Comments for shtylman.com

Comment on course/off visualization – gatech Spring 2012 by shtylman

shtylman — Wed, 02 Nov 2011 15:10:15 +0000

I think you would. Guess we just have to wait and find out I will make another one of these for fall once that time comes.

Comment on course/off visualization – gatech Spring 2012 by Dan

Dan — Wed, 02 Nov 2011 15:08:17 +0000

That’s interesting. I wonder if you’d see something similar for Fall, except with ENGL 1101 and MATH 1501 instead, for example.

Comment on The tail of MongoDB by shtylman

shtylman — Wed, 15 Jun 2011 12:34:57 +0000

This is not true when the collection has no documents in it. Also, ‘more’ only blocks for live cursors, for dead cursors you have to requery and you don’t want to do that in a tight loop (no sleep). More will also only block with the AwaitData query option specified.

Comment on The tail of MongoDB by iammutex

iammutex — Wed, 15 Jun 2011 06:19:31 +0000

hello, I have a question.

while the cursor is dead , is the c->more() method will still block for a while? if it will,I think there should be no need to call sleep as the cursor dead.

so can u gave me some info about that.

thx

Comment on The tail of MongoDB by Remon

Remon — Tue, 14 Jun 2011 16:31:19 +0000

Interesting stuff. We recently wrapped up a Kestrel driven messaging layer but since we use mongodb for the vast majority of our persistence needs this is definitely something we can look at. That said the tests should probably use more real world use cases and setups and see how it compares in terms of throughput and durability compared to some MQs (ZeroMQ, Kestrel, ActiveMQ, Kafka, etc.).

Comment on programming is hard by Mathieu D'Amours

Mathieu D'Amours — Tue, 14 Jun 2011 16:11:59 +0000

You are so right. My comment comes late, but I just wanted to let you know, I’m totally in agreement.

Comment on The tail of MongoDB by dwight

dwight — Tue, 14 Jun 2011 16:10:47 +0000

a couple notes:
- it almost goes without saying, but for low latency the capped collection should be smaller than RAM. it can be far larger than ram if you don’t mind disk IO, it’s just a different latency scenario.
- while multiple consumers were not tested above, that should work very well too.

Comment on The tail of MongoDB by dwight

dwight — Tue, 14 Jun 2011 16:00:58 +0000

cool post. yes that doc page should be better.

Comment on re-tipping the cab by shtylman

shtylman — Sun, 09 Jan 2011 18:20:44 +0000

That still doesn’t quite answer why 15% is considered “good”. I realize the current system trends towards 15% being good, but why couldn’t the range have been 0%, 5% and 10% for (poor, good, beyond). Or even 5%, 10%, 15%? Was it society that settled on 15% being good? or some other guiding hand?

Comment on re-tipping the cab by 411 New York

411 New York — Sun, 09 Jan 2011 16:06:34 +0000

Why 15% was chosen versus 10% or some other arbitrary number?

It is a starting point as the range is 10% to 20% with 10 for poor, 15 for good and 20 for beyond.

A driver who help with the bags, drives at good speeds and help with a car seat install would more likely to get between 15% and 20%.

A driver who speeds, yells out the window, breaks hard and doesn’t even help with the luggage would most like get stiffed with a smaller tip.