PiTiVi

OK so in order to save time I can plan this

presentation around the assumption that everybody in the room knows what they're doing things operators what open source is then what the Linux platform is so before I go any further anyone not knowing this and there's anyone need some

quick clarifications yes no OK so um for those of

you saw my talk in 2009 you might remember they compared to similar to die in a crash so yeah this talk is pretty much centered around the TV that it also going to cover state of detained probably do anything on the exam in general yeah so I won't go over the history of the city

project since I already did that I already really did that uh that 2 years ago so if you care about the history of the project you can go look at a video recording of the 2009 US by a time machine or something you know you can find a link to the video recording on this huge here or on my personal website and somewhere here here's what I'd like to and what your

appetite with today so to start up this is a

bird's eye view of the state of an extenuating can still see the the various camps American and the MLT camp all those independent video editing applications of the status of something the so yeah so that all things considered from the user's point of view you see applications maturing framers frameworks maturing a massive new weird things coming out of the case and so on and so very don't be offended if I'd been this your favorite pattern here I'm just simplifying our hearts in thinking in terms of engines so anyway you get to see all the other ads uh later on that so this pretty much looks good right it looks like there's some kind of healthy competition going on in various camps are advancing are not exactly are you guys ready for a cruel sucking through and thank you so what

happens is this welcome to the real world most people don't care about the little guys on the Linux map each similar to the cool clear kids playing happily on the table over there and they're using like I mean the moviemaker where it came the city of their can quarter or whatever they could get from a shitty Guyana shopping mall in Hong Kong and the guys in blue there are 2 guys from aided and they're too busy ringing the cash adjusted to the notes as you here and get back to their the open source standards video editors battle the other possible and in a few minutes has so for a few seconds and like to take a look at this broader view and ask the question why don't those 2 worlds intersect with each other why are we not saying saying for under Linux field there are many many many many reasons for that 1 of which is market share I guess I should

exclude smartphones from this figure because and doing serious video editing of smartphones like can silly but then at the same time who would have thought that a hundred years ago that we we would have cameras embedded in phones and that the main use of from would be not making phone calls anyway so when is this the the big my big maturity as for windows here Linux has maybe 1 to 1 . 5 per cent of market share and according to the last the trends I've seen and in 2008 we passed 1 billion personal computers and so when you think about it 1 . 5 per cent of 1 of more than 1 billion computers is still a lot of about 15 minute millions and that sounds about right I guess but

what happens when you combine that with the fact that not everybody does that you would see what is the intersection of the Linux user and if you data uh according to my rough estimate that's

according to the winters popularity contest that's about 4 % of woman to users used we should take other video-editing software into account we may have maybe 10 % of Linux users anyway as amino my data might be completely wrong there are many problems with this this this example it ionizes just about ballpark figure for me to know

that we are Indonesia markets I mean where 4 per cent of 1 per cent that you can be more nations that is there still and for this

I guess there is because there seems to be a sustained interest in PTT strange according to sites that's 6 there's like between 15 and 20 thousand visitors per month this kind of this graph is kind of outdated anyway and so we cover the

market share so that's only 1 of the many reasons why an open-source service stock here's another 1 history keeps repeating itself um

recently done some research and came up with the timeline of all the open source they do it as I could find I'm pretty sure you've never heard of 90 % of those ads in the timeline that's because they failed and anyway and look at all those independent lines they're all reinventing the wheel and even the injuring ones the ones that they're the ones that survive natural selection the all in different different at different directions and the problem with that is that it fragments the resources that is the few that that are crazy enough to try To make a video it's and so the the

lockable factors of failure I identified and brought term paper on that and targeting Windows and Mac ending with the hell would you ever want to do that and increases complexity more balanced more 1 power with the competing with commercial offerings and and having uses that generally don't give it an open source sometimes or that will complain that they can import the powerpoint file into your new is and there are many another key point of failure is thinking that it will be simple and that you will succeed where all the others failed and the uh and thinking that it will only take a couple months and if you think that you failed before you even get started in your competitors will outpace you before your present data is out so on the other hand the key factors of sexes

