If you 'matter' or not from a pro perspective does not depend on your upgrade cycle. If you do professional work on your computer you are a pro. Some pro environments require you to be as up to date as possible. Others do not.
If you don't need to update in 8 years and can still function with a Core 2 Duo in 2015 then you do not have Pro level needs. Your use case does not matter one iota in the discussion of MacBook Pro designs for use by professionals with demanding workflows.
Your needs can be met with an Ultrabook (aka MBA and MB) designed for professional users without computationally demanding workflows.
Your argument is like claiming that anyone who takes pictures with their iPhone as part of their job is a pro photographer and their use case is as valid in the discussion of pro level cameras like the Nikon D5. No, it isn't.
Your ability to upgrade to a new machine is wholly dependent on your being able to afford one, be it at a company or self employed level.
If you can afford a $4000 MBP then you have the ability to afford a 3 year upgrade cycle for a laptop that is more suited to your needs and income. That Apple doesn't build the MBP to suit your needs and caters more to the needs of professionals with demanding workflows which require more up to date equipment instead is NOT a problem.
My post was on your assertion of the three year pro cycle. You chose to jump to the extreme case of an eight year cycle but what if the cycle were four or five years? I'm seeing four plus years in critical infrastructure datacenters and the same is true of many pros and companies. Just read the boards here to get a basic idea. Yes, some people can upgrade more frequently. Others, not so frequently. The reasons are clear.
A three year pro cycle is average. Two year or four year is also common. However the assertion that the MBP must be designed with user replaceable components to support a 5-8 year lifecycle is ludicrous when the average professional refresh cycle is 3 years. That's not the same refresh lifecycle for an office worker but one that fits creatives, engineers and developers that have pro level computational requirements.
Yes, the reason is clear why some folks don't update for 5+ years. They don't have pro level computational needs where time=money. They may do professional work on their computers but the computer is not the most essential tool to completing their job.
If you 'matter' or not from a pro perspective does not depend on your upgrade cycle. If you do professional work on your computer you are a pro. Some pro environments require you to be as up to date as possible. Others do not.
If you don't need to update in 8 years and can still function with a Core 2 Duo in 2015 then you do not have Pro level needs. Your use case does not matter one iota in the discussion of MacBook Pro designs for use by professionals with demanding workflows.
Your needs can be met with an Ultrabook (aka MBA and MB) designed for professional users without computationally demanding workflows.
Your argument is like claiming that anyone who takes pictures with their iPhone as part of their job is a pro photographer and their use case is as valid in the discussion of pro level cameras like the Nikon D5. No, it isn't.
Your ability to upgrade to a new machine is wholly dependent on your being able to afford one, be it at a company or self employed level.
If you can afford a $4000 MBP then you have the ability to afford a 3 year upgrade cycle for a laptop that is more suited to your needs and income. That Apple doesn't build the MBP to suit your needs and caters more to the needs of professionals with demanding workflows which require more up to date equipment instead is NOT a problem.
My post was on your assertion of the three year pro cycle. You chose to jump to the extreme case of an eight year cycle but what if the cycle were four or five years? I'm seeing four plus years in critical infrastructure datacenters and the same is true of many pros and companies. Just read the boards here to get a basic idea. Yes, some people can upgrade more frequently. Others, not so frequently. The reasons are clear.
A three year pro cycle is average. Two year or four year is also common. However the assertion that the MBP must be designed with user replaceable components to support a 5-8 year lifecycle is ludicrous when the average professional refresh cycle is 3 years. That's not the same refresh lifecycle for an office worker but one that fits creatives, engineers and developers that have pro level computational requirements.
Yes, the reason is clear why some folks don't update for 5+ years. They don't have pro level computational needs where time=money. They may do professional work on their computers but the computer is not the most essential tool to completing their job.
If you 'matter' or not from a pro perspective does not depend on your upgrade cycle. If you do professional work on your computer you are a pro. Some pro environments require you to be as up to date as possible. Others do not.
