#python - Tue 17 Apr 2007 between 00:00 and 00:09

NY Lost Funds



jsoftwAny ideas why this might not be matching ( it did when doing line by line ): re.compile("^\s*LogFile\s+([^\s]+)\s*", re.M) ?
TFKyle(but yeah, I can understand where he gets wchar_t from to store ucs-2 chars, as in windows wchar_t is 2 bytes afaik)
daliasyes
microsoft made a risky bet hoping to clean up and take it all
and they lost
TFKylehmm, they cleaned everything up?
jsoftwmicrosoft and clean?
What?
dalias"clean up" as in walk home with all the chips
teimuhi, could anyone recommend a feature-rich picture editing module?
TFKyleoh, meh
jsoftwOh they ate all the chips allright. And the pies. And it shows in windows's weight
daliasthat is, they were hoping to redefine "plain text" as a sequence of little endian 16bit values, and throw away unix and all the standard internet protocols and file formats in favor of their own proprietary ones
if they failed, they would be the one with the crappy, broken, incompatible system
but if they won, they'd have no competition
of course they lost, thanks to the genius of ken thompson
and the rest is history
TFKylegoing by your definitions they neither won nor lost
daliasall data interchange is done in utf-8 or sometimes still legacy encodings
no one uses ucs-2/utf-16 for data interchange
TFKyleI'll admit that windows does have a lot of problems related to unicode, but nothing there is related to their decision to use ucs-2
daliasonly for internal windows/java representations
TFKyledunno
daliasand the replacement of ucs-2 by utf-16 turned into a big nightmare for them
TFKyleI imagine some people store unicode strings in pickles, which are
daliasbecause windows programs still have to be aware of variable-length characters, just like programs using utf-8 would
and what's worse they don't have any clean api/type to deal with complete 21bit unicode codepoints
TFKyle21bit?
daliasyes, actually only about 20.1 bits
deltabenough to encode up to 0x10ffff
jsoftw20.1 bits? lol
actionTFKyle thought it was either 16-or-32(or 31) with utf-16
TFKylethought it was either 16-or-32(or 31) with utf-16
daliasunicode codepoints range from 000000 to 10ffff
TFKylewhich could be expanded at anytime
daliasno
TFKyleoh, so all possible codepoints then?
daliasutf-16 cannot support more than 10ffff
actionTFKyle figured you meant just the current range
TFKylefigured you meant just the current range
havn't read the spec but why not?
daliasbecause utf-16 2-word sequences have a range of 1024 values of each of the 16bit words

Page: 2 9 16 23 30 37 44 51 58 65 72 79 

IrcArchive

NY Lost Funds