#tyndur · 2010-02-05

154 Nachrichten · Seite 1 von 1 · Original-CSV herunterladen

  1. eddybmorn'
  2. eddybhi, pcmattman :)
  3. pcmattmanHi eddyb
  4. eddybi'd like right now to kill the hardware designer and virtual machine solutions developer...
  5. eddybhope is not like this for you :P
  6. pcmattmanHeh
  7. eddybs/er/ers/
  8. pcmattmanWell, the opportunity is always around to join forces with... one of the CDI guys (forgot his name :( ) and pool your knowledge
  9. eddybwhy don't they work like in datasheets?
  10. eddybwhy does it always have to be tricks to make it work?
  11. eddybthey want only windoze and linux on their hardware?
  12. pcmattmanWell, technically if a datasheet says something like "should" then it's up to the hardware developer
  13. eddybyou mean XanClic?
  14. eddybor who?
  15. pcmattmanMax would be his RL name iirc
  16. pcmattmanAnd as for hardware, usually the developers take shortcuts to reduce cost of production ;)
  17. eddybshortcuts?
  18. eddybmaybe...
  19. pcmattmanWell, hacks
  20. eddybbut why something simmilar happens in virtualization solutions with closed code?
  21. pcmattmanTo emulate the real hardware ;)
  22. eddybXD XD XD
  23. eddybi think i'll make a linux code parser...
  24. pcmattmanCareful, GPL
  25. eddybhmm, last time i was able to revert minifiation of a javascript file...
  26. pcmattmanWhat we need is a glue module that turns Linux driver interface calls into CDI ones ;)
  27. eddybthe whole usb thing is too big...
  28. eddybstill, it'd be a boosting feature
  29. eddybah, pcmattman: kwolf started on the german keymaps :)
  30. pcmattmanI noticed
  31. pcmattmanHe seems to be enjoying it ;)
  32. eddybhehe
  33. eddybalso, what can we do about the "include the keymap i want" thing :) ?
  34. pcmattmanWell, why not add it to your .bashrc?
  35. pcmattmanIsn't that called every time your bash session starts?
  36. eddybah, that may be a solution...
  37. eddyband as for that, can we make in releases a change to login, so it would ask for keymap, first?
  38. pcmattmanIf you want
  39. pcmattmanThe next release will be a while away though ;)
  40. eddybmaybe even language, so it'd set an env var?
  41. eddybheh
  42. eddybwhat do you think it'd look like?
  43. pcmattmanMy plan for the next release is to have a GUI
  44. eddybsounds good... :)
  45. pcmattmanSo maybe we'll have keymap selection as part of the release load thingy
  46. pcmattmanKind of like you see in Ubuntu installations
  47. eddybdammit, i need to do something useful
  48. eddybpcmattman: every installation app has it :P
  49. pcmattmanIf you know German you could start translating the German-only comments in CDI code to English as well
  50. pcmattmanOtherwise, http://pedigree-project.org/projects/pedigree/issues ;)
  51. pcmattmanLots of things you can do if you want to do something useful
  52. eddybOpenSuSE's installation DVD got a little animation(at booting, IIRC) that says Wellcome in a bunch of languages :)
  53. eddybthen asks for language and keymap
  54. eddybif Max Reitz is XanClic, then i could ask if http://www.pedigree-project.org/issues/97 is still valid...
  55. pcmattmanI was meaning more like #40 or testing that #46 is in fact fixed :)
  56. eddybpcmattman: http://pedigree-project.org/issues/46 << would VirtualBox count :P (well, is very close to realHW, considering my problems with it) ?
  57. pcmattmanCould
  58. pcmattmanYou'll need to re-enable the floppy driver in src/modules/drivers/SConscript first
  59. eddyb#http://www.pedigree-project.org/issues/40 is a feature, so it can't be fixed :P i'll have a look and see what is there :)
  60. eddybdon't tell me what it shows up
  61. eddyb#http://www.pedigree-project.org/issues/40 << this?
  62. pcmattmanYes
  63. pcmattmaneg, strip(" hello ") gives "hello"
  64. pcmattmanlstrip(" hello ") gives "hello "
  65. pcmattman(you get the idea)
  66. eddybi meant...
  67. eddyb[09:03] <eddyb> #http://www.pedigree-project.org/issues/40 is a feature, so it can't be fixed :P i'll have a look and see what is there :) << this is how it showed up?
