Jump to content



Photo

Mac or Hack


  • Please log in to reply
72 replies to this topic

#46 protocol7

protocol7

    It's Fumbles. It was always Fumbles.

  • Joined: 06-May 09

Posted 31 January 2014 - 13:33

If I was going to buy a Mac desktop I'd go for a Mini too. For the size they're great and with a cheap RAM upgrade (does ANYONE actually buy RAM from Apple?) it'll fly.

 

I have a MacBook here and two desktop hacks. Hackintoshing has gotten easier as time goes on. People who run into trouble almost always got there by relying on hacked-up distros or installation tools like the ones from phoneymac. I haven't had an update break my installs.

 

At the high end, I think a Hack makes more sense. Especially with the new MacPro using external TB expansions. I built a hack from scratch last year (for the first time, usually I just hacked my existing PC) because it made more sense than buying a 3 or 4 year old Mac Pro for the same price with lower performance and no warranties.




#47 #Michael

#Michael

    Neowinian Senior

  • Tech Issues Solved: 1
  • Joined: 28-August 01

Posted 31 January 2014 - 14:23

Yes, there is.  There are threads on most Mac fora about drive corruption on OS X.

 Let's be honest - every OS has drive corruption issues - without exception.  There are methods in each OS to lessen the impact of corruption - however, no operating system has any method of blocking corruption.

Even OS X third-party utilities can only go so far at reducing the impact (to be fair, the same goes for first-thru-third-party utilities for any OS) -- however, only the BSDs and their derivatives - including OS X - utterly lack non-CLI utilities for dealing with it when it happens.

 

There are utilities out there for dealing with drive issues.  Disk utilities and onyx are first on my list.  Neither require any use of terminal. 



#48 ndoggfromhell

ndoggfromhell

    Master of the 6th Speed

  • Joined: 20-March 02

Posted 31 January 2014 - 15:56

I'd say the chances are it's something you're doing over it being an OSX issue. Two completely different drives in one year is either just random plain coincidence or something wrong you're doing.

My H/D in this macbook died whilst I was using it (click of death) and was pretty outraged that a 4.5 year old drive had died already, but you can only blame the hardware company for that (and as apple don't make hard drives the blame doesn't lie with them)

Neither was mine. The dealership I work for is owned by a pretty large family and besides working on the business equipment, they have me working on their personal equipment.  Both were however from the same set of users.  



#49 threetonesun

threetonesun

    Neowinian Senior

  • Tech Issues Solved: 1
  • Joined: 26-February 02

Posted 31 January 2014 - 18:41

Neither was mine. The dealership I work for is owned by a pretty large family and besides working on the business equipment, they have me working on their personal equipment.  Both were however from the same set of users.  

 

Yeah... I've heard of Macs locking up on reboot with permission errors, and Macs dying while being used due to hardware issues, but just dropping dead while running due to some weird software / OS conflict I've never heard of. 



#50 Shadrack

Shadrack

    Neowinian Senior

  • Tech Issues Solved: 3
  • Joined: 20-December 01

Posted 31 January 2014 - 23:29

It wouldn't be my first, but the others I bought hardware that was specified in a tutorial and followed it.  Never applied updates etc - and to be fair, it wasn't the best experience (dual monitor never worked, for example, had to use an external soundcard)...

 

I don't want to use this for "learning" or "tinkering" - it is to be my daily computer I use for tasks.  When a software update comes out, I want to be able to apply it and carry on working - not worry about "Oh damn, my soundcard no longer works" and so on.

 

Then get the real deal.  Hackintosh setups can work reliably for months but (in my experience) some issues can crop up and you have to be ready to take the gloves off (i.e., spend some time researching the issue, asking for advice on forums, etc.).  My experience is from years ago, so it *IS* probably better today (and if you follow a hardware guide so that drivers are well supported).

 

I'm considering going with a Hackintosh setup to replace my OLD 1st generation PowerPC Mac Mini, but I'm also OK with spending time on the tinkering of the setup if an update comes out or w/e.

 

That's not to say that having an Apple product will be bullet proof.  Obviously, there can always be technical issues that might crop up and need to be addressed.  Components can disfunction and fail and software installs may become corrupted.  Difference is: size of community and install base.  Hackintosh users are an extreme minority compared to Macintosh users (and even Linux users, probably).  That's not to say that there isn't help available out there, if you are willing to take the time and ask.

