On Sun, Oct 25, 2009 at 12:13 PM, Allan McRae <allan@archlinux.org> wrote:
Laszlo Papp wrote:
On Sun, Oct 25, 2009 at 11:58 AM, Allan McRae <allan@archlinux.org> wrote:
Laszlo Papp wrote:
* The alpm_rmrf function is available from the api, which does the same as this function did, with a small sanity check.
* It was worth to establish alpm_rmrf for _alpm_rmrf for pacman frontend as a wrapper to be able to use it in the future or for other frontend, so the function declaration was deleted in the frontend, and the new alpm_rmrf wrapper function was established for future usage with SYMEXPORT modifier.
Signed-off-by: Laszlo Papp <djszapi@archlinux.us> ---
What about this approach ?
What is different and how does this address the comments Xavier and Dan made earlier?
Allan
Allan, as you see this is another approach for avoiding the unneccesary function definitition duplication between the library and the frontend. Tt's not exactly the continue of the previous theory.
That's what I tried to do it, just making a wrapper and 'visible' function for the frontend to avoid the unneccesary duplication in the codebase, using the existing, working internal _alpm_rmrf function, without introducing a new insecure function or something. I can't mention easier solution for it to avoid the unneccesary replicating, maybe you've got better idea.
But Dan and Xavier both pointed out that exposing this function publicly was not a good idea. You seem to have just exposed it in a different way.
Allan
Hm, you're the pacman developers :) I can't mention better idea to avoid the duplication then, because I think it's a typical common used function by frontends, so it's not a frontend specific function for pacman-frontend only. (e.g. mkdir -p like _alpm_makepath_mode too) I don't understand Xavier's sentence: "alpm is not a general system utility." Just see _alpm_makepath_mode or similar functions. (mkdir -p). Isn't this a system utility like function ? And I've never said it's a general system utility. Otherwise Dan's right maybe, if there is such an issued function. Best Regards, Laszlo Papp