[Web] New CAM article released | ANN.lu |
|
New CAM article released : Comment 1 of 31 | ANN.lu |
Posted by Peter Gordon on 03-Jun-2003 07:56 GMT | Nice article! Shame about the mistakes, though. I liked this one:
"232 bytes (about 4.2 billion bytes), or exactly 4 Bytes"
Which should, of course, read:
"2^32 bytes (about 4.2 billion bytes), or exactly 4 Gigabytes" |
|
New CAM article released : Comment 2 of 31 | ANN.lu |
Posted by Jupp3 on 03-Jun-2003 08:06 GMT | It surely is better than old FFS, but guess I'll be still using SFS instead on my Pegasos... |
|
New CAM article released : Comment 3 of 31 | ANN.lu |
Posted by Anonymous on 03-Jun-2003 08:06 GMT | great article but...
FFS2 needs to be *really* good if they want it to be better than PFS3 and SFS. I have used PFS2 and PFS3 years now and I have never had any loss of data etc. Only problem is that PFS3 is m68k only, there is no PPC version... I think I would rather use SFS with OS4 than FFS. PFS3 has been so great and so much better than FFS, that I really would not like to go back to FFS. |
|
New CAM article released : Comment 4 of 31 | ANN.lu |
Posted by Peter Gordon on 03-Jun-2003 08:13 GMT | In reply to Comment 3 (Anonymous): I don't think FFS2's point is to "beat" SFS or PFS, rather to provide a slightly improved, PPC native FFS for OS4. |
|
New CAM article released : Comment 5 of 31 | ANN.lu |
Posted by Anonymous on 03-Jun-2003 08:41 GMT | In reply to Comment 4 (Peter Gordon): I undertsand that, and it's okay for me :) I have nothing against FFS2 and it's ofcourse good that FFS is improved too. But I would like to see OS4 native SFS too, author of SFS or someone else might do that... OS4 native PFS3 might be unrealistic unfortubately :( it's development has stopped anyway :( |
|
New CAM article released : Comment 6 of 31 | ANN.lu |
Posted by Ben Hermans/Hyperion on 03-Jun-2003 08:42 GMT | In reply to Comment 5 (Anonymous): There will be an OS 4 native version of SFS.
This is already mentioned in the features list we released months ago. |
|
New CAM article released : Comment 7 of 31 | ANN.lu |
Posted by Anonymous on 03-Jun-2003 08:49 GMT | In reply to Comment 6 (Ben Hermans/Hyperion): Great :) I did forget that :) |
|
New CAM article released : Comment 8 of 31 | ANN.lu |
Posted by mahen on 03-Jun-2003 09:16 GMT | I heard FFS2 is provided with MorphOS since last year. Can anyone confirme that ? (in addition to SFS) |
|
New CAM article released : Comment 9 of 31 | ANN.lu |
Posted by Christoph Gutjahr on 03-Jun-2003 09:50 GMT | Thanks Olaf, that was quite an interesting read.
Talking about TriPos, it seems you can download a version of TriPos here:
http://www.cl.cam.ac.uk/users/mr/Tripos.html |
|
New CAM article released : Comment 10 of 31 | ANN.lu |
Posted by SlimJim on 03-Jun-2003 11:10 GMT | Impressive and very interesting article. Well written too. Although I'm not the
man to have any comments on the technical issues, it gives a good insight into the
problems inherent in the design. I've always used FFS all the way up to OS3.9 and
I am yet to loose any data with it (I have to revalidate the drive now and then,
though, but that is another matter ;-) )
.
SlimJim |
|
New CAM article released : Comment 11 of 31 | ANN.lu |
Posted by Olaf on 03-Jun-2003 12:07 GMT | In reply to Comment 1 (Peter Gordon): > Nice article! Shame about the mistakes, though. I liked this one:
>
> "232 bytes (about 4.2 billion bytes), or exactly 4 Bytes"
>
> Which should, of course, read:
>
> "2^32 bytes (about 4.2 billion bytes), or exactly 4 Gigabytes"
It used to be 2^32 bytes when I submitted the original article.
Apparently, the document was changed and then changed again when
it was re-exported from the CAM layout. Practically all text
markup is gone (such as the tags which would have raised the
number '32'), as are line formatting codes and the occasional
word. Maybe I could get the original article published on
amigadev.net. I'll ask. |
|
New CAM article released : Comment 12 of 31 | ANN.lu |
Posted by Olaf on 03-Jun-2003 12:15 GMT | In reply to Comment 3 (Anonymous): > great article but...
>
> FFS2 needs to be *really* good if they want it to be better than PFS3
> and SFS.
As the article states, that wasn't the point for the whole odd exercise.
Within its narrow limits, FFS is a 'robust' (*shudder*) solution for
which a working set of tools for optimization, repair and data recovery
exists. It's a rather well-known design and functionally does everything
an Amiga file system should do. It's tough to state it like that, but
nothing, I repeat, nothing, does exactly what the FFS does. That it's
a performance nightmare, lacks scalability and is rather easy to break,
all this speaks against it, though.
