On 04-05-2008, Stefano Zacchiroli <[EMAIL PROTECTED]> wrote: > On Sat, May 03, 2008 at 11:20:21PM +0200, Stefano Zacchiroli wrote: >> > - If you have patches either to make bin-prot big-endian or to excise >> > those bits of core that depend on bin-prot, then I'll take them, or >> > try to help. >> I haven't yet attacked the big-endian bits, but I'm working on a patch >> to make bin-prot dependency optional and detected at compile time. Will >> mail you back when I've something fully working. > > So, here is my current "solution". Beware, it's really "hackish" :) > > The main part to get rid of the bin_prot dependency is emptying the > Binable interface so that other modules are not require to implement > serialization stuff. This is easy. The tricky part is a lot of other > places are using "with bin_io" modifiers and have in their > pseudo-headers declared that they want to be processed by camlp4 with > the bin_prot syntax extension. > > For this reason the patch removing support for bin_prot is rather large, > and hard to make compile-time conditional. For the same reasons is not > something I will push to Jane St. Hence what I did for the Debian > package is to use a $DVCS (git in this case) to fork a branch in which I > maintain a "no-bin-prot" branch, in the hope that this will make it easy > adapting to future upstream changes. >
What about patching bin-prot syntax extension to produce "nothing" when used... This way, everything still get compiled with "fake" bin-prot... I think it will make the required patch smaller (i.e. you will have only to patch where the serialization is used and not everywhere). Regards, Sylvain Le Gall -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