 

If uptime is important to you, just go with the official Apple product (whatever it might end up being).  They are pretty consistent with their customer service.  If you have an Apple Store in your town you can likely get an issue resolved relatively quickly.  Some people think that their out-of-warranty-service fees are high, and they do have a markup beyond the cost of a part.  But they also seem to support services for older Macintosh hardware for a pretty decent span of time IMO.  I've seen them helping people out with really old iBooks, for example.



#51 OP +Nik L

Nik L

    Where's my pants?

  • Tech Issues Solved: 2
  • Joined: 14-January 03

Posted 31 January 2014 - 23:51

Agreed.  Apple hardware hols it's value very well too. :)



#52 Brian M.

Brian M.

    Neowinian Senior

  • Tech Issues Solved: 11
  • Joined: 07-January 05
  • Location: London, UK

Posted 01 February 2014 - 14:04

If I swap out the HDD to an SSD - is there anything special I need to do?

 
Enable trim. By default OS X doesn't allow trim on any SSD that isn't Apple supplied. It's an easy thing to so though.
 

And there are problems (specifically, issues with drive corruption) that occur with surprising frequency on real Macs that Hack users can fix far easier than real Mac users.  (I mentioned elsewhere that fixing drive corruption - which even OS X is not immune to - requires dropping all the way to the command line.  There is no GUI-based fix for this issue - which all versions of OS X are vulnerable to.  Worse, it's not the only problem that OS X has that requires CLI experience to fix.  If you have any experience at all with a command-line (such as Linux or Windows, or especially BSD), there's surprisingly little different in the OS X CLI/terminal environment compared to BSD; however, despite that, it's something that Apple discourages.)

 
Not really sure where you've got this from. Hackintosh uses the same OS as Macs - so if it's fixable on one then it's fixable on the other.

And as for drive corruption, there is no risk running on Apple hardware. Both use HFS+, which, from my experience, is much more immune to corruption than other filesystems.

 

By the way - check the refurb store - there's quite a few Mac Mini's on there today - including an i7 with a 1TB drive for £579, or a previous gen i5 (dedicated graphics) for less than a current base model :). You can also get a refurb current base model for £419. http://store.apple.c...pecialdeals/mac



#53 OP +Nik L

Nik L

    Where's my pants?

  • Tech Issues Solved: 2
  • Joined: 14-January 03

Posted 03 February 2014 - 23:14

What does enabling TRIM do?  And how easy? LOL



#54 Shadrack

Shadrack

    Neowinian Senior

  • Tech Issues Solved: 3
  • Joined: 20-December 01

Posted 04 February 2014 - 21:58

What does enabling TRIM do?  And how easy? LOL

 

DISCLAIMER: I am not an SSD expert.

 

I had TRIM enabled initially (following the hack) on a 80GB Intel SSD (320 series) that I upgraded to in my 2009 MBP.  When doing some research I came across some folks who suggested that this could actually lead to problems with the drive.  Upon upgrading to the new Mac OS X (at the time I think it was Lion), I decided not to perform the enable TRIM hack.  I really noticed no difference performance wise in day-to-day routines (comparing TRIM and no TRIM).  After a few years of usage without TRIM, I performed a benchmark test (I reported it here, I can try and find it if you want) and found that my Read and Write were nominal with Intel's datasheet for my drive.

 

TRIM is explained on a lot of different web sites, and I recommend researching it.  If in doubt, however, I would recommend not enabling TRIM because from what I've read, newer SSDs probably do not need TRIM OS support to function correctly for years.  It seems like it was more of an issue at the birth of consumer level SSDs that is not as much of an issue anymore.  That's my opinion, but re-read my disclaimer.



#55 +snaphat (Myles Landwehr)

snaphat (Myles Landwehr)

    Electrical & Computer Engineer

  • Tech Issues Solved: 29
  • Joined: 23-August 05
  • OS: Win/Lin/Bsd/Osx
  • Phone: dumb phone

Posted 04 February 2014 - 23:06

DISCLAIMER: I am not an SSD expert.

 

I had TRIM enabled initially (following the hack) on a 80GB Intel SSD (320 series) that I upgraded to in my 2009 MBP.  When doing some research I came across some folks who suggested that this could actually lead to problems with the drive.  Upon upgrading to the new Mac OS X (at the time I think it was Lion), I decided not to perform the enable TRIM hack.  I really noticed no difference performance wise in day-to-day routines (comparing TRIM and no TRIM).  After a few years of usage without TRIM, I performed a benchmark test (I reported it here, I can try and find it if you want) and found that my Read and Write were nominal with Intel's datasheet for my drive.

