not sure if this is new (probably not): with the latest windows snapshot it seems that running the 'open' command in a script closes the current datafile without asking the user even if the file has unsaved changes. I find this problematic because it has just caused me some (minor) dataloss.

Am 18.12.2007 15:50, Allin Cottrell schrieb: > On Tue, 18 Dec 2007, Sven Schreiber wrote: > > >> not sure if this is new (probably not): with the latest windows >> snapshot it seems that running the 'open' command in a script closes >> the current datafile without asking the user even if the file has >> unsaved changes. I find this problematic because it has just caused >> me some (minor) dataloss. >> > > No, I don't think this is new. And it would be problematic to change > this behavior, IMO. > > A common (for me, anyway) cycle is: open script (which contains an > 'open' command); run script (may generate new variables); inspect > output; revise script; run script again. I would find it very tedious > to be prompted for a data save each time round. > True. However, the possibility of dataloss as I experienced myself makes me feel a bit uneasy about this. As a little safety net, couldn't the 'open' command check whether the current datafile has unsaved changes, and if so, create an automatic backup copy before discarding it?

> Probably a note to this effect should be added to the manual. > > That would also be good.

On Thu, 17 Jan 2008, Sven Schreiber wrote: >> As a little safety net, couldn't the 'open' command check >> whether the current datafile has unsaved changes, and if so, >> create an automatic backup copy before discarding it? > I have been bitten by this again... is there anything wrong > about the idea of automatic backups? Especially now that the > workdir reforms should have ensured that the directory is > writeable. Could you give me some context on this -- under what circumstances does the problem arise?

The general idea is that scripted actions don't ask for confirmation, and I'd imagine that if one wants to save changes to a datafile in a scripting context one would use "store". The backup idea seems a bit backwards to me. Generally a backup represents the prior state of a file that has been modified: the backup is logically "older" than the non-backup. But in this case the backup would be newer than the regular file, containing changes that the user had not saved.