Re: persistent vector

From: michel.andre_at_[hidden]
Date: 2001-07-27 02:29:59

--- In boost_at_y..., Greg Chicares <chicares_at_m...> wrote:
> [Have you examined the persistency stuff in the files area?]

Yes I have. I've implemented and use a similar persistence framework 
like Jens Maurers here but with more complex describers (includes 
property name and such stuff to be able to generate xml).

> > > Of course there are restrictions on the data contained in such a
> > > vector it may not contain any dynamically allocated data or 
pointers.
> 
> This seems to focus on the container rather than its elements. Is 
this
> intended mainly as a C++ wrapper around memory-mapped files, rather 
than
> as a more general object persistency framework?

Yes maybe i put the wrong subject on the initial posting. Actually 
it's more to be used as an c++ wrapper around memory mapped files. 
And actually I havent thought about endian issues and alignment 
issues at all. Maybe it's a to simple library for boost? But it's 
greatly needed here ;).

> How great a speed advantage do you get by using a binary format?
> '0x4000c90fdaa22168c235' is faster, but '3.14159265358979323846L'
> works on more machines and produces a human-readable file. Would
> it make sense to use a portable character representation first,
> then examine performance enhancements later?

First and foremost we use mapped files to be able to recover from 
crashes in our program. If the the program crashes and the files 
arent closed the os will write the dirty pages to the file and we 
will have correct data (asssuming the os doesn't crash on us or we 
crash in the middle of a write to the mmapped area of course). 

/Michel