Log of #mupdf at irc.freenode.net.

 <<<Back 1 day (to 2020/02/12)Fwd 1 day (to 2020/02/14)>>>20200213 
sebras angry (for the logs): so the Makefile will describe exactly what files to remove.02:33.47 
  at the moment our Makefile do not provide an uninstall target. it might be something we should consider.02:34.08 
ator openbsdtai123: using "getchar" is *not* as trivial as you make it sound. sure, add that line, but where? the program is already busy doing GetMessage() or XNextEvent().08:18.26 
  ...if it even has a controlling terminal. it could have been started with & or on windows with no terminal at all.08:18.54 
openbsdtai123 hi nice to hear from you...09:07.55 
  I am looking forward to talk technical... ok getchar()09:08.15 
  XNextEvent waiting for it yeah.09:08.35 
ator openbsdtai123: provide a patch, and I'll look at it09:08.52 
openbsdtai123 the possibility is to add a loop for the terminal waiting for input. I guess09:08.55 
  I guess that I have an example for it getchar and fltk app running there.09:09.12 
  it can be done - this is sure.09:09.43 
  If I can give you a simple example, I will share with you the method. I need to look at my harddisk where it is. maybe next weekend.09:10.41 
sebras ator: this is what he posted at the end of january https://termbin.com/2jvb10:19.23 
ator sebras: doesn't help integrate into the event loop though10:29.58 
sebras i know10:33.40 
  maybe openbsdtai123's opinion is that we should ditch tge event loop.10:34.14 
avih ator: do you intend to do something about cesu-8/utf-8? e.g. for js_loadfile, or make it behave like you described originally (str.charCodeAt(N) can be above 0xffff, or maybe something else?11:03.07 
  another option is making the internal representation utf8, and handle that on the fly. it's already variable length encoding anyway, so maybe it's not too hard?11:04.28 
  also, does mupdf does convertion to/from cesu-8 when it interacts with mujs?11:05.36 
  does mupdf do* ...11:05.57 
  ator: i noticed that cesu-8 takes more space than utf8 for valid strings. maybe js_tostring can convert it to utf8 if required without needing realloc? (and utf8 converted to cesu8 by mujs on input to mujs). i.e. make the API utf8?13:01.03 
  it would still leave an issue of inputs which are invalid utf8, and which i think js _should_ be able to represent and preserve.13:02.17 
  fwiw, there's also WTF-8 which is not supposed to be used for API/transport but could still be helpful in resolving issues: https://simonsapin.github.io/wtf-8/13:04.22 
  ator: also (from wikipedia about CESU-8): "It should be used exclusively for internal processing and never for external data exchange"16:06.05 
  i think library API should be considered as "external data exchange" here16:06.36 
 <<<Back 1 day (to 2020/02/12)Forward 1 day (to 2020/02/14)>>> 
ghostscript.com #ghostscript