Any one ever try/use IFreeMem?
in [Mac Applications]
|
From: John W on 18 Feb 2007 01:43 On 17/2/07 18:05, in article 170220071205041771%dave(a)N_O_T_T_H_I_Sbalderstone.ca, "Dave Balderstone" <dave(a)N_O_T_T_H_I_Sbalderstone.ca> wrote: > In article <xrGBh.1026$EU.1001(a)trnddc07>, Wes Groleau > <groleau+news(a)freeshell.org> wrote: > >> Tim Lance wrote: >>> On Fri, 16 Feb 2007 22:26:52 -0600, Dave Balderstone wrote >> >>>> But that doesn't explain the age component. Why does "old" RAM acquire >>>> cruft? >>> >>> No joke. That's nuts and I thought so at the very first. >> >> Well, IF the developer is correct, AND you never reboot your machine >> (as some people here will advise you not to do) then small "memory >> leaks" will add up over time. And if they really are small leaks, >> then you get what is called heap fragmentation, which makes the >> slowdowns worse. > > Never rebooting would assume you never upgrade your OS, apply security > patches, suffer a kernel panic, experience a power outage... > > Nope. I don't buy it. Hi Guys, Back again, sorry I been quite I'll after food poisoning, I suspect it was the oysters from valentines day. Right, back to my flame, Firstly RAM does not age, or grow barnacles or anything like that. The primary goal of the application is to tackle a very noticeable delay (several seconds on MacBook Pro) that the system exerts when 'Free' memory becomes very low and it sets about paging out or stealing back inactive memory, I tackle this problem forcing an increase of the free memory. This prevents this stutter happening in the middle of my using an application. This is repeatable, every time, anytime an application needs to grab more memory than is marked Free (needs to reclaim from inactive memory). By not having a low memory situation this stutter absolutely goes away. My guess was some (but not all) people would actually appreciate this smoother running of their applications. Also, although one could argue that there is zero performance overhead to having a huge amount of inactive memory that's not needed lying around in ram. I would humbly suggest there can be some overhead when you have got over 500mb of memory fragments in the inactive memory list and the pager balances the queues.. See "Page Lists in the Kernel" http://developer.apple.com/documentation/Performance/Conceptual/ManagingMemo ry/Articles/AboutMemory.html The application is for those who want the maximum amount of free un-fragmented memory to avoid noticeable performance hits mid-application, and a way to reduce inactive memory without rebooting. You have 2 choices 1. Try it yourself, make a decision if it benefits you at least some of the time 2. Don't bother, because that guy accuses the author of producing something that don't work that tackles imaginary system delays I don't mind, I'm still very happy with the application doing its job well :)
From: Clever Monkey on 20 Feb 2007 12:11 John W wrote: > On 17/2/07 18:05, in article > 170220071205041771%dave(a)N_O_T_T_H_I_Sbalderstone.ca, "Dave Balderstone" > <dave(a)N_O_T_T_H_I_Sbalderstone.ca> wrote: > >> In article <xrGBh.1026$EU.1001(a)trnddc07>, Wes Groleau >> <groleau+news(a)freeshell.org> wrote: >> >>> Tim Lance wrote: >>>> On Fri, 16 Feb 2007 22:26:52 -0600, Dave Balderstone wrote >>>>> But that doesn't explain the age component. Why does "old" RAM acquire >>>>> cruft? >>>> No joke. That's nuts and I thought so at the very first. >>> Well, IF the developer is correct, AND you never reboot your machine >>> (as some people here will advise you not to do) then small "memory >>> leaks" will add up over time. And if they really are small leaks, >>> then you get what is called heap fragmentation, which makes the >>> slowdowns worse. >> Never rebooting would assume you never upgrade your OS, apply security >> patches, suffer a kernel panic, experience a power outage... >> >> Nope. I don't buy it. > > Hi Guys, > > Back again, sorry I been quite I'll after food poisoning, I suspect it was > the oysters from valentines day. > > Right, back to my flame, > > Firstly RAM does not age, or grow barnacles or anything like that. > > The primary goal of the application is to tackle a very noticeable delay > (several seconds on MacBook Pro) that the system exerts when 'Free' memory > becomes very low and it sets about paging out or stealing back inactive > memory, I tackle this problem forcing an increase of the free memory. This > prevents this stutter happening in the middle of my using an application. > > This is repeatable, every time, anytime an application needs to grab more > memory than is marked Free (needs to reclaim from inactive memory). By not > having a low memory situation this stutter absolutely goes away. My guess > was some (but not all) people would actually appreciate this smoother > running of their applications. > > Also, although one could argue that there is zero performance overhead to > having a huge amount of inactive memory that's not needed lying around in > ram. I would humbly suggest there can be some overhead when you have got > over 500mb of memory fragments in the inactive memory list and the pager > balances the queues.. See "Page Lists in the Kernel" > > http://developer.apple.com/documentation/Performance/Conceptual/ManagingMemo > ry/Articles/AboutMemory.html > > The application is for those who want the maximum amount of free > un-fragmented memory to avoid noticeable performance hits mid-application, > and a way to reduce inactive memory without rebooting. > > You have 2 choices > > 1. Try it