 covered basically this already earlier you're on in the engineering student media I've been repeating myself and I'm quite bitter of the last few days but anyway yes so after the WI conversion stuff is complete there will be very little left in the existing bit existing SRC file format what remains is basically strings string lists images and image lists so the question is should we continue to go from the dot SRC file to the PO file format back into a whole bunch of other file formats and then back into the final binary dot RES file format if all we're doing is basically transporting strings around the place to be translated so at the you are there there is the area worked by and Russia looking into this and the line at the time there was that the size of the current PO files approximately that's without RES files that may not be actually true so we may have to think about that a bit more but if we assume that we can't keep the size about the same then the process of how we take all the existing translations we have in the dot SRC file to move them back into the dot CXX file so that developers can see when reading the source code what the string actually is instead of having a little enumeration or a hash to find that you didn't have to look up on the SRC file to see what the content is to be right there in your dot CXX we do some markup like you get from the standard get takes maybe an underscore or something else and then as a developer you no longer have to read what translation actually is it's a different view use the standard X get text application to extract translations into that PO file that already has support for dot UI files already that already exists so the standardized format would allow Deckard which gives you an in place view of what you're translating when you are translating something that's based in the dot UI file format of your translator we have to see in the website of what they're translating then use the standard dot PO dot MO process which then leaves how we load the result to translations get text itself isn't licensed properly but boost does have a boost dot locale which comes with a loader for the MO file format and icon the placement which is apparently suitably licensed and supportably safe we do that instead that would then mean that the entire orc SC directory could be deleted hopefully quite a large amount of the L10 directory could go as well in the conversion tools from one to the other and then all the parts of tools above the race manager could all go away as well a lot of that race manager stuff has got some very nasty life cycles where it all gets deleted on application exit which causes all sorts of complications and static analysis tools so you can place all of them with some new more exciting words that would have similar problems from boost and locale but at least there is a possibility that because it is used by Blender which is a large application that would be suitable for large application like a row