Quote:

Bad news from Macrosystems. I guess the Apple stack is the only source to use unless anyone else has a better plan.

mmm... then now we would need to know who are the developers...

Acill

Nice try and thank you I didnt have much hope since ppl have tried to contact them in the past couple of years with similar result..

I do find it hard to believe however that the company lost all their source code base, or technical documents on their IP!

magnetic

Assuming we pay for the device with FireWire on it (and not whether it has the software on it) and have already paid our thousand dollars or so to the MPEGLA for past devices, Pegasos is already covered.

EFIKA doesn't have a Firewire port last I checked, neither does the HDB. I am not sure if there is a cost associated with the SOFTWARE (I think these patents are on wire interfaces and signalling, not algorithms or protocols) but I will check for you.

I would go ahead and start the bounty and start coding now if you are serious about it..

Quote:

Thank you. Would be sad if one spends the time and effort to do this, only to find out that the thing cannot be used due to some legal/licensing aspects. Better safe than sorry.

@tmhg,

I would procede on the basis that legal issues are surmountable. We do already have a license for the devices - Pegasos is all checked up and ready to go. If you want to develop for the Pegasos, go ahead. Genesi will cover you guys.

If you want to ship it for EFIKA or some other MorphOS device, therein lies the rub; these devices may not have FireWire without expension cards, and as such we are not paying royalties when we build them. I would procede on the basis that the MorphOS Team have told you guys that for the forseeable future these devices are not going to be supported by a public MorphOS, and therefore you will never be able to run the code on them anyway.

If a developer codes a Firewire stack and it is good, it could be used in commercial projects with MorphOS, if such things are embarked upon, and if so there is plenty of potential that the author could gain royalties for their code, on top of the bounty, when used on these devices.

You know, I would rather see bounties for further support of Reggae though. You guys expect the world from the other developers but can't even code a bunch of image loaders? Windows icons and bitmaps are great but what about PNG, JPEG, TIFF etc., coordinate with the MorphOS team so you can not step on their heels (or at least step on them in a way they are comfortable with re class naming, API quirks and so on).

Hi all,

I don't fully understand all that Neko said in that last post, but no matter.

I just want to re-iterate again that I'm a willing donor of money for the firewire bounty. I also have a DV Camcorder here that has a firewire connection that I am more than willing to plug into the Peg II and test with any code that someone writes.

Bifford.

Bump...

So whats up with Firewire, will there be a Bounty for it? Maybe we could open it to AMIGA OS and AROS Users so that the Amiga (like) Operating Systems can team up on this...

@Donar

making the Firewire stack OpenPCI compatible may be a good idea. I have some ideas (probably some of these ideas are quite crazy, but everything can be changed) about the bounty requirements:

1.-Write a Firewire stack capable of accessing HardDisks, CDs and DVDs (this could be tested using a Firewire2IDE box)

2.-Drivers for each firewire card must be modular although the only mandatory chip to support would be VIA ones (like the one used in all Pegasos1&2).

3.-The firewire stack must be modular (just like Poseidon). At least 1 subclass must be provided: One for mass storage devices like HDs, DVDs, CDs... A TCP/IP class may be dessirable but it's not mandatory. The class will be able to load RDBs and in MorphOS and AmigaOS3 it also will be able to automount HDs with FAT16, FAT32, ext2 and NTFS filesystems using Marek Szyprowski's Filesystems. On operating systems without existing support for any of the listed filesystems it's not required to add support for it. That goes for AROS as it currently lacks support for FAT32/NTFS/EXT2 and also for OS4 (as it lacks support for NTFS/ext2).

4.-The stack must be portable: SDI headers and openpci.library will be used. AROS would require additional work for the PCI calls but a small wrapper for OpenPCI could be written for AROS.

5.- An easy MUI/Zune GUI is mandatory. It can mimic Poseidon one (Trident).

6.- Documentation and some developer info for writting new classes and drivers to support additional Firewire cards must be provided

7.- The stack should work in all pegasos (including Pegasos1), 2 pcs with aros, amithlon using OpenPCI and it would be dessirable that at least one type of classic amiga already supported by OpenPCI (maybe Mediator, maybe G-Rex, maybe Prometheus with the latest firmware update.). The devices to test will be a Firewire2IDE adapter box

8.- (optional) There will be an option of saving the preferences in the executable that launches the stack (like poseidon) to allow storing the binary in a FlashRom or directly as a kernel module. It should be resident

9.- (optional) The sourcecode will be released under the AROS license (that would allow MorphOS/OS4 team to include it as standard with next MorphOS versions)

I am glad to tell that AmiX donated me a VP-D20 MiniDV camcorder for the IEEE1394 implementation. It is here at last and i'm starting the work.
I've studied two availible opensource solutions: Apple reference stack and Linux stack. In BSD family IEEE1394 support is incomplete and supports only masstorage.
I've left out Apple's stack as soon as i've read its license. It's totally unfriendly and in fact it doesn't allow me to do anything valuable. In addition, Apple's implementation is not enough because the implement only transport level, no higher-level protocols (means no device classes).
Linux implementation is quite new, but it already works. In addition, keeping its API intact will allow easy porting of existing applications. I already have experience in ripping alien kernels (MOSNet ) so i've decided to base on Linux stack.
First stage will include raw IEEE1394 interface and IEC61883 class which will support MiniDV cameras. I don't know if some configuration will be actually needed (Poseidon is completely plug-n-play and doesn't require any setup, IEEE1394 is going to be the same).
Also i have a PCI IEEE1394 card in unknown condition. If it works i'll be able to install it into my PC and try out an IP-over-IEEE1394.
I don't have any other hardware so i'll be unable to support it.

