View Single Post
Old 07-06-2023, 03:09 PM   #66
DNSB
Bibliophagist
DNSB ought to be getting tired of karma fortunes by now.DNSB ought to be getting tired of karma fortunes by now.DNSB ought to be getting tired of karma fortunes by now.DNSB ought to be getting tired of karma fortunes by now.DNSB ought to be getting tired of karma fortunes by now.DNSB ought to be getting tired of karma fortunes by now.DNSB ought to be getting tired of karma fortunes by now.DNSB ought to be getting tired of karma fortunes by now.DNSB ought to be getting tired of karma fortunes by now.DNSB ought to be getting tired of karma fortunes by now.DNSB ought to be getting tired of karma fortunes by now.
 
DNSB's Avatar
 
Posts: 46,666
Karma: 169712392
Join Date: Jul 2010
Location: Vancouver
Device: Kobo Sage, Libra Colour, Lenovo M8 FHD, Paperwhite 4, Tolino epos
Quote:
Originally Posted by ottischwenk View Post
These are values of the latest technologies; but Kobo uses SD cards - more than 10 years old technology, you can't get them in normal shops anymore.
Also: the standard SD delivery is that for the 2 per thousand GB there is a non-partitioned space for which the controller can swap defective areas - the rest is exFAT in huge blocks.
But that is not the case with the Kobo partitioning - the entire space is partitioned and sometimes in ext4 a very write-intensive file system because it is logging.
As soon as the first cell becomes defective, the card is defective.

And, of course, Rakuten wants to keep selling devices.
And you have proof of your claim about 300 write cycles? A reference to where you read that number?

Perhaps you should take the time to read on how the spare blocks are treated in solid state drives. Gentle hint: they are not subtracted from the capacity the drive shows nor do they show as non-partitioned space. Formatting the entire space exposed by the on board controller space does not affect the spare blocks. The only time a spare block becomes visible is when it is swapped for a failed block by the on-board controller.

As for ext4 being a write-intensive file system? Another of your "interesting" claims. Yes, ext4 does journaling but then so do most modern file systems. And yes, there are log files being written. But most of the writes are relatively small blocks so contribute little to the disk wear.

Looking at a Kobo Clara 2e µSD card, I see 3 partitions. The system partition and the recovery partition are ext4 and show as 300 and 384MiB respectively. The FAT32 partition clocks in at 14,469MiB (~14GiB). Now that partition is where most of the wear would be seen since the databases stored in that partition are written to very often. There is also a 16MB unpartitioned space which is most noted for being where the serial number is stored though it also has the boot information and other bits and bobs. There is no exFAT partition used on a Kobo ereader though I have seen people reformat the exposed partition as ext4 on a couple of occasions.

I am rather surprised by some of the total crap you have been posting. Have you ever taken the time to actually read up on the technology you are so enthusiastically dumping on? Perhaps a quick search on bad block management would be helpful. I did note multiple manufacturers claim that 5-7% of the blocks are reserved for bad block replacement not 2 blocks per 1000GB. 2 blocks of either 256KiB or 512KiB (commonest block sizes on a newer SD card) would be 512K or 1MB in 1000GB in contrast to 50-70GiB which is would be common. Take the time to check on erase blocks in flash storage for more information.
DNSB is online now   Reply With Quote