and a the open-source run on and has been easy to install and support as many formats as possible uh having a nice new idea is intuitive and pleasant to use don't crash all the time I haven't seen of knowledgeable core developers to do the maintenance and the complex tasks attract

contributors the contributors to work on the cool features and foster a healthy community for 1

thing it to be the default literature review it around a little that's not saying much I guess but it kind of provides some sort of legitimacy and validation or there is not complete graph and I know the general ecosystem canonical estimates user base at 12 million users that's might take taken up with the recent events around anyway what this means is someone who wants to contribute to the uh the video editor with the broadest potential for impacting the lives of as many people as possible yeah needs to look no further than could given that PTVs since it's installed on millions of this of computers that is very and that's probably not going to change anytime soon due to architectural legal regions fighting

naively hope that we would see patches contributions coming from canonical and that means already for meaning that we would get more contributors I guess I was kind of wishful thinking and then finally what makes Batavia were quite

project and the fact that it uses standard technologies like this year of which we all of my guests yeah it's pretty much the biggest just minute but much of biggest minutes made a platform framework diner thing on Linux and and then on the part of the guests and as the upstream 1st means that when you spot a significant problem you fix it directly into a streamer which means it's better in the long term but it also means it takes a really long time to get there so I think it has more chances of survival in the long term another good thing is it is basically a torture test for just in the sense that all the other applications benefit from whereabouts we might find in PTT and fixed so in terms of user interface I spent a lot of time thinking on PCs designed to make it really simple and intuitive while not limiting what you can do with it yeah just remember that this work and for

those of you who are actually not the you're just is basically a toolkit for software developers to deal with sounds images videos and special effects and all that stuff so what

this means propidium abilities wraps is that you can get access to a huge combination of the 2 major projects begin encode and decode pretty much anything except that ABC and continuous files if she had the right 1 is installed in many this means that you

can use of wide verity of formats from the 100 dollars pocket flip FIL like video concordant thing to point and shoot and that there's a lot of what's quite interesting these days is that those cheap point of view that she point-and-shoot cameras allow you to record high-definition footage that it's much much better then what I could do with that Dedekind can core a couple years ago that custard twice as much at least so great compatibility great cost but not everything

design is important to and it's not just fast based effects a matter of thinking about how you use users well interact better with your software and trying to do better than that let's do a clone of Adobe Premiere our let's add an option for everything approach and otherwise you end up with creating features

and that's an actual screenshot from Sony Vegas anyway I described my point of view of options in 2009 sinking use a time machine goes to my talk I won't be going over this in detail and we all those all those things are completely useless I tried to find a useful option here is an fourier Anatoly need to choose if I wanna show a splash screen at start of mining million anyway in

terms of you I design I guess it's quite good thus far the problems are the usual bugs are hard to shoot and take a long time effects and when presented with with this user interface most people think like pretty aims to be a clone of Windows Movie Maker or by movie and that it's that but um that's not true

and mean it aims to be much more than that I just remote rewrote the not on the wiki recently so that people and get the wrong idea yes of course it doesn't do much at the moment and that's because these things take a lot of time to implement so the vision behind the project in terms of in terms of designing feature set is much much greater than what you can see in the current implementation uh and the instrument state the software and I believe you can design a user interface that is supposed both intuitive and power so that would suit both Joe Pumbaa and that cool kid was going to make the next indeed that blockbuster thing and so you really don't want to artificially limit states creativity uh and that some of the that you're not in the commercial world things you don't want to segment the market into low and high and and try to sell to both at different prices and anyway uh

is that not not exactly because you need a strong foundations and in strong software foundations to attract new contributors but getting those foundations builds is real hard just like this 1 said don't expect people to randomly just jump in and help you do it from the beginning you need to get it almost working for them to be interested so the way I see it basically it's like

maintaining such a big project is like meeting the International Space Station you professional core contributors that they care of the maintenance and complex internal development well in

order to attract external contributors that will tackle exciting stuff and and 1 of the great thing about being in a niche market is the we had our

uses a chain of hardcore I guess you can see that as a good or bad thing uh where are we not

