Akiba-Online Tracker [down]

[snakeboy]

Originally Posted by chompy

We toyed with the idea of a private integrated torrent tracker in the early days, but it didn't work out.

I've been thinking about revisiting the idea, but I'm not sure yet.

Well, that is good news. It's hope at least... vBulletin (although BY FAR the best forum software around) has been somewhat neglected by developers of integrated trackers until quite recently. There are a few available these days that are actually pretty good.

Most of the private trackers I am on are custom-built, as is the forum software that integrates with it. Unfortunately way beyond my coding skills to even conceive of something like that, but I certainly do have lots of great feature suggestions to someone with the skills who would listen.

I've long been whining about the hit and run factor here, and would LOVE to see the community trimmed of leeches in favor of the sharer's. Those who are pulling their weight already will be uneffected by maintaining a ratio. Only the abusers will suffer. Great! More bandwidth to share amongst contributors that way. Makes rewarding the contributors really easy and keeping the leeches still participating, but not reaping the benefits that sharing users can enjoy.

If I can ever be of any assitance, you know where to find me.

[chompy]

Originally Posted by snakeboy

If I can ever be of any assitance, you know where to find me.

Okey dokey.

[Rollyco]

I don't recommend a private tracker if your webserver is already having performance issues (which I assume it is, based on the frequent downtimes.) Personally I'm fine with no a-o tracker, even a public one. There are plenty of public trackers out there already.

[banderaz]

This post is very useful, I'll surely add it to my seeding

[Summer-Time-Fun]

Originally Posted by snakeboy

I've long been whining about the hit and run factor here, and would LOVE to see the community trimmed of leeches in favor of the sharer's. Those who are pulling their weight already will be uneffected by maintaining a ratio. Only the abusers will suffer. Great! More bandwidth to share amongst contributors that way. Makes rewarding the contributors really easy and keeping the leeches still participating, but not reaping the benefits that sharing users can enjoy.

This is not even a question of if, It needs to be done to deter hit and run cheaters. As far as running a private tracker, I would think that the AO server performance issue would not be as much a concern if such a ratio (as mentioned above) was in place.
I have no expertise on this technical stuff, I’m simply backing Snakeboy on this matter because his concerns are, with out a doubt, in the right direction! Clean house!

I don’t mean to sound heartless, but a hit and run is a crime.

[snakeboy]

Originally Posted by Summer-Time-Fun

This is not even a question of if, It needs to be done to deter hit and run cheaters. As far as running a private tracker, I would think that the AO server performance issue would not be as much a concern if such a ratio (as mentioned above) was in place.

Regarding the server performance issues - yes, it's no lie that a tracker can tax a server, BUT given the the tracker is currently open and handling many MANY queries from all the clients out there that have a torrent where the Akiba tracker is listed (and that is probably in the hundreds of thousands by now), it would be logical to assume that if the tracker was made private, it would no longer need to respond to the requests of all these rogue clients trying to do a scrape every few seconds.
It would ONLY need to respond to "legit" requests from registered uses, so in fact the server's performance would probably not suffer at all, but actually improve as a result of closing the tracker.

The tracker would have "less work" to do is the bottom line...

Aside from performance issues (moot point) I have yet to see a compelling argument that supports the fact that the tacker should be public. I mean where is the downside (for people who actually seed) if it's closed?

The ONLY thing I can think of as a "benefit" to having an open tracker is that it allows the torrent to be much more widely distributed on other boards (and we know already this is happening), and it could be argued that the large amounts of users jumping on the torrent early is a good thing. But since they all disconnect when they hit 100%, what is the real benefit? Aside from possibly having good speeds early while initial seeding is still underway and they have no choice but to stay in on the torrent until it completes during which time they're seeding to other peers.

The argument against a closed tracker would mean torrents would possibly end up being slower overall, but will continue to remain seeded for LONG periods of time allowing everyone to enjoy them. And isn't that the purpose at the end of the day? Rather than a week or two of fair speeds followed by months of re-seed and DDL requests...?

