Hello,
First post - please don't kill me.
We are in the process of launching an embedded product based on a Jetway mobo (http://www.jetway.com.tw/jetway/system/productshow2.asp?id=383&proname=J...), and need LinuxBIOS support for this setup.
Since we are a bit short on time, I would like to offer USD1500 to anyone who can get support in place according to the specs below. The final result obviously goes back into the project with full GPL. If this offer violates any project rules / etiquette, or it is way off the mark, I sincerely apologize.
Our requirements, which are open for discussion: 1) Linuxrc executing three seconds after power on. Jetway can offer 6-7 seconds for POST.
2) Always power on when power is applied
3) Blank screen. (We will splash from initrd)
4) Loads 2.6.20.4 kernel directly from flash (in order to minimize chance of a bricked system and shorten boot time).
5) Initrd is read from USB memory stick, if available.
6) If no USB is plugged in, initrd is read from /dev/hda1, where root is also available.
7) The USB initrd can optionally flash a new kernel into firmware a part of the upgrade process.
8) Done by Apr 20.
We have a motherboard to send over as a development platform.
The idea with the initrd scheme is to provide a master USB memory stick to restore bricked units by having the kernel mount the stick's initrd, which would then load additional drivers and reinstall everything on /dev/hda1.
If you have any questions or flames, please reply to the list. If you want to take on the job, mail me and we'll talk off list.
Sincerely,
/Magnus Feuer CTO Magden Technologies magnus _dot feuer _at magden _dot us.
Hi,
On Wed, Mar 28, 2007 at 09:43:30AM -0700, Magnus Feuer wrote:
If this offer violates any project rules / etiquette, or it is way off the mark, I sincerely apologize.
I for one think this is an excellent way to improve LinuxBIOS and your product at the same time!
I just wish I had more free time.
//Peter
On 3/28/07, Magnus Feuer magnus.feuer@magden.us wrote:
Hello,
First post - please don't kill me.
Thanks for posting this, we need people to know such opportunities exist. I will also ask around for people who might help you.
thanks
ron
There have been a few takers. I will follow up on them today.
One question about a VIA NDA: If we sign an NDA with VIA in order to obtain specs and developer docs, will we not be prohibited from releasing the finished LinuxBIOS port under GPL since the source would show the secrets covered by the NDA?
/Magnus F.
ron minnich wrote:
On 3/28/07, Magnus Feuer magnus.feuer@magden.us wrote:
Hello,
First post - please don't kill me.
Thanks for posting this, we need people to know such opportunities exist. I will also ask around for people who might help you.
thanks
ron
On Thu, Mar 29, 2007 at 10:58:42AM -0700, Magnus Feuer wrote:
There have been a few takers. I will follow up on them today.
One question about a VIA NDA: If we sign an NDA with VIA in order to obtain specs and developer docs, will we not be prohibited from releasing the finished LinuxBIOS port under GPL since the source would show the secrets covered by the NDA?
/Magnus F.
If it's from the datasheet only, then there is no problem with the NDA documents i've seen (but never signed). VIAs Linux NDA addendum only prohibits redistribution of (only) GPLed code VIA handed to you directly but hasn't made generally public (yet). It therefor directly forces you to breach the GPL.
Daft, clueless, useless and will probably not stand in court.
A typical example of VIA stupidity.
Luc Verhaegen. http://unichrome.Sf.net/
Luc Verhaegen wrote:
On Thu, Mar 29, 2007 at 10:58:42AM -0700, Magnus Feuer wrote:
There have been a few takers. I will follow up on them today.
One question about a VIA NDA: If we sign an NDA with VIA in order to obtain specs and developer docs, will we not be prohibited from releasing the finished LinuxBIOS port under GPL since the source would show the secrets covered by the NDA?
/Magnus F.
If it's from the datasheet only, then there is no problem with the NDA documents i've seen (but never signed). VIAs Linux NDA addendum only prohibits redistribution of (only) GPLed code VIA handed to you directly but hasn't made generally public (yet). It therefor directly forces you to breach the GPL.
