Thank you!

I am looking forward to see further developments on PFS3!

@MichielP: thank a lot!

@Gulliver: thanks for all!!!

I see there few assemler parts

There are only very few assembler parts and these are easy to replace with pure C.

I propose that anyone committing changes must always make sure that the 68k version still builds with the changes. Also we should set some guidelines defining the goals: I'd propose having a stable PFS3 that doesn't see any filesystem format or major functional changes and then PFS3-NG that would be have all the major changes and new features.

Currently my ideas fall into these categories:
PFS3-stable

• Fix "beta warning #1" requester issue when deleting a lot of data (related to RTBF_CACHE_SIZE in ResToBeFreed function)
• Add a low priority resethandler to shut down the filesystem and commit changes to disk just before the system is allowed to reboot (this should fix the loss of data when rebooting "too fast" after the final writes to the FS).
• Fix softlink support to work similar to FFS, making softlinks work in a well defined way with ixemul for instance
• Add support for native dospackets for each amigoid OS incarnation, but only when no functional changes result for adding them
• Other bug fixes

PFS3-NG

• Support for files larger than 2GB-1 bytes

unclear

• Support for partitions larger than ~137GB (*)

(*) If no format changes are required for the >137GB support then this falls into PFS3-stable else PFS3-NG.

Please, make it romable too ! Some GVP accelerator cards use the filesystem into the rom and not on the RDB...


@MichielP

What about the changes from the 18.3 to 18.5 ??

The sources are 18.4. The changes from 18.4 to 18.5 are lost. I will make a list of all we are missing. It is not much really.

There are only very few assembler parts and these are easy to replace with pure C.

I propose that anyone committing changes must always make sure that the 68k version still builds with the changes. Also we should set some guidelines defining the goals: I'd propose having a stable PFS3 that doesn't see any filesystem format or major functional changes and then PFS3-NG that would be have all the major changes and new features.

Currently my ideas fall into these categories:
PFS3-stable

• Fix "beta warning #1" requester issue when deleting a lot of data (related to RTBF_CACHE_SIZE in ResToBeFreed function)
• Add a low priority resethandler to shut down the filesystem and commit changes to disk just before the system is allowed to reboot (this should fix the loss of data when rebooting "too fast" after the final writes to the FS).
• Fix softlink support to work similar to FFS, making softlinks work in a well defined way with ixemul for instance
• Add support for native dospackets for each amigoid OS incarnation, but only when no functional changes result for adding them
• Other bug fixes

PFS3-NG

• Support for files larger than 2GB-1 bytes

unclear

• Support for partitions larger than ~137GB (*)

(*) If no format changes are required for the >137GB support then this falls into PFS3-stable else PFS3-NG.

Sounds good.
Extending the max filesize to 4GB should not be much of a problem. The directory entry already lists the size as an ULONG which implies 4GB. To go beyond 4GB requires more work.

Let's start with collecting bug and feature requests on sourceforge.

It will become ROMable (if it isn't already) for sure as placing it inside the MorphOS boot image has the requirement as well.

So what do you think of the licensing of PFS3?

It's the 4-clause BSD license, the one with the advertising clause... basicly means all MorphOS "ads" must also have a note about PFS3 and Michiel Pelt in it

Dunno about anyone else but i'd love to see a http://www.amigabounty.net/index.php
page setup to help serious development and time spur this project on and also, to give a little back to the author MichielP and people who help work on it! just like Piru

Doesn't have to be AMIGA bounty, but i'd love to help out a little with the development time/costs to keep this going, it's a great product and well worth exploring and I agree with whoever said it needs to get ported to ALL AMIGA flavours.

PayPal donation etc. etc. We are a small community now and to have applications of this caliber on such an old platform really deserves some attention, just like firefox port.

Just a suggestion. I still use this product and would have a lot less data on my AMIGA if it wasn't for this product. Really very grateful.