> I have used PFS2 and PFS3 years now and I have never had any loss of data etc.
Congratulations. Consider yourself a very lucky person. |
|
New CAM article released : Comment 13 of 31 | ANN.lu |
Posted by Don Cox on 03-Jun-2003 12:57 GMT | In reply to Comment 12 (Olaf): For me, never losing data is the absolute top priority.
Speed comes a long way after that, especially as drives are getting faster and faster anyway. |
|
New CAM article released : Comment 14 of 31 | ANN.lu |
Posted by Ferry on 03-Jun-2003 13:15 GMT | In reply to Comment 12 (Olaf): > > I have used PFS2 and PFS3 years now and I have never had any loss of data etc.
>Congratulations. Consider yourself a very lucky person.
I have lost information with PFS3 in a very strange way: moving files from a PFS3 partition to another PFS3 partition over the same HD. The system crashed but the task in charge of reading and deleting files was still active, while the writing task was not. Since I was waiting for HD activity to stop before rebooting the machine, many, many files were "moved" (actually they were being only deleted...) until I suspected that something was going wrong and rebooted.
I could only recover some files, those on the .deldir. I lost many important files...
Saluditos,
Ferrán |
|
New CAM article released : Comment 15 of 31 | ANN.lu |
Posted by MarkTime on 03-Jun-2003 13:58 GMT | In reply to Comment 6 (Ben Hermans/Hyperion): >>There will be an OS 4 native version of SFS.
>>This is already mentioned in the features list we released months ago.
Yes, you mentioned you will do it.
Still that leaves us no closer to knowing whether it will happen or not.
perhaps you should take a poll to guage interest?? |
|
New CAM article released : Comment 16 of 31 | ANN.lu |
Posted by Anonymous on 03-Jun-2003 13:59 GMT | In reply to Comment 13 (Don Cox): "For me, never losing data is the absolute top priority."
You are using the wrong OS then, I think. You can get relatively cheap solutions that are much, much better in terms of not losing your data.
Perhaps you meant to say "For me, never losing data is a relatively high priority, but not anywhere near as important as my unthinking love of AmigaOS" |
|
New CAM article released : Comment 17 of 31 | ANN.lu |
Posted by Peter Gordon on 03-Jun-2003 14:02 GMT | In reply to Comment 15 (MarkTime): MarkTime, I really don't understand your attitude. They NEVER said that BPPC OS4 was a "done deal", so your sarcastic comments about the SFS announcement are meaningless. |
|
New CAM article released : Comment 18 of 31 | ANN.lu |
Posted by Anonymous on 03-Jun-2003 14:43 GMT | In reply to Comment 16 (Anonymous): This has nothing to do with lack of mory protection similar thins, this is about filesystems.
You don't know what you are talking about. eg. PFS3 and SFS are much more reliable than some windows FATxx filesystems. What i have noticed is that even NTFS's seems to get corrupted much easily. Just close windows-computer "wrong way" and you can then have "fun" repairing disk. :P |
|
New CAM article released : Comment 19 of 31 | ANN.lu |
Posted by Anonymous on 03-Jun-2003 14:43 GMT | In reply to Comment 18 (Anonymous): mory=memory |
|
New CAM article released : Comment 20 of 31 | ANN.lu |
Posted by Don Cox on 03-Jun-2003 14:55 GMT | In reply to Comment 16 (Anonymous): "You are using the wrong OS then, I think. You can get relatively cheap solutions that are much, much better in terms of not losing your data."
I don't lose any data with FFS.
The most likely way to lose data in FFS is to delete the wrong file(s) by mistake. Fortunately, DiskMonTools has a good Undelete for these occasions. |
|
New CAM article released : Comment 21 of 31 | ANN.lu |
Posted by Anonymous on 03-Jun-2003 15:07 GMT | I've been using SFS for a long time, and
never experienced any crashes/problems.
I would like to know if anybody here has
been having a problem with SFS in the
past, because to me, SFS is rock solid :) |
|
New CAM article released : Comment 22 of 31 | ANN.lu |
Posted by MarkTime on 03-Jun-2003 16:05 GMT | In reply to Comment 17 (Peter Gordon): @Peter
oh its not meaningless to everyone.
The true believers will believe what they want.
To them all facts are meaningless. |
|
New CAM article released : Comment 23 of 31 | ANN.lu |
Posted by CnlPepper on 03-Jun-2003 17:02 GMT | In reply to Comment 8 (mahen): mahen why the hell do you consider it necessary to "advertise" morphos in every bloody thread. Its pathetic. We all know morphos is out there, we all know it has a lot of features that have been implimented for a good while now - SO STOP RAMMING IT DOWN OUR THROATS WHERE ITS NOT RELAVENT. The only thing you are doing is pissing people off and guess what they'll remember associated with that annoyance - mahen and morphos. Skill and subtle advertising are something you demonstrate a whole lack of. Tell your best buddy Buck if he wants to advertise then he can pay for it to be put on ann.lu rather than sending the fanboy morphos lackeys to post irrelavent crap in each and every bloody thread.
