I hadn't actually realized you'd released another binary.

What happened to "Oh and if there are any descrepencies between the options in the Linux version and the other two versions, they will definitely be fixed before I upload any more updates." ?

Understand it from our perspective:

October 8: Release 0.9.1
October 9: sr55 discovers a bug in how "Fast" deinterlacing is set..
October 9: sr55 fixes a bug in how "Fast" deinterlacing is set and refreshes his binary release.
October 13: You release your alpha.
October 24: I point out to you that you're still using the wrong deinterlacer, that was fixed before you released anything.
October 29: You assure us you'll fix it before the next release.
November 9: You release with the bug not fixed and call it 1.0.1, as if it's stable.

This does not exactly build trust.

As far as your users are concerned, you're part of HandBrake. We get people wanting to know when your wrapper will be available from the m0k site. You've asked for and received a forum section. You use the HandBrake icon. When your application does not work correctly, it reflects badly on HandBrake. A regular user isn't going to understand why the WinGUI deinterlacing runs faster than the LinGUI, they'll just assume the HandBrake team doesn't care about Linux enough to make it work fast. They're not going to understand why the Classic preset in OS X is so much faster than the Classic preset in Linux. They'll just assume the HandBrake team doesn't care to optimize for Linux users. It makes all of us look bad, hurting the reputation of a project with hundreds of thousands of users.

Do you *want* the HandBrake team to have your code? We've been working under the assumption you don't. Generally people who want to contribute to the team follow the advice on the Development page on our website and the associated FAQ. They get in touch with team members, come on IRC, contribute code, and ask for svn access. Because you decided to run your own show over at SourceForge without discussing it with any of us, we've all been thinking that you want to work by yourself as a fork. You blindsided everyone by working like that -- when someone sets up their own site infrastructure, it's a bit of a slap in the face: a very direct way of saying "I do not want to work with you." It's your code. Do with it what you want. But when you copy stuff wholesale from the project, at least mention it in a comment. For example, where HandBrake lifts material from MPlayer or mp4creator or whatever, it's noted next to the code. Just common courtesy.

It's really about two things: having faith that bugs can be fixed quickly, and being able to do support in a uniform manner.

I set "Fast" deinterlacing to -1 in the MacGUI on July 27th as part of the initial check-in of huevos_rancheros' filters. Scott typo'd it and had it wrong for over two months. Now it's almost been four months, and it's still not right in Linux. That is some serious lag. Do you understand what this does to core developers? It means that every time someone is considering modifying something, they have to stop and ask themselves: "But it could be over a hundred days before this change is reflected everywhere." Or: "What if this is wrong? What if I have to fix it in two weeks? Am I sure it will be fixed in all parts of the project?" That kind of thinking creates resistance to the very notion of making changes. But as an open source project, HandBrake has to encourage people to make changes. This requires an agile project, with developers who are on the ball, follow the timeline, and make necessary modifications as soon as possible. That's why communication is important -- why the team hangs out on IRC and reads the development forum. Developers have to have faith that when a change is discussed on irc and the forum, everyone else is picking up on it. That when a fix appears on the Trac timeline, everyone else who needs to know is reading it.

The other thing is support. HandBrake is too small to have a throng of well-meaning, knowledgeable users to act as volunteer support staff. We have no firewall separating the devs from the users. Developers interact with them directly, giving tech support themselves. On the other hand, HandBrake is just big enough to have a constant stream of users asking for support, and is easy enough to use that that most of them don't know how to accurately describe their problems. Often, users don't even bother mentioning what platform they're on or what version they're using. As a result, when helping this people, everyone has to assume that each interface works exactly the same -- to the extent this is technically feasible.

Listen. In the end, you're making a GUI for Linux and no one else is. You've stepped up to a challenge other people were ignoring. So no matter what I or anyone else says, you're performing an extremely valuable service and filling a void. It is a great thing for Linux users to have an easy way to run HandBrake. One of the beautiful parts of open source is you can ignore everything I'm saying. There's nothing wrong with you running your own show on SourceForge, or even doing a whole fork off of 0.9.1. But I think it'd be better for everyone if there was coordination, communication, and collaboration.