XPost: comp.os.linux.advocacy, comp.sys.mac.advocacy   
   From: recscuba_google@huntzinger.com   
      
   On 1/20/26 22:14, Gremlin wrote:   
   > -hh  news:10kp9kv$1nbgc$1@dont-email.me Wed,   
   > 21 Jan 2026 01:20:31 GMT in comp.os.linux.advocacy, wrote:   
   >   
   >> Blown capacitors are another common failure point on older PCs...how   
   >> many of these are socketed?   
   >   
   > That's not just PCs...Electrolytics do have a life span.   
      
   Everything has a finite life.  The question of enabling repairability   
   will logically focus on those components which have lower lifespans.   
      
      
   > They're used in   
   > practically everything. I don't know why you'd want to spend the effort to   
   > socket a two pin device? It's very easy to desolder them and replace them   
   > with a fresh component and solder whether it's SMD or THT. It's not even a 5   
   > minute job once you have the necessary access to the component.   
      
   Soldering is either "fine to do", or an anathema:  you're trying to make   
   it situational, based on if you want to criticize a PC manufacturer.   
      
      
      
   > Don't you think you might be trying to compare Apples and Oranges here   
   > though? a Capacitor is hardly an SSD drive or RAM module.   
      
   Its still a common enough age-base failure mode which bricks an older PC   
   which usually results in it being sent to the trash.   
      
   And this was what I was alluding to with my "Not my dog in this hunt":   
      
   1a)  Even if true, is it unique to only one computer brand so as to   
   merit a criticism of that brand, or is it as a universal issue for all   
   computers which use SSDs?   
      
   2a) Same question for this failure mode as the above.   
      
      
   >>>> Its been 40+ years since I've owned a computer whose RAM was installed   
   >>>> in sockets - - are you suggesting that this is what we all should be   
   >>>> "just in case" going back to?   
   >>>   
   >>> What sort of computers have you been using for the past 40+ years that had   
   >>> their RAM soldered in place? And what do you mean by going back to? Socket   
   >>> based RAM is still very much a thing.   
   >>   
   >> No, the stuff you're referring to is a hybrid, because even though there   
   >> was a SIMM/DIMM socket, the RAM chips on its board was soldered.   
   >   
   > Most desktops do not have ram soldered onto their mainboards.   
      
   The last time that I can personally recall a "RAM upgrades" line item in   
   our departmental budget was to extend the lives of some 386 desktops.   
      
      
   > ...Some laptops/mini/all in ones do...and not all of   
   > those have any options to add more. I wasn't disputing that.   
      
   I can't recall the last time that I had an x86 laptop that had socket   
   based RAM that was worth bothering with:  they've become appliances   
   where if anything went wrong, IT would deploy a new set of hardware to   
   replace the whole thing and reimage one's data (if recoverable) over   
      
   Plus most personal computing today isn't via laptops, but smartphones:   
   how many smartphones have user-replaceable RAM cards?   
      
   Personally, even the RIM Blackberry I had in 2005 had soldered-on   
   electronics...only thing user-replaceable was its battery.   
      
      
   > None of your   
   > comments take away from the fact that if a NAND chip which makes up the   
   > soldered on SSD shorts to ground; the Apple is going to be a paperweight. It   
   > won't boot from external media; it won't even turn on.   
      
   So?  What happens with the same hardware failure on an x86 PC running   
   Windows or Linux?   
      
      
      
   >> I was referring to earlier PCs, where there was no RAM soldered at all,   
   >> because each RAM chip was installed into its own dedicated socket.  This   
   >> was in the early Apple ][ series.   
   >   
   > Ahh. You were including ancient by todays standards gear.   
      
   Yup.  And I can recall times where the troubleshooting for a problem was   
   to open the case and squish/wiggle every single RAM chip in its socket.   
   It just wasn't the panacea that you're trying to spin it as.   
      
      
   > My Coco3s ram was   
   > also soldered to the main board. However, you could replace it for larger   
   > chips as well as solder on additional ones. I opted for the 512k expansion   
   > board. The chips on the board were socketed.   
   >   
   > I don't typically include my experience with it or the Apple computers I   
   > used at school in the early 80s. Green screen 5.25 floppies; some had dual   
   > drives. You still use the old Apple? Is it for nostalgic reasons or   
   > something?   
      
   Nah, I got rid of it in the 1980s, for better/faster technology.   
      
   That better/faster has been replaced too.  And again.  Ditto for my   
   business PCs too, because they are merely tools.   
      
      
   > You really do seem to be pulling interesting things to try and argue about   
   > here...A capacitor and some very old computers that few people use anymore?   
      
   No, because I'm noting that from a hardware reliability perspective that   
   various components have varying reliability curves, with capacitors   
   being an example of one which statistically fail earlier than things   
   like RAM chips.  So if your argument for RAM socketing is for sake of   
   hardware reliability, then because capacitors have an objectively higher   
   failure rate, the hardware designers should prioritize the socketing of   
   capacitors before RAM (especially if it incur warranty repair expenses).   
      
      
   > I was talking about much more modern gear and I don't think you weren't very   
   > aware of that.   
      
   I'm aware that you're sliding around to whatever suits your argument:   
   its why I asked the "no dog in this hunt" question, to see if you're   
   honestly focusing on hardware points, or if you're dishonestly just   
   looking for an excuse to selectively bash just one supplier.  So how   
   about answering the dog hunt question?   
      
   > I wouldn't have used them as PC examples in the context I've been   
   > discussing a PC or Mac, either. I don't think the typical user of   
   > either would consider those families along the same lines or having   
   > much of anything in common with what's sitting in front of them today.   
   >   
   > Especially considering that when you use the term PC computer - it's   
   > generally assumed that you are referring to an IBM compatible.   
      
   There was a period where that was a product differentiation, sure.   
      
   > The term has meant that to the mainstream since the late 1980s.   
      
   And things change ... and continue to change.   
      
   > And neither the coco   
   > series or the Apples of that era were considered a PC in that context. This   
   > seems like it might be a bit of a red herring on your part?   
      
   No, because the differentiation you're referring to was from ~20 years   
   ago, in Apple's "Get a Mac" marketing with Justin Long & John Hodgman.   
   When Apple shifted from PPC to Intel in 2006, the basis for that   
   differentiation ended, shifting to supporting & promoting dual-booting.   
      
      
   -hh   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)