Hi guys, I'm working on the configuration file format for the DiscFerret. As I might have mentioned earlier, this is based on the Lua programming language -- in fact it *is* Lua. To load the config, the DiscFerret tools fire up a Lua virtual machine, compile the config file into bytecode, then run it and pick the config values out of the VM's global state. What this gives you is an absolute ton of flexibility: you can do things like use a FOR loop to generate repeated Drive or Format specs. Something like this: for kv in { "pc35a", "pc35b" } do drivetypes[kv] = {} drivetypes[kv]['friendlyname'] = kv drivetypes[kv]['crossed_cable'] = true end In the DiscFerret implementation, you have two config files: - DriveSpec: these tell the software how to talk to a given drive. Basically, what to do with the I/O lines. - DiscSpec: these tell the software what the parameters of a given disc are. Minimum/maximum track numbers, double stepping required (or not), hard/soft sectored, and so on. ( There are also DecodeSpec files, which MagDAS uses to decode transition data into disc images -- and which I'll document when I've actually designed them! ) Here's an example DriveSpec file: --------------------8<-----------8<---------------------------- --[[ ####################### # DiscFerret Disc Drive Specification File # # PC 3.5in Drives A and B, twisted cable or CBL-01A cable kit ####################### ]] -- DriveSpec version flags drivespec_version = 1.0 -- Drive types recognised by this drivespec drivespec = { pc35a = { -- Passed to Drivespec functions drivetype = "pc35a" -- Shown to user friendlyname = "PC 3.5\", twisted cable (CBL-01A), drive A" -- Default step rate in milliseconds steprate = 3.0 -- Spin-up time in milliseconds spinup = 1000 } pc35b = { drivetype = "pc35b" friendlyname = "PC 3.5\", twisted cable (CBL-01A), drive B" steprate = 3.0 spinup = 1000 } } --[[ Given the drive type, track, head and sector, return a list of output pins which need to be set. Called once per sector on hard-sectored media, once per track on soft-sectored media --]] function getDriveOutputs(drivetype, track, head, sector) pins = 0 -- 3.5in FDD settings are really easy to set up. First start with drive selects. if drivetype == "pc35a" then -- Shugart DS0 = motor enable A, DS2 = drive select A pins = pins or PIN_DS0 or PIN_DS2 else if drivetype == "pc35b" then -- Shugart DS1 = drive select B, MOTEN = motor enable B pins = pins or PIN_DS1 or PIN_MOTEN else error("Unrecognised drive type '" .. drivetype .. "'.") end -- Handle side selection if head == 0 then -- do nothing, Head 0 is PIN_SIDESEL (p32) inactive/floating high else if head == 1 then pins = pins or PIN_SIDESEL else error("Head number " .. head .. " out of range.") end -- That's pretty much it, unless we need to provide TG46 on the DENSITY pin. return pins end --[[ Given the current drive status flags, identify whether the drive is ready for use. ]] function isDriveReady(drivetype, status) -- 3.5in drives don't generally have a working READY output, and we don't give a damn about DISK CHANGE. -- If this were a Winchester drive, we'd be checking READY and SEEK COMPLETE. return true end --------------------8<-----------8<---------------------------- So in about 70 lines of code (a fair few of which are comments), we've added full control support for two 3.5in PC floppy drives connected via a crossed cable. With a few more LOCs we could add support for 8-inch and 5.25-inch drives which require a 'TG46' signal on pin 2, or an 'IN USE' signal to mount the heads or switch on the LED. FormatSpecs are even simpler: --------------------8<-----------8<---------------------------- --[[ ############################# # DiscFerret Format Specification File # # 40/80 track generic, soft sectored ############################# ]] formatspec_version = 1.0 formatspec = { gen40ds = { -- format name friendlyname = "40 track, double-sided, soft-sectored, generic" -- minimum track number mintrack = 0 -- maximum track number maxtrack = 39 -- track stepping -- 1=singlestep, 2=doublestep trackstep = 1 -- minimum head number minhead = 0 -- maximum head number maxhead = 1 -- sectoring; 0=soft-sectored, or number of sectors if hard-sectored sectors = 0 } gen40ds = { -- format name friendlyname = "40 track double-stepped, double-sided, soft-sectored, generic" -- minimum track number mintrack = 0 -- maximum track number maxtrack = 39 -- track stepping -- 1=singlestep, 2=doublestep trackstep = 2 -- minimum head number minhead = 0 -- maximum head number maxhead = 1 -- sectoring; 0=soft-sectored, or number of sectors if hard-sectored sectors = 0 } gen80ds = { -- format name name = "80 track, double-sided, soft-sectored, generic" -- minimum track number mintrack = 0 -- maximum track number maxtrack = 79 -- track stepping -- 1=singlestep, 2=doublestep trackstep = 1 -- minimum head number minhead = 0 -- maximum head number maxhead = 1 -- sectoring; 0=soft-sectored sectors = 0 } } --------------------8<-----------8<---------------------------- In about 60 lines of code, again mostly comments, we've added support for three types of disc. Not bad. You could even go as far as generating all of these in code when the script loads. So on to the point of my message: does anyone have any comments to make on this type of config file format? Is there anything you'd like to see added, or done differently? Thanks, -- Phil. classiccmp at philpem.me.uk http://www.philpem.me.uk/