yourself, make a decision if it benefits you at least some of the > time > > 2. Don't bother, because that guy accuses the author of producing something > that don't work that tackles imaginary system delays > > I don't mind, I'm still very happy with the application doing its job well > :) > This explanation does not match up with the verbiage on your web page. How, exactly, does caching files in memory have _anything_ to do with the age of your Mac? You say: "Over time you may notice your Mac takes a little longer to startup and there appears to be more disk activity? This is due to more applications starting up and reading files, and those files are kept in memory, just in-case you need to access them quickly again (the cache)." How is startup time affected _at all_ by the in-memory cache used by the filesystem in the _previous_ system state? We have rebooted, so therefore the cache items are being created anew. Furthermore, if you run the same startup items as you have done since the day the OS was installed, how is it possible for this in-memory cache to change at all? Are you saying that in-memory caches are persisted across restarts? I cannot for the life of me find the logic in this: "Simply use iFreeMem when you wish to reclaim the optimal amount of 'Free' memory and have smoother running applications ..... Just like when your Mac was brand new!" What does a newly installed OS have to do with anything you've said, above? How, in the name of all the gods, does memory become "cluttered" over the lifetime of the installed machine? You are making specious comments that have no basis in logic. Whether you are doing this on purpose or not, the net result is that the app starts to look like snake-oil. I've no doubt that there may be a place to exploit the memory manager to have it cough up unused pages upon demand. You _may_ even see a change in _some_ behaviour (though you might want to consider how you are measuring this -- wall clock measurements are totally bogus). Whether or not this perceived improvement offsets running an app in the doc all the time or could be duplicated as a general improvement on all machines remains to be proven. I think you need to come up with a more reasonable technical explanation on your web site, citing the basis for your research. Some metrics might be nice, too.
From: Wes Groleau on 21 Feb 2007 00:00 Clever Monkey wrote: > John W wrote: >> and a way to reduce inactive memory without rebooting. > How is startup time affected _at all_ by the in-memory cache used by the > filesystem in the _previous_ system state? We have rebooted, so > therefore the cache items are being created anew. Furthermore, if you > run the same startup items as you have done since the day the OS was > installed, how is it possible for this in-memory cache to change at all? >.... > Are you saying that in-memory caches are persisted across restarts? Is this National Straw Man Day? Maybe you have rebooted, but it seems pretty clear to me he said "without rebooting" How about arguing with what he _said_ instead of the opposite? Better yet, take his advice--IF you notice a slow down, consider it. If not, ignore him. Or take the in-between route--reboot. -- Wes Groleau "Two things are infinite, the universe and human stupidity. But I'm not so sure about the universe." -- Albert Einstein
From: Tom Harrington on 21 Feb 2007 11:54 In article <olQCh.1604$h8.1216(a)trnddc05>, Wes Groleau <groleau+news(a)freeshell.org> wrote: > Clever Monkey wrote: > > John W wrote: > >> and a way to reduce inactive memory without rebooting. > > > How is startup time affected _at all_ by the in-memory cache used by the > > filesystem in the _previous_ system state? We have rebooted, so > > therefore the cache items are being created anew. Furthermore, if you > > run the same startup items as you have done since the day the OS was > > installed, how is it possible for this in-memory cache to change at all? > >.... > > Are you saying that in-memory caches are persisted across restarts? > > Is this National Straw Man Day? > > Maybe you have rebooted, but it seems pretty clear to me he said > "without rebooting" Not on his web site-- which continues to make the ludicrous statements that Clever Monkey nicely deconstructed. -- Tom "Tom" Harrington MondoMouse makes your mouse mightier See http://www.atomicbird.com/mondomouse/
From: Clever Monkey on 21 Feb 2007 12:23
Wes Groleau wrote: > Clever Monkey wrote: >> John W wrote: >>> and a way to reduce inactive memory without rebooting. > >> How is startup time affected _at all_ by the in-memory cache used by >> the filesystem in the _previous_ system state? We have rebooted, so >> therefore the cache items are being created anew. Furthermore, if you >> run the same startup items as you have done since the day the OS was >> installed, how is it possible for this in-memory cache to change at all? > >.... >> Are you saying that in-memory caches are persisted across restarts? > > Is this National Straw Man Day? > No, it is National Read What I Wrote Day. > Maybe you have rebooted, but it seems pretty clear to me he said > "without rebooting" > > How about arguing with what he _said_ instead of the opposite? > How about you read what I wrote. I quoted the product web page. > Better yet, take his advice--IF you notice a slow down, consider it. > Here we are, talking about this. It's what people do, especially when they see something that makes no sense. Personally, I measure uptime in weeks. I'm not allergic to rebooting when I have to. I just rarely have to. |