It seems to me that the majority of users who advocate the public tracker concept do so out of fear of actually HAVING to seed and pull their weight
if it should get implemented, as that would mean the end of their hit-and-run days.

I have yet to see other legitimate objection or valid argument against private tracking.

[guy]

Argument against hosting a private tracker?

The majority of initial seeders from AO simply wouldn't use it.

For the purposes of a private tracker (preventing hit and run), it is effective only when two criteria are met:

1. there are a lot of peers registered on the tracker who would download the file, because it will still suffer if there are fewer peers
2. it is the only tracker included in the torrent, because allowing it to be tracked elsewhere (even with Peer Exchange) defeats the purpose of the private tracker

The reason private tracking works well at sites like Demonoid and AsianDVDClub is because most torrents see a huge number of peers within the start of a torrent (and more seeds = it takes longer for a peer to reach 1.0 = peers will stay active on the torrent for a long time), and both sites see a large member turnover (old users deactivated, new users joining = new users are continuously joining the same torrent).

For AO, it will be very difficult to create the same circumstances. Most torrents that are initially seeded here get <500 peers during initial seeding, and that includes the torrent being redistributed. If you privatize it, you will see the swarm drop in size significantly.



To better explain, put yourself in the shoes of an average peer. Most of our ISPs give us large download bandwidth but comparatively small upload bandwidth. If you join a large swarm, you have lots of peers to download from (= fast download), but few peers will download from you since they too have lots of other peers (= slow upload). Fast download + slow upload = very low ratio, and in a private tracker (with a 1.0 ratio obligation) that means you will help keep the torrent alive very long -- great news for people who join late.

On the other hand, if you join a small swarm, you have fewer peers to download from (= slower download), and a higher percentage of peers will download from you (=faster upload). So while it might take you a little bit longer to finish downloading, it takes you less time overall to reach your 1.0 obligation. Assuming the majority of peers will drop out after seeding to 1.0, the torrent will diminish quickly -- bad news for people who join late, and if there is no user turnover, it's especially bad news for initial seeders who then have to work harder for the latecomer's sake.



Again, it works for the bigger private trackers because they can provide a large number of initial peers and also because there's constant user turnover. But U15 just doesn't have that many peers to begin with (honestly, our swarms are tiny compared to music, TV shows, and movie torrents): allowing people to hit-and-run isn't exactly a "benefit", but restricting anyone from hit-and-run would more than likely hurt the torrent more than it helps.

As an initial seeder, I'd rather have 1000 peers seed to just 0.2 than 100 peers who seed to 1.0. Think about it: if we only want to share files with other "contributers", that'd pretty much be about a dozen of us, and that would actually mean there's less bandwidth for us to share! (We'd each get a greater percentage of bandwidth, but the actual bandwidth would be much smaller.) If that's the case, I'd rather just upload to Rapidshare and be done with it.



So yeah, private tracking sounds like an ideal solution. But the argument against it is that it simply wouldn't make much sense for our situation. I'd much rather have no tracker if it means we can speed up the forum's load times, and then use that performance advantage to increase popularity and get more peers on the torrents.

[snakeboy]

These are all valid points that you've made, but also one's that have been addressed by the creators of various (private) tracker options. Ever heard of "seed points"? A system whereby you can achieve your seed-back ratio on torrents that have no peers, it works something like this:

