From mboxrd@z Thu Jan 1 00:00:00 1970 Message-ID: <3e8f8aa4f9ce3e3a909dcc1aaa87f694@plan9.escet.urjc.es> To: 9fans@cse.psu.edu Subject: Re: [9fans] Tags for files From: "Fco. J. Ballesteros" Date: Wed, 13 Oct 2004 19:25:17 +0200 In-Reply-To: MIME-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit Topicbox-Message-UUID: ed31deb8-eacd-11e9-9e20-41e7f4b1d025 > I'm trying to understand this. It looks to me like in addition to > what you've said above, you can write filters that get run every > time a file is noticed to have been created or changed, so > that tags can be automatically attached. Yes. The thing is that the scan script generates an initial set of tags for the files. The tags are actually generated by any /bin/tag/f* script that is found. This is to save you the work of defining tags that could be defined by a program. For example, I have used the muid and permissions from the output of ls -l and the name of elements in the full path name as the initial set of tags. This is an example. This entry is for $home/src/9/block.c /src/9/block.c muid=[nemo] perm=--rw-r--r-- usr nemo src 9 block.c > Can you give examples of what kind of searches you do? This would allow me to run "F 9 block" from acme. Other searches I do are F planb net slides F tag sys src F rwxrwxrwx sys F os course slides and the like. So far, I have not defined tags by hand. It seems that in my case the names in the path suffice. But, for example, I added the permissions as a tag soon after I started to use the program.