On April 15, 2003 07:16 pm, Alexandre Julliard wrote:
Well I see absolutely no point to the whole exercise, but I'm clearly not going to convince you. Feel free to try to persuade the Unix world to insert a bloated Win32 layer under all their toolkits; I don't think you will get very far with that idea...
Hey, "absolutely no point" is a bit strong :) But don't get upset, I was more brainstorming than anything else. I don't have time to sleep now, let alone start such a project. ;)
But we've lost track were we started from. I was advocating to make Wine easier to use as a library, and not encourage people to copy & paste code. There are a few cases where that may work: mplayer, *maybe* Gimp. But even in those cases it seems the plugins pull way to much of the API with them to have a snowball chance in hell to get something decent. Linking with Wine should give a decent result. Right now that's not the case, but it seems some simple improvements we plan to do anyway (the configuration work, lazy initialization, etc.) will greatly help in that regard.
So my end goal for this debate is to understand better what steps we need to take toward a EzWinelib (TM). And none of the problems listed seem all that hard, let alone insurmountable.