HNHacker News
TopNewBestAskShowJobs

pdovy

244 karma · joined June 18, 2010

I work in finance doing high performance C++ development. I'm always happy to talk to other developers, so feel free to drop me a line - my email is paul.dovy at gmail.com
submissionscomments
pdovy··on Harvard Search of E-mail Stuns Its Faculty Members
Does the university have an established policy for email retention and access to that data? It seems unlikely that they wouldn't, but I don't see it mentioned anywhere. It's never a good idea to rely on goodwill to prevent access to sensitive data - if it's there, a need arises and it's not illegal, it's eventually going to happen.
pdovy··on Microsoft owes Denmark $1B in taxes
How is it that the initial (2002-2013 interest excluded) tax bill is 5 billion kroner, when the acquisition cost was only 10.8 billion kroner? Is the tax rate really that high? I understand that the tax is not on the acquisition itself but rather the transfer of the rights, but surely the market value of those rights should be less than the market value of the entire company at the same point in time.
pdovy··on Farewell, neighbors
This is really too bad, this was a great site and there doesn't really seem to be anything else like it out there. I'm surprised they couldn't come up with a way to make this profitable - and the closure is pretty sudden, wonder what the back story is there.

My neighorhood in Chicago used this pretty extensively and it really helped keep people connected and aware of what was going on. Sad to see that go.

pdovy··on Nest learning thermostat gets refreshed with a slimmer design
I believe it should. If the systems have separate thermostats you'll need two nests though I think (they apparently can communicate with one another).
pdovy··on Why Tarsnap doesn't use Glacier
Based on the article it seems like the motivation for the deduplication is as much bandwidth conservation as reduction of storage costs. What you're proposing would make sense if storage costs are your only concern, but it shuts out bandwidth limited users.
pdovy··on Ubisoft: PC Piracy Levels 95%, and 95% of Free-To-Play Users Do Not Pay
Taking them at their word on their statistics (seems reasonable as citricsquid pointed out based on interviews with the World of Goo developers and CD Projekt), I wonder how that would look broken down by country.

That is to say, is piracy really that rampant in the US, or are these numbers inflated by distribution in countries where games/movies are pirated and sold in hard copy form on the street or in shops? Which is worse for piracy, the ability to anonymously download the game via bittorrent, or being able to feel psuedo-legitimate about paying real money to a real person on the street or in a shop for a burned copy?

pdovy··on UBS sues Nasdaq over $357 million IPO loss
It's better to miss an opportunity than potentially expose yourself to a loss. If the exchange side system has become obviously unreliable (not responding to requests would count here), it's generally wise to stop trading.

I can see the temptation to just forge ahead on a one time only event like an IPO, but if you do so then I think you need to be prepared for the consequences. I also imagine that part of the usage terms for direct access include something that limit Nasdaq's liability in the event of a system error, so I'm not sure how UBS thinks they will get around that.

pdovy··on UBS sues Nasdaq over $357 million IPO loss
Yeah this seems like a baseless claim to me. I work on these kinds of applications and we always assume that an order is eligible to execute from when it's sent until the exchange confirms it has been canceled/deleted. This is pretty standard and it's surprising that a big bank would have such an obvious hole in their risk system, and disappointing that they would then fail to take responsibility for their own mistake.
pdovy··on Tagstand Hits 1M Actions, Launches New Tag Writer To Bring NFC To The Mainstream
So ... what the heck is NFC?
pdovy··on HFT using neutrino physics: Stats Jackassery
The use case for this would be communication between different colocations, which is where the networking "race" really is for HFT. For example, to arb products on the CME vs. the equities exchanges in the New York area (really colocated in various places in NJ).

At any rate, as the author mentions this isn't really a feasible technology.

pdovy··on A High Frequency Trader's Apology, Pt 2
Exactly. Even more direct benefit to retail investors - low liquidity costs mean that we can invest in mutual funds and especially ETFs that have extremely low fees. Those funds have to shuffle around their holdings as investors come and go, and they'd have to charge a lot more if they were paying $0.10 to cross the spread.
pdovy··on Apple and the risks of trading 29,000 times per second
As others have pointed out below, the value of the company may change at a fast rate, however that is far from the only reason that you might see many deals in a short period.

