January 2008 - Darwin-dev - lists.apple.com

Re: Cancel Sleep
by Perry Smith 02 Jan '08

02 Jan '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On Jan 1, 2008, at 9:38 PM, Andrew James wrote: Andrew, Luke: Use the SOURCE... _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… Ok everyone calm down, yes this may not be something everyone will agree on, the people that use this KNOW the risks! How does one get a notification of the lid state changing (When the lid closes, nothing to do with sleep) that does not also mix up with the power state changing, you will notice i dont mention a single thing about forcing the system to stay awake or anything here, so please can we drop the debate and stop flooding the mailing list with the debate that will never end and just get to the point. which just as safe measure ill say once again... How does one get a notification of the lid state changing (When the lid closes, nothing to do with sleep) that does not also mix up with the power state changing Sorry for not knowing more. How much of the code is open? My approach would be to suck down what is open, feed it to cscope, and then try and follow the logic of the lid closing. If you know emacs, hooked up to cscope, you can ferret this out in a half a day tops. Judging from the quality of the code snippet you provided along with the dribble "I'm right -- You're wrong" I've watched over the past two days, these guys don't know the answer to your question. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Cancel Sleep
by John Scheiderer 02 Jan '08

02 Jan '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Cheers, John On Jan 1, 2008, at 10:24 PM, Steve Checkoway wrote: On Jan 1, 2008, at 7:00 PM, Michael Smith wrote: Sometime or other, Steve Checkoway <s(a)pahtak.org> wrote: Closing the clamshell is an explicit instruction from the user to sleep the system. -- Steve Checkoway "Anyone who says that the solution is to educate the users hasn't ever met an actual user." -- Bruce Schneier _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/john_scheiderer%40mac.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com Is there a way to make my computer wake up when the battery is dead, unplugged from power and the lid is closed? :-) (just joking) Now that I've deleted a bunch of emails on this topic, if I close the lid on my computer, I expect it to sleep, just like most people do. The 80% rule applies here on feature/functionality. The horse is dead, and now has flies. On Jan 1, 2008, at 5:22 PM, Michael Smith wrote: I don't buy this. Other laptops running other operating systems (Windows and Linux are the two that I've seen) can handle closing the laptop without sleeping. Windows != Linux != Mac OS X. I did not ever claim that was the case. Please do not misrepresent me. However, it is now the case that all three of these operating systems can run on the same hardware and two of the three can do something that Mac OS X cannot. It's as simple as that. This has wandered quite a bit off the original topic at this point, and become another one of those "I don't care what anyone else thinks, I reckon it should do <blah>" threads. You're right, from the very beginning, you've said "here's my reason this is a bad idea" and I think the reason you gave was poor. You belittled others' opinions and reasons for wanting such a feature. Saying "this is how it's always been done," while accurate neither helps anyone nor explains why it is a good idea and changing it is inherently bad, as you apparently think you've made clear. You have offered no technical reasons why this is a bad idea and have contributed nothing to this thread except your opinion, which you've made abundantly clear, as usual. This email sent to john_scheiderer(a)mac.com smime.p7s

1 0

Re: Cancel Sleep
by Andrew James 02 Jan '08

02 Jan '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:received:received:cc:message-id:from:to:in-reply-to:content-type:mime-version:subject:date:references:x-mailer; bh=6vTY0/nY3jXHUsIfpFaJhdpN5XTzjfWERPbppoPZbNE=; b=FrXtNZGcQGzoYMsM3/6GbyoOCUT2vEwrUne7bMPo7PPGp4X3iV91KvxHMmThTZe48B2+kXHzjAn5l6uumj2ZJ5JtH/qgPj5hItOyMvS53JxsUhoOa6xcNZQpx66056s/e6Fno/vol1FzqR2YaqmKnu8PhNKsT5v7ljHpqUH7+mg= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=cc:message-id:from:to:in-reply-to:content-type:mime-version:subject:date:references:x-mailer; b=D5sW4F61SM/MJfdaDiYRyqDoL9xJ/9mplti5VecdAEdJYgj/YgcEcZ7uBasaPZSjHGh70tVs8/OLAjKDqrcqQj7UdUZn0678/nuA9gJzIp1VNrSTOqKNJSYBlMUbHA+DuM4K3ETLZy7rrEUHms3CQvUF8nu70F1/nljzFlaNEEM= _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com

