On Sat, Sep 07, 2002 at 11:33:11AM +0200, Stefano Zacchiroli wrote: > On Fri, Sep 06, 2002 at 07:15:14PM +0200, Sven LUTHER wrote: > > > [ BTW I've also sent to the BTS the ancient patch for ocaml-ldconf, so > > > you can see if there is still the case to apply it or not ] > > > > you forgot to attach it or something. > > yep, as usual, resent! > > > Note: i also applied the ocaml-source move. > > great. > > > > > Notice, the third option (doing nothing) is also valable, since the > > > > paths added by /var/lib/ocaml/ld.conf come _after_ the default ones, and > > > > it would just search a few more path in the unlikely case that you try > > > > loading an unexistent library. It is not that bad. > > > > > > see above. > > > > Well, i still think it is something that needs to be around forever. > > Why you think so? I'm trading off usefullness of the fix with cleanness > [ it this word really exists in an english dictionary ] ot the package. > Do you agree with me that the user base affected by the problem after > woody + 1 is really tiny, if not empty? Then, why keep this trick > forevere? We have the bug now and I've seen few people arguing about it, > moreover the bug isn't directly seen by the user, he can safely continue > in compiling and using other ocaml related apps without even knowing > that it have a bugged ld.conf. _We_ want to solve the bug because we > hate a useless entry in ld.conf (those damnet computer scientist guys! > :-)
Mmm, yes it only is a tiny subset of people, i think. > > Mmm, does adding a dh_ocamlld -r do the thing we want ? I don't think > > so, we maybe need to call ocaml-ldconf -R -plibname-ocaml explicitly in > > a hand written postinst, until woody +1. > > > > Mmm, if there is no -r <version_number> option that we could pass to > > dh_ocamlld, then a centralized solution would be better, are you sure > > you are not saying that so i go for your solution ? :)))) > > I was just thinking to explicitely call ocaml-ldconf -R ... until > woody+1, but if you change your mind and agree with a centralized > solution .... :-) But we can do the -r <version_number> thingy, it only needs (i guess) a little change to the dh_ocamlld stuff (well, i am no perl guy, so ...) it would just mean analyzing the @ARGV to see if the first of them is "-r version" or something such. Perl is supposed to be able to do this easily, is it not ? > > Mmm, maybe i will learn perl finally and write it myself ... > > don't do it, avoid risks for your brain as most as you can! :-)) :))) Friendly, Sven Luther > > > Cheers. > > -- > Stefano Zacchiroli - undergraduate student of CS @ Univ. Bologna, Italy > [EMAIL PROTECTED] | ICQ# 33538863 | http://www.cs.unibo.it/~zacchiro > "I know you believe you understood what you think I said, but I am not > sure you realize that what you heard is not what I meant!" -- G.Romney
