MediaTomb

I am also very interested in multiple renderer support such as in post: [This post][1] I have spoken with a developer, and it can be achieved by IP address distinction - which means that if programmed, all of the media playback devices need to be statically IP adressed (or DHCP statically leased), so they can be differentiated for different endpoints. This will allow several things: 1. The playback from different devices. I have for example several directv receivers that only allow mpegps streams, and also have a PS3 which allows more formats. I also have a mix of wireless clients and wired clients. Wired clients obviously allow higher bitrate transcoding than wireless (802.11g real throughput is about 5Mbytes - not very clear transcoding bitrates).. 2. Price - if peolpe interested could donate to the mediatomb donation page [here][2], it should get implemented with a higher priority. I am a linux user who would be willing to create packages for multiple distros if we can get this done. I would be willing to almost match donations, since this is a feature I would really like to have. We're looking for a target of $150 euros, and I have donated 34 euros, so not much more to go if you would like this feature. 3. I would be willing to provide scripts for transcoding and linux distribution packages if you could help me get this implemented - every euro would help! And - at the end of the day, mediatomb would be a better product for the future! Thanks! Greg Oliver email me with any questions - oliver.greg@gmail.com [1]: https://sourceforge.net/projects/mediatomb/forums/forum/440750/topic/3478056 [2]: https://sourceforge.net/project/project_donations.php?group_id=129766

i guess i'd would just run mediatomb serveral times if i were in the same situation. i wont produces a big overhead as you can share the database using mysql.

Let me comment on this. I have been approached by Greg regarding this feature, while we had plans for it we could not say when we will find the time to implement it. He asked if we were interested in sponsorship, i.e. if we would implement it ASAP if we were payed for this work. Now, we all know the difference between hobby and work, so naturally we have to do payed projects first so we can pay our bills and then the spare time is divided between various things like "real life", friends, hobbies, etc. Since MediaTomb is not generating any profit it falls under "hobbies", we love coding for this project and work on it will continue, but as you have seen, especially during 2009, the amount of time that we can spend on it can significantly vary. Being a freelancer I can decide which project I do next, so Greg's idea would simply allow me to spend more time on MediaTomb. For example, if I was payed for coding a particular feature, I could simply put other "daily job" projects aside without jeopardizing my financial situation. A nice example for this is the inotify based autoscan feature which was financed by Iomega. Although we did have plans for this feature, I doubt it had become a reality in such a short amount of time without their sponsorship. Now, I have no idea if something like that is viewed as "good" or "bad" by the community, especially when it concerns truly free, open source software. I already had bad feelings when I put banner ads on our website, but this was a necessary step - we did not expect that the project would generate any profit, this was never the goal, but we hope that the site can at least sustain itself (i.e. domain costs, etc.). On the other hand, an idea like this opens new perspectives. Version 0.11 was downloaded over 47000 times from SourceForge alone (so we are not counting downloads via distro repositories here), imagine each of these downloaders would donate 1EUR to the project - even if we take taxes into account: this could finance *full time development for the duration of one year*. How does that sound? We already have a very rich feature set and we got here by coding only in our free time, imagine how far we could go if we could code 8 hours per day, 5 days a week for a whole year? Being a realistic person, I do not expect anything like that to happen :) We will continue our work the same way as before (and hopefully better), trying to improve things, bring you new features, and the progress will depend on how much free time we can spare, just as it has been the case during the last 5 years. ---------- Nevertheless, I have a question to all of you: would you guys be interested in something that we could call "sponsored features"? Here's how it could work: - anyone can submit a feature for review, once it has been discussed on both, technical- and wish- sides, the feature would be accepted into the list - each feature in the list would get an effort estimate, making it transparent how many hours it would take to implement, resulting in a fixed cost calculation - the list would be presented on the web page, allowing to donate for the particular feature (alternatively this could be done via something like pledgebank) - once the target amount for a feature has been reached, it will get implemented - other, none sponsored features will be implemented on a regular basis, i.e. just as it is done now ---------- This scenario would allow us to shift parts of MediaTomb development into our "regular work" time, adding it to the continuing "spare time" efforts, and thus allow us to spend more time on the project. Let me know what you think about this. That being said... Happy New Year! Let's get 0.12 out this time :) P.S. > i guess i'd would just run mediatomb serveral times if i were in the same situation. i wont produces a big overhead as you can share the database using mysql. Never ever share MediaTomb databases, bad things will happen! Either clone the whole database, or even easier: decide which instance will be your "master" server, and from time to time terminate other instances and copy the sqlite database to get in sync.