If you don't need to update in 8 years and can still function with a Core 2 Duo in 2015 then you do not have Pro level needs. Your use case does not matter one iota in the discussion of MacBook Pro designs for use by professionals with demanding workflows.
Your needs can be met with an Ultrabook (aka MBA and MB) designed for professional users without computationally demanding workflows.
Your argument is like claiming that anyone who takes pictures with their iPhone as part of their job is a pro photographer and their use case is as valid in the discussion of pro level cameras like the Nikon D5. No, it isn't.
Your ability to upgrade to a new machine is wholly dependent on your being able to afford one, be it at a company or self employed level.
If you can afford a $4000 MBP then you have the ability to afford a 3 year upgrade cycle for a laptop that is more suited to your needs and income. That Apple doesn't build the MBP to suit your needs and caters more to the needs of professionals with demanding workflows which require more up to date equipment instead is NOT a problem.
My post was on your assertion of the three year pro cycle. You chose to jump to the extreme case of an eight year cycle but what if the cycle were four or five years? I'm seeing four plus years in critical infrastructure datacenters and the same is true of many pros and companies. Just read the boards here to get a basic idea. Yes, some people can upgrade more frequently. Others, not so frequently. The reasons are clear.
A three year pro cycle is average. Two year or four year is also common. However the assertion that the MBP must be designed with user replaceable components to support a 5-8 year lifecycle is ludicrous when the average professional refresh cycle is 3 years. That's not the same refresh lifecycle for an office worker but one that fits creatives, engineers and developers that have pro level computational requirements.
Yes, the reason is clear why some folks don't update for 5+ years. They don't have pro level computational needs where time=money. They may do professional work on their computers but the computer is not the most essential tool to completing their job.
I'm going to disagree.
My last machine lasted five and a half years -- though it spent the last year or so often absolutely at (and often enough OVER) its limit. And making me a living. I was actually prepared to replace it after three years, but then Apple released Mavericks and breathed an extra two years of life into it (memory compression, wow!). At the point where it started dying, rumours of impending updates had congealed to the point where I decided to nurse it along for another six or seven months until new machines were released.
Going by that experience, I went in for the new 15" machine, hoping that it, too, will last me around five years. Yes, it will be written off after three, and I'll probably be able to afford a replacement at that time, but it would be nice not to HAVE to replace it.
I absolutely have "pro" computational needs (this is a production and stage performance machine, among other things), but having funds available to other business needs is a good thing. So if this thing lasts me five years, I'm not gonna replace it after three. I have enough other gear investments I can make if I need to for tax reasons.
My case doesn't boil down to 'this' or 'that'. They could have released machines for both markets. One playing safe and one taking risks.
Resources are a problem but more a management problem than anything else. Although Apple denies it, everything indicates that engineers have been assigned to non Mac areas and are perhaps spread too thin on the ground. ... That's why I vouch for separating the two divisions. I wonder if any shareholders will bring this up in February.
Your capacity for doublethink is astounding. In your world Apple's management should have had engineers create two distinct lines of otherwise similar Macs yet those engineers are too thin on the ground to make successful iOS and Mac products at the same time.
Have you seen many cases of the SSD and RAM socket connections to the main board failing? Have you seen many cases of solder failing (especially but not limited to graphics cards)? I bet you've seen far more of the latter.
Tell me how much care you put into soldering quality and I will tell you how reliable your machine will be. Tell me about your thermal design and I will tell you even more.
You're really not thinking things through: Aren't the electrical connections for SSD and RAM sockets SOLDERED to the main board? If soldering was a greater disaster than socketing then the RAM/SSD connections will be failing at least as frequently as any other soldered component, if not more as you've added the extra failure point between the socket and the RAM/SSD component. If components were more likely to fail being soldered than socketed then most components on a motherboard would be socketed. The fact that the vast majority of components are soldered onto motherboards THROUGHOUT the electronic industry implies soldering is superior to socketing.
but Apple left a lot out of that elegant box. For starters, it comes with no floppy drive (though you can buy an external USB version for $70 or so), no free drive bays, and only one available expansion slot. Apple tells users to "think different." We advise Apple to "think again."
