Recent changes: - Search removed; apparently getting quality search results is serious business
- Refund will work on all uncomplete torrents, no matter their percentage or number of seeds
- You can now add SHA1 hashes of torrents
- Price dropped to 0.001 BTC / GB
|
|
|
|
gadsdengraphics: They're using MongoDB. You can search for "coinbase" here to verify: http://www.mongodb.org/about/production-deployments/ Using a non-ACID compliant database for a financial application seems like a poor choice; hopefully they aren't using it keep logs of their financial transactions. I'll be damned... I am big fan of Mongo but I'd never ever use it for a financial backend. Please, someone from Coinbase come out and refute this for the sake of your company: Are you actually using mongo for the storage of the financial records themselves? |
|
|
|
Wikipedia already missed an opportunity to fund themselves for the next few years, which is a shame considering how often they beg for money. People were emailing them in droves asking to donate bitcoins a few months ago but they weren't interested. I'm guessing they will get on board soon.
So true. In a few months they will no longer be able to ignore it. |
|
|
|
|
Dammit Bitcoin, where are you going?
Cost per GB has been reduced to 0.003 BTC.
|
|
|
|
ok, you're right regarding dropbox integration.
what about an api instead?
if often use your site from work to add new torrents, and i would love if my homeserver would download them after bitfetch has finished them (so i can use them as soon as i get home).
You can already do that. Click on Settings (top right) and enable Automatic Browser Downloads. It will work on Chrome (up to 2 files, any more than that and it will popup 'This website is attempting to download multiple files. Allow? Yes/No' to which you click Yes) while Firefox/IE/Safari will require changing of the default download action in the browser settings. |
|
|
|
hello, i just love your service  just two suggestions: - dropbox integration (just place torrents in one dir and let it download with bitfetch) - dropbox: place completed files in dropbox regards flower Thanks! I think that Dropbox integration goes against the privacy oriented nature of this service. Dropbox API access is not open to everyone - you have to apply for it and get approved. I also don't see the point of watching a dropbox directory for .torrent files, since you can upload them to bitfetch just by clicking a button and they are typically <50kb anyway. Magnet links beat all other methods for convenience though, all it takes is a copy/paste. I will definitely be adding multiple file upload support (ctrl + clicking multiple files) for browsers that support it. |
|
|
|
Have you possibly considered hosting a torrent of the blockchain? I know that it has been done, but having someone to be a dedicated seeder would probably help a great deal.
I am seeding Jeff Garzik's blockchain torrent. It is also available as a direct download here. Did you know about burnbit.com?
It does the opposite of what Bitfetch does. It turns an http downloadable file into a torrent.
It's been useful to me a number of times when I needed to download a big file over an unreliable connection.
I have seen it. Download managers generally do the trick of downloading big files over HTTP just fine. Corruptions are not an issue since everything is checksummed over noisy channels (ADSL, 3G, etc) and files are almost guaranteed to stay uncorrupted on my server (ECC, verification). SSL also adds an extra integrity check to each packet. My service is for people who can't or don't want to run a torrent client. |
|
|
|
I could sit down and write a software component that could easily generate billions of transactions without breaking a sweat once it is deployed to a few thousand boxes, if I so chose Someone doesn't know how fees and transaction priorities work. |
|
|
|
Could you please elaborate about significant overhead?
Javascript, among other things. Don't get me wrong I love node but it just can't handle memory copying / allocating data structures as fast as a compiled language, which is understandable as it is using a JIT compiler with GC. In any case the difference with native C implementations isn't that big but it's there. I do realise that I'm comparing apples with oranges here but it's a valid comparison since the only stable socket.io server implementation is in javascript. I forgot to add that the big advantage of socket.io is its (almost) seamless horizontal scalability as long as your redis box is beefy. https://github.com/LearnBoost/socket.io/issues?page=1&sort=comments&state=openIts creator is busy working on 1.0 which is a complete rewrite (uses engine.io as the abstraction for the underlying transports). it might also be the way MtGox is using it or how they hooked it up to their other software
They are doing it wrong. |
|
|
|
PS: I know a good PRNG is Mersenne Twister, maybe you could look into seeding that with your clicks.
No, just no. Mersenne Twister is a terrible RNG for cryptographic purposes. You can actually jump to an arbitrary position in the stream given enough bits to calculate the state: http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/JUMP/index.html |
|
|
|
|
On chrome, if you paste a URL in the bar (or look it up by typing its first letters) it will be prefetched. So the following can happen: you paste your non-https instawallet URL in your bar with the intent of changing it to https, but it's already too late since it has been sent in cleartext by the prefetch mechanism.
|
|
|
|
Answers to emails I get a lot: Connectivity & uploadThe server is on a gigabit connection so your torrent will be downloaded at the highest possible speed. Each torrent has a 8 MB/s upload allowance with 20 upload slots. Slow or stalled torrents are the result of low or nonexistent amount of seeds. Slow downloads from the USThis applies to browser downloads only. As a cost-cutting measure some US ISPs (Comcast, Verizon) will limit the throughput of transatlantic TCP connections. To get around this you can use a multithreaded download manager such as DownThemAll. RatioRatio is not guaranteed to be 1, in some cases it will be much higher or lower. It all depends on the amount of leechers on the network. Average ratio of all torrents ever downloaded is currently 1.569 (BT total bytes out / BT total bytes in). I am open to suggestions on how to handle ratio for people that use private trackers. |
|
|
|
|
Free market goes both ways and if the miners choose not to relay spam then SD will have to adapt. This is not about shutting them down, it is about them using proper netiquette. We're all in this together using a shared resource which we would like to see boom over the following years. Good manners would go a long way towards that outcome.
I consider 1 satoshi* transactions that are basically equal to "LOL YOU LOSE" as spam.
* Recently been changed to 5000, but it's the same thing.
|
|
|
|
One is that on zipping a file, I changed my mind and wanted to download the files individually, but there wasn't a way to do this.
I guess I could add a cancel button which would kill the zip process, but this such a rare use case. It doesn't help that zipping is painfully slow and the reason for that is that I've limited it to 30 MB/s so that it doesn't hog down the entire server. Another is that zip as an archive format has the index at the end, so there's no way with standard tools to access files in a partially downloaded zip while it's still downloading. If you offered tar as well, then, due to its streamy nature, customers would be able to access files in the tar file so far in their partially downloaded tar file.
This is a great idea. What other archive/compression formats keep their file list in the beginning and allow opening of partial files? |
|
|
|
|
You can extract unbiased output from biased input (Von Neumann whitening is one such technique), but creating a hardware RNG by yourself is a bad idea.
If you really really want to use your geiger counter then your best bet is to feed /dev/random with entropy from the counter and use it's output.
|
|
|
|
|
Show Posts
