# What does lyx munge about the file format to the convertors?

(I'm using Lyx 1.6.2)
A lyx file of mine has this text (when I look inside the raw lyx file)
*
--8<---**--8<---**--8<---**--8<---**--8<---*
*\begin_layout Section
The ntvfs model
\end_layout

\begin_layout Standard
Many vfs modules may be defined, as types of share backends.

\end_layout

\begin_layout Standard
Each vfs module could be considered to subclass an abstract [[struct
ntvfs_modul
e_context]], and each vfs module should define a struct to define
per-connection
instance data.
\end_layout*
*--8<---**--8<---**--8<---**--8<---**--8<---*

Apart from the puzzlement of why Lyx breaks the line in the middle of a
word, it all looks fine.

But this is NOT what is fed to the first stage of the Noweb->Lyx noweave
convertor.

If I prefix my convertor with:
cat i | tee /tmp/pre
and look inside /tmp/pre, I see this instead:

*--8<---**--8<---**--8<---**--8<---**--8<---
**\section{The ntvfs model}

Many vfs modules may be defined, as types of share backends.

Each vfs module could be considered to subclass an abstract {[}{[}struct
ntvfs\_module\_context{]}{]}, and each vfs module should define a struct
to define per-connection instance data.
**--8<---**--8<---**--8<---**--8<---**--8<---*

Whats going on? Apart from the annoying and bogus quoting of the [ ] and
_ which I'm trying to fix-up, it means that noweave when run from Lyx
will give different answers to noweave when run from the command line -
say a Makefile which tries to generate the pdf.

Sam