@wawa

Quote:

wawa wrote: @Hans Quote: A quick look at PFS' sourceforge project shows that the last commit was over a year ago. So not very active at the moment. is there any problem with it you want to have fixed or why should that be active?

I've never used PFS myself, so I really wouldn't know. I was just giving a direct answer to the original question.

That said, I would like to see an OS4 native version, and an update to support huge disk/partition sizes (as in >1TB, seeing as disks are that big now). I do like the idea of having a multi-user filesystem.

Hans

_________________
http://hdrlab.org.nz/ - Amiga OS 4 projects, programming articles and more. Home of the RadeonHD driver for Amiga OS 4.x project.
https://keasigmadelta.com/ - More of my work.

Wrong remark here. Should not have been posted.

Last edited by OldFart on 21-Nov-2015 at 10:09 AM.
Last edited by OldFart on 21-Nov-2015 at 10:07 AM.

_________________
More then three levels of indigestion and you're scroomed!

@Hans

Quote:

Hans wrote: I've never used PFS myself, so I really wouldn't know. I was just giving a direct answer to the original question. That said, I would like to see an OS4 native version, and an update to support huge disk/partition sizes (as in >1TB, seeing as disks are that big now). I do like the idea of having a multi-user filesystem. Hans

i simply wonder that a completed, freely available and apparently stable project is declared dead upon no commit activity in the repository? you are a coder yourself, so i think you would share the opinion that the sparce resources should be spent where actual problems are, instead commiting to something just for the sake of it. as a pfs3 user myself i can confirm that it is working rather reliably. i have my observances, but nothing definite, since the drives i use were involved with heavy testing, swapping, replacing, repartitioning and the like, i dont have anything i could claim to be a pfs3 issue (atm).

Last edited by wawa on 21-Nov-2015 at 12:01 PM.

@Hans

I consider PFS3 i pretty complete at the moment. If you want to add new features, well, the source code is out there.

After all, FFS and SFS have not seen much development in last 15 years. Bugs were fixed but no new features added.

_________________
Amiga Developer
Amiga 500, Efika, Mac Mini and PowerBook

If a filesystem has a good set of options, and it's quite stable, it's normal for it to have little updates.

It's one of the most important system components, which should ensure data preservation and integrity, so changes to it has to be done wisely and with a lot of testing.

@wawa

Besides supporting large files 2Gbyes and SSD, I don't see need for a lot of changes to filesystems, anyway file notification is not that great on AmigaOS. Fix file notification API, and maybe there will be interest in improving the filesystems.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

@Raffaele
Quote:

That's my two cents

there is no kind of coordination you can impose on the scene. that has been repeatedly proven. as long as the projects have their own repositories and binaries appear on aminet all si fine i guess. but if you want to be on a sure side none will stop you from mirroring a repository on your own server, as long as sorces are open, that is, which they are in this case.

@NutsAboutAmiga

Quote:

Besides supporting large files 2Gbyes and SSD

may well be pfs supports already larger files, dont remember, toni has removed a lot of limits. what concerns ssd im not sure how to support the extra features, but so far flash and ssd drives were working fine for me.

@wawa

SSD disks have optimal block size, and alignment.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

@wawa

Quote:

wawa wrote: @Raffaele Quote: That's my two cents there is no kind of coordination you can impose on the scene. that has been repeatedly proven. as long as the projects have their own repositories and binaries appear on aminet all si fine i guess. but if you want to be on a sure side none will stop you from mirroring a repository on your own server, as long as sorces are open, that is, which they are in this case.

Once there is a coordination by a committee of users, then they can advice developers to follow certain guidelines.
If committe gains reputation then developers will follow their advices.

On a longer time a committee can became an Authority.

I beg your pardon. Not sure what you intend by mirrorong existing projects on site of my own. We are speaking two different topics here.

About imposing.

There is no necessity of imposing anything.

If committee decides that we need a new set of Amiga Filesystem readers into Linux for example, all they need is to hire a bounty on Linux sites advicing Linux Developers we need a PFS media reader/writer and indicating the amount of money we offer.
If someone in Linux camp will accept the task, we will gladly pay him for creating such a program.
The software will be hosted on common existing Amiga and Linux repositories. No needs to creating new ones.

