I was one of the early adopters of µTorrent, having started used the first public beta 1.1 all the way back in 2005. When Bittorrent acquired it in 2006 I was concerned but decided to stick with it as it was still the best Bittorrent client available.

I even tolerated the ads added in the August 2012 build of µTorrent which I needed to update to thanks to the write cache bug uncovered by Pleasuredome users (of which I am one). However, recent bullshit has prompted me to look elsewhere posthaste.

After evaluating my options (namely Halite which I’ve looked at before, Deluge, Tixati, qBittorrent, and Transmission), I decided to set myself a few criteria:

  1. must be open source
  2. updated regularly

1. eliminated Tixati unfortunately, which I like the look of. 2. did away with Halite, which I really wanted to succeed, but it seems a bit of a dormant project.

I was down to 3: Deluge, qBittorrent and Transmission. I tried all three, but they didn’t really suit my workflow. And then, the client-daemon architecture of Transmission to the rescue. I found Transmission Remote GUI, which I found suits me just fine (i.e. it works like old µTorrent!) and now I have my new BT client.

Transmission BT

So, I guess this is goodbye µT. It has been great, thanks.

I still see this bandied about as a possible factor in slow P2P speeds. It is easily misunderstood, but the half open connections limit imposed by XP SP2 and Vista does not matter for P2P speeds. It limits the rate at which connections can be made, not the total number of simultaneous connections.

Now, 10 half open connections seem a little low, but most ordinary users do not need more than this. Even heavy P2P users such as myself.

I usually run eMule and Bittorrent at the same time and I do not usually partake in very popular torrents (my usual haunts: Pleasuredome and DIME; what would I do without you guys?) but I usually have no problem maxing out the bandwidth available to my 3 Mbps connection. Sure, the clients spin up a little slower than before but how does it matter? I wasted maybe two or three minutes? The other negative side effect would be the occasional 4226 error in the Event Log, but I don’t care about that.

I am not denouncing LvlLord’s patch as evil as I really do not care if everybody installs it, but it won’t increase your P2P speeds. It won’t make your computer run better. It won’t solve the global poverty problem and it definitely won’t make you more popular with the gents/ladies.

Woo! No protocol encryption yet, being dependent on libtorrent’s implementation, which I hope is underway. Can’t find a changelog, but 0.29 has got to be 0.01 better than 0.28, right?

Get it from the official Halite site, as usual.

New version of Halite released with a snazzy new icon. Changes from 0.27 to 0.28:

  1. IP filtering support eMule style ipfilter.dat files.
  2. DHT support, thought it is turned off by default.
  3. Can select alternate save directory for torrents.

Long time readers of my blog (yes, all two of you!) will know that since BitTorrent Inc announced the acquisition of µTorrent, I have been keeping an eye on the development of Halite, and the library it was based on, libtorrent.

The major stumbling block to more widespread adoption is, I believe, the lack of support for protocol encryption (PE) in libtorrent. Well the bad news is there is still no support. There is good news though. Eóin, author of Halite, is considering implementation of PE and contibuting it to libtorrent. Most excellent. Pity that I am no C++ hacker or I would have given it a shot myself.

I cannot recommend replacing µTorrent with Halite right now, but once PE is implemented, add features and polish from a few more iterations and I should be all over it.

Announced a few days ago. I am not optimistic about the future direction of uTorrent, given that Bittorrent, Inc has a deal with Warner Brothers (the enemy!). While I will continue to use the current version of uTorrent, I am already looking out for alternatives.

BitComet is definitely out, as it is still banned on a number of trackers due to it being considered harmful to the network, I am also not comfortable with the interface. BitLord, no, just no. Azureus is just too heavy for my poor little machine. ABC, BitTornado and the bevy of clients implemented in Python, lacking on some crucial features. Halite seems promising, but the libtorrent library it is based on lacks some features I need.

Since libtorrent has gained a higher profile thanks to the uTorrent situation, and I am hoping this will spur someone to get some heavy development work done to fill the gaps in its feature list, such as protocol encryption.

Introducing µTorrent. It is a pretty young client, and already it is gaining a groundswell of rabid users. I have been using it for a while and, barring an unfortunate tracker stat incident with an earlier version, am very pleased with it.

Why?

  • The size of the executable is under 100 KB
  • CPU and memory utilisation are minimal, as in just about insignificant
  • Very fast iterations for bug fixes/new features (though this means users are all essentially beta testers)
  • Enough features for most users

The only negative I can think of is that it is not open source, but probably a moot point given the amazing work done so far.