\[Kickstart\] Link - The File Sharing Blockchain Protocol
-
[s]Also. Can we please pin/sticky this?
Here is fine or maybe in the development category?[/s]
Nevermind, Forgot there’s all ready one in Tech Dev… That actually all ready contains the tech stuff…
-
Tonight I am:
[list]
[*]Writing the FAQ and about page, and fixing the footer
[*]Registering a domain
[*]Firing up a host in the cloud
[*]Installing Feathercoind, Link, and Link-server on it
[*]Testing everything in “production mode”
[/list]We’re that close. I’ve been testing all the main functionality and site navigation. It’s solid enough for beta-testing. Mnstrcck said he’d probablly get to the Search Result CSS tonight as well.
If all goes as planned, you donators may be receiving a link in your inbox to Link sometime tomorrow… Although Tuck will get his before anyone else. Watch this space for more…
-
[quote name=“Kevlar” post=“47223” timestamp=“1387849065”]
Tonight I am:[list]
[*]Writing the FAQ and about page, and fixing the footer
[*]Registering a domain
[*]Firing up a host in the cloud
[*]Installing Feathercoind, Link, and Link-server on it
[*]Testing everything in “production mode”
[/list]We’re that close. I’ve been testing all the main functionality and site navigation. It’s solid enough for beta-testing. Mnstrcck said he’d probablly get to the Search Result CSS tonight as well.
If all goes as planned, you donators may be receiving a link in your inbox to Link sometime tomorrow… Although Tuck will get his before anyone else. Watch this space for more…
[/quote]all that in 1 night
-
[quote name=“Kevlar” post=“47223” timestamp=“1387849065”]
Although Tuck will get his before anyone else.
[/quote][img]http://www.reactiongifs.com/wp-content/uploads/2012/06/c9X6n.gif.pagespeed.ce.mqTuBwn0na.gif[/img]
-
I will be watching, can’t wait. If I can use it anyone can. Hopefully I won’t break it.
-
[quote name=“BikerGSX” post=“47228” timestamp=“1387850276”]
Hopefully I won’t break it.
[/quote]I hope I do break it because that’s what beta testing is.Really, Tuck, how much did you donate to get so much love? :p I thought I read that there was >500 FTC that Kev was using for testing and re-dispersing and I know 100 of it came from me and some other user matched it.
-
[quote name=“HopeStillFlies” post=“47229” timestamp=“1387850547”]
Really, Tuck, how much did you donate to get so much love? :p
[/quote]Not enough that’s for sure. I hope to continue contributing as the project grows.
I think Kevlar will accept just about any coin out there as a donation and since this project can be used on any blockchain it could assist in testing the various coins or be converted to any coin for testing or other needs.
-
[quote name=“HopeStillFlies” post=“47229” timestamp=“1387850547”]
[quote author=BikerGSX link=topic=5663.msg47228#msg47228 date=1387850276]
Hopefully I won’t break it.
[/quote]I hope I do break it because that’s what beta testing is.Really, Tuck, how much did you donate to get so much love? :p I thought I read that there was >500 FTC that Kev was using for testing and re-dispersing and I know 100 of it came from me and some other user matched it.
[/quote]Look, this is going to be beta quality software. Period. I mean I’ve not even done unpublished expiration. You can break this 60 days from Sunday if you’re clever. You could spam the blockchain with Junk if you wanted to. It would cost you, but you could do it.
I’m trusting a select group to do the right thing. If you’ve donated, you’ll get the link to Link. If you post it on the forums, the secret will be out and you’ll ruin the exclusivity. I ask you don’t do that. If you click the publish button 666 gazillion times, you’ll fill up the HD with unpublished information. I ask you don’t do that. If you even try and send the payment in multiple transactions adding up to the final amount and there’s not another coin in there to send it, it’ll break. It needs whole spends equal to or greater than the amount to spend. It’ll break all over the place.
