Jan Harkes wrote:

> . . . As Coda caches whole files, it is not very useful to actually have
> larger than 2GB files because it takes a lot of time to fetch/store such
> a large file (I know networking is getting faster every day, but still).
> However, being able to run Coda on 64-bit architectures and supporting
> larger than 2GB venus-caches are very useful indeed.
>
> Jan

I'm not to concerned about file transfer rates of large files since most of the
data will be generated locally on the client and aggregation doesn't need to
occur rapidly anyway. The performance numbers we found while testing using
adaptive mode looked really good (understandably because of the asynchronousness
of adaptive mode) but the limits on file sizes is concerning enough to pursue
working on enlarging those. With Linux's upcoming 64bit filesystem support even
for 32bit platforms, getting 64 bit support into coda would make coda more like
irreplacable rather than just great ;-)

Is anybody already working on 64bit support then for coda for even 32bit
platforms? If not, how hard do people think it'd be to go 64bit? Anything we can
do to help?