At 42Floors, we experienced a similar need for auditing. Initial we did a
similar approach as described in the post. Over time the audit tables began
to be incorporated into every aspect of the product, which over time caused the
table to become quite large, and ended up causing a lot of downtime. Our initial
implementation only audited what was inserted by our Rails application. If a
change was made by another application or manually, it wouldn’t appear in the
table.
Over the course of a few months we extracted the system and built a standalone
application that today has become [changebase.io](https://changebase.io/). We
now use CDC, change data capture, to capture all events in the database and a metadata table to
capture the context of why the change occurred and who made the change. The
changes are then structured to be easily queryable.
Initially, we encoded the data as JSON, but over time have opted for encoding the
data in the database’s native format. This creates a more complex query system,
but at the end of the day we don’t have to worry about type conversions and
querying is done in the same format the database that was replicated.
Teams have benefited from the system by being able to recover data that was
accidentally deleted, or deleted maliciously. Notifications have also been used
for high value datasets where someone on a sales team wants to immediately know of
a change in the system.
Why? The battery swap only makes sense when you are on a road trip longer than the typical range of your battery(hundreds of miles), so it's not like Tesla owners are going to be doing a battery swap every week. Most likely, tesla owners will charge at home for over 99% of their trips, and maybe choose to do a battery swap if they are on the road for a long time and don't want to wait the 30 minutes for a super charge.
My guess is that this was just an easier story to sell because they can just approve accredited investors and not have to deal with those who are not due to regulations. They may have planned to make it open to everyone but had to step the plan back a bit.
So while this is an astute observation, a properly run exchange would go through a lot of effort to separate the matching engine (the inherently single-threaded part of the exchange) from the portions of the system that accept orders and spew data. The registration system and web site also don't belong anywhere they could possibly contend with the matching engine. Real exchanges process orders of magnitude more volume on a single symbol than mtgox does, yet have latencies measured in microseconds rather than tens of minutes.
The matching engine should be separated completely from the rest of system as this is the one component where you want to align with cpu cache and all other kind of crazy low level optimizations. I do agree that mtgox system is slow but we should keep in mind the resource difference both in time and money the large stock exchanges have had time to build their systems. Most of the stock exchanges also don't build their own system but rely on buying one from a independent organization or another stock exchange.
A small nitpick:
Volume is one metric but it does not matter in reality the important part is the amount of transactions on a specific symbol.
A small nitpick: Volume is one metric but it does not matter in reality the important part is the amount of transactions on a specific symbol.
That's quite right, thanks :)
Real venues incentivize participants to post orders in round lots (generally 100 shares) and some of them penalize participants who add liquidity with extraordinarily low fill rates. These rules help reduce load on the matching engine, and afaik mtgox has no analogous rules.
I really do believe arguments about the 'freedom' of speech should be kept to the context of government versus the populace.
The government censors people through force, sometimes in autocracies this quite literally means a gun to a dissenter's head.
It is a separate thing all together, On the other hand, if individuals freely choose to criticize, flame, or just simple disassociate themselves with someone else because they disapprove of their opinions. We should all be free to do at least this without fear that merely by agreeing with the majority we will be accused of 'censoring' the minority.
The point is, some types of free speech should be encouraged and some types discouraged. I think you should be allowed to flame people, but I'm not obligated to provide you a venue to do so, and I'd really rather you didn't.
No, thats defamation/libel, and it is a civil offense, even a criminal offense in some cases. Free speech is not an unconditional right, it places responsibilities on those who exercise it.
Defamation laws are one of those laws that skirt on a thin-line of interfering with free speech rights (similar to hate crime laws).
It's also not simply offending language that equals defamation.
From wikipedia:
"Opinion is a defence recognized in nearly every jurisdiction. If the allegedly defamatory assertion is an expression of opinion rather than a statement of fact, defamation claims usually cannot be brought because opinions are inherently not falsifiable."
also
"In Common Law it is usually a requirement that this claim be false."
>It's also not simply offending language that equals defamation.
Based on the post, many of the tweets were insult - direct attacks against the author, and not simply criticism/opinions.
All but 4 states in the US also recognize cases of defamation per se, where the truth of the claims is irrelevant, and the damages are predetermined. Claims which impute the following fall under that category (I've omitted 2 of the 4):
1. Allegations or imputations "injurious to another in their trade, business, or profession"
2. Allegations or imputations "of loathsome disease" (historically leprosy and sexually transmitted disease, now also including mental illness)
Many of the tweets directed at him fall into that category, and hence are defamation per se.
