Hi Arthur, thanks for reconsidering your opinion and adding the patch, first.
On Wed, Oct 01, 2008 at 09:41:42PM +0200, Arthur de Jong wrote: > Perhaps for group names it could be not very unusual but I would not > call having spaces in user and group names very common, especially if > you look at the history of *nix systems. well, yes, I'm not a big fan of spaces in user- or groupnames either. But these days it's quiet common to combine two worlds: Windows and *nix, whereas in the first of both worlds spaces in groupnames are very established. In fact the group "Domain Users" and alike are standard groups in a windows environment and it seems that Windows partly requires it (we had some problems before they were created with users unable to work on a windows terminal server and after some research I heard rumours that some of those groups are obligatory for Windows. Sorry, I have no source for this anymore, but adding these groups and adding our users to this group solved problems so I don't care that much). And as far as I am concerned I don't see a good reason in not supporting such "features" just because it was always done different. That is not a convincing argument, too, in my opinion. > > 2) It breaks compatibility with libnss-ldap, which does not have the > > problem. Thats bad, as your solution is advertised as a replacement for > > it, because of the better design. However to be standards-compliant, > > while breaking functionaliy and compatibility is not really a good > > thing. > > I don't find this a very convincing argument. nss-ldapd is not a 100% > bug-for-bug compatible replacement for nss_ldap. nss-ldapd is missing a > number of features (most of which aren't very well documented in > nss_ldap). Right and it does not have to be a "bug-for-bug" compatible replacement. But for the reasons above I wouldn't consider spaces in group names a bug. In fact forbidding spaces in group names is just a restriction, which does not seem to be neccessary and annoys people who are used to it. > > Anyway: As Andreas already said: It is a common setup, which works > > flawless with libnss-ldap (except the non-related flaws this has), so: > > No it does *not* break stuff. > > Very well, I will apply your patch. Not because your arguments are very > good but because it will not hurt nss-ldapd. Well, that makes me happy (and others as well) and I can live with you disagreeing with me. However I hope I had a chance to bring this a bit nearer to you. Again: Thanks for including the patch. Best Regards Patrick -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