Daft, clueless, useless and will probably not stand in court.
A typical example of VIA stupidity.
Luc Verhaegen. http://unichrome.Sf.net/
I have, erm, acquired the vt8237r datasheets (not through via, either). Email me privately for anyone who wishes to work on this. If you aren't a usual poster on the mailing list, or someone I know, don't expect a reply.
-Corey
On Thu, Mar 29, 2007 at 02:58:28PM -0400, Corey Osgood wrote:
I have, erm, acquired the vt8237r datasheets
I would strongly like to ask everyone to refrain from using any knowledge gained from documents that they do not have a legal right to read, at least while working with LinuxBIOS.
LinuxBIOS' reputation among vendors would suffer greatly from even a single incident if the relevant agreements were not in order with the vendor before acquiring documentation.
This kind of mistake is even more undesired, unneccessary and uncool when the vendor is willing to work with us, as my impression is for VIA.
We would only be hurting ourselves.
Worst case scenarion:
Vendor Exec: "If they can't even do it with VIA who practically give away their docs, why should we even give them half a chance?"
//Peter
Peter Stuge wrote:
On Thu, Mar 29, 2007 at 02:58:28PM -0400, Corey Osgood wrote:
I have, erm, acquired the vt8237r datasheets
I would strongly like to ask everyone to refrain from using any knowledge gained from documents that they do not have a legal right to read, at least while working with LinuxBIOS.
LinuxBIOS' reputation among vendors would suffer greatly from even a single incident if the relevant agreements were not in order with the vendor before acquiring documentation.
This kind of mistake is even more undesired, unneccessary and uncool when the vendor is willing to work with us, as my impression is for VIA.
We would only be hurting ourselves.
Worst case scenarion:
Vendor Exec: "If they can't even do it with VIA who practically give away their docs, why should we even give them half a chance?"
//Peter
Afaik, the docs I've got on this chipset (vt8237r) are not covered under any sort of NDA, there is no watermark or other indication that they are. Searching for "NDA" in the document turns up sta"nda"rds and seco"nda"ry, and "disclosure" and "confidential" both come up empty. The server that I found them on is accessible through Google, and they've been there for quite a while, so Via has obviously not cared enough to have the owner take them down. Personally, I don't see why sucking up to a hardware vendor for non-confidential datasheets is necessary.
I have personally contacted Via 3 times now about datasheets for different chipsets, and never been given a response, and 2 other people who I've given this datasheet to said the same thing. With my most recent attempt, I let them know both about Magden's product, and about the potential for enthusiasts to purchase newer products instead of having to stick with older ones should they want to use LinuxBIOS (this was my personal case). Still no word. If they practically give them away, I'm practically Brad Pitt!
But, in the true spirit of I-don't-wanna-piss-anyone-off, I will stop distributing them, and I won't commit back any code I may write until after I've been (if I ever am) given the docs officially by Via.
-Corey
Hi,
2007/3/29, Corey Osgood corey_osgood@verizon.net:
I have personally contacted Via 3 times now about datasheets for different chipsets, and never been given a response, and 2 other people who I've given this datasheet to said the same thing. With my most recent attempt, I let them know both about Magden's product, and about the potential for enthusiasts to purchase newer products instead of having to stick with older ones should they want to use LinuxBIOS (this was my personal case). Still no word. If they practically give them away, I'm practically Brad Pitt!
Yes, I contacted VIA too with no response: http://linuxbios.org/pipermail/linuxbios/2006-September/015597.html
They will stop to ignore us when they lost money because their hardware don't support LinuxBIOS.
-Corey
Alan
* Corey Osgood corey_osgood@verizon.net [070330 01:00]:
Afaik, the docs I've got on this chipset (vt8237r) are not covered under any sort of NDA, there is no watermark or other indication that they are. Searching for "NDA" in the document turns up sta"nda"rds and seco"nda"ry, and "disclosure" and "confidential" both come up empty. The server that I found them on is accessible through Google, and they've been there for quite a while, so Via has obviously not cared enough to have the owner take them down. Personally, I don't see why sucking up to a hardware vendor for non-confidential datasheets is necessary.