  68. pcmattmanOh, right, yeah
  69. eddyband i know what strip is, i worked with php :P
  70. eddybthere you go, #40
  71. eddybregexes always make problems
  72. eddybnope, not even multibyte support
  73. pcmattmanThat's why the feature request hasn't been closed yet ;)
  74. eddybdo you want me to add them?
  75. pcmattmanIf you want
  76. eddybthe problem is not that i can't
  77. eddybthe problem is that i don't know the best and easiest way of doing it :)
  78. pcmattmanJust make sure you do multibyte right. I'd rather strip/rstrip/lstrip first then multibyte after that's done, to be honest
  79. pcmattmanAs for now, I'll be back later
  80. eddybwell, i was saying that for (l|r)?strip :)
  81. kwolfMoin
  82. eddybmoin, kwolf
  83. eddybwho
  84. eddyb*how's keymap going?
  85. kwolfI think I've found some bugs, but in general I have a German keymap which should be usable for all West European languages
  86. eddybbugs?
  87. eddybdo tell
  88. eddybhave you started Usb?
  89. kwolfNope, just the normal one
  90. eddybyou'll be a bit blind while doing that, UHCI doesn't work with qemu anymore
  91. kwolfDead keys on AltGr do work as dead key, but if you type them twice you get the normal character without AltGr
  92. kwolfFor example, AltGr-+ is ~ in the German layout. So if I type ~ twice, I get a +
  93. eddybbecause is set to that
  94. eddybyou see, set-combinator got two arguments
  95. kwolfYep, and it's right
  96. kwolfIt's just ignored
  97. eddybit is?
  98. eddybfunny
  99. kwolfIt works for all keys where no AltGr is involved
  100. eddybah, i know
  101. kwolfAnd the other thing is that your bash isn't UTF-8 aware. Leads to funny results with backspace. ;-)
  102. eddybthe second retry disabled both combinator and altgr
  103. eddybwhat user?
  104. eddyb*disables
  105. kwolfjames, all other users don't even display most characters
  106. eddybyou can override that by adding a ALTGR COMBINATOR bla bla to they combinating key, but it's a bit of work
  107. eddybi'd better add another level of retry, with only combinator disabled
  108. eddyb(the retry is to ensure we find what to show even if there are all possible modifing keys pressed :))
  109. kwolfYou don't need the second entry without AltGr, so I consider it a bug that it doesn't work with AltGr. So yes, just fix it. ;-)
  110. eddybit should work fine on james :|
  111. eddybyup
  112. eddybdoes altgr+> work for you?
  113. eddybalso, we need to get more combinated values even in *Us
  114. eddybs/\*Us/\*EnUs/
  115. FreakyPenguinMoin
  116. eddybmoin, freakypenguin :)
  117. eddybi'm about to test the floppy driver...
  118. eddybdoes it work in Virtualbox?
  119. kwolfMoin FreakyPenguin
  120. FreakyPenguinHm don't know, I don't think I ever tested it there
  121. FreakyPenguinAh floppy is another thing that would need some work before the release...
  122. eddybokey
  123. eddybinvalid opcode: FAIL
  124. eddybjust after it does, informing partitioner of our presence
  125. eddybCdiDisk::read apparently does a jump to 203
  126. eddybor has an 203 opcode
  127. FreakyPenguinSounds like a broken function pointer
  128. eddybmaybe
  129. eddybi think is int cdi_storage_read(struct cdi_storage_device* device, uint64_t pos, size_t size, void* dest)
  130. FreakyPenguinDid you test it in qemu?
  131. eddybnope
  132. kwolfI'm not sure how mature Pedigree's cdi/storage implementation is
  133. eddyb#PF now
  134. eddyband it gets to cdi_storage_read
  135. eddybmeh, without mangling, i can't get to the symbol...
  136. eddybwhatsoever
  137. eddybalso
  138. eddybi'm checking changes to usb-uhci.c in qemu
  139. eddybseems like just simple PCI ones until now
  140. eddybbingo
  141. eddybreturn driver->read_blocks(device, block_read_start, block_read_count, dest);
  142. eddybthis causes big problems :)
  143. eddyblet's disable DMA, maybe it works
  144. eddyboh, no, not monologue again :((
  145. FreakyPenguinHehe
  146. FreakyPenguinIt's some time ago I tested cdi/floppy in tyndur, but I don't think It should be that broken...
  147. eddyb0xe0ff5eee is the address where it #PFs
  148. eddybsomething funny about it
  149. eddybgiving up, i don't know that german to debug it
  150. eddyb*that much
  151. quokmoin
  152. kwolf__Moin quok
  153. eddybmoin quok :)
  154. FreakyPenguinMoin quok