Joining a torrent late (that has a fair amount of seeds, which it will have if the concept is employed appropriately) will mean that once you have your copy of the file, if nobody else downloads, you cannot achieve your 1:0 ratio. You may sit seeding that torrent for months in order to get to reach a 1: ratio and that IS ridiculous and unreasonable I agree. Seed points will firstly set an obligatory seed time that needs to be met. Say 5 days. Once you have been seeding for 5 days (which is easily monitored) you will start to accumulate seed points at a set rate, and these will cause your upload ratio to be "faked-up" until you reach 1:0. Depending how it's configured, this may mean being active on the torrent for an additional 3 days or so, at which time the system lists your ratio on the torrent as 1:0 and you're done, you'be pulled your weight and can move on. In reality you may not even have uploaded anything to anyone during this time, but still you will achieve 1:0. If you uploaded 20meg (and it's an ISO you're seeding), what you have "actually" uploaded will simply be added to your 1:0 ratio when you achieve it, so your ratio will actually be more than 1:0 if you've uploaded ANYTHING at all. Of course you can still stay on the torrent and accumulate more upload ratio as people download it, but you can also stop it having done your part to ensure longevity of the torrent. It's a good system that addresses the "I came late, I cannot seed back to 1:0 in this lifetime" arguement.

I agree that the amount of peers/seeds will be drastically CUT when applying private tracking, but again, this is not all bad. 50 seeds 0 peers is a better torrent to join than 5 seeds 300 peers, ALL of who are throttling upload because there is no consequence for doing so or their upload bandwidth is totally overtaxed because they are not managing it correctly. I've seen sites using this system or similar techniques based on "active time" or "last announce" time benefit simply because the people seeding are seeding "to the max" and not simply cruising through the torrent forums and dropping every single torrent into their client and clicking "force start" thinking they are beating the system by having uTorrent download 10 torrents simultaneously regardless of their bandwidth when they should be downloading only 2 or 3 at the same time to be efficient (in uploading). They just see data coming in and are simply not concerned about what's going out. MAKING people aware of what's going out will force them to download with moderation which actually ends up making things FASTER, not slower. Because the people in the swarm actually give a shit about their outgoing bandwidth and configure so that it benefits the swarm, and therefore them in the long run.

It's also true that Akiba does not have gargantuan numbers on u15 torrents, but I don't believe that's the main reason for poor performance. It's because those peers/seeds that ARE participating simply don't (need to) care about giving data back. I've seen publically tracked torrents around the 2000 seeds mark give DISMAL download speed (around 15k/sec) versus private torrents of the same file with around 30 seeds and it FLIES down around 2meg/sec!) It isn't just about the numbers on the torrent, it's about what bandwidth those numbers have and how they are (being forced to) apply it.

Unfortunately this mentality exists that poeople seem to think they're somehow "entitled" to an instantly fast download to completion. And while that's a nice ideal to pursue in la-la-land, it's not a reasonable expectation in reality. ESPECIALLY given the fact that every torrent HAS to be initial seeded, so all users WILL be at the mercy of that lone users bandwidth at SOME point. But once seeding takes place and downloading begins people want an ISO to be on their drive within hours, which is possible in well populated torrents. But I would rather an ISO take a week to arrive on my drive, but know that it will be available for others to get for months into the future. Rather than the selfish mentality of "I have to get it down now, as fast as I can and as long as I have it screw the others, not my problem" .

Briefly well populated torrents vs Long Lasting torrents at the price of having to wait a few days for the file instead of a few hours... I'm for the latter. I have patience, and while a fast download is always nice, I am not at all bothered if there are 2 seeds and I have to wait a week but I know that it will still be available for those who arrive a month or two from now. To quote Blade Runner: "The light that burns twice as bright burns for half as long." And with the idea of torrents and "community", that's not a good thing.

The community that buids around the concept of equal sharing and joint contributiont as opposed to a "free-for-all-greedfest" tends to be a lot closer and people gladly help each other out by contributing to others they know are also pulling their weight. People feel better about firing up a re-seed if someone asks because not only does it help their ratio, it helps their conscience that they are accountable to for a change.

How many re-seed requests do you think the hit-and-runners are prepared to fill? imho we can do without people like that, even if it means loosing the tiny benefit of having those leeches on the torrent during initial seeding. Since they're gone shortly afterward I don't see that the benefit of having them around outweighs the benefit of NOT having them around.

