From mint-bounce@lists.fishpool.fi Sat Aug 29 18:59:13 2009 Subject: Re: [MiNT] symlinks and hostfs From: Petr Stehlik To: MiNT Mailing List In-Reply-To: <4A99AB5E.5080008@freesbee.fr> References: <4A97D65D.6090100@freesbee.fr> <1251532769.3902.23.camel@joy> <4A98E9A6.4040207@freesbee.fr> <4A99972B.90808@freesbee.fr> <1251581126.6084.10.camel@joy> <4A99AB5E.5080008@freesbee.fr> Content-Type: text/plain; charset="UTF-8" Date: Sun, 30 Aug 2009 00:57:06 +0200 Message-Id: <1251586626.6084.77.camel@joy> Mime-Version: 1.0 X-Mailer: Evolution 2.26.1 X-ecartis-version: Ecartis v1.0.0 Sender: mint-bounce@lists.fishpool.fi Errors-to: mint-bounce@lists.fishpool.fi X-original-sender: pstehlik@sophics.cz Precedence: bulk List-help: List-unsubscribe: List-Id: X-List-ID: List-subscribe: List-owner: List-post: Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by mail.sparemint.org id n7TMxCjA020288 Vincent Rivière píše v Ne 30. 08. 2009 v 00:27 +0200: > Ok, you're right, I was confused. > Cygwin's / is mounted on MiNT's /d (because of D = / in ARAnyM's config > file) so MiNT's /d/cygdrive/c is actually Cygwin's /cygdrive/c, that > path is perfectly valid. glad we have come to the same conclusion :) > Anyway it is very confusing to see ARAnyM tweaking the symlink target > that way. It is quite insane that ARAnyM tries to find symlink targets > across hostfs mounts. I don't know why HostFS does that (->Standa). It might be another cool feature... > I hope you did that in your "relative symlink" patch. no. Don't even ask what I have done there :-) > Now about the totally different feature named "Standa's real file > simulator". If the target of the symlink is located outside the same > hostfs mount, it is reported as a real file (so MiNT will never know the > value of the actual symlink target). the way I understand it's that MiNT never knows it was a symlink. > I think it could be a good idea, in > order to make some files appear magically in the hostfs, while being > physically located elsewhere. right. It's working that way. > However I'm not sure this option is really useful in the real world. imagine some hot-plug stuff on the host (if it was creating symlinks). Or just imagine someone being lazy to quit ARAnyM and reconfigure their config file just to mount something they forgot to mount earlier (like a flash disk or something). They just symlink another part of the hostfs to the already mapped directory and they're done. Again, not sure what has Standa in mind originally but see my A/B/C choices in a mail to Alan. This still looks to me like the best of the choices. Anyway, we don't need to judge this feature - what we are trying to do here is to find out why you can't see normal symlinks as symlinks, right? Or can you see them now? You report both success and failure in various mails so I am getting confused. > Finally for intermount symlinks I'm not sure... But seriously, does > someone do this in the real world ? It becomes really weird when / is > mounted on a TOS letter like in AFROS... Remove the D:\ mapping if you don't like it. I added it to AFROS just to show that ARAnyM can access the whole host filesystem, if needed. > > I have just removed the D: mapping in AFROS and it's still working > > correctly. > > Either you don't use the official ARAnyM 0.9.8beta, or we don't test in > the same way. I've just seen your other post, I'm going to continue there. I use the CVS version, naturally, but I have just disabled the "relative symlink" patch I cooked at lunch time and the worst what happened was that samantha.prg now points to /c/olga.app. Still a valid symlink, though. Petr