Defamation per se is very limited under the First Amendment, and mostly historical at this point, dating from an era when courts would entertain prosecutions for calling someone's honor into question. To the extent it survives, it's been limited mostly to where "regular" defamation would also apply, to false statements of fact. For example, "injurious to another in their trade" can't be used to prosecute any comment that could harm someone's business, but only false factual statements that could do so, such as accusing someone of selling counterfeit parts. Nowadays the only real distinction is in showing damages: in a per se category, once defamation is proven, the plaintiff doesn't have to further prove that they were harmed, because harm is assumed in those categories.
I think mostly we're discussing something else here, angry and vitriolic "piling-on" responses, which is a different issue from defaming someone.
It is a shame to see such hate against people. @antirez was attacked for openly sharing his thoughts about sexism. The sad part is that those who attacked him seek respect for their points of view yet do not give it to others.
I'm not that great of a writer. My papers were usually in the B range and I almost never got an A. Then I started using LaTeX and now my papers are used as examples in classes.
When I was in Uni I had a similar experience. I never collected enough data to really show a conclusive relationship and obviously it was possible my writing improved around the same time that I switched.. but it was still something that I noticed.
One of my theories was that professors expected rushed papers to have poor or missing formatting, but my rushed papers wouldn't. The formatting was therefore sending a signal to the professor saying "He didn't write this in the morning, so give him the benefit of the doubt." I never considered that the typeface that I was using could have been involved.
In the case of (La)TeX I'd actually also be curious if i.e. the rules governing hyphenation/justification makes any difference as well, not just the typeface.
I'm pretty sure they do. Justified text looks more "professional" than ragged right one. The problem with it is, that it can create rivers without proper hyphenation. TeX is pretty good at this stuff and that's one reason why text set with TeX looks so good. I wish browsers would catch up on this.
I'm pretty sure the grader for my algorithms class just assumed that anything typeset in LaTeX must be correct, as I routinely got full credit for completely incorrect things, while classmates not using LaTeX got dinged for every minor mistake.
I was a TA for some computer classes at Stanford a little while ago. Part of that meant that we would spend many hours grading exams at the end of the term. Every time I wondered to myself how much student handwriting affected their grades. How could a student write almost entirely illegibly and expect to get a good grade on the exam? Sometimes I couldn't find the correct answer even if it was there on the page.
Other students had very clear handwriting, while others had very crisp and confident handwriting. I'm sure there are numerous subtle cues in handwriting that give it the air of professionalism or confidence.
Handwriting differences could easily account for a few points on an exam, which might translate into a few crucial percentage points.
Soon we'll just have to force everyone to type their answers in a standard format and then the graders can view the results in the editor of their choice.
I wonder if that has ever affected my grades. I have truly horrendous, mostly irregular handwriting but it is readable to most people. I can, however, write in calligraphic styles but it isn't automatic.
Does handwriting say anything more about a person than that they spent a thousand hours or so in their early schooling repeating symbols on paper?
Latex commands so much authority over people who have never used it- things typeset with it carry incredible gravitas! I just now realized that must be why I was drawn to it. You can say the same thing, and sound/look smarter :)
I used to dislike CM when it was printed on laser printers, but discovered that when it was printed on a Linotype--the Art of Computer Programming books, for instance--it seemed to take on an entirely different character. It looks clean and pretty, rather than spindly.
Clearly that's subjective, and you might hate it no matter the output device. :)
I'm a Scottish teenager. Perhaps the only possible reason I managed to scrape a 2 (2 or 1 is high enough to do Higher level course the year after) in my Standard Grade English was because all my work was in Georgia.
Over the course of a few months we extracted the system and built a standalone application that today has become [changebase.io](https://changebase.io/). We now use CDC, change data capture, to capture all events in the database and a metadata table to capture the context of why the change occurred and who made the change. The changes are then structured to be easily queryable.
Initially, we encoded the data as JSON, but over time have opted for encoding the data in the database’s native format. This creates a more complex query system, but at the end of the day we don’t have to worry about type conversions and querying is done in the same format the database that was replicated. Teams have benefited from the system by being able to recover data that was accidentally deleted, or deleted maliciously. Notifications have also been used for high value datasets where someone on a sales team wants to immediately know of a change in the system.