The precedent of time is against you avon b7. The iMac had similar criticisms levelled against it and twenty years down the line it seems Apple has done okay! The conditions for the MacBook Pro adopting Thunderbolt 3/USB-C are even more favourable now than those for the iMac and USB[-A] in 1998.
Nobody else remembers an entire series of PowerBook with failing lower RAM slots? If memory serves, they even instituted an Extended Repair Program for that at the time.
Also, googling "MacBook RAM slot failure" confirms my personal recollection from years in sales and support that RAM sockets would fail ALL THE TIME. Not systematically, but certainly a common point of failure.
My case doesn't boil down to 'this' or 'that'. They could have released machines for both markets. One playing safe and one taking risks.
Resources are a problem but more a management problem than anything else. Although Apple denies it, everything indicates that engineers have been assigned to non Mac areas and are perhaps spread too thin on the ground. ... That's why I vouch for separating the two divisions. I wonder if any shareholders will bring this up in February.
Your capacity for doublethink is astounding. In your world Apple's management should have had engineers create two distinct lines of otherwise similar Macs yet those engineers are too thin on the ground to make successful iOS and Mac products at the same time.
Have you seen many cases of the SSD and RAM socket connections to the main board failing? Have you seen many cases of solder failing (especially but not limited to graphics cards)? I bet you've seen far more of the latter.
Tell me how much care you put into soldering quality and I will tell you how reliable your machine will be. Tell me about your thermal design and I will tell you even more.
You're really not thinking things through: Aren't the electrical connections for SSD and RAM sockets SOLDERED to the main board? If soldering was a greater disaster than socketing then the RAM/SSD connections will be failing at least as frequently as any other soldered component, if not more as you've added the extra failure point between the socket and the RAM/SSD component. If components were more likely to fail being soldered than socketed then most components on a motherboard would be socketed. The fact that the vast majority of components are soldered onto motherboards THROUGHOUT the electronic industry implies soldering is superior to socketing.
November 9, 1998, but Apple left a lot out of that elegant box. For starters, it comes with no floppy drive (though you can buy an external USB version for $70 or so), no free drive bays, and only one available expansion slot.
Apple tells users to "think different." We advise Apple to "think again."
The precedent of time is against you avon b7. The iMac had similar criticisms levelled against it and twenty years down the line it seems Apple has done okay! The conditions for the MacBook Pro adopting Thunderbolt 3/USB-C are even more favourable now than those for the iMac and USB[-A] in 1998.
No.
Apple engineering is seemingly spread thin on the ground but the difference between a stealth fighter and bomber would not have been all that big for engineering. The real challenges were on the machine that they already released. Thinness, hinges etc. That was all done. Remember, that according to the rumours (and I haven't checked it, I'm just speaking from memory so I may be mistaken) the stealth bomber was even knocked up into a prototype. Apple might have trouble bringing a Mini, iMac and MBP to market at the same time, but not two variants of an MBP and, as I said before, this is more a management problem than anything else.
Solder points and stress points.
Why do think I asked you how many connection failures between RAM and SSD sockets on the main board you had seen?
I imagine you have seen far fewer connection points between socketed elements and the main board fail than solder points on other areas of the board. There are different reasons but one of them is heat and that is just one of the reasons why graphics cards tend to suffer in higher numbers than other elements on the board. What I want to say is that having less so called 'failure points' makes great sense on paper but it is no less true that the entire board is a failure point if your thermal design doesn't keep the components from excessive stress. There is more to this but I won't go into detail here. Time will tell with these machines but I will never put my hand in the fire for Apple on thermal management.
