From: mutazilah@gmail.com   
      
   On 22/02/24 17:57, Richard Kettlewell wrote:   
   > Paul Edwards  writes:   
      
   >> Cygwin is a correct example, I believe.   
   >>   
   >> And Cygwin doesn't change the behavior of Linux ELF   
   >> executables either.   
   >   
   > AFAIK it can’t run them at all, so I don’t know why that would be   
   > relevant.   
      
   That's exactly the point - it doesn't run them.   
   No behavior is being changed by Cygwin either.   
      
   >> Nor does it change the behavior of existing POSIX source code.   
   >   
   > Your use case was C90 programs, not POSIX programs, last time you   
   > mentioned it. Has that changed now?   
      
   I am the vendor of a C90 runtime library for Linux (PDPCLIB).   
      
   fopen() necessarily does an open syscall.   
      
   >>> If, in fact, you don’t want to add translation to read() and write()   
   >> I do.   
   >   
   > OK, so an ELF executable (that had been modified to use your   
   > hypothetical O_TEXT flag) would get newline translation from read() and   
   > write() when run on your toy OS, but not when run on a real Linux   
   > kernel. There’s the change in behavior.   
      
   Well you can call it that if you want. But as I said,   
   I think it is odd to call something a change in behavior   
   when the software that would be "changing" hasn't even   
   been written yet. Or at least, the INT 80H open() processing   
   of PDOS/386 hasn't been written yet. Only write to stdout   
   has been written so far. And exit.   
      
   Also, when you say "the ELF executable that has been   
   modified", I would use the word "built" rather than   
   "modified". I am building new ELF executables according   
   to a hypothetical version of POSIX/Linux that includes   
   a new O_TEXT definition.   
      
   > At any rate we’re back to the issue that you say want to add   
   > Cygwin-style translation to read() and write(), but you also say that   
   > you will only be running programs written in C90, which doesn’t have   
   > open(), so nothing will ever pass O_TEXT and the translation will never   
   > be activated.   
      
   PDPCLIB will do that.   
      
   I clearly haven't done a very good job of explaining   
   what I want.   
      
   I am organizing contact with the "Austin Group" (I've   
   become a member and downloaded the latest POSIX draft).   
      
   Is there any wording I should use to avoid the several   
   days of confusion we've had here?   
      
   When I wrote my original message I thought it was a very   
   simple suggestion, using the widely-known Cygwin prior art.   
      
   Instead, I'm not even sure my proposal is understood.   
      
   >> I guess it's just a semantic debate as to whether this constitutes a   
   >> change in behavior.   
   >   
   > This would go a lot quicker if you didn’t engage in “semantic debates”.   
      
   I wasn't attempting to. I was just trying to identify   
   the point of contention as semantics, not a technical   
   issue.   
      
   It's still not clear to me that it really is semantics,   
   because you are using language that I wouldn't use.   
      
   BFN. Paul.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)