HNHacker News
TopNewBestAskShowJobs

karl_gluck

325 karma · joined June 22, 2012

[ my public key: https://keybase.io/karlgluck; my proof: https://keybase.io/karlgluck/sigs/gzNpgSBv9GjAUGBD7XIGlEynRqHc1MJKlCESBgSWDjw ]

karlgluck.com

submissionscomments
karl_gluck··on Show HN: Code Kingdoms – a game that teaches kids to code
First off, great idea! The screenshots look friendly and I think it would be something I'd have used when I was a kid.

That said, you have some stability/UX issues. The loading delays are very long. I can easily see kids losing interest because it takes >2 minutes to load the Quick Play game. Then, once I was in the game, the graphics rendering was laggy and made my computer stutter / the screen flash. Shrinking the window made it run fast enough. I walked around a bit but as soon as I encountered the first duck the game went into loading mode again... and has been stuck there ever since.

karl_gluck··on Show HN: Code Kingdoms – a game that teaches kids to code
Same. Hit it again and it works.
karl_gluck··on Facebook Buying WhatsApp for $16B in Cash and Stock Plus $3B in RSUs
I would agree, but if you don't update WhatsApp it disables itself after a week or so.
karl_gluck··on Open Libernet: a Bitcoin-based fully encrypted mesh networking protocol
Not too long ago, the internet measured top speeds in tens of kBps. So first off, give it time. Engineers are good at solving problems like this.

Furthermore, mesh networking is a fundamentally different architecture solving a different problem than the internet. The current internet works great for high-bandwidth applications like commercial video streaming, but is terrible in terms of protecting itself from concerted attack by powerful actors. Mesh networks solve the single-point-of-failure problem.

http://gizmodo.com/5912383/how-to-destroy-the-internet

karl_gluck··on Can't remember iPhone stock photo site submitted to HN (don't upvote)
Hahah yes!!!! Thank you SO much. My only regret is that I have but 1 upvote to give.
karl_gluck··on GBA Emulator in HTML5 (Alien Hominid)
Wow! That was pretty fun. I just wish it had a description of which buttons were which; I died after only a few minutes because I couldn't figure out which button was L.
karl_gluck··on Time-lock encryption
Good point! Still, I think the best hash inversion I've seen is slightly under 1/2 the hash space (2/5ths?). Using a 512-bit hash like SHA3, even at an ungodly hash rate (1 TH/s), you still get approximately 2^141 seconds = 2^133 years maximum chain length before it becomes more efficient to invert the hash.
karl_gluck··on Time-lock encryption
The issue is that you can parallelize the "unlocking" operation. So, if one person is attempting to unlock, they solve in 10 years. If two people attempt, it's solved in 5 years. If the NSA puts all their computers on it, it takes 1 second.

The author is proposing methods that are forcibly sequential.

karl_gluck··on Time-lock encryption
I like the parallelized hash chain construction idea; I've never seen that before.

One could improve a chunk of the chain by having checkpoints along the way. As the author mentioned, it would suck to be 2 years into mining a 35-year computation, only to make a mistake that you can't detect until the very end.

To add checkpoints, one could release both the original seed of the chain A, and a number of pairs of hashes (x0,y0) (x1,y1) ...

Let's say you wanted to do 1-month chains. Hash the seed A for a week, then take the current value x0 such that H(B)=x0. You know the value of B, since you've been computing the chain. Pick another random value y0, and continue the chain with H(B^y0). Write (x0,y0) in the output, and hash for another week. Do the same for (x1,y1) (x2,y2) and (x3,y3). Each chain then has a seed value and 4 pairs of 'checkpoints'.

When unlocking the crypto puzzle, these checkpoints can't be used to jump ahead in the computation, but they can tell you that you're on the right track.

I think that you could even use a secondary hash chain for the y_n values, so y_n+1=H(y_n). If you also derived y0 from A (e.g. y0=H(A^const) ), you would just need to publish the seed value A and each checkpoint hash x_n in order to have a fully checkpointed crypto puzzle.

karl_gluck··on Don't use Hadoop when your data isn't that big
I love how whoever mods this site now doesn't even have to follow their own rule about not editorializing headlines.

Mods: Don't be hypocrites. If you're going to enforce your "only use the source title" trash on us, follow it yourself.

Original title: "Don't use Hadoop - your data isn't that big"

Mod-invented title: "Don't use Hadoop when your data isn't that big"

karl_gluck··on It's A Terrible Time To Buy An Expensive House
You assume the 5% per annum is going entirely to equity. Consider that property taxes, HOA fees, utilities, mortgage interest and maintenance all contribute to the ~8% figure, and your p/a equity looks closer to 3%.

Now let's compare if you were to purchase stock/ETFs with that equity investment instead of investing in your house.

Let's do a housing-optimistic comparison. Let's say you get a 2% dividend (nothing special) on the stock and the stock price declines at 0.5% per year. For the house, let's say you gain 3% equity per year and the home price stays flat. Rent costs are 20% higher than your expenses for the house, so you can only gain equity in the stock at 80% the rate you do in the house.