I want to make this really clear ahead of time: [b]It WILL lose your coins[/b]. I don’t plan on returning them, even if the data never posts. When it works, it is destroying them, and giving them away as miners fee. I’m sorry in advance for all the bugs, and for your lost coins. As way of consolation, I’ll use whatever I sweep up off the floor to fix the bugs that caused to you lose them in the first place. This comes with zero warranty except the following: I’ll do my best to not make a total ass-hat of myself. All other illusions about quality should be dispelled now.
I estimate the average cost to be less than 1 FTC.
The fee breakdown is as follows:
15 bytes = 0.00000001 data encode + 0.005 miners fee.That means that for a message with 700 bytes (which approximates to a name, a magnet link, a 100 word description, and 5 keywords), the cost will be 0.94000047. That’s 0.00000047 coins destroyed, 0.94 for miners, and 0 for me. This is of course trivial to verify in the blockchain.
-
[quote name=“Kevlar” post=“47237” timestamp=“1387854063”]
Look, this is going to be beta quality software. Period. I mean I’ve not even done unpublished expiration. You can break this 60 days from Sunday if you’re clever. You could spam the blockchain with Junk if you wanted to. It would cost you, but you could do it.[/quote]Don’t get defensive. I want to put it through the paces. Abuse it if necessary so that anything that needs to be addressed by primetime can be. I’m on your side on this, do not take it as a criticism of your ability.
[quote]I’m trusting a select group to do the right thing. If you’ve donated, you’ll get the link to Link. If you post it on the forums, the secret will be out and you’ll ruin the exclusivity. I ask you don’t do that. [b]If you click the publish button 666 gazillion times, you’ll fill up the HD with unpublished information. I ask you don’t do that. If you even try and send the payment in multiple transactions adding up to the final amount and there’s not another coin in there to send it, it’ll break. It needs whole spends equal to or greater than the amount to spend. It’ll break all over the place.[/b][/quote]
Good, this is information that we need to know as ground you’ve already covered. That said, I’m willing to still find ways to make it not work for a [i]dumby[/i]. I want this thing to work, I want it successful, I want it to eventually be as trivial an end user process as someone putting and pulling a magnet link on a centralized website right now. I’m just saying, sorry if not tactfully enough, that I’m willing to dedicate the time and resources to it even if it means losing both. I know you’ve got tons of hard work invested in this, but I also know my expectation isn’t that of this being a shrinkwrapped product because that’s just the way these things work.
[quote][b]I want to make this really clear ahead of time: It WILL lose your coins[/b]. I don’t plan on returning them, even if the data never posts. When it works, it is destroying them, and giving them away as miners fee. I’m sorry in advance for all the bugs, and for your lost coins. As way of consolation, I’ll use whatever I sweep up off the floor to fix the bugs that caused to you lose them in the first place. This comes with zero warranty except the following: I’ll do my best to not make a total ass-hat of myself. All other illusions about quality should be dispelled now.[/quote]
That’s a given and I understand the process of how it works from an end user perspective. I’m just saying that we (I! Me! The royal we!) understand the risk and are willing to put it through it’s paces.
Please don’t ever take what I say as some sort of personal attack or a slight towards work I haven’t even put hands on yet. The expectations are simply realistic and I have confidence that this is going to shift in the way life, value and distribution work and that’s not to be taken lightly.
[quote]I estimate the average cost to be less than 1 FTC.
The fee breakdown is as follows:
15 bytes = 0.00000001 data encode + 0.005 miners fee.That means that for a message with 700 bytes (which approximates to a name, a magnet link, a 100 word description, and 5 keywords), the cost will be 0.94000047. That’s 0.00000047 coins destroyed, 0.94 for miners, and 0 for me. This is of course trivial to verify in the blockchain.
[/quote]sounds like the pricepoint sweet spot. For keywords, since there’ll be a logical limit, is there any sort of “keyword cloud” that people can follow or even “click to label” their stuff in order to minimize redundancy? -
I will sit this one out at the moment and leave the testing to the experts; you know who you are!
As a layman I will only get in the way, But I can help if it needs more funding. Let us know if more s needed to move this along. -
[quote name=“Kevlar” post=“47237” timestamp=“1387854063”]
[quote author=HopeStillFlies link=topic=5663.msg47229#msg47229 date=1387850547]
[quote author=BikerGSX link=topic=5663.msg47228#msg47228 date=1387850276]
Hopefully I won’t break it.
