Path: utzoo!mnetor!uunet!husc6!necntc!ima!think!ephraim From: ephraim@think.COM (ephraim vishniac) Newsgroups: comp.sys.mac Subject: Re: Odd Finder bug... Message-ID: <19707@think.UUCP> Date: 14 Apr 88 14:57:25 GMT References: <8WKQ4ZI1DU-0QO008d@andrew.cmu.edu> <1415@daimi.UUCP> Sender: usenet@think.UUCP Reply-To: ephraim@vidar.think.com.UUCP (ephraim vishniac) Organization: Thinking Machines Corporation, Cambridge, MA Lines: 40 In article <1415@daimi.UUCP> jnp@jupiter.UUCP (J|rgen N|rgaard) writes: >In article <8WKQ4ZI1DU-0QO008d@andrew.cmu.edu> jv0l+@andrew.cmu.edu (Justin Chris Vallon) writes: >>How about this one: >> Please insert the disk: >> RamDisk >>Not a chance... >Isn't it the driver for the ramdisk that is responsible for not >ejecting it ? There's a field in the drive queue element which indicates (among other things) whether the drive holds an ejectable medium. Generally, Mac software is pretty casual about observing this flag. _Eject, for example, doesn't check it. Instead, you're supposed to check it *before* trying to eject a disk. Fortunately, the standard file dialogs do check it, so the problem doesn't come up too often. A well-written ramdisk needs several lines of defense. (A poorly written one, like the atrocity that comes with Aztec C, hasn't got any defenses at all.) First, it should set the "not ejectable" flag. Second, it should post a disk inserted event whenever it receives an eject call. Third, it should check periodically to see whether it's mounted (i.e., whether it's volume is on-line in the vcb queue), and post a disk inserted event if it's not. Some programs are incredibly stupid about ejecting disks. Old versions of Microsoft Multiplan were the worst. They expected to find the key disk in the internal floppy drive. If it wasn't there, they ejected *the current system disk* (not the floppy) and put up a modal dialog asking for the key disk. They didn't care whether the disk that they "ejected" was a ramdisk, hard disk, or floppy, they just blew it away. Ephraim Vishniac ephraim@think.com Thinking Machines Corporation / 245 First Street / Cambridge, MA 02142-1214 On two occasions I have been asked, "Pray, Mr. Babbage, if you put into the machine wrong figures, will the right answers come out?"