To take some emotions out of this discussion:
The 8237 data sheet can be downloaded from the via.com.tw site.
It's just not even 10% of what is needed to do a port.
Stefan
* Stefan Reinauer stepan@coresystems.de [070330 02:31]:
To take some emotions out of this discussion:
The 8237 data sheet can be downloaded from the via.com.tw site.
http://www.via.com.tw/en/downloads/datasheets/chipsets/VT8237R_SouthBridge_R...
On Thu, Mar 29, 2007 at 07:00:18PM -0400, Corey Osgood wrote:
Afaik, the docs I've got on this chipset (vt8237r) are not covered under any sort of NDA, there is no watermark or other indication that they are.
Thanks for confirming that!
The server that I found them on is accessible through Google, and they've been there for quite a while, so Via has obviously not cared enough to have the owner take them down.
Unfortunately that may not be how legal departments reason, had it been a document covered by an NDA. :(
Personally, I don't see why sucking up to a hardware vendor for non-confidential datasheets is necessary.
It isn't! I just wanted to remind everyone to stay aware of the pitfalls.
I have personally contacted Via 3 times now about datasheets for different chipsets, and never been given a response, and 2 other people who I've given this datasheet to said the same thing.
That's bad. Maybe they're only easy to work with if there's a company involved.
With my most recent attempt, I let them know both about Magden's product, and about the potential for enthusiasts to purchase newer products instead of having to stick with older ones should they want to use LinuxBIOS (this was my personal case). Still no word.
I'm sure all documentation will be readily available for the taker of the CN700 port.
If they practically give them away, I'm practically Brad Pitt!
Hehe. You got the point at least. Sorry for over-dramatizing. :)
But, in the true spirit of I-don't-wanna-piss-anyone-off, I will stop distributing them, and I won't commit back any code I may write until after I've been (if I ever am) given the docs officially by Via.
That wasn't my intention at all, I just wanted to post a reminder that it's important to be careful.
//Peter
Quoting Peter Stuge stuge-linuxbios@cdy.org:
On Thu, Mar 29, 2007 at 02:58:28PM -0400, Corey Osgood wrote:
I have, erm, acquired the vt8237r datasheets
I would strongly like to ask everyone to refrain from using any knowledge gained from documents that they do not have a legal right to read, at least while working with LinuxBIOS.
LinuxBIOS' reputation among vendors would suffer greatly from even a single incident if the relevant agreements were not in order with the vendor before acquiring documentation.
This kind of mistake is even more undesired, unneccessary and uncool when the vendor is willing to work with us, as my impression is for VIA.
We would only be hurting ourselves.
Worst case scenarion:
Vendor Exec: "If they can't even do it with VIA who practically give away their docs, why should we even give them half a chance?"
//Peter
Is this why Intel is so uncooperative??
Thanks - Joe
On Fri, Mar 30, 2007 at 01:48:56AM -0400, joe@smittys.pointclark.net wrote:
Quoting Peter Stuge stuge-linuxbios@cdy.org:
Worst case scenarion:
Vendor Exec: "If they can't even do it with VIA who practically give away their docs, why should we even give them half a chance?"
Is this why Intel is so uncooperative??
Who knows?
My guess is that Intel has a policy to not bother with too small companies in order to minimize risk. Another policy would be to not work with any project that project sales below a given limit.
Either makes at least some business sense but sucks for !megacorp.
//Peter
On 3/29/07, joe@smittys.pointclark.net joe@smittys.pointclark.net wrote:
Is this why Intel is so uncooperative??
no, not at all. We have not in almost 8 years had any complaint related to IP or release of NDA or any such problem.
Intel's lack of cooperation is driven, from what they have told me, by their concerns related to releasing chipset information that they consider (their words) "crown jewels".
