Date: Thu, 3 Jan 2002 02:11:40 +0100 From: Cliff Sarginson <cliff@raggedclown.net> To: freebsd-questions@freebsd.org Subject: Re: SMBFS Unix to Windows End of Line Problem Message-ID: <20020103011140.GD13841@raggedclown.net> In-Reply-To: <20020103000932.GA1577@student.uu.se> References: <3C338EC8.D20C9A67@unisys.com> <3C33904B.1029227@T-Online.DE> <20020102234006.GA13841@raggedclown.net> <20020103000932.GA1577@student.uu.se>
next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, Jan 03, 2002 at 01:09:33AM +0100, Erik Trulsson wrote: > On Thu, Jan 03, 2002 at 12:40:06AM +0100, Cliff Sarginson wrote: > > > Doug Fee wrote: > > But since just about nobody has actually implemented the OSI-model it > doesn't really matter what it contains. Well, that is not quite true, but I know what you mean. > > Also whatevr happened to net-ascii then, isn't it supposed to be the > > level at which this is sorted out ? I expect if I look at a text file > > for it's appearance to be independent of it's origin. > > net-ascii is basically used for on-the-line representation of data and > not for data stored on the disk. Yes but it is a platform neutral representation. Therefore it should behave appropriately for whatever platform it is being accessed on. And If I am accessing files on a Samba share I am accessing them on-the-line.. > WHy would you expect a textfile to look the same regardless of where it > comes from? Remember that Unix does not differ between textfiles and > binary files. Ah, I knew someone would say this. And of course it is true that Unix has no concept of file-type in that sense. However of course for all practical purposes it *does* have a concept of file type - think of the 100's of Unix tools that make assumptions about files as being text files (grep, sed, vi etc etc) on the basis of being lines of representable text ending in a newline. > One certainly does not want binary files to be translated > this way, especially since it is a non-reversible translation. Well that is not being suggested. > Note also that it is not possible to write a program that can > distinguish between binary files and textfiles with 100% accuracy. > You can get pretty damn close to it. > Having such a translation take place automatically would therfore be useless > unless on binary files whatsoever were stored on the disk. That sentence didn't make sense (I mean it isn't sensible english, so I don't understand it). > > > > Surely something wrong here ? > > Not really. Ok, your argument, while I don't completely disagree with it, may just as well be applied to byte ordering, ebsdic, or whatever. There is an expectation that a text file looks like a line of text as it would on a typewriter (quaint old Unixism). I would argue that the point of cross-platform file sharing (and I mean plain-text) should be completely transparent to the user. She should not have to use a program for a file from one origin and another program for a file from a different one. Having thought about it though, I am not sure I can think of a foolproof way of doing it. -- Regards Cliff To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20020103011140.GD13841>