![]() |
| bullshit 7GB space for me .. when i give in 10gb and stay online for 70% of the time... thats just crap i dont see commercial viability |
| it uses a bunch of OSS and they plan to open it in pieces... |
| wren22coop dies81need afar61neighbor i like it ! |
| or codes: size96camel grove57hit |
| snug89crow wall87name print20hug washtub85apron different14swallow redbreast86read grasshopper9board leader43lead window34tablet meant35possible loaf37beet soon91fiddle |
| These are fresh and working invitation codes: lend46eager yard63their meow57party animal63gown dash27drum mrs.90sunflower dance42wolf magazine50cow |
| How does this improve on AFS/OpenAFS? If it's not open source, how can you hope to verify that it's secure? |
| What do you mean? For me it's very useful. |
| I'm using this and I don't like the fact that it's commercial/closed source, but it's still free, and I think it's the best thing around for exchanging things with friends. And it works perfect on linux. Can you share specific things with specific people on freenet? Or just backup your stuff only for yourself? Please let me know.. |
| In the end, it doesn't really make anything easier -- people just use it to say they use it. It's a status thing. |
| this misses the point. there is no such concept of "a copy" in this network - at least not in a sense visible to the user. copies of the various packets are sent and purged based on demand. |
| Could you just do this using FXP? |
| 48:32 (!) |
| On the point you make at 6.37 about data movement - I have always wanted someone to develop a distributed file system that supports an unlimited file size but more importantly when i select a file from another machine to be copied to yet another machine i dont want it to do 2 copies, one to me and then another to the destination machine, just one copy from the source to the destination. Does your file system support this? If so, do you have any downloads? |
| Wow, a commercial freenet |
| No. Copyright and ease of sharing are two different issues. This simply lets you easily access files from anywhere, including the sharing of files with other users. |
| we only see files that have been made public. all private / shared files we don't see and we can't encrypt, since your password never leaves your computer. |
| i think i didn't explain that point well enough. let's say you revoke access to a folder to marc. then marc doesn't have access to that folder anymore immediately. lazy revocation is only on a technical level: only if marc had a hacked client which would keep the access key to that folder, he could decrypt the files he had access to before as long as there are no changes to the folder (add, edit, remove). see our cryptree paper or lazy revocation in general for details. |
| This imo is the beginning of the end of copyright.... right? a huge scale professional p2p network and easy to use? bomb! |
| Hi Dominik, Nice work, I have a little concern though. How will you manage the balance between the stored and the uploaded data on the long run (when the google servers are out of the picture) if let's say every user wants to upload 7g, which represents 35g in the system, but on the other hand "only" 10g data is stored at every peer? |
| I have a question about privacy. As you said you have (or want to have) flags for copyrighted/inappropriate material. However, as the files in question are encrypted how do you 'legally' gain access to them and review their content? Saving encryption keys on your servers sounds dangerous from a privacy,if not a legal, stand point (unless you explicitly mention such a situation in the EULA). Either way, I'm excited about such technology and really liked the video. Nice one! :] |
| It is true you don't get as much as you share but what you get is a different kind of storage. Universally accessible from any Wuala client and fault tolerant. Also you are only deducted what you add into to Wuala once, not for every replication. So 7gb would only be 7gb not 35gb. |
| Hi Dominik, This was an interesting talk. I have however some questions. You said that if someone offers 10 Gb of its own disk space to the system, and is online 70% of the time, will get 7Gb of public disk space. This does not seem to be a good trade, as I give more than I receive. And does this mean that I can store really 7Gb in the system, which actually eats up 35Gb if all data is replicated 5 times? |
| how fast would an folder-access-change propagate thru the network? (e.g. removing a 'friend' would still allow him to 'get' files for some time) how secure are the routing tables? a malicious user could manipulate them to redirect traffic to his storage |
| Do you have no possibility to see or access the files shared with a group of friends? Let's say you get a hint of someone, saying "this user shares these illegal files with these people", what can you do about it? |
| Zitat: "For the other parts: We don't see the files so there is nothing we could do there." Ist das wirklich dein Ernst? Ihr seid euch hoffentlich klar darüber was das für rechtliche Konsequenzen für euch haben kann... |
| yes, you're right of course. |
| The encrypted files are kept in a cache. You can define how big that cache should be, e.g. 1 or 10 GB, very similar to a browser cache. If a file becomes very popular, a content distribution protocol similar to BitTorrent comes into play, making use of encrypted fragments in the cache. As for the other question, we don't see what people store or share if the files are private since we respect the user's privacy. |
| I have to cut my first statment short: this was very interesting to me. But I have a couple questions Once a file is downloaded does the client keep the decrypted file, the encrypted pieces, both or none? So if one particular file became exceedingly popular would there still only be a few copies of it lying around? Also it is pretty clear that you do not want any pirating, but if there is pirating going on in a private group how will that be handled? |
| Just to clarify, doesn't d points uniquely specific a polynomial of degree d-1? For example, 1 point uniquely specifies a polynomial of degree 0 (constant), 2 points uniquely specifies a line, 3 points specify a parabola, etc. |
| And thank you for the invite, I'm currently working with one of your engineers to make it work on my Fedora box.. Can't wait to play with this cool technology |
| ok, another try: you can find the paper at dcg dot ethz dot ch (search for cryptree and havelaar) |
| bah, youtube doesn't allow me to post URLs? whenever i post a URL, it fails... |
| Wuala for Linux is here! :-) Since yesterday. |
| Wuala sounds promising! I can't wait for the Linux client :-) |
| Any hope you can link the papers you referenced? Also can you expand on the licensing the code will be under, I could see a lot of interesting integration work being done e.g. with Nautilus and the GNOME Online Desktop project to make Wuala an integrated part of the users desktop and computing life but this would only be possible if the license was compatible. |
| This is Dominik. Thanks for your interest! I assume you have received your invite by now. If not, please let me know. |
| I'd love to be in the alpha, and since I watch these engedu films, I would assume that I'm the right type of person for it :) |
| Oh, we have to wait 48 hours to see if that one worked? |
| "All antendees will also get an invitation code to join the early alpha version" Would you consider giving invitations to the first 100 or whatever who respond to this video? think about it. I mean how many people are watching the techtalks as they come out? There's gotta be some kinda payoff for that. |
| Otomobil ve Araç Resimleri |
| Komik Resimler |
| Eğitim Resimleri |
| Eğlence Resimleri |
| Film Animasyon Resimleri |
| İpucu ve Stil Resimleri |
| Müzik Resimleri |
| Haber ve Politika Resimleri |
| İnsan ve Blog Resimleri |
| Hayvan Resimleri |
| Bilim ve Teknoloji Resimleri |
| Spor Resimleri |
| Turistik Yer Resimleri |
| Bağış Aktivitesi Resimleri |