HNHacker News
TopNewBestAskShowJobs

dkuebric

417 karma · joined June 18, 2011

submissionscomments
dkuebric··on Why Microsoft Word must die
Monopoly or not, it's the only wsywyg text editor I've ever found truly usable. I now prefer Google docs for convenience but it's not quite there yet for composition/edit experience. Hoping for more good alternatives, if anyone has suggestions.
dkuebric··on Github For Everything
This isn't what I thought it would be. I think "Github for everything" is something everyone should be looking for in their lives. Version control is useful for so many things.

I think the understand this already though. With the release of stuff like 3d diffs now it's the "GitHub of CAD", I could imagine a useful future for the "GitHub of legal documents", "GitHub of _____"....

dkuebric··on Twitter's Tool For Debugging At Scale
Latency isn't the only thing--if that's all you wanted, you wouldn't need the tracing, just log the latency at each step. The real amazing thing is the structure of requests. Having a record of distibuted N+1 problems is pretty amazing, for instance.

As for where you hook into the stack, it's definitely a tradeoff. The lower level you go, the wider your coverage, but it also becomes more generic and potentially less actionable data.

I've worked on building a similar system at AppNeta (it's basically commercial Google Dapper). Here's some slides from a lightning talk I gave about distributed tracing at Surge last year: http://www.slideshare.net/dkuebrich/distributed-tracing-in-5...

dkuebric··on Deploying Python Without Downtime
One thing worth noting about these rolling restarts that I didn't see in your post: if the new code isn't completely backwards-compatible, you can end up with bad states from having a mix of workers running. This negates a lot of the value of the rolling restart because it creates other failure modes.

For example, if you introduce a new ajax endpoint in the release, and a client hits a new worker generating a HTML page that calls it, but 90% of your gunicorn workers are still serving the old version of the app, 90% chance that you're going to 404 that request.

dkuebric··on GitHub Pulse
Feature request: pulse for milestones--show me a burndown chart, comment activity on hot issues!
dkuebric··on How To Survive a Ground-Up Rewrite
Data migration becomes a lot easier the closer to "raw" data you store. The farther away you are, the tricker it will be to write those migration scrips. And, if what you're storing is rolled-up enough, you may not even have the information you want/need any more. Note that the source data doesn't have to be what you're serving queries from; it could be an archive with batch scripts to re-process that's only dusted off in case of disaster/migration.
dkuebric··on Heroku 2X Dynos in Public Beta
I've been told they're encouraging add-ons to move away from a per-dyno price. (Looked like New Relic is the only one right now).

The guys I talked to cited both the future (now present) "dyno is not a unit" problem, as well as reporting user dissatisfaction with non-flat-rate billing.

dkuebric··on Heroku 2X Dynos in Public Beta
Yeah, I had the same thought when the discussion came around last time. I actually took the source to the simulations (thanks for sharing, rapgenius!) and did some experiments on how random routing performs if it's routing to backends that can handle various #s of concurrent requests.

Here's the writeup: https://www.appneta.com/2013/02/21/the-taming-of-the-queue-m...

(Spoilers: turns out it gets a lot better very quickly.)

dkuebric··on Show HN: Trip.js, a jQuery plugin for product landing page navigation
pageguide.js - http://tracelytics.github.com/pageguide/
dkuebric··on My Apology
Hard to dislike a guy that writes so well, and I agree with the general premise of recognizing the human capacity, and necessity, for error.

But I do think it's a bit disingenuous to compare the "mistakes" of scientists to fabrication and deceit. Would Niels Bohr have been referring to the manufacture of evidence when he spoke of "all the mistakes that can be made in a very narrow field?"

dkuebric··on Charts of alcohol consumption, obesity, and physical activity in Europe
I interpreted this as the country self-reporting, not the individuals.
dkuebric··on Ask HN: Who Is Hiring? (October 2012)
Boston, MA - AppNeta (aka Tracelytics)

Full-Stack Software Engineer, Front-End Engineer

We're working on an awesome product that helps devops run their web apps more efficiently. We do distributed tracing, ala Google Dapper / Twitter Zipkin, then turn it into actionable SAAS-delivered performance insights.

Our team is super-smart, very autonomous, and well-funded. We get to work with cool technology and solve high-data-volume problems in real time.

Technologies: Python, D3, backbone.js, C/C++, Cassandra, MySQL, EC2, git

More info: http://tracelytics.com/jobs

(we've also got hackers located in Providence, RI and Vancouver, BC -- inquire within for details!)

dkuebric··on Debugging node.js memory leaks
Nice! Reminds me of gdb-heap, which takes what sounds like a similar approach to identifying python objects in the heap: https://fedorahosted.org/gdb-heap/wiki
dkuebric··on We Are The 99% Of Startups
It's also very much for the 99%--not the best example for this post.
dkuebric··on Tracelytics: Google Dapper for the Rest of Us
Both Dapper and Tracelytics borrow ideas from the X-Trace project, which has a bit of source available, though not much.

http://x-trace.net/wiki/doku.php

dkuebric··on Ixoth
Neat. "ixoth" is a pretty unique name, pronunciation-wise. Anybody know if there have been studies done exploring the sound-space of voice recognition? E.g. if you could map typical sounds of (english) speech to some sort of dimensional space, which areas are dense and which are sparse?
← PreviousPage 3 of 3