.. in tcsh/bash
Date: 20 Sep 2005 13:38:57 GMT
Suppose symlink is a symbolic link and consider the sequence
cd symlink
cd ..
It seems that you end up in different places depending on whether one
is using tcsh (parent of the symlink destination) or bash (symlink
source). Wouldn't this be defined in the Posix spec? If so, which is
correct? If not, why not? How could something this basic be left
unspecified?
Bob T.
Relevant Pages
- Re: Moving include/asm-* [was: Re: Moving sound/* to drivers/ ?]
... logics that creates a symlink for each of those. ... depending on target. ... All the dependency checks than "just works" and if arch changes ...
(Linux-Kernel) - Re: A Logging data structure
... When you want to write one, read the symlink to ... filesystem operations to read or write a block, ... will come straight from the cache, ... which would add more directory writes to the block-write sequence, ...
(comp.lang.java.programmer) - Re: diff mishandling symlinks
... |> BTW, a symlink reference to a non-existant file, that may, or may not, ... |> come to exist later on, is a valid symlink. ... | comparing the metadata of the symbolic links. ... | time of the symbolic link, but unlike other files, the contents of the ...
(comp.unix.programmer) - Re: A Logging data structure
... Don't reinvent wheels - use the filesystem. ... When you want to write one, read the symlink to ... which would add more directory writes to the block-write sequence, ... The solution i suggest requires no user-level changes to the disk except ...
(comp.lang.java.programmer) - Re: A Logging data structure
... When you want to write one, read the symlink to find the current highest revision number, and then make a file with the right block number and a revision number one higher, and re-point the symlink. ... filesystem operations to read or write a block, which are likely to be something like Oin the number of files in the directory, but pretty fast - taking a handful of block reads, which will come straight from the cache, for each operation. ... If it turned out to be a problem, you could change the arrangement of blocks on disk so that the frequently-accessed ones were kept in a different, and so much smaller, directory than the old revisions, which would add more directory writes to the block-write sequence, but speed up read accesses. ... Although you could still use this approach to manage the version history, and keep the current state of the file in a single real file. ...
(comp.lang.java.programmer) |
|
