the VM. I spend all my time in an Ubuntu VM running XMonad, mainly so that I can use nice os x apps (mac end-user apps
238.
by jamwt 16y ago
Yeah, I remember that from freakanomics. I'd imagine there's an age where the monotonically-decreasing F and monotonically-increasing P line cross on the plot--maybe middle school years...
239.
by jamwt 16y ago
Yes, you're right. Additive is what I meant, fixed.
240.
by jamwt 16y ago
I'm not sure it suffers that flaw--I may not have articulated it well, but my point wasn't that teaching cannot be improved, or these methods do not improve it. Or that it's not worthwhile to improve it... it certainly is! (implied: and y

More ›