<azonenberg>
For the short term, i merged my master a little bit so that you can now pull from master into it
<azonenberg>
from upstream master*
<azonenberg>
and it has all of our research work in it
<azonenberg>
I'm going to continue doing research in master and then make feature branches ex post facto we can send to upstream (and refactor as needed)
<balrog>
hey welcome back azonenberg
<azonenberg>
o/
<azonenberg>
back after a fashion
<azonenberg>
family is still here today, i leave tomorrow for a week onsite with a client
<azonenberg>
but will have internet and be working after business hours chicago time
<azonenberg>
(client isn't there, but same TZ)
<openfpga-github>
[yosys] azonenberg created greenpak-portfixes (+1 new commit): https://git.io/v53Fl
<openfpga-github>
yosys/greenpak-portfixes e623622 Andrew Zonenberg: Fixed bug causing GP_SPI model to not synthesize
X-Scale has joined ##openfpga
m_t has quit [Quit: Leaving]
pie__ has quit [Ping timeout: 248 seconds]
X-Scale has quit [Ping timeout: 248 seconds]
n00blet has joined ##openfpga
X-Scale has joined ##openfpga
theMagnumOrange has quit [Ping timeout: 246 seconds]
test123456 has joined ##openfpga
pie__ has joined ##openfpga
[X-Scale] has joined ##openfpga
X-Scale has quit [Ping timeout: 240 seconds]
[X-Scale] is now known as X-Scale
[X-Scale] has joined ##openfpga
X-Scale has quit [Ping timeout: 252 seconds]
[X-Scale] is now known as X-Scale
pie__ has quit [Ping timeout: 276 seconds]
X-Scale has quit [Ping timeout: 248 seconds]
digshadow has joined ##openfpga
X-Scale has joined ##openfpga
test123456 has quit [Remote host closed the connection]
test123456 has joined ##openfpga
digshadow has quit [Quit: Leaving.]
X-Scale has quit [Ping timeout: 248 seconds]
pie__ has joined ##openfpga
GenTooMan has joined ##openfpga
test123456 has quit [Remote host closed the connection]
test123456 has joined ##openfpga
X-Scale has joined ##openfpga
test123456 has quit [Read error: Connection reset by peer]
MrY has joined ##openfpga
MrY has quit [Remote host closed the connection]
MrY has joined ##openfpga
MrY has quit [Remote host closed the connection]
enriq has joined ##openfpga
enriq has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
X-Scale has quit [Ping timeout: 240 seconds]
n00blet is now known as n00blet_
n00blet_ is now known as o_0
o_0 is now known as n00blet
X-Scale has joined ##openfpga
<rqou>
wat
<rqou>
the "high level bitstream" for ice40 project is GPLv3+
<rqou>
not BSD like the rest of icestorm
<balrog>
rqou: which part?
<balrog>
looks like it'
<balrog>
looks like it's whatever was contributed by Roland Lutz
<balrog>
clifford: ^
<balrog>
I see these are recently merged PRs... not sure that was overlooked
<rqou>
hmm, the "asc2hlc" script is really slow too
<rqou>
about 10 seconds to process one file
<rqou>
it's kinda neat
<rqou>
er, except it's busted
theMagnumOrange has joined ##openfpga
<pointfree>
put my word in.
<balrog>
pointfree: that's not the problem here, the problem here is the way GPL snuck in
<balrog>
I'd be okay with modules being [L]GPL
<balrog>
or the whole thing even, but not like this :p
<pointfree>
balrog: agreed. There should be discussion before converting anything to GPL.
<pointfree>
like that.
specing has quit [Ping timeout: 240 seconds]
specing has joined ##openfpga
dingbat has joined ##openfpga
<balrog>
pointfree: I see he responded... anyway pushing license changes this way is generally seen as extremely rude and inappropriate
test123456 has joined ##openfpga
<cr1901_modern>
"The open source fpga tools are unlikely to get monentary support or future corporate contributions without any form of copyleft." I am... not sure how this follows.
<pie__>
its not gonna get any anyway \o/ :D
<pie__>
jk i hope it does gud work adoin
awygle has joined ##openfpga
<qu1j0t3>
cr1901_modern: because BSD is a black hole, there are almost zero obligations attached.
<cr1901_modern>
qu1j0t3: _without_ any form of copyleft?
<awygle>
to quote the zeromq guide, which i spent most of yesterday reading: 'BSD is like food. It literally (and I mean that metaphorically) whispers "eat me" in the little voice one imagines a cube of cheese might use.'
<awygle>
regardless of license debates, obviously sneak-changes like that are unacceptable
<qu1j0t3>
cr1901_modern: by which i understand to mean 'BSD-like'
<awygle>
qu1j0t3: i also don't really understand how "copyleft" leads to "corporate support". can you elaborate?
