krazyson Posted February 19, 2006 Report Share Posted February 19, 2006 i have been using the 1.4 stable and i havent got that much bad data ( hashfails). well now i updated it to the 1.4.2 beta 426 and now sudanly i get a lot of bad data (112.6 mb ) how do i fix this http://img75.imageshack.us/img75/2381/14228yf.jpg Link to comment Share on other sites More sharing options...
jlaw84 Posted February 19, 2006 Report Share Posted February 19, 2006 I dont know if its just me, but when I was using 1.4.2 beta 225, I was using UPnP Port Mapping and everything was working good. Now that I updated to 426 I get a NAT error. Is this just me or has anyone else noticed this? Link to comment Share on other sites More sharing options...
avatarl Posted February 19, 2006 Report Share Posted February 19, 2006 Strangely enough, I'm getting many offline tracker reports with this and the previous beta version. Anyone has the same experience or is it just me? Link to comment Share on other sites More sharing options...
rafi Posted February 19, 2006 Report Share Posted February 19, 2006 rafi: that's been asked for a lot, and been shot down repeatedly That was a theoritical question and for Martin... It was not a request (yet...) . A second opinion will not heart anyone, right ? It's an open forum... Link to comment Share on other sites More sharing options...
Ultima Posted February 19, 2006 Report Share Posted February 19, 2006 Yes, it's possible. But it hurts the swarm -- the file(s) don't get distributed as widely as it possibly can, and it ends up becoming closer to HTTP transfer, thus losing most of the BitTorrent advantages. Link to comment Share on other sites More sharing options...
Technarch Posted February 19, 2006 Report Share Posted February 19, 2006 I'm in a torrent and I see a client that says:[FAKE] µTorrent/142BReally?? Link to comment Share on other sites More sharing options...
Ultima Posted February 19, 2006 Report Share Posted February 19, 2006 Yeah... there are people who fake their client ID... Link to comment Share on other sites More sharing options...
dAbReAkA Posted February 19, 2006 Report Share Posted February 19, 2006 yeah Fake Bitspirit and [FAKE] µTorrent/xxx is commonly a seen thing Link to comment Share on other sites More sharing options...
Nefarious Posted February 19, 2006 Report Share Posted February 19, 2006 why would anyone fake bitspirit anyway Link to comment Share on other sites More sharing options...
Martin Levac Posted February 19, 2006 Report Share Posted February 19, 2006 Rafi, as far as I know based on my observations, uTorrent has the following capabilities using these two settings bt.prio_first_last_piece/bt.always_prio_rare:1. Request first/last + 50% availability pieces2. Request 50% availability pieces3. Request first/last + rarest only (not sure because bt.always_prio_rare may override bt.prio_first_last_piece)4. Request rarest only (that's what I use)Not sure about the following but I think that regardless of settings, it will always request the very first piece on 50% availability, once that piece is complete, settings kick in as normal. Also based on my observations, it appears that first/last will request 3/4 first/last pieces.uTorrent also has the ability to request individual files. It does not have the ability to select individual pieces. The closest you can come to is select the file you wish to DL, unselect the files you don't want, while both settings are set to first/last + no rare. Link to comment Share on other sites More sharing options...
canuckerfan Posted February 19, 2006 Report Share Posted February 19, 2006 I've also had 34 hasfails on a torrent running for about a full day. Quite weird. Link to comment Share on other sites More sharing options...
rafi Posted February 19, 2006 Report Share Posted February 19, 2006 Thanks Martin. I was aiming at a case of a single file (avi). Do you think that such a feature ("prioritize pieces closer to start-of-file", IF added to uT) will cause degradation in performance ( = DL speed) ? Link to comment Share on other sites More sharing options...
