Changes between Initial Version and Version 1 of Ticket #1437, comment 8


Ignore:
Timestamp:
2014-12-20T04:17:59Z (10 years ago)
Author:
daira
Comment:

Legend:

Unmodified
Added
Removed
Modified

  • Ticket #1437, comment 8

    initial v1  
    1 I tried to post this at https://twistedmatrix.com/trac/ticket/5203 but couldn't get past
    2 the spam filter:
     1I tried to post this at https://twistedmatrix.com/trac/ticket/5203 but couldn't get past the spam filter:
    32
    43----
    5 It doesn't particularly surprise me that this issue has gone for three years without
    6 comments, because the Description is quite rambling and doesn't focus on what Zooko wants
    7 done. (In fact I'm not clear on what Zooko is asking for either.)
     4It doesn't particularly surprise me that this issue has gone for three years without comments, because the Description is quite rambling and doesn't focus on what Zooko wants done. (In fact I'm not clear on what Zooko is asking for either.)
    85
    96What I would like in order to be able to use `FilePath` in Tahoe-LAFS is either:
    10  * for `FilePath(unicode_path)` to be specified to work, and to return Unicode strings
    11 from methods/attributes that currently return path components;
     7 * for `FilePath(unicode_path)` to be specified to work, and to return Unicode strings from methods/attributes that currently return path components;
    128 * or alternatively, a `UnicodeFilePath` class that works that way.
    139
    14 Currently I think the status is that `FilePath(unicode_path)` mostly works by coincidence
    15 of the stdlib file APIs accepting Unicode paths, but that doesn't fill me with confidence.
     10Currently I think the status is that `FilePath(unicode_path)` mostly works by coincidence of the stdlib file APIs accepting Unicode paths, but that doesn't fill me with confidence.
    1611
    17 Note that representing paths as byte strings can't possibly work correctly in general on
    18 Windows. It is not sufficient to only be able to represent paths that have characters in
    19 the "ANSI" encoding.
     12Note that representing paths as byte strings can't possibly work correctly in general on Windows. It is not sufficient to only be able to represent paths that have characters in the "ANSI" encoding.
    2013
    21 The behaviour around undecodable paths is a secondary issue that we could work around; I
    22 wouldn't care if we needed wrapper functions to get the behaviour we wanted there.
     14The behaviour around undecodable paths is a secondary issue that we could work around; I wouldn't care if we needed wrapper functions to get the behaviour we wanted there.