Over the same time period that you would pay off your mortgage and get 100% equity in a $100,000 house, thanks to compounding interest, your stock is now worth $107,000. That means that you have 7% MORE value in stock--even though you paid 20% more rent for the entire period of investment AND the stock price tended to decline (slowly) over time.

If we adjust to something more balanced (3% equity per year, 3.5% dividend, stock price stays flat, rent costs are 15% higher) the result is even more dramatic: 53% more value in the stock option.

Of course, all these numbers would be better if backed up with real-world statistics, but it's something to think about.

karl_gluck··on It's A Terrible Time To Buy An Expensive House
You make a pretty big assumption here: that anything you can invest in to earn money passively CAN be guaranteed. What you mean by "generate guaranteed income" is "to allow me to grow my investment". After all, investing $100k in a house that loses 10% of its value is worse than just keeping the $100k in dollar bills. In fact, if it is still worth $100k nominal down the line means you lost 2-3% per year on your investment. So after mortgage, property tax, maintenance, and everything else, your renters would have to cover about 7-9% of your home's value every year for you to make a 2-3% margin.

What I've found is that the reward is always proportional to the risk--otherwise people invest in it until it isn't proportional anymore.

Please investigate this for yourself, but personally, I take a two-way split-risk strategy to grow my investments:

- Very high risk, very high reward investments that I thoroughly research and believe extremely strongly in, using money I can afford to entirely lose.

- Very conservative investments in indexing ETFs that have rock-bottom annual fees and decent dividends.

Here's an article you might find interesting:

http://www.npr.org/2013/06/05/188306471/resisting-the-tempta...

karl_gluck··on It's A Terrible Time To Buy An Expensive House
Ding ding ding. You are correct! The number of people showing up here (disproportionately from California, actually) is pushing housing prices through the roof. Houses sell about 10% over ask pretty much anywhere.

Rent has doubled for most of the area over the last 5 years. In my dismay a few months ago, I did some research into buying a place but quickly came to the same conclusion as the author.

I just did a quick look around my area over on Zillow. A condo comparable to my apartment (townhome style, 2-car garage, similar sq ft / bedrooms, location, etc) is still listed for 23 times my current annual rent, even though rent went up 15% last year. I'm renting for 4.3% the price of an (old!) condo. Anyone who thinks this is sustainable is going to be in for a surprise.

[edit: rent calculations annual...]

karl_gluck··on 64-bit ARM chips in iPhone 5S serve up taste of Intel-free future for Apple
"Work" is pretty broad. Let's define it as "have a productive experience" because any non-period matched programs have issues other than pure compatibility. For example, bring many old DOS programs up on modern hardware and they will "work" but be entirely unusable unless you do something to slow them down.

So, 2013 programs that don't "work" on 2005 hardware? Off the top of my head: SolidWorks 2013, VMWARE running Ubuntu 12 on top of Windows, Avid, Visual Studio 2013.

karl_gluck··on S3for.Me – cheap alternative S3 cloud storage
This is great news! I'll be checking this out for sure.
karl_gluck··on S3for.Me – cheap alternative S3 cloud storage
Where are they located? (Outside of the US?)

I can't find anything on their website that lists physical contact information.

karl_gluck··on 'StarCraft' Gameplay Boosts Mental Flexibility, Says Study
If you really believe that, I'd suggest you check out the WCS Season 2 Grand Finals from last night. Really, all of the WCS event. Proxies do quite well against standard build orders--that's one of the things that makes pro play so much fun to watch.
karl_gluck··on Quipu
Exactly! I agree.

However, all the current system does is penalize people who follow the submission rules (since their content doesn't get up-voted), and generate annoying comments about the "original topic" for the rest of us to sift through.

I've now clicked on this link twice because the title changed and I thought it was about something else. Then, having not looked at the comments last time because the subject matter didn't seem interesting, I saw "12 comments" I went and looked to see if I had missed something. Nope, just a lot of unnecessary meta-discussion caused by mods changing the title.

To the mods: it's annoying. Enforce the rules at submission, or knock it off and let the community decide what's interesting.

karl_gluck··on Quipu
Mods invisibly changing titles really bothers me.

It's even worse that the mods decided it was a great idea to silence debate by [dead]ing the topic when HNers discussed possible alternatives.

Makes me wonder how much longer I want to stay in this community.

karl_gluck··on 3D printing will explode in 2014 thanks to the expiration of key patents
I'm not sure if there's a cause-effect relationship here. MakerBot was only founded in 2009 [1], RepRap released their first project in 2007 [2] and the project I was a part of, Fab@Home, open-sourced in 2006 [3].

I can't speak to the other projects' motivations, but I know that the fact that patents were expiring made little to no difference on our decision to make a printer. If you can believe it, the Fab@Home was almost an afterthought. We had top-of-the-line commercial 3D printers in the lab, but they didn't do the kind of custom material jobs required for the project that Evan Malone was doing. His PhD project was to create a 100% printed-from-scratch robot, and no technology fit the bill. So, he and Hod built one in order to get that project done. Like the Oculus, technology had advanced to the point that doing this was actually feasible on a shoestring budget. It ended up being a project of its own, but it wasn't like they saw the patents expiring and all of a sudden decided it'd be great to open-source a 3d printer design.