In the same way they push for thinness, they also push for quietness and sometimes thermal management and quietness don't mix well.
You are correct that the USB-C switch is better suited now than the USB switch then, but that is not the question. That is irrelevant.
People are not complaining about the switch but how it was done. Both switches received the same complaints. You would have thought Apple would have learnt. It seems they haven't and now we have the backlash.
The difference between now and then was that Apple had some serious issues as a company. Now it has more than 200 billion dollars sitting in the bank. That changes things a lot although what will never change is the fact that this wholesale switch was just unnecessary.
I know this thread is off topic but the topic itself doesn't have much substance to it on a posting level. Or the stores have the diagnostics and repair infrastructure or they don't. Not much to actually discuss beyond if the situation is real or not and a couple of users have suggested it may be so.
That said, if it has to be sanitized because of the rules, so be it. I understand that.
Nobody else remembers an entire series of PowerBook with failing lower RAM slots? If memory serves, they even instituted an Extended Repair Program for that at the time.
Also, googling "MacBook RAM slot failure" confirms my personal recollection from years in sales and support that RAM sockets would fail ALL THE TIME. Not systematically, but certainly a common point of failure.
No more.
I don't remember the PowerBook issue. If they extended the warranty to cover the it, it was surely a design or manufacturing problem and not the fact the RAM was socketed. The same design and manufacturing issues could affect non socketed RAM.
As for failing RAM slots. With regard to the connection to the board I have personally seen or heard of none. I Googled what you said but the pages I looked at were largely inconclusive but most involved the lower slot, yes. Possible causes were the RAM itself, firmware or OS upgrades and of course the connections (both to the board and the pins and the socket. That said, the numbers, or the sensation you get while Googling shows that the possible biggest cause of failure involves graphics cards and specifically solder failure (mainly due to thermal issues but also physical stress). My guess is that we have seen fewer issues of socket to main board failure because they are designed for more physical stress and people tend to upgrade RAM once or twice only, maxing out the capacity of the machine in the process.
That's why my preference is having removable RAM in spite of any hits in speed and thinness and theoretical failure statistics. More so when the upshot is having to BTO all you need (and at Apple's pricing) at time of purchase.
My guess is that many people would opt for that option if given the opportunity. I would even go so far as saying that if Apple offered good quality non retina screens (and prices were adjusted as a result), they would also be very popular.
The problem is that the premium machines they currently offer would take a hit and currently Apple doesn't want to offer much in that market (non premium).
We will have to wait for the new iMacs and Minis to evaluate the line but I can't see them being competitive with stock machines or BTO offerings. They re-jiggled both lines to make them less competitive and get users to move up the line. Not sure if the Mini will survive and the Air looks terminal.
Thanks. A bit of a collection of reasons but the main one seems to be a logic board manufacturing issue. That kind of issue doesn't distinguish between socketed or non socketed design.
Other causes were the ones I mentioned and of course the reseating which is by itself basically a non-issue. We've all plugged something in at some.point and realised later that it wasn't completely in.
More problematic perhaps is getting Apple to admit machines have a problem in the first place.
Thanks. A bit of a collection of reasons but the main one seems to be a logic board manufacturing issue. That kind of issue doesn't distinguish between socketed or non socketed design.
Other causes were the ones I mentioned and of course the reseating which is by itself basically a non-issue. We've all plugged something in at some.point and realised later that it wasn't completely in.
Knowing what the cause is does not make something a "non-issue". Erratic behaviour in critical situations is not made less erratic by the fact that you might have a good chance of fixing it.
Eliminating the point of failure is still the better option.
Thanks. A bit of a collection of reasons but the main one seems to be a logic board manufacturing issue. That kind of issue doesn't distinguish between socketed or non socketed design.
Other causes were the ones I mentioned and of course the reseating which is by itself basically a non-issue. We've all plugged something in at some.point and realised later that it wasn't completely in.
