On Sunday, January 9, 2022 at 4:00:57 PM UTC-5, Alan wrote:
> On 2022-01-09 9:19 a.m., Thomas E. wrote:
> > On Thursday, December 30, 2021 at 1:18:51 PM UTC-5, Alan wrote:
> >> On 2021-12-30 8:36 a.m., Thomas E. wrote:
> >>> On Wednesday, December 29, 2021 at 6:21:37 PM UTC-5, Alan wrote:
> >>>> On 2021-12-29 3:13 p.m., Thomas E. wrote:
> >>>>> On Tuesday, December 28, 2021 at 8:58:13 PM UTC-5, Alan wrote:
> >>>>>> On 2021-12-28 5:51 p.m., Thomas E. wrote:
> >>>>>>> On Saturday, December 25, 2021 at 9:50:47 PM UTC-5, Alan wrote:
> >>>>>>>> On 2021-12-25 7:05 p.m., Thomas E. wrote:
> >>>>>>>>> On Saturday, December 11, 2021 at 12:37:59 PM UTC-5, Alan wrote:
> >>>>>>>>>> On 2021-12-11 9:04 a.m., Thomas E. wrote:
> >>>>>>>>>>> On Saturday, December 11, 2021 at 8:21:38 AM UTC-5, -hh
> >>>>>>>>>>> wrote:
> >>>>>>>>>>>> On Friday, December 10, 2021 at 11:33:07 PM UTC-5, Thomas
> >>>>>>>>>>>> E. wrote:
> >>>>>>>>>>>>> On Thursday, December 9, 2021 at 7:16:29 AM UTC-5, -hh
> >>>>>>>>>>>>> wrote:
> >>>>>>>>>>>>>> On Wednesday, December 8, 2021 at 11:33:53 PM UTC-5,
> >>>>>>>>>>>>>> Alan wrote:
> >>>>>>>>>>>>>>> On 2021-12-08 8:24 p.m., Thomas E. wrote:
> >>>>>>>>>>>>>>>> On Wednesday, December 8, 2021 at 3:44:15 PM UTC-5,
> >>>>>>>>>>>>>>>> -hh wrote:
> >>>>>>>>>>>>>>>>> ... While we also know that the degree of SSD
> >>>>>>>>>>>>>>>>> overprovisioning is another factor which will
> >>>>>>>>>>>>>>>>> affect the product's useful lifetime, it is
> >>>>>>>>>>>>>>>>> nevertheless a good point that the
> >>>>>>>>>>>>>>>>> "SSD-as-RAM-surrogate" aspect to the M1 designs
> >>>>>>>>>>>>>>>>> will have an impact at some point in the
> >>>>>>>>>>>>>>>>> product's lifecycle, and for some percentage of
> >>>>>>>>>>>>>>>>> its users will probably be the determinant for
> >>>>>>>>>>>>>>>>> when its useful life has been curtailed.
> >>>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>>> Pragmatically, the design is such that it does
> >>>>>>>>>>>>>>>>> indeed become a disposable item at that point,
> >>>>>>>>>>>>>>>>> and while Apple does offer trade-ins, the degree
> >>>>>>>>>>>>>>>>> to which SSD wear will become another criteria
> >>>>>>>>>>>>>>>>> (or already is) by which they adjust their
> >>>>>>>>>>>>>>>>> trade-in price offer is a good question.
> >>>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>> Pragmatically Apple should not solder in parts
> >>>>>>>>>>>>>>>> known to wear out over time. There is no reason to
> >>>>>>>>>>>>>>>> do so unless you want to limit a product's useful
> >>>>>>>>>>>>>>>> life.
> >>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>> Every part wears out over time, Idiot.
> >>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>> So according to you, nothing should be soldered in at
> >>>>>>>>>>>>>>> all.
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>> Its also ignoring that there's benefits to soldering in
> >>>>>>>>>>>>>> the trade-off, such as:
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>> a) reduced cube b) reduced costs c) increased
> >>>>>>>>>>>>>> reliability
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>
> >>>>>>>>>>>>> Reliability? Sources please.
> >>>>>>>>>>>>
> >>>>>>>>>>>> Besides the Master’s Degree I have in Engineering? Well,
> >>>>>>>>>>>> there’s the MILSPEC/HNDBK on it at work that the EE-centric
> >>>>>>>>>>>> teams use on TRL 6+ designs.
> >>>>>>>>>>>>
> >>>>>>>>>>>> Basically, its the design principle of fewer parts, as
> >>>>>>>>>>>> provisioning a removable connector adds two parts (each
> >>>>>>>>>>>> side of the connector), a solder joint, plus the
> >>>>>>>>>>>> reliability hit from the connector pin connection too.
