jgord

31 ms·

181.
by jgord 8y ago
yes, and yet "perfect is the enemy of good" ... I think the article encompasses your objections, to paraphrase : "algorithms can never be perfectly 'fair', because logic itself can never be perfectly fair" Can
182.
by jgord 8y ago
Have to pipe up here and say an enthusiastic / visceral / emotional 'thankyou' to BurntSushi for all your great contributions - xsv, rust-csv, ripgrep etc - not to mention your superb blog articles. Im using xsv to wrang
183.
by jgord 8y ago
two kinds of indexes, for different purposes ... a) basically a preprocessed static inverted index on keywords [ using GIN / tsv_ etc ] b) geo location index on GIS geometry field - relying on postGIS to be performant over geo queries
184.
by jgord 8y ago
anecdota - We used fast csv tools and custom node.js scripts to wrangle import of ~500GB of geo polygon data into a large single vol postgresql+postGIS host. We generate svg maps in psuedo-realtime from this data-set : 2MB maps render sub-s
185.
by jgord 8y ago
Tesla's core innovation is that they don't give a fuck about nonsense opinions, they care about facts - science and engineering. Plenty of other viable technologies exist, we just need some rich science-aware people to help us rol
186.
by jgord 8y ago
You comment seems to suggest there is some rational decision making emerging from an open back-n-forth of facts and ideas. I wish that were the case, but the simplest explanation of political reality both in AUS and USA is that our leaders
187.
by jgord 8y ago
Performance aside, also consider Ramda.js Although Ramda has forEach, I augment it with a version of each(func, data) where data can be an array or map, and func(val, key) where key is the key of the map item, or the index of the item in th
188.
by jgord 9y ago
I wonder if there is a model whereby you can have many investors donate small amounts in crypto and earn deferred interest on their investment. Like how people purchase a mining contract for BitcoinCash .. except you get a delayed payout, a
189.
by jgord 9y ago
read Anathem, if you haven't already.
190.
by jgord 9y ago
.. and some paintings : gordart.wordpress.com
191.
by jgord 9y ago
I built a system that takes map data for whole UK topography and produces maps in html/svg, pdf/jpg .. and autocad DXF. Built with postgres+postGIS and node.js. Around 400Gb dataset, maps generate in around a second. A lot of tim
192.
by jgord 9y ago
High BTC Fees are hitting lots of legit businesses who were previously making the move to support Bitcoin. If they are halfway thru the rollout, its not a lot of work to switch to supporting BitcoinCash instead, due to the similarity of t
193.
by jgord 9y ago
thx for the tip. just tried it .. very vim, very nice.
194.
by jgord 9y ago
In your opinion, are high level commanders well educated? field experienced ?
195.
by jgord 9y ago
Some people see that as a feature - a deflationary rare resource, which a government can't simply print more of. If I just want to send funds from A to B, the current value of bitcoin in fiat [ USD / GBP / JPY ] is less impor
196.
by jgord 9y ago
yes .. but! we also simply need a better form of currency. While bitcoin has its implementation issues and politics, it solves that problem in a way that wasn't solved before. A better token for Cash is incredibly valuable in and of it
197.
by jgord 9y ago
Something that hasn't been done much yet, but probably will as the space matures, is for a blockchain to reuse another blockchain as its "timestamp" generator. eg. A 'MyCoin' miner grabs the latest 10 blocks off the
198.
by jgord 9y ago
I think that its quite possible that we can see scaling of 1000x on the linear blockchain, without fundamentally changing the economic/security properties. Increasing blocksize is just the first thing to do of many engineering optimiza
199.
by jgord 9y ago
Its a transient equilibrium during growth, not a zero-sum game. We could see lower fees and miners earning more : Hypothetically, if blocksize went up tomorrow by 8x, then potentially fees might drop by 4x and volume of transactions go up
200.
by jgord 9y ago
Its a commonly voiced, but false dichotomy - increasing blocksize 4x or even 8x will have negligible impact on the security or decentrality of Bitcoin. It will just allow more transactions through, which will allow fees to fall to levels wh
201.
by jgord 9y ago
mempool size over the past 30 days : https://blockchain.info/charts/mempool-size?timespan=30days Its almost never below 15MB - ie 15x as many transactions waiting to be processed, as can possibly fit in the next block.
202.
by jgord 9y ago
Everyone can look at the numbers and decide the most plausible explanation : https://blockchain.info/charts/avg-block-size?timespan=all To my eyes, thats a graph of user growth, not a spam attack. Similarly, this graph
203.
by jgord 9y ago
Even the current linear architecture can certainly handle more that 3 transactions per second on modern hardware .. there is clearly a bottleneck.
204.
by jgord 9y ago
Not a great article. Perhaps best to downvote until a better one comes along. Things to consider, which most people can agree on : - SegWit2X [ segwit feature, followed by 2MB block size ] is seen as a compromise worth taking to move thing
205.
by jgord 9y ago
I have no bitcoin on either camp - but for engineering reasons, came to the conclusion that the 1MB block is the real bottleneck. I think its possible Jihan and many others are acting in the best interests of Bitcoin community, on behalf of
206.
by jgord 9y ago
This misunderstanding comes up often, but the code and the blockchain are different domains. You need consensus on the blockchain, but you need a benevolent dictator, coupled with good communication, to make great opensource software. While
207.
by jgord 9y ago
I'm not sure why SegWit is put forward as a "scaling solution". It does make some room by moving signature data out of the main block, which may allow 2.5x as many transactions - but thats a one time improvement, afaict. The
208.
by jgord 9y ago
for a while at least, yes : ]
209.
by jgord 9y ago
I think miners acting in their self interest would realize that a much larger number of transactions with smaller fees would result in a larger total fees profit per block solve. Almost all of their resources would be the same - the vast ma
210.
by jgord 9y ago
The max blocksize is a code constant, so it requires upgrading software on a lot of machines to a version with the fix, which is incompatible with older versions. Called a "hard fork", meaning a fork in the blockchain - because o

More ›