One of the primary benefits that market makers (which are often HFT firms) provide is the efficient transfer of risk. That is to say, the value of the company may not change thousands of times a second, but the willingness of existing holders of that stock to continue to do so may change rapidly. Or similarly some participant may suddenly need a hedge and buy this stock because it has the correlation their looking for. This is all to say that there are valid reasons for buying and selling a stock that have no direct (but some indirect) relationship with the value of that company. This is not a game played by retail investors, but it's not just played by HFT firms either. Institutional investors (e.g., my 401k money) will also use financial products in this way.

Or put a completely different way, I may sell my Apple shares because I need the cash back to help pay for a downpayment on my house. Did I sell those shares because I think the value of the company has suddenly changed? No. There are a myriad of reasons participants may move in and out of a position.

pdovy··on The real reason you can't hire developers....
Definitely. I did a lot of applying and interviewing when I was first looking for a job out of school, and it's amazing how disorganized some companies are. I will always specifically remember the companies that may have said "no" but were timely and polite, and I'd recommend them to my colleagues. On the other hand I also remember the companies that were rude or disorganized and recommend against them when asked.

Bottom line: Candidates should always like you even if you don't like or need them, especially since you'll likely reject far more people than you hire.

pdovy··on This 28-Year-Old Is Making Sure Credit Cards Won't Exist In The Next Few Years
That isn't necessarily true. I have and frequently use a credit card, but I do not actually need the line of credit. I have it to establish a credit history so that someday when I actually do need credit (e.g., for a home loan), I'll be able to get it.
pdovy··on Steve Jobs: Not a very nice person to know
I don't think there is necessarily any correlation between being a great businessman and a great human being. Was Jobs the former? Certainly, I don't think anybody would dispute that. Was he the latter? I don't know, and I don't think it's really our place to judge since the facts are pretty murky.

What bothers me slightly is that much of the praise for Jobs seems to equate one with the other. Let's celebrate the man as a visionary entrepreneur, and leave the rest to those that actually knew him.

pdovy··on What Would Happen if Amazon Ruled Publishing?
This is a cool idea. I doubt you'd even need to offer a rebate at all, though it'd be an interesting incentive. If reporting errors was as easy as highlighting a passage and clicking a button on your e-book reader, I bet a lot of people would do it.
pdovy··on The demise of the low level programmer
The major point that I've discovered from doing this stuff for years (on considerably more complex cases than the example above) is that programming efficiently for a modern architecture is qualitatively different ...

I think this is a great point. Low-level programming has always required an in-depth understanding of the underlying hardware and that hardware is changing as fast as ever today. It may be true that a large percentage of development can live happily in a land of high-level abstractions, but there will always be a need for low-level optimization and that space is still rapidly changing in very interesting ways.

pdovy··on I am an HFT Programmer
Argh, guys like this give the rest of us a bad name. It's ridiculous to claim that we're somehow universally better than developers at Google or anywhere else.

One of the big arguments in that Slashdot thread seems to be backlash over developers bragging about how they change software that handles millions of dollars in the middle of the day without any testing. Let me tell you, any HFT firm worth it's salt is very conscious about risk controls. I mean honestly, what kind of business are you running if you're routinely exposing yourself to potentially massive losses because of one developers error? Not one that would be around very long.

I can only speak to where I work, but we are not coding by the seat of our pants. Yes, traders do make intraday changes to their strategies, but they can do that because their software is backstopped by a tiered risk infrastructure that limits the damage their software can do.

pdovy··on The most useful GCC options and extensions
Has anybody used the __builtin_expect feature to good effect? We had been using this extensively in our codebase in places where there is an obvious common "fast path".

I tried an experiment where I did a comparison of the __builtin_expect version against a version with our likely/unlikely macros defined to the identity, and I actually saw no difference, which was disappointing. When I did some digging I found this blog post: http://bitsup.blogspot.com/2008/04/measuring-performance-of-... which suggests the same result for the Linux kernel.

FWIW GCC also has a reasonable set of PGO options (not mentioned in this article), and those actually can (anecdotally) make a noticeable difference, although you're subject to the overtraining problem if you're not careful.

pdovy··on The need for good vocational schools for programmers
Universities can and should do more to provide CS students exposure to real software development, but I'm not sure that a vocation school is really the way to go.

