Try streaming a DivX or Xvid torrent with the DivX browser extension .coolies

I wonder If Travis has had lunch with Jordan Greenhall lately and if not he should

Hi KnowBuddy, I’ll let Tom handle the detailed response but thanks for the excellent feedback.

As for (4), I’m curious to hear more. Do you need to have these options presented at install time (which we’re loathe to interrupt) or would it be ok if we just provided reasonable defaults and let you change them via the control panel?

Also, in (9), what specific status would you like? Basically, we agonize over the balance between giving “just enough” feedback without giving “too much”. With this in mind, rather than give separate STUN and UPnP status, what if we gave a basic red/yellow/green “connection quality” indicator (and perhaps when you mouseover/click it’d give instructions on what the color means and how to improve it)?

Garrr. I had another huge reply and I somehow managed to toast it. Oh well. Here’s attempt #2.

So you press the Stream button and nothing happens?

Yep. The machine sits there like a stump.

Does the button work after the file has completed downloading, when the button says “Play”?

Dunno. Good question. I’ll try it.

If that doesn’t work, can you tell us if just double-clicking on the downloaded file in Windows plays it?

It definitely does, but I haven’t reimaged my home computer in a while, so I imagine the file associations are a bit crufty. I’m also using MPC and QT Alternative instead of WMP and the real QT, so that might also account for some strangeness.

Would displaying the port that FoxTorrent uses (9420) on the WebUI be sufficient

As long as it is constant and is represented as such. Maybe some hover text that says “RS/FT always uses port 9420 to talk to other people - you may need to set your firewall to allow traffic on this port”. I know that’s a bit verbose …

In what situations do you need to change the port that it uses?

I have multiple layers of firewalls in addition to outbound traffic monitoring. Any unconfigured applications trying to access the ‘net make red lights flash and a large robot wave its arms and shout at me. (Figuratively.) But, to answer your question:

1. Port scans target application vulnerabilities based on port numbers. If RS is always on port 9420 and (I hope not, but) if a buffer overflow or whatnot is discovered, having the app running on the default port makes it very easy to target. Even if there’s nothing wrong with the code, DOS attacks are easier to target if the port is always the same, etc. I change the ports on all of the applications I can because of this. It sounds like paranoia, but I’ve been the victim of enough targeted port scans and brute-force attacks that I know it works.
2. Some hardware routers have a limited number of port forwarding options available. Allowing users to aggregate ports into ranges makes it easier on people with dumb hardware.
3. Not that I think it’ll happen any time soon, but I don’t need my braindead ISP waking up tomorrow and deciding that all RS/FT traffic should be denied and locking down that port. They do it on enough ports already.

Do you need to have these options presented at install time (which we’re loathe to interrupt) or would it be ok if we just provided reasonable defaults and let you change them via the control panel?

I think the defaults are fine, and only serious edge cases like myself are ever going to want to change them. Hiding them away in the Control Panel or a config file would be fine. I agree that the seamless install is vital.

The planned use-case was [..] you either press Back and check on the download later, or you click Save As (to bring the download into Firefox’s Download Manager), and then press Back.

While you and I know that RS/FT will continue downloading in the background, my grandmother is trained to not close or navigate away from any page that has a progress indicator that isn’t at 100%. As are many of my non-tech-savvy coworkers. (Many of whom can’t even figure out how to open a new tab.) It’d be best if RS/FT links just magically worked inside the Download Manager, as they know what that is, but I know that’s serious XUL/XPI mojo.

Think the progress meter is too big?

Not so much. I think my primary problem was that it was a single progress meter at the top of an otherwise-unused page. It just seemed like a hella waste of space.

Are the double set of [peer/swarm] numbers valuable, or would the simplicity of one set be an acceptable step in this direction?

That’s a very good question. I know what I use them for, but I have to wonder what the average user thinks when they see them. Maybe if you just made it specifically clear that we are looking at the swarm size or the neighbor count? How about a tooltip over the one number that then provides a breakdown into the two? Or a “Show Details” twisty that shows a bunch more info? I do not envy the dumb-user versus guru-user decisions that you folks have to make.

we could place a “Current internet connection is limited” warning link (linking to an explanation/help page) on the WebUI if the port hasn’t been forwarded within a reasonable amount of time.

That works. I’m sure you know, but uTorrent uses a similar method and it works well. I think the average user couldn’t care less about STUN or UPnP. But again, I’d love to be able to drill down and see this type of information - even simple stuff like “Not Needed/Connecting/Negotiating/Successful” indicators for the STUN subsystem.

Have you guys ever used Shareaza? It’s a good example of a case where the default UI is very, very basic, but there’s a ton of information that can be drilled down into for the power users.

PS. Do you mind if we post this feedback on our FoxTorrent mailing list?

I know this is now a week old, but I don’t mind at all.

-R