[Bug 33410] Do on-disk avatar cache in CM
bugzilla-daemon at freedesktop.org
bugzilla-daemon at freedesktop.org
Mon May 14 16:10:32 CEST 2012
https://bugs.freedesktop.org/show_bug.cgi?id=33410
--- Comment #9 from Xavier Claessens <xclaesse at gmail.com> 2012-05-14 07:10:32 PDT ---
(In reply to comment #6)
> + <tp:contact-attribute name="AvatarURI" type="s"
> + tp:name-for-bindings="Avatar_URI">
>
> Should be called avatar-uri.
fixed
> The documentation should say that:
>
> * it must be a file URI starting with file:/// as defined by RFC 1738
added
> * it points into a cache maintained by the connection manager
added
> * whenever the content of the avatar changes, the URI must change
I already tell that:
<p>Connection Managers MUST choose token in a way that an avatar update
always result in a different URI. For example using the image sha1 is a
good
idea, but not using the contact's ID.</p>
It's not enough?
> It should have rationale and probably also examples.
>
> Rationale for file:///: if we allowed other URI schemes (such as HTTP) we'd
> have to specify which schemes clients were expected to support, and for decent
> performance those clients would have to do their own caching anyway.
>
> Rationale for not just using filenames: on Unix systems, local filenames are
> arbitrary byte sequences which are not necessarily UTF-8, or even any
> consistent encoding at all.
added
> Example 1: on XMPP with XEP-0084 (PEP avatars) or XEP-0153 (vCard avatars),
> avatars are announced by their SHA-1, so CM implementations could base their
> filenames in the cache on that SHA-1. Similarly, in AIM/ICQ, avatars are
> announced by their MD5.
>
> Example 2: in protocols where a change to a contact's avatar is announced by a
> "last updated" timestamp, CM implementations could base their filenames in the
> cache on the pair (contact ID, timestamp).
added
--
Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the QA Contact for the bug.
You are the assignee for the bug.
More information about the telepathy-bugs
mailing list