In my mind the greatest benefit of a strong CS degree is that it gives you a broad knowledge base to draw on that lets you easily pick up new technologies and concepts. The exposure to particular programming languages or technologies is less important, in large part because our industry moves at a fast enough pace that focused knowledge can quickly become outdated.

All that aside, the best "hands-on" experience you can get is actually working in the industry. I'd much rather see existing CS programs step up their efforts to place students in good internship and co-op experiences than add more software engineering courses.

pdovy··on Sitting for hours can shave years off life
If you bothered to look at the above linked article/video prior to replying you'd find it is not some stirring endorsement of standing desks. He actually explicitly makes the point that sitting or standing poorly all day will be detrimental to your health, and in reverse taking a few minutes to learn to do either properly can have a significant positive impact.
pdovy··on Sitting for hours can shave years off life
Kelly Starrett did a talk at Google recently on (roughly) this subject: http://www.mobilitywod.com/2011/06/episode-276365-mobilitywo...

He's got a PhD in Physical Therapy, so I'd rather listen to what he has to say about being healthy behind a desk than a story with a sensationalist headline.

pdovy··on Skype’s Worthless Employee Stock Option Plan: Here’s Why They Did It
It's unfortunate that this happened to these guys, but I'm glad it's been made public. I'd like to think I'd always carefully read a contract before signing, but this is one gem we can all watch out for now in particular.
pdovy··on I'm the guy who bought 259684 Bitcoins for under $3000 yesterday
An order with 1/100th penny precision isn't usual for currencies, but I see your point. Either way, this was a very unusual situation. Why were there so many people sitting at $0.01? If you're going to play that game, why not jump on at $0.011 or $0.02? Hell, pretty much anything < $1.00 you could probably safely assume would give you a profit as long as you were willing to bet you weren't watching the demise of the BTC.
pdovy··on I'm the guy who bought 259684 Bitcoins for under $3000 yesterday
I'm not sure where you're coming from with #2. Most exchanges use price-time priority to determine execution order. From the description the best bid was $0.01, so obviously he would get filled first at $0.0101.

The real problem is that the market was able to descend to that level to begin with. There should have been some sort of circuit breaker limits in the Mt. Gox system that would have halted such a massive decline, similar to what exist in the US equity and futures markets.

As to #3, really the only thing that matters in terms of exchange access in the US is that you satisfy the appropriate compliance burdens (broker-dealer status) and can pay the fees required for naked access. That may be good selection criteria for "informed traders" but it certainly isn't any guarantee.

pdovy··on Okay, Okay, Maybe Netflix Is a Problem for Cable After All
I got rid of my cable service about a year ago and haven't regretted it. It seems to me there is a lot of room for cable companies to easily improve their service, yet IMO it remains consistently poor.

- On demand selections are limited, and the interface is usually pretty clunky. If I have a DVR, I have to remember in advance to record something, rather than being able to just grab it after the fact.

- You need to have a physical box, which either arrives outdated or quickly becomes outdated, and they never replace it.

- It costs $60-$70/month for decent cable service where I live, versus $8/month for Netflix. Combine that with Hulu and free on-demand streaming from major TV networks, and it's a great bargain.

The only thing I miss about having cable is being able to channel surf when I'm bored. Honestly, I am probably a more productive human being without wasting my time doing that, so I'm not too disappointed.

pdovy··on Why I won't be using Groupon again.. A consumer perspective.
I don't doubt that people that use coupons are probably statistically poor tippers, but this seems like a self-fulfilling prophecy to me.

If you expect someone to tip poorly, so you give them poor service, of course they are going to tip poorly!

pdovy··on Cool, but obscure Unix tools
Yeah, most of these weren't anything new, but I hadn't heard about ack. Very useful.
pdovy··on Want speed? Pass by value.
This would introduce another performance consideration though, wouldn't it? You may not be paying the cost for the explicit copy, but now you are paying the cost of allocating space for a Foo on the heap instead of the stack.

It would depend on the situation, but this seems like it might just be wash at the end in terms of performance?

← PreviousPage 4 of 4