Hi everyone,

Before I left to become a gambler, I was in the business side of the software industry. I have quite a bit of experience negotiating contracts related to software use, source-code escrow, reseller and OEM agreements, etc. I've been giving the empeg source code situation a lot of thought recently.

It seems to me that the empeg software may be considered (but some) old enough to not contain significant trade secrets at this point (this is no way a disparagement, just a comment about how business-folks may see it). I can imagine some business reasons that could be used to justify making the source accessible, with certain limitations, to the empeg enthusiast community, possibly as a non-profit donation, or some other approach.

I am very comfortable approaching the executives of the corporations controlling the IP rights to the source and having a business conversation with them about the idea of this. Before doing so, I would want to understand the ownership and licensing situation in some detail. My understanding is that the empeg software used libraries from at least one third-party company with whom the current owners of the empeg source do not have a license agreement. In principle, that doesn't really represent an impossible problem, it would "just" mean that a separate agreement would need to be made with the third parties.

An attorney friend of mine (also an empeg enthusiast) could assist me with creating a non-profit organization that could act as the legally responsible entity for the licensing agreements. That organization should consist of a board ideally composed of those that would control and monitor access to the source.

My questions for those on this BBS:

1. Assuming such an agreement could be reached, is this desireable? I believe there are several very talented software engineers in this community who could do real magic if they got their hands on the empeg source. What I don't know is if they'd want to... I am certain that access would mean that the IP could be used exclusively for enhancing functionality of the empeg and not for developing anything of one's own.

2. What is the ownership status of the IP, and what third parties are invovled?

I know from personal experience that things like this are possible. I've been on both sides of similar negotiations. They can be a bit complex, but they are possible. I am willing to volunteer my time and experience to setting up an organization and then representing it to obtain access to the code.

The key is to approach the problem from a business point of view rather than get bogged down in technical details. The starting point is the position that the software is technically obsolete, so why not make it available on a semi-restricted basis to enthusiasts, or even an open-source basis (like I said, this is the starting point)? Especially if one could take a tax deduction for donating this access? I don't particularly want to argue the technical points of whether this is true, and I'd prefer that this thread not go in that direction.

That said, it *is* important to understand exactly what would be required from a technical point of view to make the result usable by the community. Exactly what access is required, from whom, and to what degree? In other words, what would an empeg community development enviroment look like, and who holds the keys to making it come true?

The rest is just finding out if they are willing. That's not simple, but it's the part I'm good at...

FWIW,

Jim

...and we don't need to "open source it", but rather to have an organization gain access to it, and to extend that access to a few community developers via NDAs if necessary.

Just whatever it takes to get a few hands onto fixing up the v3 release, and perhaps adding the odd UI tweak or two.

Cheers

How would you get round the fact that the licence for the mp3 decoder has disappeared, would you try and plug something like MAD in its place?

I'm resigned to the fact that the v3 alpha currently running on my player is probably the last update I'll ever see.

Upgrades could take the form of a "patch" that includes all player software except the decoder... so anyone wanting 3.0 Community Release would need to first have a 3.0 alpha or 2.00 release and then patch it to get an mp3 decoder... ?

Great thread btw, I hope this goes somewhere.

I'm pretty sure that the player application is monolithic, instead of linking to a separate decoder library, which makes that sort of patch virtually impossible. However, as long as the only publicly available version was a compiled monolithic application, I don't see how the original licenser could have a viable complaint, depending on the terms of the license.

This has been covered before; the code is still very much in use and not obsolete - yes, the branch used to build the empeg is now orphaned - and the branch point is quite a while ago - but there is much commonality with the code that Sigmatel are now licensing to people buying their chips (ie, if you took the empeg v3 code, you'd find not insignificant parts unchanged in the current codebase, such was the excellence of the sw done in the early days). I'm not saying it could never happen, but as has been noted in the past, the empeg codebase is not abandonware - it's still in very active development (just not for the old platform) - which complicates the business case for letting people licence an old branch of it.

It's sort-of akin to asking Microsoft "seeing as NT3.51 is obsolete now, can we have the source?" - WinXP is a current product and is based on the same code, so there are plenty of commercial issues.

IMO the best option is just to start from scratch and GPL or BSD license it, which then means anyone can work on it, no agreements necessary, and it could also be used on other platforms. Sort of like the Rockbox software except ideally more elegant and modular. One thing that makes this much easier is that the empeg-car is wildly overpowered, so you can do things more elegantly/expansively than you might have to do if you were working on a portable device.

Just my 2c, and obviously I'm even less involved now than I was earlier this year

Hugo

What about the tuner ?

Hugo, I don't disagree with any of this. I still think there is a possible business case to be made to let a small non-profit group under NDA support their player.

I think its worth a try, we have absolutely nothing to lose by asking...

Is sigmatel the only party that needs to be involved? What about the decoder software? What about the visuals software?

Best,

Jim

Yes. Since Hugo points out it is not considered abandonware, part of the business case could be that this effort could be the property of the current owners. The owners have nothing to lose, and possibly something to gain!

Jim

MPD would be a nice base for an open source car player. It's client-server based; the server part does a lot of the goodies we like, like gapless playmode and crossfading. The server could almost be used as-is, only a new client would be needed.

The nice thing is that MPD can talk to several clients simultaneously. So you could have one client doing the regular UI, one do the remote control handling, one handling the stalk controller, and others that will handle your custom control methods.

Pim

Oh, and while we're at it, we could try to make an open source car player that also runs on generic car PC hardware, like one of these babies.

Pim

Apart from the parametric EQ stuff, all the docs are available - someone even managed to get the 7705 programmers manual as a softcopy, which is something that empeg never managed to prise out of Philips. I could well believe that official sanction to get the parametric eq programming code wouldn't be a lot of hassle.

Leaning heavily on open source stuff would make things progress faster; using some existing client/server code (and just running both on there) would be pretty nice too. This way you'd also decouple the part which needs to keep the disks spun down from the whole UI section which might be useful for debugging too. Looking at how far people got with new code for the Rio Receiver, which was a far, far more resource-limited platform, in a relatively short time, it doesn't seem to be an unattainable goal to get something going which could be built on.

Hugo

I have quite a lot of documentation on the hardware...

pca