[/quote]I hope I do break it because that’s what beta testing is.Really, Tuck, how much did you donate to get so much love? :p I thought I read that there was >500 FTC that Kev was using for testing and re-dispersing and I know 100 of it came from me and some other user matched it.
[/quote]Look, this is going to be beta quality software. Period. I mean I’ve not even done unpublished expiration. You can break this 60 days from Sunday if you’re clever. You could spam the blockchain with Junk if you wanted to. It would cost you, but you could do it.
I’m trusting a select group to do the right thing. If you’ve donated, you’ll get the link to Link. If you post it on the forums, the secret will be out and you’ll ruin the exclusivity. I ask you don’t do that. If you click the publish button 666 gazillion times, you’ll fill up the HD with unpublished information. I ask you don’t do that. If you even try and send the payment in multiple transactions adding up to the final amount and there’s not another coin in there to send it, it’ll break. It needs whole spends equal to or greater than the amount to spend. It’ll break all over the place.
I want to make this really clear ahead of time: [b]It WILL lose your coins[/b]. I don’t plan on returning them, even if the data never posts. When it works, it is destroying them, and giving them away as miners fee. I’m sorry in advance for all the bugs, and for your lost coins. As way of consolation, I’ll use whatever I sweep up off the floor to fix the bugs that caused to you lose them in the first place. This comes with zero warranty except the following: I’ll do my best to not make a total ass-hat of myself. All other illusions about quality should be dispelled now.
I estimate the average cost to be less than 1 FTC.
The fee breakdown is as follows:
15 bytes = 0.00000001 data encode + 0.005 miners fee.That means that for a message with 700 bytes (which approximates to a name, a magnet link, a 100 word description, and 5 keywords), the cost will be 0.94000047. That’s 0.00000047 coins destroyed, 0.94 for miners, and 0 for me. This is of course trivial to verify in the blockchain.
[/quote]why dose it cost so much can’t anyone already embed any message on any blockchain simply by pushing the appropriate tx, each TX has loads of extra space for meta data, no?
-
[quote name=“adamstgbit” post=“47336” timestamp=“1387902828”]
why dose it cost so much can’t anyone already embed any message on any blockchain simply by pushing the appropriate tx, each TX has loads of extra space for meta data, no?
[/quote]It costs so much because otherwise miners don’t seem to want to confirm the transactions.
The cost is strictly variable. The more data you enter, the higher the cost. 700 bytes is a large message in my opinion, but maybe in the future it will be small. I don’t know.
The 47 Satoshi’s in the example is required for 700 bytes. The .94 miners fee is a number I arrived at through trial and error, and equates to 0.05 per output (47 Link addresses + 1 change address = 48 outputs * 0.05 = 0.94).
Yes, you can push the TX yourself. And assuming you’re a miner, you could mine it into the blockchain yourself with zero fee. I’m still contemplating ways to expose that to the user (list of send addresses?), but for this first pass you’ll just have to pay the high miners fee.
I tried it with 0 fee, and I tried it with less fees. The transactions never confirmed. So I’m sticking with what the QT client seems to insist upon, since I’m sure everyone’s using that for their miners, including pools. We can always revisit this.
-
[quote name=“HopeStillFlies” post=“47239” timestamp=“1387855795”]
sounds like the pricepoint sweet spot. For keywords, since there’ll be a logical limit, is there any sort of “keyword cloud” that people can follow or even “click to label” their stuff in order to minimize redundancy?
[/quote]These are all things planed for a future release.
The list is extensive. I just figured it would be better to get SOMETHING out there now, sooner rather than later.
The list goes something like this:
- File attachment support
- Automatic torrent/magnet link generation from files
- Searching across multiple coins
- Publishing across multiple coins
- Live data feeds of freshly published data
- API support (read and write, REST and streaming)
- Auto-suggestion for form fields
- Dedicated search result item page that exposes all meta-data for an item
- Better indexing
- Publish request expiration
- Better date handling
- Tag clouds and category management
- Capatcha and security hardening
- SEO
All of these things are coming next year over the course of the next few months. But I want to get this up and working, and demonstrate viability of the solution itself.
