Torrent render for 3ds max — almost everyone will have to redo

Continue my series of articles about the project torrender.net, because the current situation is again far from my previous plans, and the most interesting for me, perhaps, will begin in the near future.

Previous articles:

History torrent render for 3ds max
the History of progress torrent render for 3ds max

Background:

So, it's been four months since the release, during this time, I finished the project to the stage where using MAXScript already nothing substantial to add was impossible, and it introduced the breakdown of the render single frame in different seeds and the collection of pieces into one picture on the side of the Lich via ImageMagick, this feature was not ready at the time of release and I pinned great hopes on it, because after all, max is used primarily for architectural visualization, and there is that rendering a single frame and lasts three or four hours, that's really then that wish will appear and the system will finally become active, as willing to render static is many times more than those who render the animation. But I was wrong, the feature breakdown of statics passed by user, not increased, neither the gob nor any significant number of new nod, despite the fact that I posted info about the static on one of the most popular forums "Maksimov-interarmco" — 3ddd.ru didn't help. And, if so, it is itself a thought — and what I sense to implement a local network render, attach custom render farm (it was the next and last stage of development), even if the static no one wants to render? And is there any sense at all to continue if the users just do not need it, then what do they generally have?

On this question, there the answer is simple — the button "render nicely"

After reading here and there (and on habré in particular) various comments, I came to the following disappointing conclusions for themselves:
the

the
• System will not be massively used in its current form
the
• If you continue, I will have to go through a much more complex way and alter everything except the tracker
the
• there are No guarantees of the success of the new system

In the end, a torrent client should be similar to existing applications for local network rendering (such as backburner) and work as a local render farm, only the nodes located in the next room, and somewhere on the Internet. Get: any number of node to job, jobs of any size and complexity, the breakdown of the single-frame according to different seeds (you can render heavy static), rating system, a mode for friends to connect local farms, and it's all free to the end user. Well, by analogy with torrents: a regular torrent user downloads the file he needs and begins to distribute. In torrent render it distributes the jobs file and downloads files of frames (or frame fragments) — distributes one shakes more.

What I have to do:

the

the
• Completely rewrite the client from MAXScript to C#
the
• the Client must be in the system tray of Windows and start 3ds max (or another 3D package) for rendering
the
• Scripting part I will deal only with the creation of the job and this process needed to be simplified for users (to make Samobor all supporting files)
the
• Protection of scenes to remake such a way that does not transfer the scene files for this actually have to invent a new format for the jobs, i.e. to create scripts that generate a custom scene (or at least part of it)
the
• you Need to make a new client was useful to the user even if he renders the scene, without the involvement of other volunteers, namely: to make auto-sending a finished image by email, fill in ftp, deferred rendering, multi-jobs (i.e. the easy run on the render multiple scenes in a row) and other such stuff as a bonus

Why C#? Actually I first wanted to go the easy route and try to do it on AutoHotkey, but then I remembered that on my macscripter is full dotnetskin inserts in C# with them will be much easier, and besides, if I had to pick a new language to learn, better still, something bigger than AHK.

After considering all the pros and cons, remembering previous years of designs and picking in PHP, I realized that the idea of remaking the client essentially from scratch, plus the study of another language I don't like in case, if again it will only be in one of my pure enthusiasm and without any guarantees of success of the future system. There would need to insert a picture, where the gloomy genius in the head lamp lights up because the thought of "Russian crowdfunding" about the way arose immediately materialized in the next post on 3дцентре:


currently I have no more opportunities to continue to develop entirely on its own, because the money and free time have come to an end, and so the project goes to crowdfunding, a specific platform will choose on the day. What in the code project you need to me personally, I already did, but the users give very different features, I suspect, so the node offline and sit. The actual collection of funds and the need to implement the user desires (its something I did for free), and if you do not employ the means necessary to nobody, will save time and I will not continue to wallow away for free.

The new features will require a lot of time to implement and I don't want to risk Pragati in such a scale on one enthusiasm:
1. The client will copy from MAXScript to another development language, it will work in the Windows system tray discreetly and independently of max (torrent-client)
2. Max is used and only starts to create job and to render the whole logic of the client, not directly related to max and render will work without max — it is also a Foundation for other packages
3. Samobor stage all related files (textures etc.) a machine for creating jobs.
4. 100% protection of scenes — scenes not))) ie, no max files. If anyone remembers, there was a script from Bobo that the current scene is translated into a set of commands MAXScript, i.e. generate a script that creates when you run the same scene — that's about will. Without a client to access this job would be physically impossible, and because the client will not macscript, hackers to crack it and rip the parser job quickly fail (if at all possible).
5. Scene and the final render will be possible to distribute not only the torrent, but upload to FTP or send via email.

z.s. If crowdfunding is successful, the project will rebranding will be a new name and a new domain.


Well, more a matter of technique, determined to launch a crowdfunding campaign, I actually launched planeta.ru but within a month, the project managed to recruit only 7 thousand out of 200 thousand rubles, less than a week, so do not be surprised if this doesn't work, and again I have to puff on the enthusiasm in your free time as always, C'est La vie.

Article based on information from habrahabr.ru