1 0

Re: Cancel Sleep
by Steve Checkoway 02 Jan '08

02 Jan '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On Jan 1, 2008, at 7:00 PM, Michael Smith wrote: Sometime or other, Steve Checkoway <s(a)pahtak.org> wrote: Closing the clamshell is an explicit instruction from the user to sleep the system. I don't buy this. Other laptops running other operating systems (Windows and Linux are the two that I've seen) can handle closing the laptop without sleeping. -- Steve Checkoway "Anyone who says that the solution is to educate the users hasn't ever met an actual user." -- Bruce Schneier _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com On Jan 1, 2008, at 5:22 PM, Michael Smith wrote: Windows != Linux != Mac OS X. I did not ever claim that was the case. Please do not misrepresent me. However, it is now the case that all three of these operating systems can run on the same hardware and two of the three can do something that Mac OS X cannot. It's as simple as that. This has wandered quite a bit off the original topic at this point, and become another one of those "I don't care what anyone else thinks, I reckon it should do <blah>" threads. You're right, from the very beginning, you've said "here's my reason this is a bad idea" and I think the reason you gave was poor. You belittled others' opinions and reasons for wanting such a feature. Saying "this is how it's always been done," while accurate neither helps anyone nor explains why it is a good idea and changing it is inherently bad, as you apparently think you've made clear. You have offered no technical reasons why this is a bad idea and have contributed nothing to this thread except your opinion, which you've made abundantly clear, as usual. smime.p7s

1 0

Re: Cancel Sleep
by Michael Smith 02 Jan '08

02 Jan '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com

1 0

Re: Cancel Sleep
by Steve Checkoway 02 Jan '08

02 Jan '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On Jan 1, 2008, at 5:22 PM, Michael Smith wrote: On Jan 1, 2008, at 2:55 PM, David Elliott wrote: On Dec 31, 2007, at 10:53 AM, Michael Smith wrote: -- Steve Checkoway "Anyone who says that the solution is to educate the users hasn't ever met an actual user." -- Bruce Schneier _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com What exactly is the reason for requiring the system to go to sleep and then be woken up again? Closing the clamshell is an explicit instruction from the user to sleep the system. I don't buy this. Other laptops running other operating systems (Windows and Linux are the two that I've seen) can handle closing the laptop without sleeping. It certainly wasn't an instruction from the user to sleep when I've seen it. In the cases I've seen, there was no external power connected. It was simply, close the laptop, move from one room to another, open the laptop and continue working. Your scenario is indistinguishable at the software level from my daily routine, where I close the lid on my MBP, pull all the cables from it and stuff it into my backpack. I don't stop to wait for it to sleep, or manually sleep it; I expect that when I close the lid, it will go to sleep and stay asleep. Then presumably, you would not enable this mode of operation since you don't want it. You can't trust the OS to do it, and in fact the system doesn't; the SMC will shut the system down hard if it is headed for a thermal emergency. So what is the problem? If the user screws up and overheats his machine, you're saying that it will shut down, hard, in order to protect the machine. You mentioned class action lawsuits (are you a lawyer? I am not) but if this mode of operation is something that has to be explicitly enabled by the user with a clear warning of the consequences and there is no way an application (other than say System Preferences) can enable it, then it is pretty clearly the users' fault and not Apple's. The point here being that an x86 Macintosh portable without airflow is almost certain to do this; a mode of operation that encourages the user to put the system into an enclosed space (bag, pile of stuff, etc.) is just not a viable thing. Nothing stops me from blocking the airflow to any of my computers. I expect (and you have confirmed above) that the machine will save itself if it overheats. Adding additional constraints in one specific situation seems pointless. That aside, which part of the design of the x86 Macintosh portables is so poor that it cannot do what other x86 portables can do? Just a guess, but I'd suspect nothing. The system is designed to deal with being overheated occasionally and accidentally, but not as a routine matter of course, which is what your new mode of operation will facilitate. Facilitate is not the same as cause. I just tried plugging in a monitor, keyboard, and power into my laptop while it was closed, pressing a key woke it, unplugging either the power or the monitor caused it to sleep. Why is it that the system has to sleep first and then be reawakened? Even in the case you mentioned, "close the lid on my MBP, pull all the cables from it and stuff it into my backpack" would work just fine. It wouldn't sleep until you pulled one of those two cables. It might be reasonable to ask Apple for a preference which dissociates closing the lid from the act of putting the system to sleep, although I would strongly expect that this would be refused for several sensible reasons, but it is *not* reasonable to ask for a programmatic way to subvert the user's intent in the fashion the OP desires. p.s. If you wonder why such a preference would probably be refused, examine the Apple support forums and other web archives for complaints from users relating to systems that fail (due to bugs or hardware issues) to sleep when expected, and add the likely incidence of user error and malicious/incompetent/well-meaning-but- wrong-headed programmatic manipulation of the preference to the number you encounter. The result would not be pretty. It is reasonable to ask for such a preference. Your "sensible reason" is pretty poor. Don't let the preference be set programatically, similar to how display rotation does not seem to be settable programatically (although, it would be very handy if I am incorrect and I could set it programatically). smime.p7s

