On Monday 24 of November 2014 12:25:45 KO Myung-Hun wrote: > Hi/2. > > KO Myung-Hun wrote: >> >> >> Paul Eggert wrote: >>> Myung-Hun KO wrote: >>>> But any special reasons why binary-io module of gnulib should >>>> be used instead of pre-existent SET_BINARY_MODE() ? In addition, >>>> SET_BINARY_MODE() is being used already. >>> >>> My impression is that the proposed patch is reimplementing binary-io, >>> which would mean it's better to bite the bullet and just use it. >>> >>> >> >> I've updated the patches. > > Updated, again.
Hi Myung-Hun KO, [ad 0002] neither you nor Paul have spoken about pros/cons of particular implementation. >From me, I also think that using binary-io makes sense. Your patches are clearly portability fixes and this is what is gnulib for (pros: you'll get another portability fixes from any gnulib user in future). What about switch paxutils/tar source to use binary-io? Pavel