Knowing what the cause is does not make something a "non-issue". Erratic behaviour in critical situations is not made less erratic by the fact that you might have a good chance of fixing it.
Eliminating the point of failure is still the better option.
What is critical about about reseating some RAM? RAM doesn't unseat itself every day. For the very few people that had that specific problem, it was probably the first and last time they had it. Often because it wasn't fully clamped in the first place or some jolt released it. However, having a solder point break on a soldered module would be a bit critical. You'd have to change a large chunk of the machine.
With blanket cures like the different low level resets etc, socketed or non-socketed isn't the issue. The problem could occur on either system.
And eliminating the point of failure can never happen. Whichever route you go down (pocketed or soldered) you can run into problems. Of course you have the plus of substituting or increasing RAM at your leisure and when the price is right for you if it is socketed and accesible. There are tradeoffs in both cases but I'd go for the more flexible, cheaper option.
What's critical about it is IT FAILING.
If my machine has reliability issues during a live production, I DON'T CARE whether the issue is easily fixable after it's failed.
I find it amusing that you're seriously claiming that a soldered connection isn't inherently more secure than a plugged one. As an owner of several Hammond organs and countless vintage keyboards, I find that notion patently absurd.
And while I do try not to put words in your mouth, in this case, I don't see any other way to read what you're writing.
What's critical about it is IT FAILING.
If my machine has reliability issues during a live production, I DON'T CARE whether the issue is easily fixable after it's failed.
I find it amusing that you're seriously claiming that a soldered connection isn't inherently more secure than a plugged one. As an owner of several Hammond organs and countless vintage keyboards, I find that notion patently absurd.
And while I do try not to put words in your mouth, in this case, I don't see any other way to read what you're writing.
As I said. On paper, it's fine logic. The reality is the living conditions. How hot does your Hammond get? Is it subject to many knocks?
The number one enemy of solder points is the heating and cooling process. The points don't distinguish between where they are although socketed elements usually have some extra reinforcement due the physical stress they have to tolerate.
You shouldn't worry about RAM becoming unseated during a show and if you did, you would probably have to setups running at the same time anyway. Not just to cover one kind of failure but any kind of failure.
My MacBook pro touch 15" died yesterday. No power, won't charge. I brought it into the Apple Store on 5th Avenue and they had no diagnostic equipment and had to send it back to Apple for repairs. They could not replace it because it wasn't the standard stock model. Not sure how they are equipped to do repairs if the NY flagship store is unable to do basic diagnostics.
I'm sorry you find yourself in that situation. Is repair the only option or can you request a new machine straight from the factory?
Not even having diagnostics tools really is unacceptable more than two months after launch.
Turns our they can't repair it, because Apple doesn't have the part? They are replacing it, but I will have to wait 3 weeks for a new computer.
My MacBook pro touch 15" died yesterday. No power, won't charge. I brought it into the Apple Store on 5th Avenue and they had no diagnostic equipment and had to send it back to Apple for repairs. They could not replace it because it wasn't the standard stock model. Not sure how they are equipped to do repairs if the NY flagship store is unable to do basic diagnostics.
I'm sorry you find yourself in that situation. Is repair the only option or can you request a new machine straight from the factory?
Not even having diagnostics tools really is unacceptable more than two months after launch.
Turns our they can't repair it, because Apple doesn't have the part? They are replacing it, but I will have to wait 3 weeks for a new computer.
Sorry to hear that. Did they give you a diagnosis of the problem?
I think this type of situation will be common for bto MBPs that fail. If little can be repaired, the only solution will be to wait for a factory bto replacement. Out of warranty, repair may not be a viable due to the cost.
It would be nice if they could tell you exactly why it failed. I hope the new one goes well.
Comments
Your needs can be met with an Ultrabook (aka MBA and MB) designed for professional users without computationally demanding workflows.
