I fully agree. The ability to convert trading platform language (Amibroker's AFL, EasyLanguage, etc) to universal languages like C#, C++, Jave, etc would be wonderful! If this worked, I would buy it almost immediately.
clean database, period. I wished there was a vendor who offers clean data, data that can be accessed lightning fast, such as through KDB. A database that gets me tick data, market depth, bars on the fly without delay, data that is cleaned, backward adjusted for corporate actions. It does not exists, period. No Reuters, no bloomberg, no other vendor can offer it. Even most investment banks, despite a host of people at data groups, dont seem to be able to cut it. Sad state of affairs if you ask me, but the status quo.
Sorry guys, I'll have to be a bit negative here: Look at the history of programming language translators. It's always a complete mess and they hardly ever arrive at their purpose. Even with languages that should be really close like Visual Basic to Visual Basic .NET the translator (supplied by MSFT) is a mess and as soon as the programs translated become non-trivial the output of the translator is frequently inappropriate / needs much hand work.
I wouldn't generalize here. f2c (fortran 77 to C) compiler works without a hitch on large computational libraries, algorithms in which are by no means trivial. Where VB to VB.NET compiler fails is not the laguage is the narrow sense but "library" objects. How forms, database objects, library ActiveX controls operate in VB and .NET has subtle differences which make translation non-trivial. Microsoft, on their side, have a superimposing objective to make everyone use .NET. So, they wouldn't provide "compatibility" features and would rather live with some translated code not working properly. Regarding EasyLanguage, similarly the main challenge will be to translate visual elements of TradeStation charts (which may have no analogues in other platforms) and the like....