summaryrefslogtreecommitdiffstats
path: root/docs/mdk-9.2
blob: 1d5a0649457b76be0a0c2d66530ebfb361cd2163 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
-*- outline -*-

        todo list for mdk9.2
        (mainly gtk+ oriented for now)

* erratas
** put them in gi/perl-install/patch/mdk_9.1
** move mdk9.1 ones from gi/perl-install/patch into gi/perl-install/patch/mdk_9.0


* libDrakX
** document internals through docbook like kernel-2.5.x does

* mcc vs tools communication
** faster tools embedding in mcc (gc idea)
   move gui into gui/<sg>.pm so that mcc can just do require on them
   instead of doing slow fork+perl startup (same idea as fast_cgi,
   perl_mod and the like)
** told apps to ask user whether to save their changes or to discard them on cancel (#1437)
** to let diskdrake fit better when embedded, we should write "formated - mounted" rather than "formated\nmounted"
** "testing mode" option in options menu so that we pass --testing to our tools ?

* ugtk2 & tools guies:
** cleanups:
*** $o->flush() should be sub { $_[0]->window->process_updates(1) }
    see logdrake::parse_file()'s progress bar as an example
** saner/cleaner usage of gtk+-2.2 widgets:
*** use MessageDialog for modal windows (cleaner)
    especially for wait/error messages so that one uses std gtk+ icons for info/error/warning and the like events
*** use button boxes everywhere when we have several buttons for saner gui (as per gnome specs/recommendations)
*** gtk::combo => gtk::optionMenu (as per gtk+-2 doc)
*** use size::group to let things be smoother in some places (eg: clock.pl from mcc)

* tools
** all:
*** provide menu entries for more gui tools (?): for now, only harddrake2 and logdrake have one
*** explanations of std standalone.pm functions are broken due to common.pm overloading
*** have a button or a menu to access help (was already part of mdk9.1 specs)
    only harddrake2, mcc and rpmdrake have one for now
    policy might be: "most users access tools via the mcc, so let handle help, bug reports and co only in mcc"
*** rename and provide compatibility links:
**** adduserdrake  => drakeuseradd
**** diskdrake     => drakdisk
**** harddrake     => drakhw
**** keyboarddrake => drakkeyboard
**** localedrake   => draklocale
**** logdrake      => draklog
**** menudrake     => drakmenu [fredc]
**** mousedrake    => drakmouse
**** printerdrake  => drakprinter
**** userdrake     => drakuser [deush]
**** XFdrake       => drakx11
** drakboot:
*** should prevent the user to insert a space in a kernel label (#2826)
*** fix "Display theme under console" setting from current theme config file (#2662)
** drakconnect
*** does not show up zeroconf as a valid mode in "configuring lan" window in embedded mode
** drakedm: 
*** restart dm service ?
** harddrake: 
*** colorize the internal help
*** add raw description (or join('|', vendor, descr)) for mass medias (see drakx bug #3???)
*** use pure gtk+ instead of interactive for "fields description" or alter interactive so that it creates modal windows
** logdrake:
*** search field does not get focus when embedded
*** we may disable the search button when one deselect all explanation files
** mcc:
*** make tool names be black or blue (like in rpmdrake & harddrake2 guies) (#2695)
*** hide non executable programs instead of warning at runtime if program isn't executable
*** make more program being options rather than depandancies (ie like rpmdrake and drakwizard)
*** add scrollbars if root window is below 800x600 (#2961)
*** add scrollbars for too big translated messages (but how to detect it)
*** move drakedm from hardware to system section
*** enhance mcc descriptions (#3265)
** rpmdrake
*** save mode (sort by group, mdk choice, ...)
**** sa serait cool que tu sauve la position (aka choix mdk, tri/alpha, par
     groupe, par taille, ...) et que tu la restaure
*** m-q accelerator to quit
*** if package is unselectable, explain why
*** si conflit, proposer de deselectionner le package en question +tot que de
    devoir se taper l'arborescence pour le deselectionner (a - que la
    recherche e perde pas la selection de package en cours)
*** some windows are still non modal, like:
    o tout ce qui se rapporte aux dependances "des paquetages
      supplementaires sont necessaires", ...  tu oublie de passer "{
      transient => $w->{rwindow} }"
    o toutes les fenetres grpmi.  faudrait voir si on peut foutre le
      core de grpmi dans un package pour faire "use grpmi" et seter
      le nouveau $::main_window pour corriger ca.
*** we need access to urpmi options like "no gpg signature check"
*** bug 1
1) selectionner des packages a installer

2) rpmdrake dit que X doit etres vier

3) rpmdrake le vire

4) rpmdrake passe a l'update/isntall proprement dite

5) rpmdrake failled

rpmdrake re-tentera d'enlever le package X du 1 en cas de reessai
alors qu'il n'est + la

** scannerdrake/sane-find-scanner
*** do not detect every usb devices as scanner
** urpmi
*** --repair-system: add missing dependancies for currently installed packages


* packages
** main package updates:
*** rpm-4.2
*** kernel-2.6 (if availlable and stabilized: not probable) ?
    at least, keep kernel-2.4.x
** add new packages:
*** kernel-mosix
*** dvb related packages: vcr, vdr, ...
** icewm/windowmaker/other wm
*** export BROWSER like gnome/kde (look at /etc/gnome/gnomerc)

* hardware support
** kernel: update alsa, bttv, dvb & saa7134 drivers for better hardware support (the laters from http://bytesex.org)
*** add "REGISTER ^dvb/adapter[0-9]+/[^/]+$ PERMISSIONS root.video 0660" to devfsd
*** update oss audigy driver from pserver:anonymous@cvs.sourceforge.net:/cvsroot/emu10k1 cvs
** add cx88 sound driver from http://bytesex.org/snapshot
** isa detection/probing in ldetect ? (#118, #497) => need some old hw in order to test
   probably not