> >>>>>>>>>>>>
> >>>>>>>>>>>> Graphically:
> >>>>>>>>>>>>
> >>>>>>>>>>>> (Part A) —— sj —— (Part B)
> >>>>>>>>>>>>
> >>>>>>>>>>>> vs
> >>>>>>>>>>>>
> >>>>>>>>>>>> (Part A) —— sj —— (connector/M) —— cp —— (connector/F) ——
> >>>>>>>>>>>> sj —— (Part B)
> >>>>>>>>>>>>
> >>>>>>>>>>>> …where: “sj” = solder joint interface “cp” = connector pin
> >>>>>>>>>>>> interface
> >>>>>>>>>>>>
> >>>>>>>>>>>> It’s one (1) failure point vs five (5) failure points….
> >>>>>>>>>>>> since 1<5, 1 wins on reliability.
> >>>>>>>>>>>>
> >>>>>>>>>>>> -hh
> >>>>>>>>>>>
> >>>>>>>>>>> All true, but solder joints fail too. So, the tradeoff is
> >>>>>>>>>>> that you know that some parts are more likely to fail from
> >>>>>>>>>>> normal use. Do you accept the small chance of an easily
> >>>>>>>>>>> replaceable part's connection failing versus the making that
> >>>>>>>>>>> junction more reliable at the expense of making the failed
> >>>>>>>>>>> part difficult or impossible to replace? Does a soldered SSD
> >>>>>>>>>>> lead to trashing an otherwise functional piece of equipment?
> >>>>>>>>>> Can you really not read simple English, Idiot?
> >>>>>>>>>>
> >>>>>>>>>> Making a part removable means using MORE SOLDER JOINTS.
> >>>>>>>>>
> >>>>>>>>> Can't you understand that at least you have the option of
> >>>>>>>>> replacement of a key part that is sure to eventually fail?
> >>>>>>>> Can't you understand that adding more connections means creating
> >>>>>>>> more ways for the whole system to fail?
> >>>>>>>>
> >>>>>>>> Transistors are "sure to fail", too. Should we go back to vacuum
> >>>>>>>> tubes?
> >>>>>>>
> >>>>>>> Another Baker deflection. Transistor MTBF is several magnitudes above
> >>>>>>> a vacuum tube. A well-crafted connector for a SSD or RAM chip is not
> >>>>>>> likely to fail.
> >>>>>> Nope. A legitimate discourse on the fact that there is more to it than
> >>>>>> simply something is "sure to fail".
> >>>>>>>
> >>>>>>> Anyway, I really doubt you could build a modern laptop using 30
> >>>>>>> billion or so vacuum tubes.
> >>>>>>>
> >>>>>>> Check this out:
> >>>>>>>
> >>>>>>> https://frame.work/blog/introducing-the-framework-laptop
> >>>>>>> https://www.tomshardware.com/news/framework-announces-marketplace-for-expansion-cards
> >>>>>> It's been tried before.
> >>>>>>
> >>>>>> How's that worked out?
> >>>>>>
> >>>>>> :-)
> >>>>>
> >>>>> Pretty well in a limited sense. PCMIA was a good idea, then the USB standard replaced it. Today's Dells get high repairability scores due to their modular construction, and they are reliable.
> >>>> PCMIA is not an entire modular laptop.
> >>>>
> >>>> USB isn't either.
> >>> Which is why I used the word "limited"
> >> Expansion tools aren't a modular laptop in ANY sense.
> >
> > Lack of response to my request for info on other modular laptops is noted.
> 'Dell actually introduced a fully modular laptop (more like a desktop
> replacement) in 2019 called the Alienware Area 51M. It's a 17-inch
> behemoth of a device that comes with a desktop-spec processor and GPU.
> Best of all, you can replace these components in the future, provided
> Dell releases a GPU designed for the system.
>
> But after one year, the company stopped releasing new hardware for the
> Area 51M and instead focused on newer, more powerful laptops with a
> fixed processor and GPU.'
>
> <https://www.makeuseof.com/why-the-framework-laptop-is-a-big-deal/>
>
> From the same source:
>
> 'Some laptop makers are selling upgradeable laptops. However, they're
> not what you think. These fully modular devices are designed for
> fieldwork use. So you'll find that these devices are thick and heavily
> protected as well as modular.
>
> According to some reviewers, the Area 51M was an impractical laptop. It
> weighed more than 8 lbs, required two power bricks, one of which weighed
> more than an ultralight laptop, and the fans made a lot of noise.
>
> The classic reason manufacturers can't make a fully modular laptop is
> that space is at a premium when making portable devices. You can't just
> shoehorn a regular processor or GPU into thin and light devices, and if
> you try, you're going to have to compromise on either performance, size,
> or both.'
>
> Also from that source:
>
> 'One big question in everyone's minds is, "Will the company last?" There
> have been attempts to create modular consumer laptops that didn't pan out..'
>
> Let me set that last sentence out for you:
>
> 'There have been attempts to create modular consumer laptops that didn't
> pan out.'

OK, never heard of any of these obscure products.