[chippy]

I don't want to dilute this very cogent discussion with a mere blip of commentary, but I feel compelled to offer the perspective of one who has been guilty of more than his comfortable share of hit-and-run offenses.
I have a modicum of big-screen movie torrents that I've downloaded, largely because my geographical location makes it prohibitively difficult to enjoy cinematic releases. I'm late on the bandwagon of anything good that hits the theatres, but I don't mind much because BitTorrent still allows me to eventually see films I'd otherwise miss. I go to ISOHunt and commit an embarrassing amount of hit-and-run offenses: seeders in the thousands on what I download ease my conscience...mostly.
When it comes to IV, however, AO is my only source: it's a civil (usually) community that succeeds in providing the very best of what I love so dearly--cuteness that makes your heart melt. AO means a great deal to me, and so does the cavalcade of cute that is presented in the U15 forum. There are times when I'm downloading movies and IV videos at the same time. As patient as I try to be, and despite the fact that I always give AO torrents precedence as far as seeding goes, I do commit the crime of hit-and-run more than I'd like to admit.

In my opinion, the number of times I've seeded to 2:1 or more for AO-U15 torrents matters not at all if I've committed even one hit-and-run. That is my personal opinion...of myself.

What I'm getting at, though, is that (from the perspective of a dedicated fan of AO and U15 videos,) I would submit to a ratio system, a "points" system, or whatever system the major contributors of the entertainment I so love decided to put into place. If I ever felt comfortable enough with my skills at ripping (or found access to the kind of material posted) I'd love to join that elite group as well, but in the meantime I would be perfectly willing to stretch my patience, donate my bandwidth, and comply with a quota system in order to continue to enjoy the very best of U15-IV entertainment.

I am posting this so that you all will take into consideration the willingness to "change one's wicked ways" many (I hope!) of us can promise. You'd be justified in saying we should already be exhibiting more of a sharing behaviour in the first place, but give us a chance to prove our dedication to AO, and many of us will. Tell the community to put its bandwidth where its mouth is, so to speak. Thanks for reading my post: I hope it enlightens.

[chippy]

I don't want to dilute this very cogent discussion with a mere blip of commentary, but I feel compelled to offer the perspective of one who has been guilty of more than his comfortable share of hit-and-run offenses.
I have a modicum of big-screen movie torrents that I've downloaded, largely because my geographical location makes it prohibitively difficult to enjoy cinematic releases. I'm late on the bandwagon of anything good that hits the theatres, but I don't mind much because BitTorrent still allows me to eventually see films I'd otherwise miss. I go to ISOHunt and commit an embarrassing amount of hit-and-run offenses: seeders in the thousands on what I download ease my conscience...mostly.
When it comes to IV, however, AO is my only source: it's a civil (usually) community that succeeds in providing the very best of what I love so dearly--cuteness that makes your heart melt. AO means a great deal to me, and so does the cavalcade of cute that is presented in the U15 forum. There are times when I'm downloading movies and IV videos at the same time. As patient as I try to be, and despite the fact that I always give AO torrents precedence as far as seeding goes, I do commit the crime of hit-and-run more than I'd like to admit.

In my opinion, the number of times I've seeded to 2:1 or more for AO-U15 torrents matters not at all if I've committed even one hit-and-run. That is my personal opinion...of myself.

What I'm getting at, though, is that (from the perspective of a dedicated fan of AO and U15 videos,) I would submit to a ratio system, a "points" system, or whatever system the major contributors of the entertainment I so love decided to put into place. If I ever felt comfortable enough with my skills at ripping (or found access to the kind of material posted) I'd love to join that elite group as well, but in the meantime I would be perfectly willing to stretch my patience, donate my bandwidth, and comply with a quota system in order to continue to enjoy the very best of U15-IV entertainment.

I am posting this so that you all will take into consideration the willingness to "change one's wicked ways" many (I hope!) of us can promise. You'd be justified in saying we should already be exhibiting more of a sharing behaviour in the first place, but give us a chance to prove our dedication to AO, and many of us will. Tell the community to put its bandwidth where its mouth is, so to speak. Thanks for reading my post: I hope it enlightens.