Got that? Doubt it but what the hell.
Now go and take your fanboy behaviour to another thread while the rest of us actually discuss the issue that was relavent - filesystems.
CnlPepper - Fed up of this more recent spate of "subtle" advertising |
|
New CAM article released : Comment 24 of 31 | ANN.lu |
Posted by Olaf Barthel on 03-Jun-2003 17:06 GMT | In reply to Comment 8 (mahen): > I heard FFS2 is provided with MorphOS since last year. Can anyone
> confirme that ?
No denying that. I try not to break the promises I make.
So much for relevance in this thread... |
|
New CAM article released : Comment 25 of 31 | ANN.lu |
Posted by itix on 03-Jun-2003 18:03 GMT | In reply to Comment 18 (Anonymous): > PFS3 and SFS are much more reliable than some windows FATxx filesystems.> What i have noticed is that even NTFS's seems to get corrupted much easily.> Just close windows-computer "wrong way" Too bad Amiga is not better. Just let your Amiga crash at wrong time and you can start looking for disk salvation. Because Amiga (and clones, A/Box) lack comprehensive MP so any broken application can trash filesystem data in RAM and cause disk corruption. When it happens doesnt matter were you using SFS or PFS. |
|
New CAM article released : Comment 26 of 31 | ANN.lu |
Posted by Sigbjørn Skjæret on 03-Jun-2003 18:30 GMT | In reply to Comment 12 (Olaf): The problem with FFS is that it is *much* easier to put in an inconsistent state, something which will make it start the validation process, something many people have become acquianted with over the years .. This is usually where nasty things happen, as the validator has never exactly been foolproof, and has been known to cause nasty things like corrupted files, and sometimes even neverending validation, with more and more corrupted blocks appearing each time...
SFS and PFS have been constructed so that this won't happen as easily, but ofcourse, they have had (and probably still do) their share of other issues, but personally, out of experience, I'd say PFS3 is most likely the securest filesystem you could run in an Amiga(ish) environment these days.
- CISC |
|
New CAM article released : Comment 27 of 31 | ANN.lu |
Posted by Ray A. Akey on 03-Jun-2003 19:05 GMT | In reply to Comment 11 (Olaf): I'll look over the published document and compare with the original to see what changes were made in the published version.
There shouldn't be any need to re-publish it anywhere else as long as the proper corrections are made to the published version. |
|
New CAM article released : Comment 28 of 31 | ANN.lu |
Posted by Kolbjørn Barmen on 03-Jun-2003 19:29 GMT | In reply to Comment 12 (Olaf): Olaf, you asked for input and ideas, so here goes...
How about a buildt in file version system ala VMS's filesysten as a plugin? :)
Sure makes it even more robust and safe, hehe |
|
New CAM article released : Comment 29 of 31 | ANN.lu |
Posted by Amon_Re on 03-Jun-2003 20:02 GMT | In reply to Comment 15 (MarkTime): MarkTime, that was totally uncalled for.
Cheers |
|
New CAM article released : Comment 30 of 31 | ANN.lu |
Posted by Amon_Re on 03-Jun-2003 20:05 GMT | In reply to Comment 22 (MarkTime): MarkTime, are you pretending to know what you are talking about again?
SFS is going to be there, so it is written, so it shall be.
The BlizzPPC was never "written" down, they always said that if possible, they
would do it
If you really want to troll, why not goto usenet & bug SG?
Cheers |
|
New CAM article released : Comment 31 of 31 | ANN.lu |
Posted by Thomas W on 04-Jun-2003 06:26 GMT | In reply to Comment 21 (Anonymous): I once lost a drive to SFS.
The circumstances that caused it would be pretty much impossible to duplicate though.
Up till then SFS had been behaving very nicely.
These days I use only FFS as I've never lost a file I couldn't recover while using FFS. Not unless there was a physical error on the HD that is. |
|
Anonymous, there are 31 items in your selection |
|
- User Menu
-
- About ANN archives
- The ANN archives is powered by #AmigaZeux. It was updated daily (news last: 22-Oct-2004; comments last: 18-May-2005).
ANN.lu was created, previously owned and maintained by Christian Kemp, www.ckemp.com.
- Contribute
- Not possible at this time!
- Search ANN archives
- Advanced search
- Hosting
- ANN.lu was hosted by Dreamhost. Sign up through this link, mention "ckemp" as referrer and he will get a 10% commission on any account you purchase.
Please show your appreciation for any past, present and future work on ANN.lu by making a contribution via PayPal.
|