I found Eclipse very quickly. It is very nice, and runs circles around Xcode in terms of usability. It subscribes very heavily to the windows UI design, which I despise, but it works, which is more than I can say for Xcode. Just to summarize the things that Xcode does wrong:

• The code sense completion fails all the time
• Stopping the debugger never kills the java process (which must then be killed with a -9, causing sockets to not close nicely)
• The compiler error messages are cryptic at best (such that if I didn’t know exactly what I was doing in Java, I’d never figure it out)
• The debugger cannot find any source files unless I have previous set breakpoints in them
• The compile process has no progress indicator, since it uses an external make
• The compiler can get stuck in a circular dependency without any clue how to resolve it (I can reproduce this, 100% of the time, with code conforming to the Java spec).
• A myriad of other little things

So, if you use Xcode for Java development, do yourself a favor and stop. I would even go as far as to say if you only use the dev tools for Java development, just trash them and save yourself several gigs of space. I think this just demonstrates that Apple doesn’t care in the least about Java, so don’t expect them to do anything in terms of support. I wish I was wrong about this, but I am afraid this is truly the case.

My tests did turn up one interesting fact. I spent forever trying to figure out why an .avi file I had would refuse to passthrough using Apple’s method. I eventually concluded that the difference is that the brilliant people who decided to stuff AC3 into a format never designed to handle anything of the sort, also decided that it was acceptable to destroy the whole point of the format. AC3 is comprised of a frame of data, which is entirely self contained, and must be complete to decode. AVI (and other formats), have the concept of a packet of audio data for decode. Now the “brilliant people” decided there was no need to make the packet and frame have a 1 to 1 correspondence. Result, badly packetized AC3 frames, where several frames are in a single packet, and some frames span two packets.

Apple’s passthrough requires that a packet begin with a frame, and a frame not be split across two packets. I haven’t tested multiple frames in a packet, but I really don’t care to since the above is a deal breaker. This means that Apple’s passthrough will never work with AVI files, hence commit r858 to perian. Since I figured I had nothing to loose, I filed a ticket with Apple about this issue, number 5876598. Maybe they’ll care to fix it, but since it has no bearing on the consumer AppleTV, I doubt it. Since they haven’t done anything yet about passthrough on the desktop (see my other post on this topic), I doubt they will do anything about this either.

Guess we can always hope that they will surprise us all someday.

Well, I noticed that the page doesn’t actually authenticate the email addresses, but instead just accepts anything that is entered in. I started entering in the letter, “a” and it took it. It will not accept a blank entry though So, what’s the point of having this at all? It wasn’t until the last days of the conference that I finally figured this out. The conference hotel changes for internet access, $12 a night, if I heard correctly. So, this login screen is to prevent the hotel’s guests from leaching free wifi off the conference, and thus off the hotel itself.

That raises an interesting trend that I have started to notice. Why is it that the expensive hotels (this one was was $300 a night) require you to pay for internet access, but the cheaper ones include it free with the room? I was at a hotel which was a fraction of the cost, and it’s free internet access was 20 times the speed of what I got at the conference! So, if you want a nice place to stay with internet access, this is one more reason to not pay too much for the hotel room, because you may end up paying more just to check your email. Otherwise, the good old fashioned post office starts to look like a pretty good deal. It isn’t as fast, but it sure is a hell of a lot cheaper.

Hopefully a page like this will help shame these people into doing the right thing, but its power is limited. Somehow I doubt that the latest addition (not yet on the list at the time of this writing) will care that much coming from the country that routinely treats copyright as little more than a hinderance that should be abolished (China).

Anyway, Phoenix is weird. I walked around outside, and it felt like 70 to me. In truth, it was 90. Without the humidity, it just doesn’t feel as hot. But I already hate downtown. Here are some things I noticed:

• The lights are not synchronized so you get stopped at over 50% of them
• Central street is under construction so it’s speed limit is 25, but the old 35 speed limit signs are not covered up
• People are no where near as friendly as they are in Texas. This is evident by the fact that a police officer jumped in surprise when I said hi to her.
• There are many people begging and loitering downtown
• There are signs telling you that a sidewalk is suddenly closed without recourse for an alternate route

At least the conference is good. Hopefully my allergies don’t ruin it for me. I dislike dry climates.

Apple has enabled AC3 passthrough with Quicktime. They have enabled it on the AppleTV to support playback of their HD movies. The code for passthrough is even in the latest version of Quicktime on the desktop, but the celebration ends there. Apple has not exposed a method for enabling passthrough on the desktop either through the user interface, or a hidden preference. They have not even exposed a method for developers to use it through Quicktime (I’ve tried).

So, what is to be done about it? Tell Apple to address this. Go to http://bugreport.apple.com and fill in an enhancement request. You can reference my report id, which is 5799133, and says:

Please enable a method for AC3 passthrough on the desktop using Quicktime. There are no licensing issues for passthrough, and you already have the enabling code in place. The AppleTV proves that your code works, so please add the interface so this can be used on the desktop.

The hack in use by Perian isn’t perfect. It does not work with all receivers where the true passthrough will. Audio outputted using the encoded digital audio output of the optical interface works, as evident by the DVD player, but this method is not available to Quicktime in any documented manner.

If enough people fill in reports, Apple should realize this is important to people and enable the interface. Please use your own words in your own reports; it says more to them if the reports are not copy/paste.

Update: Apple has marked my ticket as a duplicate of 4557501.

1. Install Perian
   Just go to Perian.org and click the big download link. Double click on the pref pane, and it’ll install itself.
2. Connect your receiver
   The audio connection must be a digital connection. In all cases that I know of, this means an optical cable from your mac to the receiver. In the case of the AppleTV, this is certainly the case. I suppose this could work with the coaxial digital connection, but I don’t know of the audio hardware that provides this connection.
3. Change the sample rate to 48KHz
   • On the mac, open /Applications/Utilities/AudioMidiSetup.app
   • In the lower right, change the format to 48.000KHz, and 2ch-16 bit
   • On the AppleTV, this is much more difficult. I recommend using Sapphire instead. It’ll do this step for you
4. Enable AC3 passthrough.
   This is easy when done correctly. Open /Applications/Utilities/Terminal.app and type

   defaults write com.cod3r.a52codec attemptPassthrough 1

   To turn off passthrough, you type this instead:

   defaults delete com.cod3r.a52codec attemptPassthrough

5. Play with 100% volume
   This does not mean modify the .mov to play with greater gain than the original .mov file in the properties, but it does mean put your computer’s volume output at the top, and the .mov file should also be at it’s top with the little slider.

That’s it. Play you movies with AC3 audio and enjoy your 5.1 system.

Edit: Added volume information which was missing.

First, there were the developers, who seemed to largely agree with my comments. Then there were the users, who took offense at some of my comments, or the way that open source developers have treated them in the past.

Now that I have had some more time to think through the issue more, I have some additional comments. In addition to working on Fire, Adium, and Perian, I have also added Sapphire to the list. Here are some of the things that I have learned:

In my previous post, I listed three reasons why a developer might release their program under an open source model. These were:

1. The software uses another piece of code or library which requires the work as a whole to be released under an open source license (GPL being most notable here).
2. The developer hopes to interest more developers to join his efforts.
3. The developer fundamentally believes that all software should be free and is trying to lead by example.

One person attributed these 3 reasons as “blackmail, lonely, and zealot.” While the third is accurate of what I was trying to convey, I don’t think the first two are.
In the case of 1, I have actually known of a program that had part of it released under an open source license specifically to get around the license restriction of the extensive libraries he used. If he had his way, it would all be closed source. This is what came to my mind first, and so I listed it first.
With 2, I was referring to a reason to open source a project, not why to work on the project in the first place.
Anyway, I would like to add a few others to the list which I hadn’t considered:

1. A work for potential employers to view
2. The developer was feeling altruistic
3. The developer wishes to cease involvement with the project but doesn’t all the work to go to waste
4. The developer wants others to learn from the designs and algorithms he made

I didn’t realize the potential of the first until I interviewed with a few companies. Instead of asking me about my work experience in the interview, they asked me about my open source projects. I would surmise that they view work on an open source project as more indicative of my passions. If they were able to find a project within their company that lined up with the passions I displayed in these projects, then I would be a much more productive worker for them.
The last three are somewhat similar, but I have seen each distinct forms in the past.

Now, on to the piece that was missing in my first post. Why does a developer work on an open source project in the first place?

• The project is something fun.
• The project’s output (program) is something the developer uses himself.
• He wishes to see some sort of enhancement for his own use.
• He is being paid.

If a developer finds himself working on a project, and he doesn’t find any of the above reasons describing his situation, he should seriously consider abandoning the project. Why work on something that will not benefit you, isn’t paying you, and is simply not fun?

The other side of this issue is the perspective of the users. They can be roughly split into four camps.

1. Those who use the software and love it
2. Those who use the software, and love it minus a small pet peeve, but are appreciative of what they already have.
3. Those who use the software, love it, but demand that things go their way.
4. Those who used the software, hated it, and vowed to never use it again.

Those in camp 1 are, in some ways, the best users. On the other hand, those in the second camp are also great, because while they wish for something to be changed, they are still appreciative. Camp 2 users can actually be better than camp 1 when they come up with insightful suggestions, and when they present a bug report, more often than not, the developer will be happy to address them.
Camp 4 users may provide a “This software sucks because of …” comment and you never hear from them again. They may be a bit of a hit to the ego, but just ignore the comment and it’s over.

It is camp 3 that is the destruction of open source software. While they are often a small minority, unfortunately they are the most vocal. They will often take the attitude of the selfish jerk who things that his way is the only way things should ever be done. He will often tell the developers that something must be done in an archaic fashion against the developers thoughts, and may even go as far as to insult them hoping it’ll help get it done. “It is so easy,” and, “this is useless,” are examples of such phrases one may seen being used by such people.

If camp 3 users were eliminated, open source software would be a much richer field. So, what can be done about them? First, watch the google video: How Open Source Projects Survive Poisonous People (55min, 111MB).
In addition to the things specified in this video, I have started employing the phrase “patches welcome.” This has become my way of saying:

No. I have no reason to do what you say, and it will take a considerable amount of time that I do not have. The same is true of all the other developers, and your persistence in demanding this is becoming a nuisance. If you really want it so badly, then you are going to have to do it yourself, find/pay someone to do it for you, or spend the time learning the skill set so that you can do things on your own.

A common response to this is something along the lines of “I don’t know how to program.” This is akin to saying “I can’t do this for myself so you are obligated to do it for me.” No one is entitled to have a program designed for him just the way he likes without paying for it. Don’t bother arguing this point; you’ll just appear foolish. How many people think that you could go up to an professional tax preparing service, tell them that they don’t know how to do their taxes, and have them fill out the forms for free? Not going to happen, is it? So why do people expect the same of open source developers?

So, what is a developers response to seeing reports from users in any of the 4 camps? I cannot answer for them all, but I can tell you mine. When I am perusing through the ticket tracking system, and I see a report/suggestion, I’ll apply the following criteria in decided whether I will work on it or not:

• Is it a quick change (5-10 min. tops including debugging)?
• Did this person show any appreciation for the work I have already done?
• Is this something I want myself?
• Will this be something fun?
• Am I being paid or will this benefit me in some other way?

If the answer to all of those is no, there is no way I will work on it. Anyone who would expect me to do otherwise should get out of their communist mindset. I will not adhere to working according to my ability and pay according to my need. Too many have fallen into the trap of believing this to be a good idea; I will staunchly appose them.

In conclusion, open source software (OSS) can be a beautiful thing, if kept in the proper context. It is starkly different from commercial software in that the developers (in most cases) are unpaid volunteers. Users of OSS, treat the developers with appreciation and respect, and you may convince them to direct the program down your chosen road. If they don’t, do not be upset, insult them, or become more demanding/annoying, but instead thank them for their work and ask if it can be revisited in the future. If you have the cash, you can even offer to pay a bounty on it, open to anyone who submits the accepted solution. Those are the best roads to take, and are most likely to get the results you desire. Follow another, and take your chances.

Now it is time to sit back, relax, and get some sleep before the bug reports start piling in.