Your argument is like claiming that anyone who takes pictures with their iPhone as part of their job is a pro photographer and their use case is as valid in the discussion of pro level cameras like the Nikon D5. No, it isn't.
If you can afford a $4000 MBP then you have the ability to afford a 3 year upgrade cycle for a laptop that is more suited to your needs and income. That Apple doesn't build the MBP to suit your needs and caters more to the needs of professionals with demanding workflows which require more up to date equipment instead is NOT a problem.
A three year pro cycle is average. Two year or four year is also common. However the assertion that the MBP must be designed with user replaceable components to support a 5-8 year lifecycle is ludicrous when the average professional refresh cycle is 3 years. That's not the same refresh lifecycle for an office worker but one that fits creatives, engineers and developers that have pro level computational requirements.
Yes, the reason is clear why some folks don't update for 5+ years. They don't have pro level computational needs where time=money. They may do professional work on their computers but the computer is not the most essential tool to completing their job.
Apple doesn't do that. I suggest you move on to HP or Dell.
My last machine lasted five and a half years -- though it spent the last year or so often absolutely at (and often enough OVER) its limit. And making me a living. I was actually prepared to replace it after three years, but then Apple released Mavericks and breathed an extra two years of life into it (memory compression, wow!).
At the point where it started dying, rumours of impending updates had congealed to the point where I decided to nurse it along for another six or seven months until new machines were released.
Going by that experience, I went in for the new 15" machine, hoping that it, too, will last me around five years. Yes, it will be written off after three, and I'll probably be able to afford a replacement at that time, but it would be nice not to HAVE to replace it.
I absolutely have "pro" computational needs (this is a production and stage performance machine, among other things), but having funds available to other business needs is a good thing. So if this thing lasts me five years, I'm not gonna replace it after three. I have enough other gear investments I can make if I need to for tax reasons.
In your world Apple's management should have had engineers create two distinct lines of otherwise similar Macs yet those engineers are too thin on the ground to make successful iOS and Mac products at the same time.
You're really not thinking things through:
Aren't the electrical connections for SSD and RAM sockets SOLDERED to the main board? If soldering was a greater disaster than socketing then the RAM/SSD connections will be failing at least as frequently as any other soldered component, if not more as you've added the extra failure point between the socket and the RAM/SSD component.
If components were more likely to fail being soldered than socketed then most components on a motherboard would be socketed.
The fact that the vast majority of components are soldered onto motherboards THROUGHOUT the electronic industry implies soldering is superior to socketing.
The precedent of time is against you avon b7. The iMac had similar criticisms levelled against it and twenty years down the line it seems Apple has done okay!
The conditions for the MacBook Pro adopting Thunderbolt 3/USB-C are even more favourable now than those for the iMac and USB[-A] in 1998.
Also, googling "MacBook RAM slot failure" confirms my personal recollection from years in sales and support that RAM sockets would fail ALL THE TIME. Not systematically, but certainly a common point of failure.
No more.
Apple engineering is seemingly spread thin on the ground but the difference between a stealth fighter and bomber would not have been all that big for engineering. The real challenges were on the machine that they already released. Thinness, hinges etc. That was all done. Remember, that according to the rumours (and I haven't checked it, I'm just speaking from memory so I may be mistaken) the stealth bomber was even knocked up into a prototype. Apple might have trouble bringing a Mini, iMac and MBP to market at the same time, but not two variants of an MBP and, as I said before, this is more a management problem than anything else.
Solder points and stress points.
Why do think I asked you how many connection failures between RAM and SSD sockets on the main board you had seen?
I imagine you have seen far fewer connection points between socketed elements and the main board fail than solder points on other areas of the board. There are different reasons but one of them is heat and that is just one of the reasons why graphics cards tend to suffer in higher numbers than other elements on the board. What I want to say is that having less so called 'failure points' makes great sense on paper but it is no less true that the entire board is a failure point if your thermal design doesn't keep the components from excessive stress. There is more to this but I won't go into detail here. Time will tell with these machines but I will never put my hand in the fire for Apple on thermal management.