Tough question. With OpenSource it is quite obvious, if you invest much time it's fine, if you stop doing anything at all, it's our problem. As soon this sponsorship happens, the question will be, if you have done 100hours development, 80 out of them because you are paid, does that really mean without the sponsorship you would have done 20 hours development only? Or 60 hours. Or would you have donated 100 hours of your spare time anyhow even if not paid at all? I think that just asks for suspicions and trouble. Or in other words, how do you proof the paid time is on top of what you would have done? Even if you are really transparent with the budgeting of time, how would you proof? How do you avoid accusations? Having said that, I am so happy with the way MT is running now that I have terminated TwonkyMedia although I have a license for it. Doesn't that mean the software is worth something. That would boil down to a donation kind of sponsorship. But this software is used by too few people, you need millions of users to make donation sponsorship working. I like the arrangement with IOMega you explained, as this is a win-win situation for everybody. You have to chat with a handful of companies only instead of each and every individual user. The companies get a clear benefit. And all users win as well. Why don't you take this approach a little bit further? - Contact all companies running mediatomb on their NAS, tell them what features you are planning for next and if they would like to sponsor that development. - Contact all UPnP Client vendors and tell them that if they have a particular usecase in mind that requires a special kind of response from MT server, you are willing to cooperate. The Sony Playstation support would be a nice example. I don't expect Sony would have paid you, but this is the kind of cooperation that could happen. Microsoft: Send them a list of NAS devices - all running MT - their XBox and Win7 cannot deal with because of the UPnP implementation Microsoft has chosen. If they want to be supported... My Geocoding feature would be another example, the UPnP Client could send a search request "Gimme all high rated images within that map coordinates" to Mediatomb and the UPnP Client can put it on their Map. Would look great on e.g. the Playstation Live Globe (found under Network -> Life with Playstation). Or containers could have thumbnails as well. Or background images, e.g. a mosaic of all images in a folder. For the sponsor you would implement and test device dependent responses. - Each sponsor is mentioned on the webpage of yours and on the MT start web page. I think that could be a viable approach also. my two cents

> Tough question. With OpenSource it is quite obvious, if you invest much time it's fine, if you stop doing anything at all, it's our problem. As soon this sponsorship happens, the question will be, if you have done 100hours development, 80 out of them because you are paid, does that really mean without the sponsorship you would have done 20 hours development only? Or 60 hours. Or would you have donated 100 hours of your spare time anyhow even if not paid at all? I think that just asks for suspicions and trouble. Of course, these are valid thoughts. But then again: I doubt that the sponsorship will happen on a bigger scale (if at all), meaning that if we would fully stop development in our spare time, the project would simply die. Two, three, even bigger features can not keep it alive. And then there is maintenance and bug fixing that still needs to be done anyway. > Or in other words, how do you proof the paid time is on top of what you would have done? Right, there is no way to prove that. Just take last year as an example, 100% of my spare time went into car restoration, almost no work has been done on MediaTomb. > Even if you are really transparent with the budgeting of time, how would you proof? How do you avoid accusations? Well, I think main point of this idea is: those who decided to sponsor a particular feature want to have it. And since we will be coding it and thus presenting them with what they want - why should they bother with accusations anyway? As mentioned, if at all, I expect it to be very low volume, but still with a nice side effect that we will simply be able to add more than we could otherwise. > But this software is used by too few people, you need millions of users to make donation sponsorship working. Not so few actually and we are not talking "full development support here". Of course bigger features, like it was the case with inotify need more budget, we'd probably need a couple of months instead of a week if we coded it in our free time. But often you have things that will take 3-4 hours or maybe a day, and although that seems easy - we often can't find the time for it, because there are open bugs, problems on other distros that need to be fixed, someone wants a static package for his new NAS box, questions in the forums need to be answered, and then there is some cool feature that we are working on and that has a higher priority on our list than some of the "smaller" ones. I think in those cases this idea could do the trick and provide us with additional time to get more things done. > I like the arrangement with IOMega you explained, as this is a win-win situation for everybody. You have to chat with a handful of companies only instead of each and every individual user. The companies get a clear benefit. And all users win as well. Yes, it was great for us as well, but this is the only case in our history when something like that worked out. We had talks with some companies who wanted various features, but mostly it did not go anywhere. Either they did not have enough budget for something that was actually a lot of work, or things got stuck in the bureaucratic decision making process. We also had good experience with eXcito, who run MediaTomb on their Bubba server, but this was more a friendly relationship where we got the hardware and offered support and consulting regarding setup, etc. I tried contacting companies, both player and NAS manufacturers, but it seems that most do not seem to be interested. > Microsoft: Send them a list of NAS devices - all running MT - their XBox and Win7 cannot deal with because of the UPnP implementation Microsoft has chosen. If they want to be supported... Sorry, but that sounds funny :) If you look at the Xbox implementation you will see that *they want it that way* and they surely do not care about some open source software not working with their devices. Especially since almost all commercial UPnP servers added hacks to support their broken implementation. Working out "special" features with some UPnP client manufacturer would be nice, but I work in a consumer electronics company myself and know how difficult it is to develop something like that within the constraints of a running project. You have to be a big player with a lot of budget and be willing to invest in improvements for your customers. If we look at the UPnP players that are available today, we will see that most companies just buy a reference design, add their branding and a minimum on customizations and throw unfinished products out in the marked. Hell, I heard of devices that displayed popups "this feature is not yet finished" :) Sadly, it's not about quality in the mass CE market. Often they won't even fix a bug in their player because the server they provide with it has some hack that makes it work with their device and it does no matter if the implementation is simply wrong in regard to the spec and fails with all others. Anyway, I guess I got distracted from the topic :) Thanks for your input, I'm really curious how people see this.

