• Open Menu Close Menu
  • Apple
  • Shopping Bag
  • Apple
  • Mac
  • iPad
  • iPhone
  • Watch
  • TV
  • Music
  • Support
  • Search apple.com
  • Shopping Bag

Lists

Open Menu Close Menu
  • Terms and Conditions
  • Lists hosted on this site
  • Email the Postmaster
  • Tips for posting to public mailing lists
Re: FSCopyObjectAsync: useless and crippled
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: FSCopyObjectAsync: useless and crippled


  • Subject: Re: FSCopyObjectAsync: useless and crippled
  • From: email@hidden
  • Date: Sat, 14 May 2005 14:50:21 +1000

You want FSCopyObject to pop-up a dialog and ask the user to authenticate a copy of a file they don't have permission to read? Then switch to privileged execution mode if it needs to? Automatically?

Yes. Why not? Isn't OS X a modern operating system? Why should it be so difficult?

What I suspect the original poster was trying to say, aside from the practical issues, is that this is a security nightmare - I don't want any app which happens to copy files occasionally getting root privileges. Even if it asks me for them; users are too compliant with typing their password, since they're prompted for it a dozen times a day.


Permission issues like this are thornier than most people acknowledge... while I really appreciate the new ability in Panther's Finder to authenticate when required for a copy/move/whatever, that's only because I trust the Finder. For 3rd party apps, I'd much rather fix the permissions myself, manually, than give them dangerous privileges.

What the difference? Either I write an API from scratch, or Apple provides an API. Either way the outcome is the same. The problem is that for a modern OS, copying a file should be easy. What's insane is that the Apple APIs are so pathetic. I can't believe someone actually got paid to design and write such crap.

It would be *nice* if copying a file were easy; in reality, exactly *because* MacOS X is a modern OS, there are many potential issues with copying any arbitrary file. For simplicity, we could all just go back to MacOS 9. Hope you're not used to having private data.


Btw, please keep your tone civil on the list, particularly when talking about Apple engineers. We all get frustrated at times, but don't go saying things you'll regret. Apple's engineers do a pretty fantastic job most of the time, all things considered. They've been making a noticeable effort to become more accessible and approachable since the release of Tiger, which comments like yours are only going to discourage.

The API could and should launch it's own setuid tool and handle all the communication and pass info to the app. If the app doesn't want user interaction, then the app can turn it off. Just like the alias resolving APIs. There are options for "no user interaction"

Since it's a bad idea to have a GUI app setuid, that requires either an extra GUI app just for the interface, or the 3rd party app in question to display the dialog. The latter isn't really an option from a security point of view. The former is a fair bit of complexity - this whole process could take tens of seconds, which means you [ideally] have to design your app asynchronously around it. You also need to be able to provide precise instructions about what you're doing to the GUI, so it can display them to the user properly and so the setuid tool can perform them... that then causes further issues if things don't go according to plan - like you run out of disk space, or there is another [different] permissions error, etc etc. Lots of callbacks? Nightmare.


Irregardless, this is a Carbon question and you should probably ask on the Carbon list.

Yeah, and get more non-answers.

Please, such comments are unnecessary.

I agree that the current implementation is lacking - I'm really just playing the devil's advocate here. But there's better ways to go about solving it, than just getting cranky on mailing lists. If you have a superior system, please do submit it to Apple, in whatever form you have*. Until then, yes, your complaint has been noted, so now you'll have to give Apple some time to digest it (if you're not going to solve it yourself).

* = Note that sending code to Apple sometimes makes them cranky... some legal B.S. Either include a legally valid license (e.g. BSD, APSL) or legally valid provision of rights to your unsolicited submission.

Wade Tregaskis (AIM/iChat, Yahoo & Skype: wadetregaskis, ICQ: 40056898, MSN, audio/video iChat & email: email@hidden, Jabber: email@hidden)
-- Sed quis custodiet ipsos custodes?


_______________________________________________
Do not post admin requests to the list. They will be ignored.
Cocoa-dev mailing list      (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden


  • Follow-Ups:
    • Re: FSCopyObjectAsync: useless and crippled
      • From: Steve Gehrman <email@hidden>
    • Re: FSCopyObjectAsync: useless and crippled
      • From: Dave Rehring <email@hidden>
References: 
 >Re: FSCopyObjectAsync: useless and crippled (From: James Bucanek <email@hidden>)
 >Re: FSCopyObjectAsync: useless and crippled (From: Steve Gehrman <email@hidden>)

  • Prev by Date: How to get observe KVO changes to *every* key for an object?
  • Next by Date: Re: FSCopyObjectAsync: useless and crippled
  • Previous by thread: Re: FSCopyObjectAsync: useless and crippled
  • Next by thread: Re: FSCopyObjectAsync: useless and crippled
  • Index(es):
    • Date
    • Thread