Martin Levac Posted February 19, 2006 Report Share Posted February 19, 2006 Rafi, any function that gives priority to one piece over another except rarest will cause performance to drop, by how much depends on the extent of the function.Let's put this in perspective.Let's say I ask only for pieces 1-10 when there are 100 pieces. Because I only ask for pieces 1-10, I can only ask them from the limited number of peers that have those pieces. Also, because I will only have pieces 1-10, I can only trade them with peers that don't have them, also a limited number of peers. In short, since I only want 10% of the torrent, I can only trade with 10% of peers, perhaps even less than that. So yes, it will cause a degradation in performance.I understand what situation would call for such a function. A torrent where the content is a single file that can be sampled by playing only the beginning. I consider this type of torrent to be very inefficient because it doesn't include a short sample file that you can select individually from the complete file(s). I prefer to get torrents that include a short sample file and where the complete file(s) is compressed in multiple RAR files with an SFV check file. I suggest you look for a better source of torrents where the uploaders understand this. Link to comment Share on other sites More sharing options...
canuckerfan Posted February 19, 2006 Report Share Posted February 19, 2006 427 is out. looks like the hashfail bug is fixed. good stuff. Link to comment Share on other sites More sharing options...
Firon Posted February 19, 2006 Author Report Share Posted February 19, 2006 If you're having all trackers offline issues (or not being able to connect to most peers), remember to check your TCPIP.sys and see if it's been unpatched! Link to comment Share on other sites More sharing options...
geezer Posted February 19, 2006 Report Share Posted February 19, 2006 remember to check your TCPIP.sys and see if it's been unpatched!Potentially idiotic question (sorry!):Does this only apply to people who have overriden the default number of half-open (pending) connections in µTorrent? I thought the default prevented µTorrent from hitting the max of 10, making it unnecessary to patch. Is LvlLord's patch optional or mandatory when using µTorrent's default configuration? Link to comment Share on other sites More sharing options...
Firon Posted February 19, 2006 Author Report Share Posted February 19, 2006 Well, the problem is that if you have lots of torrents (even stopped), those take up half-open connections when starting up, and can cause the trackers to timeout Link to comment Share on other sites More sharing options...
canuckerfan Posted February 19, 2006 Report Share Posted February 19, 2006 Im guessing this release won't be labelled 1.4.1, right? There's been a bucket-load of changes, and it deserves at least a 1.5 or 1.6 marker. Link to comment Share on other sites More sharing options...
Firon Posted February 19, 2006 Author Report Share Posted February 19, 2006 Most likely 1.5 Link to comment Share on other sites More sharing options...
rafi Posted February 19, 2006 Report Share Posted February 19, 2006 How about 2.4 ? Az compatible... Link to comment Share on other sites More sharing options...
rseiler Posted February 19, 2006 Report Share Posted February 19, 2006 Rafi, as far as I know based on my observations, uTorrent has the following capabilities using these two settings bt.prio_first_last_piece/bt.always_prio_rare:1. Request first/last + 50% availability pieces2. Request 50% availability pieces3. Request first/last + rarest only (not sure because bt.always_prio_rare may override bt.prio_first_last_piece)4. Request rarest only (that's what I use)Should those two settings be used in tandem (assuming you're using one or the other)? Link to comment Share on other sites More sharing options...
Nefarious Posted February 19, 2006 Report Share Posted February 19, 2006 How about 2.4 ? Az compatible...errr... please dont Link to comment Share on other sites More sharing options...
TheDude Posted February 19, 2006 Report Share Posted February 19, 2006 Question to Martin Levac :Are you saying that "always_prioritize_rare" will not decrease speeds in normal, healthy swarms?And thus should be on by default for everyone ? Link to comment Share on other sites More sharing options...
shadek Posted February 19, 2006 Report Share Posted February 19, 2006 Question to Martin Levac :Are you saying that "always_prioritize_rare" will not decrease speeds in normal, healthy swarms?And thus should be on by default for everyone ?I have experienced the reversed effect; speeds getting lower than normal. I really believe this option should be used when there are a ratio of less than 1:20 (seeds:peers). Link to comment Share on other sites More sharing options...
Martin Levac Posted February 19, 2006 Report Share Posted February 19, 2006 TheDude, that's what I'm saying: http://www.bittorrent.com/bittorrentecon.pdf sections 2.4.2/2.4.3. In fact, it will produce normal healthy swarms. Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.