Fully understood. Btw, last input from my side before I shut up: Go to the Marketing guys at the companies. The engineers do not have control over the budget but the marketing department has. Tell them there is the opportunity to be placed prominent on your webpage with so many hits for little money, that they get named as a supporter "This MT feature was brought to you by xyz" and yet they get a beneficial feature to their customers as well ahead of time to the competition.

Hello. Just wanted to chime in for what it's worth, and say that I would be willing to put some money towards a proper Mac version of MediaTomb. I did manage to compile and install it on my Intel iMac about a year ago, but it's a lot of hoops to jump through, especially for Mac folks who are used to drag n drop solutions. It worked fairly well though, I was impressed. Anyway, I've since reinstalled my OS, so I don't have MediaTomb anymore, and have gone to other native Mac solutions. However I'm still converting/transcoding most of my files before I can play them on my PS3, no matter what media server software I use. So I'm always in the market for a better solution. Long story short, I support your asking for donations and sponsored features. I hope several people can chip in and get what they want, and the project continues to move forward. I can't donate enough for you to hire a full time Mac developer, but hopefully I'm not the only Mac user out here. :) Best of luck! I'll keep my eye on this project and maybe give it another go at some point.

> Just wanted to chime in for what it's worth, and say that I would be willing to put some money towards a proper Mac version of MediaTomb. We do recognize that even though we are available via Fink and darwinports, it's not what Mac people are used to. Our main problem is: currently we simply do not have Mac OS anywhere, so we simply can not create those images (unless it is possible without having a Mac/Mac OS?). > Long story short, I support your asking for donations and sponsored features. I hope several people can chip in and get what they want, and the project continues to move forward. I can't donate enough for you to hire a full time Mac developer, but hopefully I'm not the only Mac user out here. :) Best of luck! I'll keep my eye on this project and maybe give it another go at some point. Thanks for your input! Well, for that feature someone would have to donate a Mac OS VMware image (is that legal?) :) But basically that's the idea, if people have special feature requests that they wish to get done faster than via the regular list, they could submit the features for discussion/technical review/effort estimation and once accepted, everyone who also wants it could donate for it. Unfortunately, in your particular case we would not be able to accept it because of technical limitations on our side.

As stated, this is a tough question to answer because there's a lot of possible repercussions. I've always been one to try and donate anyway, and did so when I first downloaded MT and realized I would continue to use it. I'd like to think that everyone does that, but I'm not so naive to think that everyone does. My local public Jazz radio station has the same problem. Hundreds of thousands of listeners and I think about 4000 paid members. On the surface, however, I think it's a good idea, and I would certainly be willing to support to develop the feature(s) I want to see. I must admit I have grown impatient on occasion due to lack of updates, even while fully realizing - and understanding - that it is currently a "hobby" for you and you need to support yourself with a real job. One potential downside is that, some folks are honestly unable to contribute - their financial situation just doesn't allow it. And, unfortunately, they may be the best and brightest among us. For instance, I, for one, would like to see built-in support for transcoding to the PS3. I've done enough mucking about in MT to get my DB structure where I want it, etc., but the whole transcoding thing is a bit beyond me, and at this point in my life I'd rather just wait on you folks to do it. And, having contributed something, I don't feel guilty asking for it. So, assuming I had deep pockets, what's to prevent me from funding that to the exclusion of everything else? Maybe the PS3 is a bad example, because a lot of folks use that, but there's probably some pretty obscure features that others would want, and support even though a lot of others with fewer resources couldn't care less. At its extreme, I think you could potentially start a bidding war - "I'll give you 1,000 Euros to work exclusively on my feature requests until they are done" - great for you, probably not so great for MT or the MT community. I know you say that a pricing list would show how much cost to implement a feature, people would contribute to that, and then "once the target amount for a feature has been reached, it will get implemented " - but I'm afraid it wouldn't work that way I don't think, just because soon you would (again) have too many features to implement in a timely fashion, and then people would *really* feel justified in publicly complaining, because "I paid for it - what's the holdup?" I'm not sure you want to put yourself in that position. I'm rambling a bit here, but those are my initial thoughts.