 

TRIM is explained on a lot of different web sites, and I recommend researching it.  If in doubt, however, I would recommend not enabling TRIM because from what I've read, newer SSDs probably do not need TRIM OS support to function correctly for years.  It seems like it was more of an issue at the birth of consumer level SSDs that is not as much of an issue anymore.  That's my opinion, but re-read my disclaimer.

SSDs maintain a list of internal blocks that are marked as "clean" for writing. Writes are going to be of nominal performance if that list is mostly free. Performance degradation is really only going to crop up if you have large numbers of cyclic writes and erases to the drive. Why?

 

Because deleting files on a file system does not tell the SSD that the corresponding blocks are now invalid so the drive is doomed to treating the blocks are valid until such a point that new data is written over the logical locations where the old files were. Partial deletions and modifications to files don't have this problem because those are really writes to the file system, so the SSD is given the logical locations in such cases and immediately marks the blocks as invalid.

 

The point, I'm making here is that essentially if you aren't doing a lot of writes and deletions, your drive is going to have a pretty good idea of what is really valid block wise, but if you are, you'll eventually get to a point where there are many invalid blocks still marked as valid and so when the drive has to dish out new blocks, it has to consolidate partially written blocks into a single block before writing the data you requested to be written. At the end of the day, performance degradation is going to be largely dependent on your use-case.

 

I feel like many people don't understand properly what TRIM brings to the table that internal garbage collection can't achieve. It is not the fault of people though, it's really that what I just said about invalid blocks still being marked as valid after deletions is rarely explained properly in discussions of TRIM.

 

Wikipedia has a nice brief explanation of what I just said (in case my own isn't coming across to folk):

http://en.wikipedia....lification#TRIM



#56 Shadrack

Shadrack

    Neowinian Senior

  • Tech Issues Solved: 3
  • Joined: 20-December 01

Posted 04 February 2014 - 23:30

SSDs maintain a list of internal blocks that are marked as "clean" for writing. Writes are going to be of nominal performance if that list is mostly free. Performance degradation is really only going to crop up if you have large numbers of cyclic writes and erases to the drive. Why?

 

Because deleting files on a file system does not tell the SSD that the corresponding blocks are now invalid so the drive is doomed to treating the blocks are valid until such a point that new data is written over the logical locations where the old files were. Partial deletions and modifications to files don't have this problem because those are really writes to the file system, so the SSD is given the logical locations in such cases and immediately marks the blocks as invalid.

 

The point, I'm making here is that essentially if you aren't doing a lot of writes and deletions, your drive is going to have a pretty good idea of what is really valid block wise, but if you are, you'll eventually get to a point where there are many invalid blocks still marked as valid and so when the drive has to dish out new blocks, it has to consolidate partially written blocks into a single block before writing the data you requested to be written. At the end of the day, performance degradation is going to be largely dependent on your use-case.

 

I feel like many people don't understand properly what TRIM brings to the table that internal garbage collection can't achieve. It is not the fault of people though, it's really that what I just said about invalid blocks still being marked as valid after deletions is rarely explained properly in discussions of TRIM.

 

Wikipedia has a nice brief explanation of what I just said (in case my own isn't coming across to folk):

http://en.wikipedia....lification#TRIM

 

Sounds important enough.  Why do you think Apple has kept from enabling TRIM for 3rd party drives and MS has only recently made it a non-hack for Windows 8 (Windows 7, you must use a similar hack IIRC)?

 

Anyway, doing my own research I come across a lot of folks saying things like:

 

Most new SSDs come with garbarge collection which negates having TRIM enabled in the OS. I don't know about the Samsung 840 Pro series.

 

I have an Intel 520 Series SSD and do not have TRIM in OS X since it has garbage collection. It's performance is still outstanding.

Do they not know what they are talking about? (source: https://discussions....thread/5092567)

 

Using a 'hack' for performance reasons is something I understand.  Using a 'hack' to prevent adversely effecting hardware seems suspect to me.