suddenly overwhelmed with patches and contributions part of the answer I guess is that uh compared to other other free

and open-source software out there the users which is the filmmakers are usually not programmers only 2 % of the visitors on getting website check out the contributing page that's 2 per cent of 4 per cent of 1 per cent folks standing on the other hand suffer developers they are and usually not filmmakers of these they're not anymore because they're too busy developing software to combine that with the fact that a video editor is a complex beast and you have a vicious cycle so what is the solution then

it's it's easy to sell software right not really anything theory you can try In practice pretty much nobody does this in free software mean even asking for donations not sure that would work because most people with a brain we make a quick economic

analyzes and the women why should I pay for some incomplete apt when could spend a hundred dollars on a complete commercial application that would do everything I want probably uh without needing to wait 10 years so here come what I call the indirect competitors and the guys in the low end segments and they are the ones who were well began at the high and to do worry me because everybody's going to get it from like at time website or city in a supermarket highly in Hong Kong something so it's not really possible would think that you could have and direct revenue stream for the average user it as a

genius and a couple years ago and she I quoted her 2 years ago to be doing this thing is a complex problem that requires professional developers to work on it so

during this strange situation where the upper costs at a hundred thousand dollars per year in means you have a choice of can have 2 good a fast heating and good and fast would mean that you have a strong business model around it are really really deep pockets yeah so now you are in a better position to understand why we're still haven't seen the perfect open-source bedeutet arise in the last 10 years

now I that I did you all depressed let's look at it that's side of things PTT does have some sort

of backing in the sense that there have been multiple stoked students working on the project as part of cool summer of code program and there are some collaborate employees working on PT in their spare time those that help does that make a significant

difference yes it seems so involvement from collaborate help in the various summers of code helped a lot example for last summer a simple but since then implemented video and audio effects and they were integrated in pretty that this is not yet released in the stable version and just yet and pretend summaries of code helped a lot in implementing major features but we're always looking of course for new contributors uh there's a nice set of instructions that I wrote on the site and the wiki and there is a list of easy things to tackle for newcomers and I'm really quite excited

about what will be the result of this year's Summer of Code as we will have 4 students working on coarticulated projects and so those

are some of the things that will be working on uh G. S. stands for GC-RMA editing services and come back to that in a minute uh animated effects meaning being able to change of properties of an effect over time so you have you say for example goal from black to white to uh to full column and transformation is being able to easily rotate crop resize parents do stuff like that backflips with with the clips on the canvas titles it is being able to create text onto your video instead of using like Inkscape in exporting the PNG and having at times votes because of that and MIT and there is already some good work done for this but it just needs someone to finish um and the future is render profiles which is being able to have a default set for easy compared to export to say the video core you to e-mail or Blu-ray put smartphones are toasters anything so on a related note video uploading is also being worked on is being able to have I was that renders your project and directly opposed season and that 1 of the Summer of Code students already has a working implementation of this feature so it's just need some Polish because they you ice horrible and in any way is the needs some testing and then it can be integrated with all the rest and is going to be also will be able to post videos of adults on skateboards on you to so and we currently have some basic support for presets In project settings but not for rendering settings and we don't provide the full list of those the full set of those precepts so that's hopefully going to be fixed to so in that they assume probably been wondering what the hell is that

optical flow thing it's quite an ambitious said summer could project for stream that he did he make benefit from from my understanding being able to come out parts of the scene and do compositing on top of that and that would be managed semi automatically hence the optical flow thing that means that instead of manually annotating every single frame the computer would do part of the work for you and by analyzing individuals and that means that if you want to do the next die-hard or Star Wars movie you can big explosions in light without needing to get it every single frame give an so what about it's

basically and another at section layer that was allows you to create apps like PTT yogic closer or loan balance uh more easily yeah I can hear is

seeing what another library on top of all this why and this is why according to Edward widely we

