Le 9 avr. 07 à 20:21, Yen-Ju Chen a écrit : > On 4/8/07, Quentin Mathé <[EMAIL PROTECTED]> wrote: >> Le 8 avr. 07 à 18:28, Yen-Ju Chen a écrit : >> >>> On 4/5/07, Julien Wintz <[EMAIL PROTECTED]> wrote: >>>> >>>> Everything builds fine but when typing $> sudo make install under >>>> etoile, I get the following message: >>>> >>>> Build Project: AddressView >>>> >>>> Making all for framework AddressView... >>>> Copying resources into the framework wrapper... >>>> Copying localized resources into the framework wrapper... >>>> ln: creating symbolic link `libAddressView.so' to >>>> `./Versions/Current/libAddressView.so': File exists >>>> make[5]: *** [AddressView.framework/./AddressView] Error 1 >>>> make[4]: *** [AddressView.all.framework.variables] Error 2 >>>> >>>> I'm not familiar with the GNUmakefile system so I have no patch nor >>>> solution to propose. How can I end my installation of etoile ? >>> >>> We noticed this problem before and thought it was fixed. >> >> It seems to be a bug in gnustep-make since you can reproduce it >> outside of Etoile build process. >> I haven't yet figured out a way to reproduce it 100% reliably, that's >> why I haven't reported it on GNUstep bug tracker. >> It won't be fixed until gnustep-make itself is fixed, so a lot of >> people are probably going to report it for some time :-/ >> >> More about the issue here (patch included): <https://gna.org/bugs/? >> 8806> >> >>> A right fix may be reorganizing the directory structures of >>> AddressesKit. >> >> This won't solve the problem. It can only make it happens in a less >> recurrent manner. > > I notice that if I do 'make' in AddressView alone, there is no > problem, > no matter how many times I do it. > The problem only happens while I do 'make' in AddressesKit, > which tries to build two frameworks one after another. > Therefore, I was thinking if I can merge these two or make them more > separated, > the problem may be gone. > I maybe miss something here, though.
I have done further investigations, you can enjoy the whole story here: <http://gna.org/bugs/?8806> By the way I also sent a mail to Nicola Pero about this bug because I noted a commit in gnustep svn similar to the patch I wrote for gnustep-make. In his reply (readable on gnustep-dev list), he told that this issue seems to be the one he fixed. :-) I'm not sure to have really understood all the details, but basically the bug seems to be fixed for sure at gnustep-make level. We just need to test it and the story normally ends here and now ;-) Quentin. _______________________________________________ Etoile-dev mailing list [email protected] https://mail.gna.org/listinfo/etoile-dev
