With all due respect to Mark's knowledge and observation abilities, I believe he is mistaken in this question. mark@cheers.jsc.nasa.gov (Mark Manning/Muniz Eng.) writes: >According to Matthias Ulrich Neeracher: >>Mark: >> >The file isn't even officially created until after it's been closed. >> >> Hmm? Where did you get that idea? The Finder is slow to update sometimes, >> but the file is certainly created at open() time. > >The OPEN command does not guarantee that the file is >created immediately. But it does! >Only that you wish to open and/or >create a file if it isn't there. With a ram cache (which >is automatically engaged in Mac OS version 7.5 and later), In fact, I remember reading that a cache setting of 0K was silently interpreted as a non-zero cache size for some system versions prior to 7.5 :-) >the open command can be cached and, until a flush is issued >(ie: a close usually does this) by the user or system, the >request to create a new file can take as long as five >minutes. We have to distinguish two properties of the file system cache here: - Is a file created *on disk* immediately when it's opened? No. That may indeed take a considerable time (same as on UN*X). - Is a file created *in the file system* immediately when it's opened? Yes. This means that, although a freshly created file would not be visible to a tool reading SCSI blocks directly, it is visible and available to manipulation in the file system (since *every* call goes through the cache first). The following program: $name = "t001"; for ($i = 100; $i--; ) { open(T, ">$name") || die "open $name"; MacPerl::SetFileInfo($name, "TEXT", $name); ++$name; } created all 100 files and set their creators exactly as specified. If you have a perl program to demonstrate your claim that freshly created files may not be immediately visible, you're welcome to post it & convince me :-) >My feelings are: Better safe than sorry. :-) A fine principle, although, to be pedantic, close() on a file does not entirely flush the disk cache, so you're not safe with close either. Matthias ----- Matthias Neeracher <neeri@iis.ee.ethz.ch> http://www.iis.ee.ethz.ch/~neeri "There once was an Age of Reason, but we've progressed beyond it." -- Ayn Rand, _Atlas Shrugged_ ***** Want to unsubscribe from this list? ***** Send mail with body "unsubscribe" to mac-perl-request@iis.ee.ethz.ch