1 0

Re: Cancel Sleep
by Amanda Walker 02 Jan '08

02 Jan '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Well, it has been for most of Apple laptop history :-). --Amanda _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… On Jan 1, 2008, at 8:43 PM, Marc Epard wrote: On Jan 1, 2008, at 7:22 PM, Michael Smith wrote: Closing the clamshell is an explicit instruction from the user to sleep the system. Uh, no. It's the pulling of the cables that is my instruction to sleep. If I close the lid, but leave the keyboard, external monitor, and power connected, I intend to simply reduce my screen real estate to just the external monitor. Hmm. At best I'd say it's the combination of closing the lid and unplugging the cables. I plug and unplug stuff all the time, and I'd be very annoyed if my machine went to sleep when I unplugged a monitor or USB mouse. Going back to Andrew's question, though. Andrew, there is no cookbook answer, since there's no API for doing the particular things you want. This means you will need to use undocumented mechanisms, which may take some experimentation, and may be fragile when Apple changes the OS. This doesn't mean that you can't do so (many shipping products do undocumented or unsupported things), it just means that no one has a simple answer to give you. Too illustrate this, here's how I'd approach the problem if I were trying to change the system's behavior in this way: - Download the kernel sources and study them carefully. - Write an IOKit driver that always loads - When the driver starts, have it patch the running kernel to interpose your code (the easiest way is to patch the function prolog for a notification bottleneck, if one exists). - To be extra friendly, allow the behavior to be turned on and off via the driver's user client, so that you can hook it up to a prefpane, control it from an application, etc. The driver's workloop may also be handy as a way to do periodic processing, catch interrupts, etc. if you need it. - Your driver must never unload. - You must declare the kernel dependencies correctly so that it only loads into versions of the kernel that you've tested it on to confirm that it's compatible. If you want to know which specific calls to interpose on, and code for doing that (that is, you want someone to do the actual work involved), there are probably quite a number of people on darwin-drivers that have the requisite knowledge. The last time I did IOKit consulting my hourly rate was $150/hr, but you could probably find someone cheaper. If you are not comfortable writing code that patches a running kernel, perhaps this is not the best project :-). This email sent to site_archiver(a)lists.apple.com

