draft of a letter to the transmission network „subtitles – new direction”
this draft has 2 main parts – why (boring – 1) and how (hopefully interesting – 2) change our working group into something that can actually make a difference in our video-activist community and outside of it.
in this version the idea is mostly my but would never emerge if not for the discussion that I had with Alex and the rest of RaFilm so if it works I own them a beer. and thats gonna cost if it’s meant to be a swedish one…
As it emerged from discussion on the meeting in Lund, it might be a good idea to change the main target of our group. As setting up a work flaw for a production of a multi lingual DVD might be interesting, in case of the tools that we used up till now, its far from being a step forward, and what I’m going explain further on, might be actually a step back. I made my mind about that recently, as everybody knows I’m participating in the subtitles group since „re:transmission” in London, and till now I supported the last „plan”, so it might be not a perfect explanation, nor it shouldn’t be considered a finished idea, pleas, write back, do some research, if you get nervous, get a tee, or a beer.
The tools that we used for translating the last DVD was a mailing list and subtitle workshop (windows only, „free” (not FLOSS) software btw), which is being used probably since the rise of indymedia , which is a good few years ago. This way we don’t engage anybody besides the activists in the translation, and it’s not really a collaborative thing, as each of the translators, unless they really try hard has to work alone, as subtitle workshop is a client-side software, and there is no useful way of using it while working as a group. As well the mailing list it self, being brilliant as it is, does not help with the translation management, as its rather a tool of passing/spreading information then a project management tool (of course it can be used for this purpose, I just don’t want to waste time in pointing out whats wrong about it, to put it straight – you can eat a soup with a fork, but why do it if you got a spoon at the reach of your hand?)
As I said, that was and is used for good few years, and for us to document a work-flow on that doesn’t bring us anywhere. It’s like we would decide to describe a standard phone line. Something that I guess has not only been done, but is of no use, besides a historic interest.
The DVD it self can be created by anybody, with an image of the disk being uploaded to one of the servers that we got, to deliver it to all the other groups, it’s not hard to learn how to do it, and we can use the fact that were a network, and for example ask another group to help out with the authoring.
There would be no point in this critique if I wouldn’t have a better idea.
What I would propose, and in my consideration, would brilliantly fit into our puzzle(jigsaw?), is a server side subtitle management and translation tool. Not really creating one – they already exist, only until now, they had a different purpose, and needs only small alternations. The one beneath is an example, I don’t say that we have to use this particular one – its meant to show you how it might work. And endorse research on similar systems that I might not know about.
Some of you, probably the „geekier” ones know „launchpad/rosetta” a tool used by canonical, the company behind ubuntu linux, to manage the project, its used for bug reports, blueprints, and few other things that were not interested in. What interests us is it translation part „rosetta” which is the part used for translating not only the system itself but also the software around it and quite any that’s open source and needs a place. Including Democracy Player that all of you should know.
In this place, it would be best if all of you just log in to see how it works, I created a temporary login for this, email me and I’ll send it, or just register yourselfs, then go to translations of democracy then pick one, click it (for example the polish one), anyways what you hopefully see, is a line of the untranslated text, propositions of translation suggested by „rosetta” and then the line to put in your translation, if non of above matches. What is proposed by „rosetta” is what another translator wrote in a translation of another piece of software. When the translation to a particular language is finished (indicated in a graphical form with a color stripe and in statistics, you can see the overall translation progress to all languages for example here: again the main democracy translation site) or even before the whole translation can be downloaded as a .po file, which is a standard open-source language/translation file for a piece of software. Other things that should be noted is that there is no limitation as to how many people can work on one translation at the same time, as they never operate on more then one sentence/line at any given time, other people can see translations that are done at the moment, there progress, and the text of it, and change it, if they think that its not translated correctly.
I think that by now most of you all ready know what I mean. A movie translation, or to start with, transcription, has a format similar to the one seen in the example – it not only can be, but since its beginning is split into lines, marked with time code detailing when they should be shown. So lets say that instead of the text of a program, we get a text of a movie there. Now lets say, we already translated whatever amount of simple movies, like, activist videos from demonstrations. Now every time I got another „x number of people got arrested” or „fuck the G8, shell to hell” the program automatically tells my that somebody before, translated thous to polish as „aresztowano x osób” and „pieprzyć G8, shell idź do diabła”. There more videos get translated the more probable it is that the whole sentences are going to be translated by somebody before, and will just need marking that their right, as language has a limited number of combinations, and we tend to use the same sentences and words, that are in popular use, instead of creating new ones. That rule is used by few more different (open source) programs like translate or OmegaT. .
Going on, lets say, that I got a screening in half an hour, as the revolution is starting, or as usually i forgot that I’m organizing it, the crowd wants to see „surplus” or the „corporation” and my friend who was meant to translate them is pretending that (s)hes out somewhere in Siberia, as (s)he didn’t finish the promised translation. I end up with nothing thanks to a „client side translation” meaning that he/she is translating that on his/her computer, that i haven’t got access to. In case of a server side solution, i just log in, and download whatever is there, cuz that’s better then nothing. Did you notice the „export to .po” link? I don’t think that it would take a lot of work to change the export function to get an „export to .srt or .txt”.
What would be the biggest advantage? That we can group all the existing translators into one place. Instead of film makers and translators both searching for each other, we got one place. You make a video, upload it somewhere so people can see it, and upload the transcript in your language or/and English to one website which runs our brilliant server side translating management tool and zap! – people just translate it. just as they do with every open source program. Just as they do at the moment. But thanks to the fact thatther’s a website with all of them – they are easy to find, easy to translate, easy to download. Further more; we could connect that with the meta data standard. Further more we could decentralize that central side, if we find/write the wright solution – every group could have it’s own part of it, or the whole database on their server, and update them throe a rss/atom feed (or anything else) with all other copy’s. And last but not least – remember what Szczym dreams about? Play our trailer :]. What problem will it to integrate subtitles into Democracy player, if podcasts that people want can be translated nearly instantly, if the author provides the transcript? On the side of developers it would be just acting against their own project, not to include a feature like that…
btw. for native-english – most of You doesn’t even imagine how hard it is to transcribe (write down the whole text) a movie not being a native speaker… doing that takes around 50-70% of the whole translation time, so even if you don’t know any languages besides English, or any other that the video is published in and is your native, you can always transcribe it, and make somebody’s life way easier.
After writing all of that I had a chat with one of the people behind launchpad, trying to get to know if it’s possible to use this piece of software, generally what we got to is that yes, if we sort out the extra coding involved; time-code, export to desired formats and at least an idea how to manage a sheared database/keep it connected, (not to completely split the software one from our translation one, at least if we want to use the basis of rosetta ), he might help but it’s us who should do the work, anyway seeing this particular person interested and positive, convinced me that its a step in the wright direction.
At the moment I wonder with it wouldn’t be possible with a little coding (export feature) adding extra cells for the time next to the translation (so it can be hand edited, and assembled together when exported) and some css magic to hide the stuff that we don’t need, and then connecting that to our website… but thats rather a question for the ones with skills well higher then my own, and only if were not the only people who think that it’s a good idea.
anything to add/correct?