Permalink
Commits on Jun 1, 2015
Commits on May 14, 2015
  1. 0.5.1

    substack committed May 14, 2015
  2. update travis targets

    substack committed May 14, 2015
Commits on May 13, 2015
  1. Remove octal literals, pass in strict mode

    We all hate this.  But since this module is used everywhere, it's
    a pain to not be able to use it in environments where strict mode
    is desired, even if that desire is misguided.
    committed May 13, 2015
Commits on May 6, 2014
  1. 0.5.0

    substack committed May 6, 2014
  2. document opts.fs

    substack committed May 6, 2014
  3. failing opts.fs sync test

    substack committed May 6, 2014
  4. test suite cleanup

    substack committed May 6, 2014
  5. reformat package.json

    substack committed May 6, 2014
  6. failing opts.fs test

    substack committed May 6, 2014
  7. 0.4.2

    substack committed May 6, 2014
  8. Parsing mode from string already in node's lib.js

    Hello,
    
    Node's fs.mkdir and fs.mkdirSync already call an internal modeNum method to parse the mode from string type.  Therefore it is unneeded in this lib.
    
    See: https://github.com/joyent/node/blob/master/lib/fs.js#L324
    
    Thanks!
    
    **Charles**
    doublerebel committed with substack Sep 30, 2012
Commits on May 4, 2014
  1. 0.4.1

    substack committed May 4, 2014
  2. add minimist

    calvinmetcalf committed with substack May 2, 2014
Commits on Apr 22, 2014
  1. swap 0.9 for 0.10 on travis

    substack committed Apr 22, 2014
  2. 0.4.0

    substack committed Apr 22, 2014
  3. usage

    substack committed Apr 22, 2014
  4. --help

    substack committed Apr 22, 2014
  5. add cli

    calvinmetcalf committed with substack Apr 19, 2014
Commits on Feb 22, 2013
  1. drop 0.4, add 0.9

    substack committed Feb 22, 2013
  2. modernize readme, deps

    substack committed Feb 22, 2013
Commits on Feb 12, 2013
Commits on Aug 20, 2012
  1. 0.3.4

    substack committed Aug 20, 2012
Commits on Jun 16, 2012
  1. All non-ENOENT errors should be handled the same

    There's a lot of special-casing for EISDIR, EEXIST, etc.  It dawned on me
    that we've special cased all the errors that fs.mkdir() can possibly raise,
    so why not just make that the default: case, and cut out a bunch of code
    that has to be read?
    
    Argument against: if some error is ever raised, which *isn't* an error that
    fs.mkdir() can 'possibly' raise, then it's raising an error that is *impossible*
    and we probably want to get out asap.
    
    But even in the case of impossible errors, if the dir is there, and it's
    a dir, then who cares?  mkdir(1) ignores any errors making a directory with
    -p if the directory exists and is a directory.
    committed Jun 16, 2012
Commits on Jun 5, 2012
Commits on Jun 4, 2012
Commits on May 23, 2012
Commits on Apr 30, 2012
  1. bump for erofs

    substack committed Apr 30, 2012
Commits on Apr 16, 2012
  1. Handle EROFS errors

    A EROFS will be raised when trying to write a dir onto a
    read-only filesystem.  However, if the dir already is there,
    then it should be treated like an EEXIST (since EROFS can
    be raised by trying to create a dir over a mount point,
    where the mount point is not read-only, but you still can't
    just clobber over it).
    
    If the dir doesn't already exist, then the EROFS will be
    accurately reported as the reason why it could not be
    created.
    committed Apr 16, 2012