Hacker Newsnew | past | comments | ask | show | jobs | submit | more phpnode's commentslogin

If this trick is still really working for you, please, do yourself a favour and shut up about it. Drawing attention to it is not going to end well for you, either twitter shuts you down or people just copy your idea and it loses effectiveness. The correct time to write about these things is when they stop working.


All he's doing is smartly parsing the results of Twitter searches. This isn't really groundbreaking, it just requires some legwork.


If all he's doing is parsing search results, then nobody would be seeing his advertisements. You conveniently left out the part where he uses a bunch of different Twitter accounts designed to appear to be different people.


I agree, but the general point is that specific marketing tricks that work well should be hoarded, not shared, because sharing them destroys their value.


> "because sharing them destroys their value."

... Why?

Because it'll lower your competitive advantage? If your main competitive advantage is knowing about some cheap Twitter trick--sorry, I mean "Growth Hack"--then you need to rethink your game plan.


because sharing them makes them stop working. No one said anything about this being his main competitive advantage.


A very inspiring guy, RIP


isn't there a good PEG available for python? Compare the source code in the post to the equivalent PEG.js grammar - http://pegjs.majda.cz/online


What is the relationship between PEG parsers and Parser Combinators?


(Original blogpost author here.)

They're pretty much orthogonal.

PEG is a particular parsing technique. Combinators are a technique for implementing parsers.

You can write a parser combinator library which uses PEG or packrat algorithms underneath.

In this blog post, the library that I'm using is basically building LL parse generators, not PEG. But you could easily implement a PEG-based combinator system.

I didn't do that, because I didn't want to get bogged down with explaining too much about parsing algorithms. The LL/recursive descent approach is pretty easy to understand when you see the grammar code. For PEG, I would have needed to spend more time explaining the algorithms, and the post was already too long!


That's not quite equivalent; the JS doesn't do division or subtraction.

The Python syntax is a bit klunky. The Haskell equivalent is on par with the PEG. I don't know how they compare as the grammars become non-trivial.


wow, that's a distracting background. Also the video doesn't explain how the key exchange works, and isn't that the important/hard bit?


It's the Matrix


This is definitely true, look at the rise of John Deere's GreenStar - GPS controlled tractors and implements which dramatically improve yields. Farmers are like hackers in many ways.


I guy I knew years ago used to work for a software company that sold software to farmers, and he said he was always surprised by how incredibly high tech farming and farmers where.


> "This is like giving every MIT student a laptop thirty years ago"

Pardon?


> "That didn’t happen, but I had to make some kind of comparison."

?


Right, it didn't happen. And if it had it would have been:

1. Totally different, as bitcoins are not portable general purpose computers.

2. Two orders of magnitude more expensive than $100 per student.

So I'm wondering why on earth this person would draw such a comparison?


I'm guessing that he's probably referring to the concept of giving them access to a very new technology at a very early stage.


more business 101 - after showing initially good results and establishing confidence, drop your prices to unsustainable levels (i.e. $0) to drive growth and inflate your valuation, then make a swift exit. This strategy netted me ~$200k + $90m+ valuation.


that was great, but funny that literally every adsense ad on those articles was for herbalife products.


agree, this sounds like a bug, console.log() returns undefined so that's what the next promise in the chain should receive.


NoSQL has always been a stupid term, especially as there are "NoSQL" datastores that support SQL, e.g. OrientDB.


The term was perfect. The spectrum of NoSQL implementations out there defies an easy name to cover them all. They aren't all key-value stores, they aren't all json stores, they aren't all big tables, etc.

Indicating that they were NOT within the current paradigm of SQL databases was the only way to go.


The term is borderline moronic!

SQL is a QUERY language. NoRDBMS would be slightly less moronic...

Thbbft!


Why is it a good idea to lump such technologies together, when many of them are fundamentally different? Do we call road vehicles that aren't cars NoCars?


Ironically, we kind of do, at least in the US. We call the trucks. SUVs are a bit of an anomaly, but I've heard plenty of people colloquially call our explorer a truck.


You call motorbikes, buses, and bicycles trucks?


It's stupid if you don't realize it stands for "Not only SQL".


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: