brave [little] meme

Behind the Scenes: Interview with the Plex Team

On January 11, 2008, the following post appeared in the fledgling Plex blog:

“Videos actually sort of play now, although there’s no sound […] and videos play at strange rates. Sometime fast, sometimes slow.” - elan

That was then, and this is now. Just over a year later and Plex is already one of the most advanced and full-featured Media Center packages available on OSX. Accelerating away from its roots as an XBMC port, Plex has over a short time managed to carve out its own unique identity as a solid, rich media platform.

Today we have a chance to look behind the scenes of one of the Mac platform’s greatest labours of love, by sharing some question time with two of its key creators: Elan Feingold and Isaac Ordonez.

Q: Tell us a little bit about your background.

Elan: I was born in Pakistan to American parents, grew up in Europe, and my first computer was an MSX2. I studied Computer Science and Electrical Engineering in upstate New York, worked for Digital Equipment Corporation and then a string of start-ups.

Isaac: I’m a self-taught Server and Systems administrator for a medium sized school district in California.  I’ve been working on Unix and Mac servers for quite some time.  Since I deal with students and teachers on a daily basis I tend to understand what everyday computer users expect while at the same time also understand the nitty gritty. I try and apply this when working with Plex as I feel our target market is going to be the everyday user.

Q: Why Plex?

Elan: I was looking for a media center for my Mac, and I was trying Perian and Front Row. However, Perian had horrible support for AC3 passthough and no support for DTS passthough (at the time, at least). I was considering coding it up, but then got bored one evening and started on the port of XBMC, which I had used on an XBOX for years. More to the point, why Plex and not XBMC? I found that I had very different goals for the application, which is I guess how most forks occur.

Isaac: As with most people I started as an XBMC user on my Xbox1.  I’ve always bounced around trying to find the perfect media center.  I got an Apple TV, Played with VLC, Perian and Frontrow but none of these really stuck.  The biggest problem I had was that while I could use these systems, my wife and most of my friends couldn’t. Heck I had to pay someone to mod an Xbox1 and install XBMC on it.

When I saw Elan was porting XBMC to OS X I jumped at the idea of creating something easy yet functional.  Once we forked it’s been great only having to focus on one platform, we’ve been able to avoid having to develop features for a “hobby” audience and try and focus on the masses.  Whenever we look at a new feature we think “Is this wife friendly?”.  I usually sit my wife down in front of the feature and hand her the remote to get feedback.  If the feature is too confusing we have to take a step back and consider if it’s worth implementing.

Q: How would you define your role within the scope of the Plex project team?

Elan: I’m one of the people coding, obviously, and I try to help keep us focused as a team and set priorities. I also act as gatekeeper for the release branch of Plex and make the releases.

Isaac: So far my role has been “do a little bit of everything”.  I manage web services, keep an eye on our forums, dabbled in coding, help with plugins, project management and the list goes on.  I also try to keep developers grounded with the user base.  Sometimes I worry someone will run off and create something totally useless like an ASCII video player.

Q: With its Plex/Seven series of releases, the Plex project has been establishing a more defined sense of identity by diverging from its XBMC roots through new features such as TV Themes, the Plex Media Server, and the new App Store. What inspires you and the team as to which new features to pursue?

Elan: These things all play into our plans for world domination. Honestly, I think we have a shared vision, but the exact pieces of the puzzle change over time, and we’re lucky to have really creative people like James (author of the TV Themes feature) who’s always coming up with great ideas. Also, one of the great things about forking is that we have this powerful code base, but without any of the potential baggage that comes with it in terms of ownership, existing priorities, or politics.

Isaac: As with most projects, personal interest is what drives new features.  You have to be really excited with a feature to want to work on it.  At the same time you have to listen to the user requests and sometimes work on a feature you have little to no interest in.  As long as in the end it’s for the good of the project.

Q: There have been a few hints in the past as to a complete rebuild of the Library system for Plex. How’s that coming along, and what sorts of improvements can we expect from it?

Elan: We don’t comment on upcoming features, but if you look on the forums, you’ll see that it is one of the biggest annoyances at the moment.

Isaac: Without directly answering the question, I think you could say we’ve make good progress already. Since we added full iTunes and iPhoto support the Music and Picture library system is becoming irrelevant.  Most Mac users already have their music and pictures neatly organized in these programs.  I think Apple has provided a really great way to manage your these types of media so there isn’t much of a reason to re-invent the wheel here.

Q: Are there any new features you can hint at on the Plex roadmap?

Elan: I can’t say much beyond saying that we’ve scoped out our roadmap for the rest of the year, and that there’s a lot of exciting stuff on it. If we manage to get it all done, I think users will be really pleased.

Isaac: We’re looking at adding Gopher support and possibly streaming RAR files over telnet, other than that we have some ideas.

Q: How, if at all, is Plex’s reliance on XBMC code changing as the project increasingly develops its own unique feature set?

Elan: The reliance will continue to lessen over time. If you break down XBMC into components, you have the skinning engine, the players, and then everything else. The skinning engine is the component of primary interest to us, along with the players. The other parts will likely be of decreasing importance as time goes on. What people don’t seem to realize is that every feature comes with a non-zero support cost. More features doesn’t imply a better, more stable, or more usable product. We’ve not serving our users well by trying to be everything to everybody.

Q: What about skinning; Plex currently defaults to MediaStream. What’s your team’s relationship to the skin’s creators?

Elan: Mike Beecham (one of the authors of MediaStream) is a great guy, and he’s now part of our team. We take design seriously, and we actually have as many or more designers as we do coders.

Q: How would you rate MediaStream in terms of aesthetics and usability?

Elan: I don’t think that’s a totally fair question, because I think that a skin’s usability has a lot to do with the underlying structure of the XBMC application/UI code. Having said that, I think the appearance of MediaStream was groundbreaking in terms of both aesthetics and usability.

Isaac: I’d give it an 8 and a 7.  It’s a great looking skin, when people come over and I show them Plex they go “what’s that!?”.  My wife can use it, but sometimes she gets lost.  I’ve programmed “Movies”, “TV Shows” and “Music” buttons on our remote to help with this.  Sometimes I see people getting stuck on a scroll bar or on the options menu to the right (or top).  All in all I think Mediastream really raised the bar for XBMC/Plex skins in balancing beauty with usability.

Q: The upcoming Aeon Stark interface has been getting quite a buzz lately, and seems to also shaping up to become one of the more user-friendly interfaces available for XBMC. Are there plans to extend Plex’s skinning support to be Stark-friendly upon the skin’s release?

Elan: First off, major kudos to DJH, who’s an incredibly talented designer. We don’t have any concrete plans, although a resync with the skinning engine code is definitely in our plans for the future. We have our own priorities, and I worry about getting distracted by these sorts of things; at the end of the day we only have a limited number of spare hours.

Isaac: Aeon Stark looks like a beautiful skin from the videos and pictures I’ve seen.  From it’s looks though I wouldn’t put it in on a device and give it to the common user though.  Personally I would like to see something that way halfway between Front Row and Mediastream.  Being able to control every little aspect how your setup looks is important, but only to a small subset of people and in the end usability will always trump eye candy.

My Comcast DVR for example, the interface is ugly as dirt, but there is no need for an instruction manual. But at the end of the day millions of people use it, and it doesn’t even come with a user guide. After I created a help screencast for adding your movies to Plex someone who had used been using Plex/XBMC for years went “Woah! You can add multiple paths to a source!?” Basic functionality like this should be intuitive for even casual users.

Q: Does the team have any interest in evaluating Stark ahead of its release?

Elan: I’m interested but sadly don’t have the time. The ideal would be a Plex user with some development experience who could be an advocate for the skin, test it out, and perhaps cherry pick some of the new skinning features that are required. Also, I’m sure the (very talented) author of Stark has a good idea of what new things he needed.

Q: What about the skinning community as a whole: it would seem that, currently, most of the third-party interfaces in use by the Plex community are originally designed for XBMC. Does the team have any plans to further engage the skinning community to encourage a stronger focus towards Plex?

Isaac: Personally I would rather see one KILLER skin for plex than 500 mediocre ones.  But that is just me being a stick in the mud. ;)

Elan: I agree completely with Isaac. Having more skin choices isn’t always better for the end user who just wants to sit down and watch something on their TV. To answer the question, we definitely engage with designers who approach us, which is why we have quite a few of them working with us now.

Q: I’d now like to talk a little about Plex in relation to its siblings in the HTPC software market, namely XBMC and Boxee. In a previous interview with Jmarshall from the XBMC team, he stated they maintained no real relationship with the Plex team as a whole. Would you agree with that, and if so, is that by choice?

Elan: It depends how you define “relationship.” If you mean “share code” then yes. If you mean we email back and forth to exchange ideas, then no. I respect and like Jonathan, and I’ve always gotten along well with Vulkanr of the Boxee project. However, at the end of the day we all have different visions and different priorities.

Isaac: As Jonathan noted in his interview Scott and I had lunch with him when he was in San Francisco.  He has always been open to communicating to us and we appreciate that. Aside from Jonathan there have been a few XBMC members who have been more than willing to talk about common issues via PM or E-mail.

Q: How do you think communication between the Plex and XBMC teams could be improved?

Elan: To what end? The basic fact is that we have different aspirations with our respective projects. As long as we have open communication channels to discuss common issues (which we do), I think things are working fine.

Q: Would it be fair to say Plex has adopted a more closed-off development model relative to XBMC? If so, what factors influenced such a move?

Elan: Yes, it would be fair. Releasing the source at the same time as a release (which is also what Boxee does), reduces or eliminates the number of half-baked or “in-between” builds that might exist, which helps lighten the support load. Overall we haven’t found any downsides in doing things this way, especially since we tend to release fairly often.

Isaac: Open or closed development makes no real difference when it really comes down to the end user that I can see.  Most people aren’t going to be bothered to download the source and compile.  Bottom line is they want new features and they want them now.  This is why our auto update feature has been such a big hit.

Q: Looking at Plex’s new App Store feature, one might observe that it is taking Plex into a somewhat analogous direction to the more streaming-focused Boxee. Was that intentional?

Elan: I don’t consider it a change of direction, just an additional piece of functionality. It rounds us out.

Isaac: Streaming is just once piece of the puzzle, mediocre quality video on demand probably won’t replace higher quality files that reside on the users system.  It’s the way the world is moving though.

Q: Does the Plex team maintain any contact with streaming service providers like Hulu with regards to making their content available via App Store plugins?

Elan: No, although this is likely to change. To be honest, though, I’m really not very impressed with the major content providers (I described my feelings more here) as I don’t think that they’ve provided a viable alternative to pirated content, which is what they should be doing. I’m personally much more excited with things like TED Talks than I am with low-resolution, ad-laden old episodes of MacGyver.

We have been approached by companies who are interested in getting their content onto people’s televisions, and this is the sort of win/win relationship we are interested in.

(As a funny side note, if you look at the source of a Hulu video page and search for ‘pl’ - with quotes, I can only assume it stands for “Plex” - you’ll see the only communication we’ve received from them, in the form of a hidden div countermeasure.)

Q: Has Plex’s increased focus towards streaming content provision raised any legal issues?

Elan: We are rendering these sites in a web browser, we just happen to zoom in to fit the content inside the Plex player, so we believe we are within our rights.

Q: How does the project keep going financially without relying on advertising or other commercial tie-ins?

Isaac: We get paid in smiles… Almost literally.  We have some fantastic users who have provided hosting for our files and liquidgravity.com has provided us with an almost dedicated server for our web services.

Elan: As a team of people working on this in their spare times, there are no real financial requirements. As Isaac said, generous people have come forward to provide us with the bandwidth and hosting we need. Everything else is just fingers and a keyboard. People have been very kind and generous in providing us with beer, as well.

Q: Is Plex’s financial model likely to move in a more commercial direction in future?

Elan: First of all, we’re obviously not in it for the money, because we’ve put in thousands of hours over the last year with very little in return. Getting an appreciative email, forum post, or a beer donation is really awesome, and we absolutely love our users and think they’re great.

The second thing I’ll say is that obviously we really love working on Plex. Wouldn’t it be great if we could actually make a living doing it? The question is whether we can figure out a way to do that without compromising our vision.

Q: Where do you see Plex going over the next year?

Elan: We have big plans for the year, and I’m really hoping we can release 1.0 before 2010.

Q: Finally, do you have a message for our readers?

Elan: Thanks for reading, thanks for your interest, and if you’re a Plex forum user, thanks so much for contributing to the nicest and most productive forum I’ve ever seen on the Internet!

..

bravelittlememe would like to thank Elan and Isaac for their time, and encourages you to visit them over at the Plex home.

An Interview with djh, XBMC Skinning Legend

Anyone familiar with XBMC, the Xbox Media Centre project, or its mac spinoff Plex is likely to be acquainted with a little skin called Aeon. Released years ago but never ultimately completed, Aeon set a new vision for what XBMC could achieve in terms of user experience. 

As he prepares for the release of his latest skinning project, code-named Stark, Aeon’s creator Duncan Harris (better known as djh) has kindly taken some time out to share some insights on his work, as well as a few tantalising hints as to what users might expect in future.

Q: The release of the Aeon skin made you somewhat of a legend within the XBMC community. Despite never making it beyond alpha, Aeon and its various mods are still in widespread use by both XBMC and Plex users.

A: I think you might be exaggerating a bit there, at least as far as the first bit goes. Still, it’s nice to get so many kind words via email and the guestbook. Weird, though, as people are obviously referring to a version of Aeon I wouldn’t touch with a bargepole now. I just hope the new version doesn’t disappoint them – that’d be typical.

Q: What was it that initially motivated you to get into the skinning game?

A: The best motivators of all: frustration and impatience. Too much time spent hopping between one media centre and another and thinking, “Is this it?” It goes right back to Winamp and its library view, which never worked the way I wanted it to and bugged me something rotten. Even worse is when you find the app that does everything you want bar one really important thing. That’s how it was with the XBMC interface. In my job, you come to realise there are far too many commentators and precious few doers; I’m just doing my bit to make things work a little better, I suppose. 

Q: Why XBMC?

A: Man, why not XBMC? It’s incredible. Granted, it had a head start by being on Xbox where you could just pick up the pad and play with it, but it’s carried that robustness over to HTPC which is clearly where it belongs. From my angle, it’s obviously the skinning engine that’s the big draw – I’ve never seen anything like it in an app of this kind. People joke that the imagination is the only limit with that thing, but it really is true. That’s a great motivator right there: knowing that if you dream something up while you’re walking down the street, you can build it in just a few hours. 

Being open source is a key factor as well. I’ve seen enough interest in Aeon to know that everyone wants to jump on the HTPC bandwagon, but as soon as money’s involved you’re in shit town. It can’t do this, it can’t do that; so-and-so has to have its logo here; we need advertising there; this format’s a threat to our security; we can’t monetise such-and-such. Fuck it. No matter what the big boys come up with, XBMC will flatten it. Or at least it will if it has the looks, which is where us skinning bods come in.

Q: How has your real-life background influenced your design style? What would you say would be your major sources of inspiration?

A: Well, I work in games – or around them – so I see a new user interface every day. That counts for something, I guess. So when you go from looking at a Codemasters interface to something far less attractive for a media centre, you start asking questions. That’s really all I have as far as inspiration goes; I’m no artist. I should give credit to Mathias Mahling, though, the guy who did the original XBMC interfaces, and Smokehead (whatever his real name is) who did an older skin called “Clearity”. That was the first sign that XBMC was poised for better things. And I had a (too-) brief partnership with Emilio Ayala, a graphic designer from NYC who has tremendous vision when it comes to this kind of thing. He’s a genius. I can’t thank him enough.

Q: What’s your core design philosophy?

A: This is where I try not to sound like a dick and usually fail. In short, it’s about making stuff like movies and music real again. So you’ve had years of MP3 and now digital video steadily killing off the optical formats like CD and DVD, which themselves stripped much of the art from vinyl and VHS. Well, maybe not VHS, but you see the point. No one has any stuff anymore. We used to dedicate little corners of our homes to the things that inspired us, and now we have a bunch of files on a computer that no one will ever see. Great. So if I have any design philosophy it’s simply: enough is enough.

People harp on about Apple and the iPod ruining things like cover art, but in a lot of ways they’ve saved it. And they introduced real world textures to their interfaces, which went some way to creating a kind of “virtual collection”. Aeon’s all about taking that to its ultimate form, celebrating taste through things like backdrops, fanart, and user customisation.

Actually, I may as well ride my high horse while I’m on it. One thing that really drives the web side of things – the Aeon site and all the free backdrops, which trust me have consumed a fair bit of time and money – is the way Microsoft stiffed everyone with the user customization side of Xbox 360. After all that talk of how it would cater to “the remix generation”, all it amounts to are these trivial “gamer pics” and garish themes, few of which you can preview, all of which must be paid for. That’s shit, no matter what the financial needs. 

Q: How do usability and ease of use factor in your design choices?

A: They have to factor quite heavily, right? It’s one of those 49/51 ratio things: which do you prioritise out of usability and looks? They’re both as important as each other, but there’ll always come a crunch where you have to choose one over the other, and usability must always win.

Q: What were you primarily trying to achieve with your first effort, Aeon?

A: A statement of intent, I guess. I had a vision of what it should be from the start, and it’s the same one I stick to now. So I needed to communicate that and see if people took to it, and then think about making it work. Actually, I’m lying a bit there. The primary goal was entirely selfish: I just wanted all my stuff to look nice. But one thing quickly overtook the other and that’s why I’m still here doing it.

Q: Were you satisfied with how Aeon turned out? How did you feel about having to abandon it before an ‘official’ release?

A: I get bored of things very easily. Too easily, in fact. But I’ve come to realise it’s a good thing, and the only way of knowing if something’s truly worthwhile. So if you’re looking at something day in, day out for two years and you’re still happy with it, it’s worth seeing through. Aeon was always a prototype exercise and there came a point where I was just fed up of it. Bits of it didn’t look right with certain backdrops, the art style wasn’t focused, and it just plain didn’t work much of the time. It’s amazing anyone used it at all. So when I faced up to the need to finish it, it was clear it had to be scrapped and started afresh.