Thanks for contributing, we appreciate everyone's support, but I guess the outlined idea goes beyond that. > I must admit I have grown impatient on occasion due to lack of updates, even while fully realizing - and understanding - that it is currently a "hobby" for you and you need to support yourself with a real job. Hehe, yes, my car restoration adventure this year really stalled the whole development :) Those of you who want to see the whole action: http://www.deadlock.dhs.org/jin/UAZ/remont/index.html > One potential downside is that, some folks are honestly unable to contribute - their financial situation just doesn't allow it. Well, they would not be in a worse position than right now - the plan is to continue with regular development same way as we were doing it before. Of course no one can say how much we will be doing and when, we had years when we were able to do more, or years like 2009 when almost nothing happened, so yes, regular development is quite unpredictable. > For instance, I, for one, would like to see built-in support for transcoding to the PS3. I've done enough mucking about in MT to get my DB structure where I want it, etc., but the whole transcoding thing is a bit beyond me, and at this point in my life I'd rather just wait on you folks to do it. I'll be honest regarding that feature: I had a look at it and it's really A LOT of work to do it, especially to do it right. The fact that I do not have any experience with all the codec stuff and associated libraries does not help either. Hell, I often can't even figure out a VLC command line to transcode something, so I try stuff from our user contributed scripts too :) What I think we could do in the near future: add time based seeking support, this would allow pausing and seeking in transcoded streams. Add dedicated profiles, i.e. hand pick special configurations, for example for a PS3, and offer config.xml files that would be tuned to a particular renderer, or a command line option to generate a config file that is tuned to a particular device. I think this would greatly improve the situation, the user would only have to install the transcoding agents (like vlc) and would not have to look at the config.xml at all. Adding native transcoding will not happen anytime soon, it's a shitload of work and I currently see other more important issues. > And, having contributed something, I don't feel guilty asking for it. I think no one should feel guilty, asking for a fix or a feature. Worst thing that can happen - that we don't do it :) Then of course there are features that are more interesting to implement than others, or features that are easier to implement than others, I can't really say that there is a particular selection criteria right now. We think what we would like to add, what would be nice to have, and also at what the users want, and then pick something and do it. Often, our mistake was (and still is) that we start doing too many things at the same time, just look at the number of unfinished things in SVN. We should definitely do better there and show more discipline. > So, assuming I had deep pockets, what's to prevent me from funding that to the exclusion of everything else? You wouldn't - the idea is to generate more time for us, that we could work on the project. So assuming you had deep pockets and could sponsor couple of weeks of internal transcoding development - you'd have it. But ultimately this would mean: you allowed me to work on the project instead of doing regular work, you were not taking away my spare time. Of course one can always argue and suspect, if we would have enough motivation to add another couple of hours to the day, in addition to your sponsored feature, but I think that several years of ongoing development show that we are still having fun here. > Maybe the PS3 is a bad example, because a lot of folks use that, but there's probably some pretty obscure features that others would want, and support even though a lot of others with fewer resources couldn't care less. Well, as outlined in the earlier posts - not "just any" sponsored feature would get accepted. If you requested something, that I either don't want to implement for whatever reasons, or that I think makes absolutely no sense or goes against the project's spirit - I wouldn't accept your request. > At its extreme, I think you could potentially start a bidding war - "I'll give you 1,000 Euros to work exclusively on my feature requests until they are done" - great for you, probably not so great for MT or the MT community. Have to disagree here again, main idea is to do more development than we can afford now, so people would get your feature in addition to the regular stuff, and not instead of it. > I know you say that a pricing list would show how much cost to implement a feature, people would contribute to that, and then "once the target amount for a feature has been reached, it will get implemented " - but I'm afraid it wouldn't work that way I don't think, just because soon you would (again) have too many features to implement in a timely fashion, and then people would really feel justified in publicly complaining, because "I paid for it - what's the holdup?" I'm not sure you want to put yourself in that position. Surely a valid point, my thoughts on this were: - make sure the list is small at all times (i.e. have a fixed number of slots in the queue) - order is visible, i.e. first come first serve - workload is transparent, since price is calculated via effort estimates That means it would be visible what is being worked on, the ETA till it's done, and what's next in the queue. But still, true, *if* there is any holdup complains would be justified, meaning that the selection process should be done very carefully. I surely would not want to end up in a mess by accepting just anything :) > I'm rambling a bit here, but those are my initial thoughts. Thanks for your input, you raised some interesting and valid points here and I fully agree that this whole idea is not as simple and easy as it may sound.