With how popular 3D printers are getting nowadays, it's possible some companies are looking to take advantage of the expiration of SLS patents, but I wouldn't hold your breath for an open-source project. If someone was making one, you'd know about it already.

[1] http://en.wikipedia.org/wiki/MakerBot_Industries

[2] http://en.wikipedia.org/wiki/RepRap_Project

[3] http://en.wikipedia.org/wiki/Fab@Home

karl_gluck··on Game Programming Patterns
The early GPU Gems books might be of interest to you. They are full of stuff like frame timing subtleties and interesting ways to solve common game loop problems. These days, you can probably get them for less than $50 each, too.
karl_gluck··on Game Programming Patterns
Hi! Wow, very nice to meet the author. Most of this was written after I'd already finished my game, but the Object Pool article was part of the inspiration for my memory pool class [1]. Thanks for everything you've written, it has helped more people than you probably know.

[1] https://github.com/karlgluck/Evidyon/blob/master/Source/comm...

karl_gluck··on Game Programming Patterns
I wouldn't hold your breath. When I last used this site in 2010 [1] it wasn't that much less complete.

Still, what is has is very useful and I'd suggest giving it a read.

[1] http://web.archive.org/web/20100131082252/http://gameprogram...

karl_gluck··on Why so many people are moving to Texas
While I agree with most of these as great reasons for moving to Austin, as a resident I'd caution those interested in moving here to not take them at face value. Come visit and consider what they're not telling you when making a decision:

- It was cheap, but it's not staying cheap. The idea that 'Austin is cheap' has inspired a lot of people to start moving here recently, and the reality hasn't quite started to impede the hype. Reality? My rent has gone up 10%/yr each of the 4 years I've been here, and home values / rent prices across the area are still rising rapidly. A 640 sq ft. studio apartment that was $530/mo when I moved here now rents for $1050/mo, and I'm not even close to downtown.

- Texas has no income tax, and house prices may seem to be good at first glance, but property taxes are VERY high. 2.7%-3.5%, and some areas are taxed by multiple entities.

- If you can't find anything affordable close to work, prepare for a nightmare daily drive. Most people I work with have a 45- to 60-minute one-way commute. Austin rapidly outgrew its 2 highways, so "rush hour" is 7:45 AM - 10 AM, 3:45 PM - 7 PM every day, except Friday afternoon where it starts at 2:30 PM.

- Even as someone with a good software job and no debt, I can't come close to affording a house within a 30 minute one-way commute of work that was built after 1960 given the recent price increases.

- Austin is hot, so be ready for that. We set a record of 90 consecutive days of 100+ degree weather in 2011!

- Despite all of the above, I do love living in Austin for its food, weather (I hate gray skies and winter), and culture. I worry about how long I'm going to be able to afford to stay... but for now it's a great place to be.

karl_gluck··on Hacker News is depressing
That's not a particularly good analogy, because both roles require significant commitment to practice and study. Trying to achieve both would take would likely render you worse than most at either. Being a good programmer takes training and effort; being respectful and understanding does not take any superhuman effort. It just requires understanding that demeaning people is not acceptable behavior.

vs.

SHUT UP. Your analogy is awful and I never want to hear that kind of drivel from another HN poster EVER again. Go rtfm on English lol

karl_gluck··on Bizarre Soft Robots Evolve to Run
Viktor Zykov (one of Hod's grad students) did something like this in 2006 [1][2]. He built a "starfish" then had it learn its own shape through experimentation, then evolve a way to move.

[1] Video: http://www.youtube.com/watch?v=msw267lisow [2] Info: http://creativemachines.cornell.edu/emergent_self_models

karl_gluck··on Bizarre Soft Robots Evolve to Run
My first thought was "this is really familiar"... then I realized it's because I used to work in Hod's lab :)

I think it's interesting that most of the robots that evolved don't have rigid components. I wonder if reducing the structural support of the "muscles" themselves would favor robots with a defined skeleton.

karl_gluck··on Announcing Firepad — Our Open Source Collaborative Text Editor
Very cool technology! I really appreciate your decision to allow us to just give it a try without having to set up an account or log in.
karl_gluck··on Show HN: Fast and efficient domain checking tool
Wow! Great work. If I had a request, it would be to support more exotic TLDs and other registrars--but it's amazingly useful as-is. The price comparison is killer; I had no idea I could register a .TV domain for 10 bucks. I'll definitely be using this.

Is your VPS doing the DNS lookup, or is that client side? If it's done on your server, I would suggest that the site preview feature could be made click-activation rather than automatic. It's already pretty slow to load, and I would be just as happy with a link to open an unavailable domain in a new window. Basically, if the preview slows down the domain availability lookup, I think it should be pushed to a lower priority.

Overall, very slick! Congrats on a successful side-project.

karl_gluck··on Show HN: an app I built to help you find great beer
Looks great (and I live in Austin, so recognizing some of the places was cool!) but I only have an Android phone. Would love to know if you make a version for it.

Is the data for this app user-submitted, or do you pull from somewhere?

← PreviousPage 3 of 4Next →