5 Jan
2009
5 Jan
'09
11:01 p.m.
Themis, I have a couple of questions / comments that I think already came up when you reached out for comments a couple of months ago: I believe what you suggest is a useful feature to have. However, I don't think the fact that the file format that you read in is using XML has any particular impact on the features. In other words, you may provide an API that is agnostic to the file format, and then implement it using different 'backends', one of which is based on XML. In particular, couldn't you implement the whole library in terms of boost.serialization ? What about the boost.xml sandbox project (of which I'm the author) ? Thanks, Stefan -- ...ich hab' noch einen Koffer in Berlin...