Exactly was I was thinking... It'll be in a bios upgrade once they have 50% of the market, but at $525 they are going to have trouble against local dealers.
This clearly has nothing to do with aircraft performance but rather is designed to punish Airbus (EAS) for their 148% failure. An not in the industry but rather in the public perception which is a random and important part of aircraft success. I'm sure if one would the clock back to the launch of the 747 that the other players in that early market spread all sorts of rumors, it'll never fly, it's drop out of the sky, you'll suffocate etc etc. The 380 will sell so will the 787, both maiden commercial flights will be on slashdot.... both will crash at least once.
I scanned the headline too quickly, thought it said "...from the Netherlands"..... I mused, "Ok they're fairly smart but all of that cheese and dope can't be too good for cognative function."
I called their UK office first who basically told me to call the US, I got put through to a guy called Larry, he sounded like he was not just a phone jockey, he had me on speaker and I could hear his voice echo around his large office.
I acted dumb (not hard for me;-) and asked what I had to do to stay legal, and in his short reply he made three interestig points....
1) They are only "attacking" the US market.
2) They are only going after corporate users.
3) This position will stay the same for a least a couple of months.
i guess we would need to lean on one of the vendors or buys the BCM94710 evaluation kit or write an open letter to Broadcom denying that they do (lawyers don't like setting themselves us to be caught)
I would doubt if Broadcom could NOT distribute the compiler given the midifications that it is clear that the vendors are making
i.e. string run over the belkin and linksys kernels highlight that they are including modified header files from local directories e.g./home/hyin/cvs/BELKIN/ASKEY_31190/src/linux/linux/ include/linux/skbuff.h ghow could this be there if Belkin did not compile ther kernel themselves , how could the compile the kernel on this modified MIPS32 without the modified compiler
Also the vendors seem to be including their own open choice of packages/utilities this agin would require the modified compiler.
objdump from binutils compiled under gentoo (feel the power!!) with --enable-targets=all..... the --source option doesn't seem to work, I don't know if the objdump build is the problem or the compiled object.... or the fact I have just discovered... that the gcc compiler they are using has been modified.....
All in all this angers me.. broadcom used linux becasue it was cheaper that QNX or homegrown and it did exactly what they wanted it to do... they also had no intention of releasing user linux drivers even in binary form becasue of the risk of them being hacked (objdump) or to maintain a revenue stream....
The truth is that a) they are not playing fair on this one and b) they might never use linux again because of the hassle...... There is no doubt that they are making a bunch of money out of this 4710 chip and 99% of the code that is running on it is due to the hard work of us GPL'ers....
They deserve to be punished for this but if we even make a move towards reverse engineering the laws in the good 'ol US of A will have us hung, drawn and quatered before we can say shareholder profit....
I actually have a very strong view on what needs to be done..... ask me if you dare.....
Hi, you have recently open sourced the kernel supplied by broadcom for the WRT54G. Unfortunately after examination you have not included all of the source. In fact functions such as bcm_atoi, bcm_ctype, bcm_ether_atoe, bcm_ether_ntoa, bcm_mdelay, bcm_parse_tlvs and bcm_strtoul seem to be missing. In addition the specific BCM4710 memory management code, which is in the binary kernel, is not there. There are numerous other examples.
More concerning is the version of gcc used to compile the kernel, this identifies itself as "gcc version 3.0 20010422 (prerelease) with bcm4710a0 modifications" but no such modification have been submitted to GNU as far as I am aware. Could you please clarify this and, if I am right, correct it.
The delivery reciept gave up these names: > Your message
To: opensource@linksys.com Subject: Open source kernel for WRT54G Sent: Mon, 7 Jul 2003 11:07:50 -0700
was delivered to the following recipient(s):
Ben Sakai on Mon, 7 Jul 2003 11:09:23 -0700 MSEXCH:MSExchangeMTA:US:MAIL Vito Sansevero on Mon, 7 Jul 2003 11:09:23 -0700 MSEXCH:MSExchangeMTA:US:MAIL Chris Chapman on Mon, 7 Jul 2003 11:09:23 -0700 MSEXCH:MSExchangeMTA:US:MAIL
The driver on My Belkin is wl.o the Broadcom chip has an integrated MIPS3000 core so the binary driver is in MIPS assembler objdump wil disassemble it it is pretty easy to pick out the calls mades to the card....... We could do a simple open source driver in weeks......
Slashdot Mirror
Exactly was I was thinking... It'll be in a bios upgrade once they have 50% of the market, but at $525 they are going to have trouble against local dealers.
This clearly has nothing to do with aircraft performance but rather is designed to punish Airbus (EAS) for their 148% failure. An not in the industry but rather in the public perception which is a random and important part of aircraft success. I'm sure if one would the clock back to the launch of the 747 that the other players in that early market spread all sorts of rumors, it'll never fly, it's drop out of the sky, you'll suffocate etc etc. The 380 will sell so will the 787, both maiden commercial flights will be on slashdot.... both will crash at least once.
I'm supprised the Nokia 2110 didn't make the list, It was the true Rolls Royce of mobile phones. Maybe it was not released in the US.
I scanned the headline too quickly, thought it said "...from the Netherlands"..... I mused, "Ok they're fairly smart but all of that cheese and dope can't be too good for cognative function."
RTFT
Gentoo for PPC is better
here
well sort of
www.ibm.com/whois
I have known about this for a while, it confused me at first but i quickly realised how to turn it off....
Your Parental Control Subscription is NOT-ACTIVATED.