#57 +snaphat (Myles Landwehr)

snaphat (Myles Landwehr)

    Electrical & Computer Engineer

  • Tech Issues Solved: 29
  • Joined: 23-August 05
  • OS: Win/Lin/Bsd/Osx
  • Phone: dumb phone

Posted 05 February 2014 - 00:14

Sounds important enough.  Why do you think Apple has kept from enabling TRIM for 3rd party drives and MS has only recently made it a non-hack for Windows 8 (Windows 7, you must use a similar hack IIRC)?

 

Anyway, doing my own research I come across a lot of folks saying things like:

 

 

 

 

Do they not know what they are talking about? (source: https://discussions....thread/5092567)

Windows 7 wasn't SSD aware because they weren't so prevalent back then. If you wanted "proper" TRIM support, you had to install 3rd party tools (provided by the manufacturer) which would periodically run TRIM operations. There was an alternative way to do it via a registry hack (DisableDeleteNotify) which would force the OS to report deletes to the device. It's not really a good way to do it because it can cause the SSD to perform on-the-fly erases which cause unnecessary wear in the SSD cells. You can do the same thing in Linux via a file system discard option, but I digress.

 

Your link has a poster saying what I just said (confusingly though) to the other person who said garbage collection (GC) negates the need for TRIM. To be clear, garbage collection is an idle process that consolidates partially used blocks in order to maintain maximum contiguous free blocks and to perform drive wear-leveling. This is useful because on-the-fly erases to the SSD cells are slow and as such you want to maintain as many free cells as possible for on-the-fly writing. It works great when the drive actually has an accurate idea about what blocks are really valid versus invalid, but if you do lots of large deletions all over the place, then the drive itself will think a many invalid blocks are still valid (per the issue I said before). The garbage collector in this case would see these invalid blocks as valid blocks and treat them as such. Basically, at the simplest, the drive just has incorrect information and so does the garbage collector here.

 

I've often heard what you are saying: that garbage collection is much better these days and that it negates the need for TRIM. I think the reality is that garbage collection algorithms have gotten to the point that they do a really good job of idly consolidating used blocks when under pressure (i.e. the majority of the drive cells are marked as "used" in the view of the SSD*) so it takes more data on the drive before you get to the point where the drive needs to perform on-the-fly data consolidation when you want to write data to the drive (i.e. performance degradation starts to show). Of course, having a larger drive is going to also mitigate the issue. So essentially, in practice, you would have to peg the drive harder before degradation started showing up. So basically it shifts the real-world scenarios more toward what you saw in practice: nominal read/write speeds after years in large number of use-cases.

 

*Note: it doesn't matter if the blocks are actually really valid or incorrectly marked as valid w.r.t. this statement, I'm just saying the SSD itself thinks the majority of the space is used.



#58 PGHammer

PGHammer

    Neowinian Senior

  • Tech Issues Solved: 1
  • Joined: 31-August 03
  • Location: Accokeek, MD
  • OS: Windows 8 Pro with Media Center x64

Posted 10 February 2014 - 00:57

 
Enable trim. By default OS X doesn't allow trim on any SSD that isn't Apple supplied. It's an easy thing to so though.
 

 
Not really sure where you've got this from. Hackintosh uses the same OS as Macs - so if it's fixable on one then it's fixable on the other.

And as for drive corruption, there is no risk running on Apple hardware. Both use HFS+, which, from my experience, is much more immune to corruption than other filesystems.

 

By the way - check the refurb store - there's quite a few Mac Mini's on there today - including an i7 with a 1TB drive for £579, or a previous gen i5 (dedicated graphics) for less than a current base model :). You can also get a refurb current base model for £419. http://store.apple.c...pecialdeals/mac

I didn't say that it couldn't be done on real Macs - I specifically said that it COULD be done on real Macs; however, this is something that Apple discourages (doing anything from the CLI).

However, "Hack" usage gets you used to dealing with the CLI.



#59 vhane

vhane

    Neowinian

  • Tech Issues Solved: 1
  • Joined: 15-August 04

Posted 10 February 2014 - 04:15

I didn't say that it couldn't be done on real Macs - I specifically said that it COULD be done on real Macs; however, this is something that Apple discourages (doing anything from the CLI).

However, "Hack" usage gets you used to dealing with the CLI.

 

Apple doesn't discourage CLI usage. Those tools are actively developed and maintained so that you can make use of them.



#60 Neztea

Neztea

    DoN't Be SuCh A pLaYa HatA!

  • Joined: 10-November 01

Posted 10 February 2014 - 04:33

http://www.tonymacx8...abyte-brix.html

 

How about one of these devices?