-
Progress report:
Last night I was up until 5 in the morning getting things done.
What’s done:
- First pass at the FAQ and About pages. These will be revisited. I’d like to make a donor list.
- Searching and publishing
- Deep linking
- VPS hosting
- Installing the latest NodeJS, Link, and Link-server
- Installing 0.8.5 Feathercoind
- Syncing blockchain with txindex
- Running the server (it works on the cloud server!)
What remains:
- Purchasing a domain
- Pointing the domain at the server
- Putting the server into production mode
- Sending the link to donators
I expect it to be tonight. :)
-
Holy shiznit Kevlar, I was up at 6am today and you were up - and now you’re up again. That’s crazy!
Let me know if you need anything.
-
[quote name=“Kevlar” post=“47343” timestamp=“1387907550”]
[quote author=adamstgbit link=topic=5663.msg47336#msg47336 date=1387902828]
why dose it cost so much can’t anyone already embed any message on any blockchain simply by pushing the appropriate tx, each TX has loads of extra space for meta data, no?
[/quote]It costs so much because otherwise miners don’t seem to want to confirm the transactions.
The cost is strictly variable. The more data you enter, the higher the cost. 700 bytes is a large message in my opinion, but maybe in the future it will be small. I don’t know.
The 47 Satoshi’s in the example is required for 700 bytes. The .94 miners fee is a number I arrived at through trial and error, and equates to 0.05 per output (47 Link addresses + 1 change address = 48 outputs * 0.05 = 0.94).
Yes, you can push the TX yourself. And assuming you’re a miner, you could mine it into the blockchain yourself with zero fee. I’m still contemplating ways to expose that to the user (list of send addresses?), but for this first pass you’ll just have to pay the high miners fee.
I tried it with 0 fee, and I tried it with less fees. The transactions never confirmed. So I’m sticking with what the QT client seems to insist upon, since I’m sure everyone’s using that for their miners, including pools. We can always revisit this.
[/quote]I see.
thank you -
Wow am i ever late to the party.
Great work on this innovative feature for the blockchain. Many exciting possibilities for this in the future.
I will be donating 100 FTC to your cause now. May the progress continue at such a swift pace without burnout :)
-
[quote name=“Kevlar” post=“47223” timestamp=“1387849065”]
Tonight I am:[list]
[*]Writing the FAQ and about page, and fixing the footer
[*]Registering a domain
[*]Firing up a host in the cloud
[*]Installing Feathercoind, Link, and Link-server on it
[*]Testing everything in “production mode”
[/list]We’re that close. I’ve been testing all the main functionality and site navigation. It’s solid enough for beta-testing. Mnstrcck said he’d probablly get to the Search Result CSS tonight as well.
If all goes as planned, you donators may be receiving a link in your inbox to Link sometime tomorrow… Although Tuck will get his before anyone else. Watch this space for more…
[/quote]I have take a look in the code.
Please don’t forget about sanitation of the output because it could easy become hive of xss attacks embedding browser plugins that will look for Tor publishers/clients or easy victims for botnet.
-
Oooooohhhh good call! I hadn’t considered that, but it’s entirely possible to really use this to screw stuff up. Should be easy enough. I’ll find a good regexp tonight before I send out the links to the testers.
Ninja edit: Looks like this will do the job I want. https://github.com/theSmaw/Caja-HTML-Sanitizer
-
[quote name=“Kevlar” post=“47388” timestamp=“1387919703”]
Oooooohhhh good call! I hadn’t considered that, but it’s entirely possible to really use this to screw stuff up. Should be easy enough. I’ll find a good regexp tonight before I send out the links to the testers.
[/quote]Go, Kevlar go! There are a lot of checks to be performed but for start html encoding / stripping slashes / tags and preventing words like “bitcoin: string”, "feathercoin: " will do the job (Phishing is also in game) ! Told you I’m on Link now :)