Sonic,
When this gets written up and posted(is this being done?), I will donate!

Thanks for you work!

Bob

I was thinking that it may be possible to put an option to access the MOS hard disks from other computer as virtual firewire hard disk. I think macs also allow you to connect your mac to other computer and access the mac drives just like you would do with a firewire hard disk connected to your computer.

This is very optional but would be cool anyway

@sonic:

great! We'll have to check how much costs sending stuff to Moscow. A firewire<->IDE box could be useful to develop this. We have some spare graphic cards that may be useful for your old pentium you plan to use for AROS stuff. Maybe some motherboards too.

The actual work will begin next week. Today and tomorrow i'll be sitting at work until 21:00 because on friday i left it at 13:00 in order to help my darling wife to prepare to celebration of 1 year of our marriage.

Quote:

Now that you mentioned it...

To tell the truth i'd love to upgrade my PC. One guy from here has already promised to donate me a PC parts or maybe even laptop to upgrade, but disappeared after this.
Emulating IEEE1394 hard disk is possible, i beleive it's not a great problem. In Read-only mode it can be completely transparent, however read-write mode will lock up the system during operation because the filesystems on the drive will be inhibited in order to enable the remote host to write to the drive.

I don't need firewire (I think) but it's always nice, and Sonic has shown himself qualified for this kind of job. So I'll throw in a little money when you get the bounty specs done. Better make sure the specs are very clear this time to avoid people having different expectations.

Quote:

Well, where ate the official specs?
I see only some proposal from Crumb, however i can't meet this since i don't have firewire<>IDE adapter, instead i start with DV.
It's also a pity that MorphBounties seems to be not maintained. In some time i'll make an alternative step and open my own program. Unmorphos project can have its own web page on sf.net, i'll simply put some information about my bank account and postal address there. Everyone will be able to make a direct donation if he wishes. I just need to find a way to maintain it myself. Now my restrictions at work are relaxed a bit as you see, however i still don't have own email and https, CVS, etc are banned.

Well, can't you write yourself what it is that you want to do? If you post something in this thread, people can comment on it. I can volunteer to collect the comments into a set of bounty specifications.

But I think it's best that you post something first, because you're the one who knows best what you can and cannot do. It's no problem I think if this means a somewhat limited implementation.. better not aim to high, from what I've understood it would be quite a lot of work to make a really complete firewire stack.

Why not begin with the suggestions that crumb wrote down on page 3 (lower half of page) and before starting one big project why not make several bounties?

a) Basic Stack supporting HD/CD maybe DV. Source released under some "good" license. Gui should be addable.
b) Mui Gui -> Should work under MOS, AOS maybe Zune (Zune is MUI compatible!?)
c)add things here

As i have no clue how software is written, don't take it too serious if i made crappy suggestions.

Bye

Donar

Well, as Sonic already said he can't do what Crumb is asking for, I think it's better to start by asking Sonic what he *can* do.

It's not like we're setting up a general bounty that we hope just "anyone" will come and pick up. The need (?) for firewire has been there for years and until now no one had any realistic ambition of filling this gap. So I think we should be quite aware that this is a bounty designed for whatever Sonic is prepared to do. Therefore we should write the specifications in such a way that the people who donate will be satisfied with this specific project.

Btw, although the administration of MZ bounties has not been very regular (for understandable reasons, no blame on Targhan here), I think it would be a mistake for Sonic to set up his private bounty system. I just happen to think there will be more donations if it's done through MZ. But of course this is up to Sonic to decide.

In my opinion a Firewire stack without support for masstorage is useless. It should be much easier for Sonic to buy a FW<->IDE adapter (or lend one from a friend?), then us shipping one to Moscow... Maybe Sonic can add the cost for a FW<->IDE box to his suggested goal for this Bounty? If he get's the stack working it should be forseeable for him to get the "masstorage.device" done also.

Bye

Quote:

Right. Experiencing how russian postsystem works, I wouldn't send even a rock. Can't even count how many letters they have lost, not to speak stealing 20 euros from a well wrapped package

Excuse me for being clueless about Firewire.. but I suppose the need for masstorage would not be in order so we can connect IDE hard drives this way, but because some/most cameras support masstorage?

Quote:

Oh as i have a DVD Burner, 3,5 HD and a 2,5 HD with FW connector i sometimes would like to connect to my Pegasos. So a FW "masstorage.device" would be for hooking up IDE devices.

And that is why:
USB 1.0 is way to slow for transferring bigger packets of data, and even USB 2.0 with its "Intel Specification" of 480 Mb/s seems to be slower than Firewire. But this could be also caused by the chipset used in my enclosures.

Bye

I'm stoked about the possibility of firewire-that means the door is
opeen to 'iMovie' style apps for MorphOS.That and a real browser, and
the Mac just got waaay less useful to me

Anyhow, if you guys can wait a bit on the bounty aspect of it, there
might be a pleasant surprise in store for you in that regard

I can say that I did try to get firewire working on Linux a couple
years ago-I failed. Weeks worth of nights spent compiling kernels,
compiling ieee1394 drivers, libraries etc. what a nightmare! I sure
hope it is better there than it used to be.

Bolton