Hi,

We have had a mechanism for distributing our applications and related files to desktops that has worked pretty well for us for several years. Basically we store the files (VFP.exe, word docs, etc) in the SQL database - split up into 8000Byte chunks - and then reassemble them via a loader program on each desktop. I'm getting ready to develop a new "loader" app that is more full featured and comprehensive in scope and seek opinions on this.

1 - My biggest question is the wisdom/need/value of splitting the files up as opposed to just uploading entire (dynazipped) files into the database. The 'logic' behind the split had to do with being able to recover from partial loads if something went wrong. We would have no files more than a few meg in size - and most files are meadured in K.

2 - What is the best data-type for loading binaries? binary? image? ? We've used binary in the past which explains the 8000 byte chunks. Image looks the most promising for larger files - is this accurate?

3 - On the front end - is a VFP general field the only mechanism I have for moving the data up/down? Can I use filetostr()/strtofile()? something else?

Thanks!