In the same way they push for thinness, they also push for quietness and sometimes thermal management and quietness don't mix well.
You are correct that the USB-C switch is better suited now than the USB switch then, but that is not the question. That is irrelevant.
People are not complaining about the switch but how it was done. Both switches received the same complaints. You would have thought Apple would have learnt. It seems they haven't and now we have the backlash.
The difference between now and then was that Apple had some serious issues as a company. Now it has more than 200 billion dollars sitting in the bank. That changes things a lot although what will never change is the fact that this wholesale switch was just unnecessary.
I know this thread is off topic but the topic itself doesn't have much substance to it on a posting level. Or the stores have the diagnostics and repair infrastructure or they don't. Not much to actually discuss beyond if the situation is real or not and a couple of users have suggested it may be so.
That said, if it has to be sanitized because of the rules, so be it. I understand that.
As for failing RAM slots. With regard to the connection to the board I have personally seen or heard of none. I Googled what you said but the pages I looked at were largely inconclusive but most involved the lower slot, yes. Possible causes were the RAM itself, firmware or OS upgrades and of course the connections (both to the board and the pins and the socket. That said, the numbers, or the sensation you get while Googling shows that the possible biggest cause of failure involves graphics cards and specifically solder failure (mainly due to thermal issues but also physical stress). My guess is that we have seen fewer issues of socket to main board failure because they are designed for more physical stress and people tend to upgrade RAM once or twice only, maxing out the capacity of the machine in the process.
That's why my preference is having removable RAM in spite of any hits in speed and thinness and theoretical failure statistics. More so when the upshot is having to BTO all you need (and at Apple's pricing) at time of purchase.
My guess is that many people would opt for that option if given the opportunity. I would even go so far as saying that if Apple offered good quality non retina screens (and prices were adjusted as a result), they would also be very popular.
The problem is that the premium machines they currently offer would take a hit and currently Apple doesn't want to offer much in that market (non premium).
We will have to wait for the new iMacs and Minis to evaluate the line but I can't see them being competitive with stock machines or BTO offerings. They re-jiggled both lines to make them less competitive and get users to move up the line. Not sure if the Mini will survive and the Air looks terminal.
Other causes were the ones I mentioned and of course the reseating which is by itself basically a non-issue. We've all plugged something in at some.point and realised later that it wasn't completely in.
More problematic perhaps is getting Apple to admit machines have a problem in the first place.
Eliminating the point of failure is still the better option.
With blanket cures like the different low level resets etc, socketed or non-socketed isn't the issue. The problem could occur on either system.
And eliminating the point of failure can never happen. Whichever route you go down (pocketed or soldered) you can run into problems. Of course you have the plus of substituting or increasing RAM at your leisure and when the price is right for you if it is socketed and accesible. There are tradeoffs in both cases but I'd go for the more flexible, cheaper option.
I find it amusing that you're seriously claiming that a soldered connection isn't inherently more secure than a plugged one. As an owner of several Hammond organs and countless vintage keyboards, I find that notion patently absurd.
And while I do try not to put words in your mouth, in this case, I don't see any other way to read what you're writing.
The number one enemy of solder points is the heating and cooling process. The points don't distinguish between where they are although socketed elements usually have some extra reinforcement due the physical stress they have to tolerate.
You shouldn't worry about RAM becoming unseated during a show and if you did, you would probably have to setups running at the same time anyway. Not just to cover one kind of failure but any kind of failure.
I think this type of situation will be common for bto MBPs that fail. If little can be repaired, the only solution will be to wait for a factory bto replacement. Out of warranty, repair may not be a viable due to the cost.
It would be nice if they could tell you exactly why it failed. I hope the new one goes well.