This software has been pulled because of two really bad reports on data trashing.
Model the web in your mind into software. Link together web research, your personal journal: any kind of information.
It saves all your content in plain text. It auto-saves your content. It automatically detects links as you type them (http urls and intra links with square brackets).
The following modifications to the url proc make clickable web links possible in Mac OS X.
proc url x {
set x [regsub -all -nocase {htm} $x {ht%6D}]
#exec rundll32 url.dll,FileProtocolHandler $x &
set h [open /tmp/a.script w]
puts $h "open location \"$x\""
close $h
exec osascript /tmp/a.script
file delete /tmp/a.script
}There may be better way to do this, but this'll do for now.
I've also tweaked the render proc to handle mailto and may get it to handle ftps as well. As opaque as I find Tk at the moment I can at least figure that one out.
By the way, much as I like mindweb in it's current incarnation, the fact that I can't do drag'n'drop is one reason why there'll soon be a REALbasic incarnation. Or else, someone's going to give Mac OS X's Wish the functionality I need.
A really effective way of trashing everything in the ~/data directory with MindWeb under OSX is to give the three-fingered salute aka vulcan nerve pinch aka control-command-powerbutton while MindWeb/Wish are running. I was forced to do that today (a very rare occurance) and all of my MindWeb files were filled with garbage, completely replacing the original content. Bliss ... NOT!
Fair enough, as this is the second report, I'm pulling the software. Ro
Which is kind of sad. Is it that MindWeb is the problem or that the aqua version of Wish (that comes with Alpha) is the problem?
Mindweb writes to disk a bit too often for my comfort; saving everything, changed or not, based on a timer. I assume Rohan did this for simplicity - and despite the risk of all data being trashed in one hit, I applaud that, because it makes Mindweb a great starting point for developing the idea the way you want. For example, I'm planning on using vfs to load and save pages via http to a webserver. I'll lose some of the raw speed, and danger ;) of the all-in-memory approach - but not to worry. I too think it's a shame to take the software out of circulation. Perhaps just make a note that for the above reason; in it's current incarnation it's more suited to be used as a development starting point than end-use? Julian M Noble