FAT file system problem ... Microsoft is suing TomTom ...

Lighten up. Anyone reading the post can see that the attribution is mismatched. The last thing that this group or any other needs is the usenet police patrolling the area.

Rick

It is my understanding that this depends on how you distribute your binaries. The GPLv3 section 6 states: "6. Conveying Non-Source Forms.

You may convey a covered work in object code form under the terms of sections 4 and 5, provided that you also convey the machine-readable Corresponding Source under the terms of this License, in one of these ways: ... b) Convey the object code in, or embodied in, a physical product (including a physical distribution medium), accompanied by a written offer, valid for at least three years and valid for as long as you offer spare parts or customer support for that product model, to give anyone who possesses the object code either (1) a copy of the Corresponding Source for all the software in the product that is covered by this License, on a durable physical medium customarily used for software interchange, for a price no more than your reasonable cost of physically performing this conveying of source, or (2) access to copy the Corresponding Source from a network server at no charge."

So if you choose to distribute your binaries as covered by section 6b, you must, as the license says, give the sources to *anyone* who has the binaries. I'm not sure if you can restrict the distribution of binaries created from GPL sources.

-a

imple

n?

And what about the linux I remember which used the last .3 to index a file of long names etc so it would work on FAT

On the 'kind attitude' of windows offering a format, this is a default, I could be considered unsafe behaviour in itself. 'where did that post go' phenomenon.

A readonly lock on the embedded file containing the hidden file system would prevent medling by windows format tools. A hardware lock. Requiring special format tools which need to write checksum within a lock control sector...

Returning bad disk errors. better than the alternative. It could be argued that the notification of 'not formatted' is not specific enough to allow a reasonable user to make an informaed decision on data protection grounds, makes the the default behaviour occupy the grey area of semi-legal.

For embedded systems the whole file instead of raw sector argument, is a moot point. The only problem is filename conversion. 8 characters is enough to store an md5 or other hash value, and so for finding files this presents just a hash calc problem. the chances of 2 files being in the same directory, and requireing the same hash key are quite miniscule.

On the rare ocasion where indeterminacy of filename is a problem, the .

cheers jacko

This looks like a change with the GPL v. 3 compared to the GPL v. 2, which definitely did not require you to provide source code to anyone except those you provided with the binary code. Even with the wording in the GPL v. 3, I'm not at all sure how much you can be legally bound to give rights to a third party when distributing the code to someone else.

If you read Heath's message (which you didn't quote), I don't think he realized that. It is not policing in this case, it is simply advising someone about how the system works.

GPL v2 does require that - if you distribute under 3b, anyone (not just recipients!) could request sources under 3b, if they had the information from the 3b offer:

b) Accompany it with a written offer, valid for at least three years, to give any third party, for a charge no more than your cost of physically performing source distribution, a complete machine-readable copy of the corresponding source code, to be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange; or,

"to give any third party" - not limited to binary recipients!

(GPLv3 section 6b snipped)

The corresponding section in GPLv2 is 3b: "3. You may copy and distribute the Program (or a work based on it, under Section 2) in object code or executable form under the terms of Sections 1 and 2 above provided that you also do one of the following:

b) Accompany it with a written offer, valid for at least three years, to give any third party, for a charge no more than your cost of physically performing source distribution, a complete machine-readable copy of the corresponding source code, to be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange;"

The GPLv2 text is even more liberal in that it doesn't even state that the third party needs to be in possession of the object code. Section 3c of GPLv2 further states that "c) Accompany it with the information you received as to the offer to distribute corresponding source code. (This alternative is allowed only for noncommercial distribution and only if you received the program in object code or executable form with such an offer, in accord with Subsection b above.)"

My reading of this is that if you receive a program in binary format covered by section 3b, and if you pass along a copy to someone else for free, then that other person is entitled to the source code. The GPLv3 has a similar clause in section 6c.

If I remember correctly, these clauses were used to get the sources for Sony's Playstation 1 compiler a number of years back. I also remember that this was done with the assistance of the FSF, which would indicate that this is indeed the intent of that part of the license.

-a

You (and DJ) are entirely right here - I've misread this particular detail. You only have to give the offer in writing to people to whom you distribute the binary, but the offer is valid for any third party. Thank you for setting me straight on this.

mvh.,

David

I did realize it actually, and wouldn't have commented unless I had something useful to add. In addition, not every newsreader shows attribution indentation so clearly.

FWIW, I don't mind your injections of etiquette every so often, as long as it's not too often. Not every dill needs to be told they're a dill.

Some folks on slashdot are pointing out that Universal Disk Format (UDF) while originally for DVD type media will apparently work on flash devices or even ordinary hard drives, and is supposedly supported in most versions of windows, linux, OSX, etc.

I have not had a chance to try it yet though. I suspect I will need to do the initial format in linux and then see what windows does with it.

Here's how that discussion is going:

Busybox is a good example. if you use Busybox, there's a reasonable chance you may also want to statically link with it. Busybox is meant to be "extended", but adding some more commands to it. So combine that and you've got an extended GPL'd piece of code that requires you to supply your source code, statically linked with all your proprietary code. The GPL does not make distinctions between your minor changes to Busybox to call your propriety library, and your library itself, both are derivative works.

For Busybox, which is a relatively straight forward and simple piece of code, you can easily re-invent it yourself. But re-inventing a full ext2 file system with good enough quality for customers is a major task. Even porting it (or something from BSD, which I considered) is also a huge task if your OS isn't very unix-like on the inside.

There is a problem in some application areas where allowing the end-use ability to modify the code could violate government regulations. Ie, FCC or FDA regulations in the US.

I don't know about the FDA, but as far as FCC goes, the device's type approval is invalid if the software is modified, but that is NOT the manufacturer's responsibility any more than it is Ford's responsibility if I remove the catalytic converter from my neighbor's car.

What about putting two partitions on a USB stick for example. The first one is a small FAT or NTFS partition containing code for an ext2 or whatever file system is on the second partition. The user then installs the driver, and the second partition can be accessed. The embedded system can write the first partition as a binary block and only support the file format of the second partition.

Regards Anton Erasmus

Ant>What about putting two partitions on a USB stick for example.

NTFS is harldy the lowest common denominator; there are still Win9x systems out there (think: POS).

This notion has already been rebutted:

Nice try, dude, but I don't think it will gain traction.

One doesn't need to "think: POS." I use Win98SE -- it has distinct advantages for me as my regular system. I prefer it, actually.

Jon

... snip ...

They are covered by FAT16 and FAT32.