1 0

Re: Cancel Sleep
by David Elliott 02 Jan '08

02 Jan '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Hi Michael, On Jan 1, 2008, at 8:22 PM, Michael Smith wrote: On Jan 1, 2008, at 2:55 PM, David Elliott wrote: -Dave _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… What exactly is the reason for requiring the system to go to sleep and then be woken up again? Closing the clamshell is an explicit instruction from the user to sleep the system. No, it's not. It's only an instruction from me that I don't wish to use the built-in display and keyboard anymore. I've at this point already got external power, an external display, and an external keyboard so it should be really obvious to even the most simple code that I, the user, wish to keep the system awake. Why the hell would I plug in all my peripherals, close the lid, and want the system to go to sleep? I wouldn't. Perhaps though you mistake my scenario (wanting to avoid a needless 1 second sleep) with the scenario that other people have which is to allow the laptop to run without anything connected at all. I agree that the OS should do all it can to prevent that, unless the user installs extra software in which case.. hey.. the user knows what he's doing. That said, I shouldn't have to install extra software just to keep closing the lid from sleeping the machine when I have stuff connected to it. It should "just work" and it doesn't. Your scenario is indistinguishable at the software level from my daily routine, where I close the lid on my MBP, pull all the cables from it and stuff it into my backpack. I don't stop to wait for it to sleep, or manually sleep it; I expect that when I close the lid, it will go to sleep and stay asleep. Totally false. Your scenario involves pulling all the cables. My scenario involves plugging them all in just before closing the lid. Here is how I see your scenario playing out if lid close does not sleep the system: 1. Close the lid System realizes it has external peripherals (mains power, display, keyboard) and remains on 2. Pull all the cables. System realizes it's now no longer using mains power and goes to sleep. 3. Stuff it in your backpack. My scenario (leaving power and peripherals connected) and your scenario (disconnecting everything) are easily distinguishable at the software level. Perhaps you missed that I (unlike some of the other people on this thread) do intend to keep mains power connected. I agree with you that if no peripherals are connected then indeed the scenario is indistinguishable to the machine. And of course in that case it is better safe than sorry. Oh, and by the way: Step 2 (pull out all the cables) puts my machine to sleep already. So it's really just a matter of modifying the lid close handler to check for power and peripherals. If they are disconnected after lid close, the machine will go to sleep, and that code already exists. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Cancel Sleep
by Marc Epard 02 Jan '08

02 Jan '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On Jan 1, 2008, at 7:22 PM, Michael Smith wrote: Uh, no. -Marc _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… Closing the clamshell is an explicit instruction from the user to sleep the system. Your scenario is indistinguishable at the software level from my daily routine, where I close the lid on my MBP, pull all the cables from it and stuff it into my backpack. I don't stop to wait for it to sleep, or manually sleep it; I expect that when I close the lid, it will go to sleep and stay asleep. It's the pulling of the cables that is my instruction to sleep. If I close the lid, but leave the keyboard, external monitor, and power connected, I intend to simply reduce my screen real estate to just the external monitor. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Cancel Sleep
by Michael Smith 02 Jan '08

02 Jan '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On Jan 1, 2008, at 2:55 PM, David Elliott wrote: On Jan 1, 2008, at 3:30 PM, Michael Smith wrote: = Mike _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… What exactly is the reason for requiring the system to go to sleep and then be woken up again? Closing the clamshell is an explicit instruction from the user to sleep the system. Your scenario is indistinguishable at the software level from my daily routine, where I close the lid on my MBP, pull all the cables from it and stuff it into my backpack. I don't stop to wait for it to sleep, or manually sleep it; I expect that when I close the lid, it will go to sleep and stay asleep. On Jan 1, 2008, at 3:02 PM, David Alger wrote: On Jan 1, 2008, at 4:55 PM, David Elliott wrote: Certain Apple portables do support running with the clamshell closed, but with other constraints applied that make it clear that the user intends for the system to be awake and that more or less require that it have adequate access to cooling (keyboard attached, external monitor attached, explicit wakeup after the clamshell is closed). If it doesn't have the cooling, it will get too hot. If and when that happens, the OS should halt the machine to prevent damage. Mac OS X is certainly smart enough to do this if temps reach an unsafe temp. You can't trust the OS to do it, and in fact the system doesn't; the SMC will shut the system down hard if it is headed for a thermal emergency. The point here being that an x86 Macintosh portable without airflow is almost certain to do this; a mode of operation that encourages the user to put the system into an enclosed space (bag, pile of stuff, etc.) is just not a viable thing. This email sent to site_archiver(a)lists.apple.com

1 0