If committee decides that we need PFS improvement, by supporting 64bit, files larger of 2 GBytes (not so trivial nowadays with movie files for example), free block sizes capability (not only 512, 1024, 2048 bytes long), and full retrocompatibility with old PFS versions, and mainly asking to make this PFS version "UNIVERSAL" by make it working on ANY Amiga-Like OSes... all things committee needs, to make someone accomplish this work done, again it is to hire a bounty.

The new resulting PFS will be de facto standard without imposing anything than the features indicated into bounty specs.

1) The developer/developers will receive a substantial amount of money letting him/them continue supporting our platform.

2) Committe will improve its standards requirements by letting community adopting this new standards, that are common to all Amiga systems.

3) Community of Amiga users will benefit a new filesystem and compatible with all existing Amiga flavours, enhancing data transfer capabilities and reducing distance and differencies between the camps.

It is a Win-Win situation.

Last edited by Raffaele on 22-Nov-2015 at 06:50 AM.
Last edited by Raffaele on 22-Nov-2015 at 06:47 AM.
Last edited by Raffaele on 22-Nov-2015 at 06:45 AM.
Last edited by Raffaele on 22-Nov-2015 at 06:44 AM.
Last edited by Raffaele on 22-Nov-2015 at 06:42 AM.

_________________
"When the Amiga came out, everyone [at Apple] was scared as hell." (J.L. Gassée, former CEO of Apple France and chief of devs of Mac II-fx, interviewed by Amazing Computing, Nov 1996).

@NutsAboutAmiga

Quote:

NutsAboutAmiga wrote: @wawa SSD disks have optimal block size, and alignment.

And trim - https://en.wikipedia.org/wiki/Trim_(computing)

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@Raffaele

PFS3 already supports 64-bit packets. Extending PFS to support large files (2GB+) with backward compatibility is not trivial, it might be even impossible. Mounting large capacity HD and large files must work with older PFS3 versions without data loss. If it is not possible then PFS4 is needed. This again requires changes to PFSDoctor...

_________________
Amiga Developer
Amiga 500, Efika, Mac Mini and PowerBook

@itix

Quote:

itix wrote: @Raffaele PFS3 already supports 64-bit packets. Extending PFS to support large files (2GB+) with backward compatibility is not trivial, it might be even impossible. Mounting large capacity HD and large files must work with older PFS3 versions without data loss. If it is not possible then PFS4 is needed. This again requires changes to PFSDoctor...

I was just taking an example using PFS.

Glad to know it already supports 64 bits structure even if that's not what I was pointing in the example.

The creation of PFS 4 that you suggested it is an example that fits better the point I wanted to explain.

Sure if you create a new version of a filesystem you need robust programming, lots of betatesing to get rid of bugs and new recovery tools, but any development will be better executed with a back-up support of a committee that had already indicated minimum requirements and standards necessary to make new filesystem being universal for all community.

Is that too much complicated to starting following common guidelines that could benefit all Amiga camps?

Last edited by Raffaele on 22-Nov-2015 at 09:17 AM.

_________________
"When the Amiga came out, everyone [at Apple] was scared as hell." (J.L. Gassée, former CEO of Apple France and chief of devs of Mac II-fx, interviewed by Amazing Computing, Nov 1996).

@Hans

PFS3 is finished. It's mature and stable, all known bugs are fixed. What exactly do you expect to be commited at this point? This doesn't mean that it's abandoned or unsupported in any way, it's an integrated part of MorphOS, it's widely used by Amigans of all flavors. It's perfectly safe to use PFS3, you won't be left stranded, there are a couple of developers standing by should a new bug be found that needs fixing.

@Raffaele

There will be no "committee", there will be no merger of the development of the NG flavors.

And new features in PFS3 like those you suggested can't be added while keeping backwards compatibility, at least it would be unecesarily difficult. This is also why Hans believs its "dead". PFS3 is finished. Any new features should go in a "PFS4" that should be allowed to break compatibility with PFS3. Of course the user could have both a PFS3 and a PFS4 partition at the same time.

