john@stag.UUCP (John Stanley) (03/22/89)
When I saw the note from Michal Jaegermann in comp.sys.atari.st where he complained about how awful the dLibs system() function was, I immediatly wrote a note to him asking for details and offering to fix the problem if he would just explain what said problem was.... (I have to admit I was a bit sceptical that it was as bad as he had described because I use programs that count on the system() function working correctly literaly every day of the week and have done so for more than a year...) After some passing back and forth of messages, I received the following message. I've cut out the technical details to avoid unnecessary loading of the net and retained just the parts that are significant to the general user of dLibs. (Note, if you really really want to see all the details, write me.) [ncc!myrias!mj@uunet (Michal Jaegermann) writes...] > > John, > about my problems with system() function from dLibs. I checked > carefuly dLibs sources to find out why I had trouble with it. > I don't know. I plead a temporary insanity. I guess that.... [details omitted] > When I retraced code carefuly I found that I had to be doing something > strange. I cannot right now reconstruct what. Please accept my > sincere apologies for writing first and thinking later. > > Regards, > Michal > ...{ncc, alberta}!myrias!mj > > P.S. If you would like to post this message to the net - feel free. Thank you Michal. Consider it done. My P.S..: Please don't assume that this means the system() function in dLibs can't be improved. It can (and probably will) be improved. But in its current form, it is, to the best of my knowledge, correct. I'm sure everyone will agree that there's quite a bit of difference between a function that "can be improved" and one that "badly needs to be fixed". --- John Stanley <dynasoft!john@stag.UUCP> Software Consultant / Dynasoft Systems