Intel is also in the BIOS business, and (former) intel people have told me that Intel see LinuxBIOS as a competitor. This perceived competition could also be an issue.
thanks
ron
* Luc Verhaegen libv@skynet.be [070329 20:19]:
If it's from the datasheet only, then there is no problem with the NDA documents i've seen (but never signed). VIAs Linux NDA addendum only prohibits redistribution of (only) GPLed code VIA handed to you directly but hasn't made generally public (yet). It therefor directly forces you to breach the GPL.
Why is that? The GPL requires you to pass the sources to the users of GPLed binary code on their request. There's no need to make source code generally available.
A typical example of VIA stupidity.
I can't confirm this at all. VIA has been extremely helpful and absolutely unproblematic in NDA and IP questions where coresystems was involved.
Stefan
On Thu, Mar 29, 2007 at 10:20:59PM +0200, Stefan Reinauer wrote:
Why is that? The GPL requires you to pass the sources to the users of GPLed binary code on their request. There's no need to make source code generally available.
:) How should i know. This is VIA.
They also made it GPL specific. Under that linux addendum of theirs, you're allowed to do everything you want with any other free software license.
That suggests that this is plain stupidity and that the people who wrote that document were told some very superficial things about the GPL, and set out to mitigate its main directive.
I can't confirm this at all. VIA has been extremely helpful and absolutely unproblematic in NDA and IP questions where coresystems was involved.
Right, the times i was handed an nda directly i always got handed both standard documents. The NDA itself pretty much tells you you're not allowed to disclose anything, the addendum tells you that you're allowed to disclose everything, except that bit with gpled code VIA handed you directly.
Luc Verhaegen.
On Thu, Mar 29, 2007 at 10:58:42AM -0700, Magnus Feuer wrote:
There have been a few takers. I will follow up on them today.
One question about a VIA NDA: If we sign an NDA with VIA in order to obtain specs and developer docs, will we not be prohibited from releasing the finished LinuxBIOS port under GPL since the source would show the secrets covered by the NDA?
I _think_ an NDA is not a problem, it (usually?) only protects the datasheet itself (you may not give it to third parties), but not any code created with the help of the datasheet.
Please somebody correct me if I'm wrong.
I suggest you make it perfectly clear to VIA that you're going to use the datasheet to enhance a GPL'd project, and thus you will release your code under the terms of the GPL, too. If they _know_ this, and agree to it, there cannot be any problems.
(I personally don't like NDAs in general; it would be much nicer if all datasheets were publically available; but having datasheets under NDA and being able to release GPL'd code is much better than having no datasheets at all)
HTH, Uwe.
* Uwe Hermann uwe@hermann-uwe.de [070330 12:19]:
I _think_ an NDA is not a problem, it (usually?) only protects the datasheet itself (you may not give it to third parties), but not any code created with the help of the datasheet.
This depends on your NDA contract. One way or the other, before signing an NDA, you should tell the vendor what you need the data sheets for, and get written permission to distribute resulting code under the GPL.
I suggest you make it perfectly clear to VIA that you're going to use the datasheet to enhance a GPL'd project, and thus you will release your code under the terms of the GPL, too.
Yes, fully agreed. In the end, this is not only a matter of legal perspectives, but also of how to establish a good and stable connection between the LinuxBIOS project and the hardware vendors.
If they _know_ this, and agree to it, there cannot be any problems.
Ack! Contracts are mostly for the cases where something goes wrong, in one of the contract partner's opinions. So by making this sure before you start working, you ensure both sides can sleep well at any time and no evil blood is generated.
(I personally don't like NDAs in general; it would be much nicer if all datasheets were publically available; but having datasheets under NDA and being able to release GPL'd code is much better than having no datasheets at all)
At this moment I want to point out how well AMD has been treating us since the advent of the AMD64 platform. They have been an early supporter of LinuxBIOS, and I think it is safe to say they never had any reason for regret.