To stop PARENTAL CONTROL signup prompts, Click here:
The bit that worries me is the next part of the admin page.....
Reporting,
You are currently subscribed to reporting.
What the hell does that mean, what is it reporting and to whom?
I called their UK office first who basically told me to call the US, I got put through to a guy called Larry, he sounded like he was not just a phone jockey, he had me on speaker and I could hear his voice echo around his large office.
;-) and asked what I had to do to stay legal, and in his short reply he made three interestig points....
I acted dumb (not hard for me
1) They are only "attacking" the US market.
2) They are only going after corporate users.
3) This position will stay the same for a least a couple of months.
Is any of this new stuff?
i guess we would need to lean on one of the vendors or buys the BCM94710 evaluation kit or write an open letter to Broadcom denying that they do (lawyers don't like setting themselves us to be caught)
/home/hyin/cvs/BELKIN/ASKEY_31190/src/linux/linux/ include/linux/skbuff.h ghow could this be there if Belkin did not compile ther kernel themselves , how could the compile the kernel on this modified MIPS32 without the modified compiler
I would doubt if Broadcom could NOT distribute the compiler given the midifications that it is clear that the vendors are making
i.e. string run over the belkin and linksys kernels highlight that they are including modified header files from local directories e.g.
Also the vendors seem to be including their own open choice of packages/utilities this agin would require the modified compiler.
http://195.18.94.223/ls.str
objdump from binutils compiled under gentoo (feel the power!!) with --enable-targets=all..... the --source option doesn't seem to work, I don't know if the objdump build is the problem or the compiled object.... or the fact I have just discovered... that the gcc compiler they are using has been modified.....
All in all this angers me.. broadcom used linux becasue it was cheaper that QNX or homegrown and it did exactly what they wanted it to do... they also had no intention of releasing user linux drivers even in binary form becasue of the risk of them being hacked (objdump) or to maintain a revenue stream....
The truth is that a) they are not playing fair on this one and b) they might never use linux again because of the hassle...... There is no doubt that they are making a bunch of money out of this 4710 chip and 99% of the code that is running on it is due to the hard work of us GPL'ers....
They deserve to be punished for this but if we even make a move towards reverse engineering the laws in the good 'ol US of A will have us hung, drawn and quatered before we can say shareholder profit....
I actually have a very strong view on what needs to be done..... ask me if you dare.....
>Until/unless someone can demonstrate that they didn't publish any GPL source modifications,
errm read the latest posts, they have.... quite a bit of code is clearly missing
Just sent this:
Hi,
you have recently open sourced the kernel supplied by broadcom for the
WRT54G. Unfortunately after examination you have not included all of the
source. In fact functions such as bcm_atoi, bcm_ctype, bcm_ether_atoe,
bcm_ether_ntoa, bcm_mdelay, bcm_parse_tlvs and bcm_strtoul seem to be
missing. In addition the specific BCM4710 memory management code, which is
in the binary kernel, is not there. There are numerous other examples.
More concerning is the version of gcc used to compile the kernel, this
identifies itself as "gcc version 3.0 20010422 (prerelease) with bcm4710a0
modifications" but no such modification have been submitted to GNU as far as
I am aware. Could you please clarify this and, if I am right, correct it.
The delivery reciept gave up these names:
> Your message
To: opensource@linksys.com
Subject: Open source kernel for WRT54G
Sent: Mon, 7 Jul 2003 11:07:50 -0700
was delivered to the following recipient(s):
Ben Sakai on Mon, 7 Jul 2003 11:09:23 -0700
MSEXCH:MSExchangeMTA:US:MAIL
Vito Sansevero on Mon, 7 Jul 2003 11:09:23 -0700
MSEXCH:MSExchangeMTA:US:MAIL
Chris Chapman on Mon, 7 Jul 2003 11:09:23 -0700
MSEXCH:MSExchangeMTA:US:MAIL
worth giving them a call?
Hi, i've just ripped this out of the broadcom/linksys/belkin kernel
are these on GPL'd modifications a violation?
"Linux version 2.4.5 (hyin@cactus.localdomain) (gcc version 3.0 20010422 (prerelease) with bcm4710a0 modifications) #3 Wed Jun 4 17:12:54 PDT 2003"
it's a Broadcom BCM4702, integrated MIP3000 processor PCI 2.2....
i d= PR020910
Here's the rub.... it probably can do 802.11a also
http://www.broadcom.com/products/4702.html
here is the evaluation board
http://www.broadcom.com/cgi-bin/pr/prps.cgi?pr_
I don't get you... why wouldn't this work, it's hard yeah but not impossible....
:
here is the card reset function
000013f0
13f0: 27bdffe8 addiu sp,sp,-24
13f4: afbf0010 sw ra,16(sp)
13f8: 3c020000 lui v0,0x0
13fc: 24420000 addiu v0,v0,0
1400: 0040f809 jalr v0
1404: 8c840004 lw a0,4(a0)
1408: 8fbf0010 lw ra,16(sp)
140c: 00001021 move v0,zero
1410: 03e00008 jr ra
1414: 27bd0018 addiu sp,sp,24
The driver on My Belkin is wl.o the Broadcom chip has an integrated MIPS3000 core so the binary driver is in MIPS assembler objdump wil disassemble it it is pretty easy to pick out the calls mades to the card....... We could do a simple open source driver in weeks......
The driver is wl.o contained in the firmware i have disassembled it back into MIPS3000 assembler and I'm starting an open source driver tonight
Well the symbols anyway
http://195.18.94.223/symbols.html