Debian Codegraph @

doc/TODO

raw
1
2 ------------------------------------------------------------------------
3 - TODO List
4 ------------------------------------------------------------------------
5
6 - D-Bus interface optimizations
7 - problem: right now a lot of round-tripping is needed
8 - solution: for Foo=Device,Adapter,Expander,Port
9 - EnumerateFoo() should return a(oa{sv}) instead of just 'ao'
10 where the a{sv} contains the properties
11 - FooAdded() should be 'oa{sv}' instead of just 'o'
12 where the a{sv} contains all properties
13 - FooChanged() should be 'oa{sv}' instead of just 'o'
14 where the a{sv} contains the properties that has
15 changed
16 - this will break the D-Bus interface so we we need to bump the
17 minor number (e.g. 1.0.x -> 1.1.0 transition)
18
19 - We should probably use "Phy" instead of "Port"
20 - do this when breaking the D-Bus interface for other reasons
21
22 - When the GDBus stuff lands in GLib we should probably start using it
23 - probably do this when breaking the D-Bus interface for other reasons
24
25 - UTF-8 assumptions
26 - Some properties contain file paths. Since Linux/UNIX allows any
27 NUL-terminated string of bytes as file names (restricting file
28 names to UTF-8 is clearly insane, *boggle*), we could end up with
29 strings that are not UTF-8. We should probably use the type 'ay'
30 instead of 's' for such things.
31
32 - We probably have too many polkit actions - probably just two are enough
33 - reads should require org.fd.udisks.read
34 - this should be 'yes'
35 - write/modification should require org.fd.udisks.write
36 - this should be 'auth_admin_keep'
37
38 - Jobs should be separate D-Bus objects instead of Job* properties
39 on the Device objects. We should probably also synthesize Job objects
40 for jobs being tracked/managed by separate kernel- and/or user-subsystems.
41 For example, if the user does
42
43 # echo check > /sys/block/md0/md/sync_action
44
45 from the command-line, we should create a Job object. Things like that
46
47 - We should have a notion of events and a history of events. This
48 would include SMART data, IO errors and such things. Each event
49 would have generic metadata (type, collection date, source,
50 relationship to 0 or more devices, human readable string,
51 type-specific blobs) and UI like Palimpsest would allow searching,
52 viewing, organizing and purging any type of event.
53
54 We probably want some way of rate-limiting / collapsing multiple events
55 into one etc. Think tons of IO errors.
56
57 The type-specific blob could contain e.g. SMART data and this could be
58 used to realize SMART graphs.
59
60 - It would be nice to be able to have different back-ends
61
62 - The D-BUS API is supposed to be synchronous, but property updates are racy
63 in many cases (tests/run --no-workarounds exposes those). These should get
64 fixed by having the methods only return after the properties that were
65 affected by the call got updated.
66
67 ------------------------------------------------------------------------
68 - Future features
69 ------------------------------------------------------------------------
70
71 - iSCSI
72 - target (share drive)
73 - see notes in doc/TODO-ISCSI
74
75 - Filesystem check
76 - how can we do interactive fsck? Connect to a socket on the
77 client and then make the client includes a terminal emulator?
78 Yuck
79
80 - Partition/Filesystem resizing
81 - Maybe just enhance PartitionModify
82
83 - SES2 Enclosure Management
84 - In particular, we should do something with the LEDs
85
86 - make integration-test run in uninstalled mode
87
88 - LSM man page rewrite to docbook format