Re: [ftpext] New Version of FTP HASH, RANG, LOCK, and HOST

[Anthony Bryan <[e-mail]>]

On Mon, Jan 28, 2013 at 4:04 PM, Ángel González <[e-mail]> wrote:
> On 28/01/13 01:01, Alun Jones wrote:
>> After reading the RANG spec, I note that there isn't anything to state what
>> errors are given by RETR if the range specified is incorrect, or by STOR /
>> APPE if the range specified starts more than one byte after the end of the
>> existing file on the server (and therefore, presumably, incorrect?).

current text is:

The server-PI should reject the STOR if a client issues a RANG with a
start-point greater than the actual file size. If the start-point is
less than or equal to the file size, the file is opened for writing,
continuing at start-point.

http://tools.ietf.org/html/draft-bryan-ftp-range

> I think it should be documented that the server must fill the gap with
> nul bytes,
> with a note that it may use it as a hint to do that using a sparse file.

any other feelings on this?

on a side note, wondering if it might be better to separate RANG and
RETR/STOR/etc_with_RANG into 2 separate drafts.
HASH is also a separate draft that happens to use RANG.

--
(( Anthony Bryan ... Metalink [ http://www.metalinker.org ]
  )) Easier, More Reliable, Self Healing Downloads