michaelfairley

As one of the first hires at 1000Memories, I can definitely say that we have plenty of fun, and that the office is an upbeat place.

Part of the reason we can do this is that we don't consider ourselves as being centered around death, but rather as being centered around life. We view our site as a place where friends and family can come to celebrate, rather than to mourn. Most of the design on our site reflects this intention, and most of the memory pages on our site fit this theme.

Occasionally, there are tragic stories shared on our site, and we're certainly saddened by them, but most of our memory pages have a celebratory and upbeat tone.

If anyone's interested in coming by and seeing our office (in the Mission in SF), feel free to shoot me and email and we'd be glad to show you around.

I think the point is that structure and style should be decoupled, i.e. the way something looks (color, fonts, borders, etc.) should be defined separately from the way something is shaped.

I got rid of it. Most stories about YC companies on here tend to include that information though.

Mincemeat.py (http://mincemeatpy.com) is fairly simple, only uses standard libraries, and definitely welcomes contributions.

(I am the maintainer of mincemeat.py)

I think the problem is that the feedback loop between UTCS and the Austin startup community hasn't hit a critical level. I can't name a single product made by an Austin startup that I've used, while I can think of dozens within a few miles of Stanford. Additionally, plenty of SV startups are within walking distance of Stanford, whereas most Austin startups are at least 10 minutes (by car) from the UT campus and are located in office parks, rather than the open pedestrian environments of University Ave and Castro.

Depending on how much emphasis you put on the word, it can range from one to two syllables. If you use it in a sentence ("The owl ate the mouse") spoken at a conversational speed, the sounds that come from your mouth are likely to be indistinguishable from saying the name Al. If you say the word slowly, you'll have an [æ] (the vowel in cat) or an [a] (the first vowel in father), followed by [wl̩] (sounds like a sheep's hair).

Most Americans I know use the latter, in which the r and l do not have a vowel between the two.

It's actually not a triphthong, because the r is not a vowel. Rather, it's a syllabic consonant: http://en.wikipedia.org/wiki/Syllabic_consonant

The IPA represenation is not [ə], but rather [ɹ̩]. (There's a tick under that ɹ).

And, to be fair, your statements are accurate for British dialects of English that don't have syllable final rhotics, but the original article was about American English.

> Customer support should be called “delayed sales”

Money line from the article.

Maybe YC W11 has some finance-oriented companies in it and YC's taking care of business ahead of time. No doubt this would've been helpful to InDinero over the summer.

For the record, "Tim" and "Suhail" are Mixpanel's founders, not just customer support people.

An advertisement that a lot of HNers want to see. Last time it was $99 many people on this site were pleased to know: http://news.ycombinator.com/item?id=1288125

23andMe = http://en.wikipedia.org/wiki/Single-nucleotide_polymorphism + population genetics

But, it doesn't really give actionable results, especially if you're young and healthy.

Note that there's a new $5/month subscription fee, with a 1 year minimum commitment.

23andMe is definitely looking for top-notch engineers. They're still running small (5 engineers 3 years in), so if you want to have a big impact on a potentially world-changing product, get in touch with them.

This is too bad. I really enjoyed the YCommonApp, and hope something fills its place.

Google's implementation of MapReduce is so tightly bound to their internal infrastructure (GFS, BigTable, etc.) that opensourcing wouldn't do anybody much good.

If push comes to shove in a 2 person company, the company is already dead, regardless of the equity split.

He smashed in sales because of all the other ways he's set himself up for success. Patrick has a multi-pronged approach to his business, where even if some of his efforts fail he can make a pretty buck.

Almost exactly half (150) are engineers, leaving the other half for the businessy roles.

http://remembersaurus.com - My recently launched webapp, and also my first attempt at making something look decent on the web.

Even the SEO advantages he gets from his popularity will make a huge difference.

In tech? On HN? In YC? Definitely.

It's drama, but it's also of life-changing importance to some people on HN. It's also Sunday, so it's a slow news day anyways.

The email field isn't public anyways.

The fact that they are claiming that a fictitious someone else is trying to buy the names in order to convince you to act quickly is what makes it a scam.

Ah yes, I missed the part about the size of the dataset. I'm running with a 3GB corpus of text.

Without details on the task he's running and the setup of his cluster, it's hard to know how meaningful these results are. Using a variation of the example, I've seen roughly 80% improvement with each doubling of the cluster size (testing up to 100 workers).

I spun this our of some work I was doing for an NLP related thesis. Large scale text processing where there's already a shared file system is a perfect fit for mincemeat.py.

I basically wanted something that was much, much easier to develop for, setup, and run than Hadoop. I've heard many academic researchers complain that although they had a algorithm that would fit neatly into MapReduce, they didn't want to bother setting up Hadoop and importing all of their data for a process that would only get run a few times (and that was already coded in Python).

Mincemeat.py actually does do fault tolerance and preemptive duplication. I'm trying to get it to handle all of the MapReduce logic (and reliability, security, etc.) without the extra burdens of setting up a distributed file system (the machines I've been running it on have a hefty NFS share, which is good enough for most purposes), preallocation of the machines that will be part of the cluster (my university uses Condor, which spawns processes randomly across a large cluster), etc.