I'm also interested in this feature (multi-render), so I've donated a few euros for that. Thanks in advance.

Personally, I tend to take the offer for 'sponsored' work in the spirit that it was offered by someone who has provided me with a great product for zero cost. If I can get a couple kinks ironed out with how MT plays with my WD Live, I'll happily donate. I wonder how it would work if instead of letting people vote by donating, let people vote and then allow people to donate for the winner. If I knew that you had a cool feature that you would like to work on and I wanted, I'd happily donate a few bucks to try to get it done sooner (or done at all). If you could get a couple hundred PS3 users to each donate $10, you could put your freelance work on hold for a week (or so, I won't pretend to guess how much you make) and kick out something pretty quickly. Of course, you have the case where people will complain about how long something is taking or whatever. Some people like to complain and there isn't much you can do about it. But I think the opportunity to vote for, and have an impact on getting some cool bit of functionality that I really want, is appealing. And anyone who gets too bent out of shape because they donated $5 and don't feel like they are getting their money's worth... Well, just put up a sign telling those people not to donate in the first place. In any case, thanks for a cool bit of software. Chris

Well, based on the mails and posts I received so far - only a very small amount of people seems to be interested. This tells me, that this whole idea will not work out and we will simply continue as before. Nevertheless it was an interesting test. Personally, I tend to take the offer for 'sponsored' work in the spirit that it was offered by someone who has provided me with a great product for zero cost. If I can get a couple kinks ironed out with how MT plays with my WD Live, I'll happily donate. Well, this is a little more complicated. When coding features the most important factor is time, when improving renderer support that is just not enough: you need the player. In an ideal world players should just implement the spec correctly and everything should work - we know this is almost never the case. Of course we ask our users for logs and do remote debug sessions from time to time, but really good support for a particular player is often only possible when I have that player in front of me. And once I have that we are back at the "time" point - how much effort is it to get a particular player going? For instance, we refused an Xbox and we also refused a PS3, having to add full support for those would occupy too much of our time and we prefer to spend that time on new features that would be available to everyone and not just to a group of users of a particular device. This does not mean that we are not trying to improve support for various devices, but it's still a difference compared to focusing on one device and leaving all other stuff aside for some period. So basically, donating for better device support does not make too much sense, because often we would not be able to do anything without having the device in front of us. And donating a device would only make sense if you'd first ask us if we are willing to work on it :) Anyway, if you have problems with your WD Live, just post in the forums and we'll see what we can do. I wonder how it would work if instead of letting people vote by donating, let people vote and then allow people to donate for the winner. If I knew that you had a cool feature that you would like to work on and I wanted, I'd happily donate a few bucks to try to get it done sooner (or done at all). I think this is a good idea and this is indeed something that we could try, this approach is more simple and more straightforward. Thanks for your feedback!

Maybe this thread is now old news. I have been using MediaTomb for a long time now and it is great. I haven't tried version 0.12.1 because my TV does youtube. However, I think the sponsored feature idea is a good one. One the other hand, how much would you need? Let me give you an example of a feature a lot of German people would like (and I guess UK people as well): Many of the official TV channels now offer their programs over the internet. Here, they call them Mediathek, on the BBC it is the iplayer. If those could be streamed, it would save people the hassle of a PC attached to their TV. OK, now the downside: How many hours are needed for this feature, short term and long term support? 80 Hours of coding plus 80 hours of bugfixing? Assuming a low hourly rate of € 50/hr, you would require € 8000, so 800 Users to contribute € 10 without seeing a product straight away. So, it is difficult. I would sponsor € 20 for such a feature.

Well, it was an interesting investigation but the result was, that it's not really doable and the interest in something like that is very, very low. And now I anyway do not have the time for such experiments anymore, so that is old news indeed, sorry. Right now I can't say when 0.13 will be out, work on it has not even started.