Since the firmware is the layer between hardware and software, we need to make sure to treat hardware vendors as our allies, not our enemies on the way of producing something good.
We have no reason to support any hardware vendor against their will. Instead we should convince them that they really have a big gain by supporting LinuxBIOS.
Stefan
As some of you may already know, Magnus has given me the contract. I'm writing this because I want to make a few things known:
First and foremost, I told Magnus that if any of the senior developers wanted this project (Ron, Stefan, Peter, David, Uwe, you guys know who you are), to give it to them. You guys have been doing this a lot longer than I have, and you definitely deserve it if you want it. If no one else stepped up, I had a pretty good idea that I could do it. If any of the senior devs still want this project, I'm willing to turn it over, or if someone feels you're more qualified than me, please email me privately and we will discuss it. I don't want their to be any feelings of discontent over this.
Second, I have no intention of blowing the money on a car or some insane new gaming system. Most of this will be going straight to the University of Maine, to pay off my debts so I can go back to school this fall. I've been working a full-time job and also part time under the table trying to pay off my debts so I could go back this fall, and still wasn't going to be able to make it. This was just exactly the opportunity I needed. I intend to take at least a week off work just to work on this, and it will also help to cover that cost.
Third, all the code I make will be submitted back to the LinuxBIOS project, which means that in the future we will be able to support the latest Epia-series and other embedded boards, which we've been seeing an increased demand for. I feel this is the most important part of this whole ordeal.
Fourth, if anyone wishes to help, please email me privately. There may be some money involved, but only if the code is done by the 20th (which is a very tight schedule, IMO), and as I said before, most of it is earmarked already. We can discuss it.
The problem with the whole bounty system is that someone always gets left out. Heck, everyone but one person gets left out. It's not fair, I agree, but in the long run it's for the good of the project.
-Corey
Corey, I am glad you took the job. It's a chance to grow another linuxbios expert! Congratulations!
You should not hesitate to call on us if there is anything we can do. And, Magnus, thanks for doing this.
thanks all
ron
On Thu, Mar 29, 2007 at 07:44:39PM -0700, ron minnich wrote:
Corey, I am glad you took the job. It's a chance to grow another linuxbios expert! Congratulations!
You should not hesitate to call on us if there is anything we can do. And, Magnus, thanks for doing this.
thanks all
Full ack!
//Peter
Ron,
My pleasure. I have a problem that I am willing to throw money at, and Corey had a bit of time over. At the end of the day we all benefit. I am a strong believer in open source and am happy to help your cause.
/Magnus F.
ron minnich wrote:
Corey, I am glad you took the job. It's a chance to grow another linuxbios expert! Congratulations!
You should not hesitate to call on us if there is anything we can do. And, Magnus, thanks for doing this.
thanks all
ron
* Corey Osgood corey_osgood@verizon.net [070330 04:40]:
Second, I have no intention of blowing the money on a car or some insane new gaming system. Most of this will be going straight to the University of Maine, to pay off my debts so I can go back to school this fall. I've been working a full-time job and also part time under the table trying to pay off my debts so I could go back this fall, and still wasn't going to be able to make it. This was just exactly the opportunity I needed. I intend to take at least a week off work just to work on this, and it will also help to cover that cost.
:-) Hey, if you do the work, you should not worry on how you use the money. That's why we stopped trading cows for wheat. I think it is really cool that you stepped up. Though the schedule is really tough.
I am sure that anyone on this list will try to help you as good as possible.
The problem with the whole bounty system is that someone always gets left out. Heck, everyone but one person gets left out. It's not fair, I agree, but in the long run it's for the good of the project.
Hey,.. LinuxBIOS is a growing market. Go ahead and help it grow even more. You're helping all the others involved.
-
Alan Carvalho de Assis -
Corey Osgood -
joe＠smittys.pointclark.net -
Luc Verhaegen -
Magnus Feuer -
Peter Stuge -
ron minnich -
Stefan Reinauer -
Stefan Reinauer -
Uwe Hermann