F2FS & XFS File-Systems Updated For Linux 4.5

[phoronix]

Phoronix: F2FS & XFS File-Systems Updated For Linux 4.5

The F2FS file-system pull request is quite exciting while the XFS churn for the Linux 4.5 merge window isn't as meaty...

http://www.phoronix.com/scan.php?page=news_item&px=Linux-4.5-XFS-F2FS

[Azrael5]

Extraordinary news. Very happy for it

[shmerl]

When will F2FS become stable and get the Windows driver? It feels like their goal to replace FAT32 / exFAT on common SD cards and flash drives is unacheivable pipe dream. Don't get me wrong - I very much wish this abomination of exFAT to be replaced with free filesystem like F2FS.

[SkOrPn]

When will F2FS become stable and get the Windows driver? It feels like their goal to replace FAT32 / exFAT on common SD cards and flash drives is unacheivable pipe dream. Don't get me wrong - I very much wish this abomination of exFAT to be replaced with free filesystem like F2FS.

ETA's are not a nice thing to ask for from developers. It will be ready when its ready, that's all we consumers need to know. Just be thankful that work continues on it which is a very good sign that it wont be abandoned. And their goal is to develop a more appropriate file system for memory like storage devices, not simply to replace exFAT or FAT32. Replacing an extremely mature file system such as FAT32, one that has been adopted the entire world over for a very long time is difficult to accomplish. And with the advent of H.265 replacing FAT32 just became less important for the majority or normal users. F2FS must be absolutely perfect before it is adopted as a replacement to anything. EXT4 is the main target (I would think) to replace and I do not see any input from Microsoft on this project so I doubt they will adopt F2FS to replace NTFS (only in our dreams). NAND is changing rapidly, in fact Intel is a major driving force in the storage arena and they have plans to replace the much aging NAND all together with a new tech called 3D XPoint. So F2FS might be too late to the NAND party anyway, or need some major changes before it is even considered on the future storage tech. NAND is very old tech and thus could be at risk of retiring soon, so depending on how these new technologies take off F2FS may need some major code changes just to comply with future storage tech. Hell who knows, all of today's file system types may not even work on 3D XPoint and we may need an entirely new file system developed just for it? If 3D XPoint does turn out to be a major storage tech replacing NAND, then all current storage chip manufactures, including Samsung, will need to make drastic changes to stay competitive. And drastic changes means abandoning NAND and developing something more in line to compete with 3D XPoint. F2FS is being developed solely for NAND based tech as far as I know, so has anyone ever said that it will work with upcoming tech such as 3D XPoint based chips? I wonder...

Not only this, but also think about the upcoming changes with the communications protocols such as AHCI, which will be and is being replaced by NVMe. All these changes from less mature technologies require much more R&D, plus testing. F2FS is being developed in-house for NAND and AHCI. Change all of that and you add more years to R&D. With all the upcoming changes to storage tech I don't think we will see F2FS ready for major adoption for many years to come. Things are just moving too fast for memory type storage to declare F2FS ready for adoption. Or does it not work that way?