Path: csiph.com!usenet.pasdenom.info!weretis.net!feeder4.news.weretis.net!usenet.ukfsn.org!not-for-mail From: Martin Gregorie Newsgroups: comp.lang.java.programmer Subject: Re: A proposal to handle file encodings Date: Sun, 2 Dec 2012 23:08:42 +0000 (UTC) Organization: UK Free Software Network Lines: 28 Message-ID: References: <50aed080$0$292$14726298@news.sunsite.dk> NNTP-Posting-Host: 84.45.235.129 Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-Trace: localhost.localdomain 1354489722 1481 84.45.235.129 (2 Dec 2012 23:08:42 GMT) X-Complaints-To: usenet@localhost.localdomain NNTP-Posting-Date: Sun, 2 Dec 2012 23:08:42 +0000 (UTC) User-Agent: Pan/0.139 (Sexual Chocolate; GIT bf56508 git://git.gnome.org/pan2) Xref: csiph.com comp.lang.java.programmer:20049 On Sun, 02 Dec 2012 23:52:58 +0100, Peter J. Holzer wrote: > We were obviously talking past each other. I was only talking about > mechanisms like xattr, alternate streams or resource forks, not about > revamping the whole unix file model. > I think that's likely. I've obviously not seen the guts of OS X resource forks, but I doubt their implementation differs a lot from what I was talking about: if they are not part of the OS's file handling system[*] they'd require all that messy stuff to be implemented in system programs and user-level libraries that you've described. I'm not advocating that Linux becomes OS/400 lite, just pointing out that metadata can be used in many ways, and that once you introduce the mechanism to transparently handle one attribute, such as character encodings, that there's quite a lot more that it could be used for. [*] I'm deliberately not saying Kernel because a lot of file handling stuff has already moved out of the kernel. A certain resemblance to Mach is creeping into Linux, though so far it is not nearly so fine-grained. -- martin@ | Martin Gregorie gregorie. | Essex, UK org |