Q: Let’s move on to your upcoming project, Stark. First of all, what made you decide to come back to the skinning game?

A: I think I’ve answered that above. I did drop out of the whole thing for a few months, though, because the task of basically putting a new design together from top to bottom was just mountainous, and believe it or not I do have a day job. But you regain your energy after a while and one day it comes rushing back out, so that’s when it all started again – the unending horror ;-).

Q: The early teasers you released caused a rather vocal surge of anticipation within the XBMC community. How do the community’s continued requests and expectations affect you?

A: Dare I say it, you have to ignore a lot of it for the good of the work. It’s not that the comments are unwelcome; on the contrary, there’s just so much goodwill on the XBMC forums that you could easily get complacent, release something prematurely, or get wrapped up in replying to everything and everyone. Who can say how much anticipation there really is for Stark (which, incidentally, is just a codename for this new version)? It’s just a skin for a media centre, ultimately, so I keep that in mind and just get on with it. If the wife can find a Peep Show episode without the screen offending my eyes then that’s good enough for me.

Q: Would you consider Stark to be more of an evolution of Aeon, or are you trying to achieve something new in its own right?

A: It has to be both. Like I said, the vision for Aeon has always been the same so there’s bound to be inheritance. But the Stark version is entirely new code-wise (I’m never sure if we should say “code” or “script” when it comes to XBMC), was texturally designed from scratch and does /a lot/ of new stuff. So it’d be fair to call it a reinvention, I think. Again, the good thing about getting bored easily is that you do something really cool one day, then the next day you look and think, “It needs something else.” So you do this “virtual homepage” thing for the customisation and pat yourself on the back for that, but suddenly it’s not enough. So you have this four-way “compass” menu for changing views that totally fixes everything about that transition, but it’s not enough. So you stick in media flags for things like movie studios and ratings, but an hour later it’s not enough. So you have item lists that transform into and out of… I’ve said enough ;-).

Q: We all know you from your excellent work on XBMC, but are you involved in other projects at the moment?

A: No. Being married and paying the bills while getting this finished, along with its website and user manual, is quite enough for one human being. I’m not an octopus.

Q: What are your tools and platform of choice when it comes to design?

A: Notepad++ and Photoshop. That’s pretty much it. I’m running Windows 7 on the desktop so that’s where I test things on XBMC, and the proper “quality assurance” gets done on the HTPC plugged into my TV. I’ve actually spent so long testing Aeon on the TV that the bulb’s started to go, I have lines running across the screen, and there’s an ghostly image of the Aeon panel view burnt into it. But does my wife see the need for a new one? Of course not.

Q: Beyond XBMC your skins are also widely used on Plex, the excellent XBMC mac spinoff. Does Plex factor in any way in your work on Stark?

A: Not at all, though I welcome the existence of Plex. Supporting two different branches of XBMC would be suicide, though, given the complexities involved. 

Q: You have always maintained a very open stance towards third-party efforts to modify your work. Has that changed in any way since the Meedios incident?

A: Heh, the “Meedios incident”. That was a really unfortunate thing ‘cause I ended up sounding like a right fucking plum but was actually quite upset. I really admire that guy’s work and I appreciate his loyalty to Meedios which is, as I understand it, a bit past its prime. I just wish he’d asked about cloning Stark so long before its release and maybe acknowledged the source at some point, at least before I got an email saying (and I’m paraphrasing), “Oi, your stuff’s a rip-off.” Still, he was decent enough to sort it out and that’s that. Otherwise, my stance has never changed: if you want to mod or port Aeon, be my guest. I’m hoping the documentation and openness of Stark will make that a whole lot easier for people.

Q: What would you say to anyone thinking about getting into the skinning game?

A: Go for it. What harm can it do? Just be aware that it’s a monstrous job if you go it alone and will happily consume hundreds of hours. Alright, that’s the harm it can do.

Q: Once Stark is out, do you have any other projects planned for the future?

A: I’ll actually be glad to just sit back and tinker with it, keeping the website up to date, polishing any rough edges and inventing new stuff. It’s only when you have that stable foundation that the really exciting stuff can happen. Who knows? I might even watch some movies.

Q: Finally, is there anything you wish to tell your fans?

A: I have fans? Sorry to all those who’ve PM’d and emailed me over the last few months to no avail. I’m too wrapped up in things to answer everyone. There you go.

[ If you’d like to see more of djh’s upcoming project, due for release in March, head on over to the Aeon Project page. ]