@Raffaele

that will not work... devs mainly do it for fun, bounties are only a small incentive. No linux dev will take a bounty for only a couple of hundred dollar, to create a financial incentive you would need different dimensions. Also "amiga-like" platforms are very different API wise, modern development tools including class libraries are missing at all. In many cases components from Linux you would need are missing and would also be needed.

@TRIPOS

Quote:

TRIPOS wrote: @Raffaele There will be no "committee", there will be no merger of the development of the NG flavors. And new features in PFS3 like those you suggested can't be added while keeping backwards compatibility, at least it would be unecesarily difficult. This is also why Hans believs its "dead". PFS3 is finished. Any new features should go in a "PFS4" that should be allowed to break compatibility with PFS3. Of course the user could have both a PFS3 and a PFS4 partition at the same time.

Dunno what we are talking here...

It is obvious PF3 partition could not being magically transformed in PFS4 and vice-versa. It is a physical fact.
But there were something hidden somewhere that allows reading and writing these filesytems and this is the part I am talking about. If it is capable to handle any future PFS4 it should being capable also to read any previous PFS version formatted media...

Also the standard recovery tool that will be adopted for PFS4 should support by default any previous PFS versions.

I do not want being hassled by dozens recovery tools hanging on my hard disks, anytime it will be released a new single revision of a filesystem.

That's my point. Do you understand it?

_________________
"When the Amiga came out, everyone [at Apple] was scared as hell." (J.L. Gassée, former CEO of Apple France and chief of devs of Mac II-fx, interviewed by Amazing Computing, Nov 1996).

@OlafS25

Quote:

OlafS25 wrote: @Raffaele that will not work... devs mainly do it for fun, bounties are only a small incentive. No linux dev will take a bounty for only a couple of hundred dollar, to create a financial incentive you would need different dimensions. Also "amiga-like" platforms are very different API wise, modern development tools including class libraries are missing at all. In many cases components from Linux you would need are missing and would also be needed.

Where do you live? Linux developers are in the order of thousands and many good Linux developers live in second world countries where a couple hundred dollars is a good salary, and could help developer making the day.

Also please don't mix difficulties of a Linux developer who should only realize a reader/writer running in his well known environment that he belongs to, with the difficulties that regard the division in 3 Amiga operating systems that make difficult to make common APIs and let a single binary impossible running on all three OSes at the same time.

Last edited by Raffaele on 22-Nov-2015 at 11:11 AM.

_________________
"When the Amiga came out, everyone [at Apple] was scared as hell." (J.L. Gassée, former CEO of Apple France and chief of devs of Mac II-fx, interviewed by Amazing Computing, Nov 1996).

@Raffaele

good developers cost money and people are not stupid and know their value...

but even if you are right... neither Hyperion nor MorphOS team is interested to work together, I assume the opposite. And everyone has the own ideas about direction where it should go. If there would be a commercial market existing where commercial developers decide which platforms they support then perhaps. But there is no commercial market, 3rd party developers support just one platform and are mostly not interested in the others.

@Raffaele

Quote:

Raffaele wrote: @OlafS25 Quote: OlafS25 wrote: @Raffaele that will not work... devs mainly do it for fun, bounties are only a small incentive. No linux dev will take a bounty for only a couple of hundred dollar, to create a financial incentive you would need different dimensions. Also "amiga-like" platforms are very different API wise, modern development tools including class libraries are missing at all. In many cases components from Linux you would need are missing and would also be needed. Where do you live? Linux developers are in the order of thousands and many good Linux developers live in second world countries where a couple hundred dollars is a good salary, and could help developer making the day. Also please don't mix difficulties of a Linux developer who should only realize a reader/writer running in his well known environment that he belongs to, with the difficulties that regard the division in 3 Amiga operating systems that make difficult to make common APIs and let a single binary impossible running on all three OSes at the same time.

I don't know in which world are you living.

Linux is developed 80% by (BIG) COMPANIES! Non-commercial contributions are the VAST MINORITY.

That's why you cannot compare its development with any other hobby-o.s..

Last edited by cdimauro on 22-Nov-2015 at 12:11 PM.