DGS basically because hacking appear to be still feels too complex and I guess because there's a lot of libraries that so yeah let's recap

basically picturing situation still sucks from was users hopefully after GS is implemented in PTT will be able to write ponies on rainbows and be had ever after an where an interesting situation interesting times ahead because we're all when that we I am I guess living under the impending doom of being wiped from the surface of the Earth by the crushing superiority of like works that's proprietary at this was supposed to be open source they said that a year ago or 2 and then if it is in these released as open source and if it is the pretty good or maybe and I will be really interesting to see and also what comes out of this year's Summer of Code projects for PTD so any yeah there's a website hasn't really nice set of pages for new contributors if you have problems getting started you can always feel free to to come and see me and yeah that was kind of a nonstop

stream of blob of although I don't know if you are there any questions any insults or high maybe I can just really should be about a nation source configurations just work theory and yes we got about 70 thousand euros from 2 nations lost due to their interest to have 2 people work beautiful time for about 15 months in total creator and that question that came to my mind was when the deal with the nation's I keep them with the come and and I had a lot of money to the developers when they have completed a move toward and this is also important in this patent to a a few points and releasing the

Austin only in will easily travel or maybe make come delicious tendentious much yes just check again guys how much of their monthly donations income comes from when shoes I don't understand the economic reasoning OK I can get it the role of professional and put 200 dollars an hour so why should I pay 10 dollars for source of the like to manage that did it really works and then as that can have I thought about that then I was wondering

well 1st the PTA is don't have the manpower to maintain a port to Windows and Mac and self training the barely managed to keep the but data but it's not non-funded if there could be a Windows version I mean Europe the people on Windows reading 1 of pretzels until staying at when there's like so many choices of perhaps there any might strategy was to think we might want to concentrate on where there is a huge painful need in the Linux platform is of the economics for sure for free they can any which for wonderful for free and they mostly do but I don't know how many millions of contextual windows but I do know that get about wants requires a weak and create a summary of the new project always create a fiddle with no shit I would really like to usually my wife would like to use it might go off and would like to use it like that would like to use it it's there there definitely is demand for people who don't want to steal software and don't have the money to to pay for fraction at the heart trust professionals so that they give the missing Akedah well as I mean whenever you have a can cause there's another comes on the city to to decide the sentencing a little something graphic step which usually comes from the labs will nettles In addition to manage demand for it so if you manage to reduce the winners was getting more deletions yeah to 1 and then they put mapping project and we also have about it wasn't users from using Windows and we do not but a lot of effort into them in this part we may have a guy meaning and installers about that surprisingly little developed time basically needs to go into it to make it workable the we to decay and patterns and and would be that much member of the theory the complexity is not high so it is higher and another thing is that I have an experimental group concept much work data integration and German so on the 1 discussed you later so happened Adam how did you guys you guys are going to have a video at a right he figure out some economic model generally does a couple things on bags of get play was that no the available and that's not where your arm for those who don't know long were from your book because a shallow so what you're belongs when those of lines I guess sort of as a counterexample on the some of you may know we actually on had a Windows boarded shall for 1 on and it was because actually potential donors from Google's or stop by your bones in a excited about the 3rd manager and he want certain pictures but he said he really wanted to work on Windows so we actually did good on that path what on and we had it working on wonders and for releases in succession and eventually we think runs about 6 months ago we set it to discontinue the Windows port just as the dominant disappeared nowhere in Brussels wasn't so trivial and vesicular we felt we go actually had ports on but going on run on windows and we think work the Mac for like a day and obviously experience was she k on 1 axis wonderful only 1 when there kernel hobbles along with the word there were many small glitches it has a summary understand much portion on on on that day on Mac OS is

should provide the we we can get to work on so are conclusions there would be if we could have kept alive again the the the money only would have invested in that we thought that would have been at least I don't know 1 developer worth the time or something to work it you was also worked on in the past on in order to the letter on also based on gene on all the ends of but therefore were shot will has been so successful what sort of has taken all of our attention in the video and audio projects on hold on March use there were all worship nongenomic worked pretty well for video on audio in some ways is harder just because you have to think a lot about low latency on so trying to recorded of his strictly and we talk more about that later the the thank you much of the sun but it's still warm of official and and in the whole listen