<qu1j0t3>
awygle: I just did.
<qu1j0t3>
i mean reading both licenses should make the difference crystal clear.
<cr1901_modern>
BSD isn't copyleft
<qu1j0t3>
cr1901_modern: Exactly.
<qu1j0t3>
cr1901_modern: that is the contrast i am making, in order to help illustrate why somebody might say that.
<cr1901_modern>
The open source fpga tools are _unlikely_ to get monentary support or future corporate contributions _without_ any form of copyleft
<qu1j0t3>
cr1901_modern: Yes, is it still unclear why somebody would make that statement?
<awygle>
i know what the difference is. my understanding is that most corporations prefer non-copyleft licenses (especially Apache, or "BSD in a suit").
<cr1901_modern>
That sentence reads to me as if it's saying "unless you have copyleft, you won't get monetary support"
<qu1j0t3>
awygle: yes they do, because there are practically zero obligations attached.
<awygle>
because corporations have lots of FUD around GPL-style licenses
<cr1901_modern>
which doesn't make sense
<qu1j0t3>
awygle: And yes, they do.
<awygle>
i agree with cr1901_modern, the sentence says "copyleft licenses lead to corporate money", which is the opposite of the argument you seem to be making.
<cr1901_modern>
That FUD is the exact reason I prefer GPL. Too bad I haven't written anything worthwhile lol
<awygle>
(not trying to be agressive, just trying to understand)
* qu1j0t3
also prefers GPL and has made reasonable amounts of money dual licensing.
* qu1j0t3
would never have seen a cent of that if i'd chosen BSD.
<awygle>
ohhhh i see. so you're referring to dual-licensing, rather than "apple funds LLVM" style corporate money
<qu1j0t3>
so, from where i stand, that statement is correct both in theory and reality.
<qu1j0t3>
awygle: both, really
<qu1j0t3>
awygle: although perhaps in one case, it's not mandatory
<cr1901_modern>
qu1j0t3: The sentence as is reads to me as it's saying "unless you have copyleft, you won't get monetary/commercial support". Which is markedly untrue. Copyleft repels people.
<qu1j0t3>
cr1901_modern: I can't agree. the license requirements are what matter here, not taste. If copyleft means a corporate looks for an alterantive product that's BSD, well, too bad
test123456 has quit [Quit: Leaving]
<awygle>
i mean let's work through the scenarios. it's pretty clear to me a Xilinx or a Synopsys is never going to support open source tools in any capacity. whereas someone like a Gowin or an Achronix might support regardless of the license. in a BSD scenario, to undercut Xilinx's licensing revenue. in a GPL scenario, to avoid having to pay to develop their own toolsuite (similarly to how they might pay Synopsys).
<cr1901_modern>
qu1j0t3: Oh, I think I get it now
<qu1j0t3>
if there is no alternative product, then you are in a much more favourable position
<cr1901_modern>
w/ BSD license, corporations will just take
<cr1901_modern>
w/ GPL, they'll contribute back, pay for a different license, or do nothing at all :P
<qu1j0t3>
cr1901_modern: yeah.
<awygle>
well you're leaving out "or pay for a commercial license for a closed source product, since now the open source nature benefits them not at all", or "or develop their own"
<qu1j0t3>
and if the copylefted code is expensive to re-engineer, you are likely to benefit in some way.
<qu1j0t3>
awygle: that can be more costly than dual licensing.
<qu1j0t3>
awygle: easily.
<awygle>
in the specific FPGA scenario, there are many incumbent closed source products, so if the options are "pay $$$ for GPL'd code" or "pay $$$ for commercial code" then it comes down to how many $s
<qu1j0t3>
of course the BEST case is not dual licensing, it's "oh hey, let's support open development."
<awygle>
i just don't think it's a total slam dunk in either direction
<qu1j0t3>
that is even cheaper for them, potentially.
<awygle>
true
<qu1j0t3>
awygle: Right, it's not, there are calculations involved, but the license differences are a huge factor, because BSD is a "take take take" short circuit.
<awygle>
if they can bring themselves to do it (which again, an upstart like Achronix has more likelihood of than an incumbent like Xilinx or Synopsys)
<awygle>
yeah i don't actually disagree with your conclusion, just your degree of certainty lol. i'm seriously considering GPL for my current work and will probably go that way.
<awygle>
thanks for the discussion, enjoyable and informative
<qu1j0t3>
:) ... glad I could help in any small way...
<awygle>
on a somewhat-related note - i'll be at the open source hardware summit in Denver in October, if anybody else will be there please get in touch
<qu1j0t3>
awygle: When I picked GPL for my work I had no idea dual licensing would eventuate.
<qu1j0t3>
awygle: that was years and thousand of hours later.
<qu1j0t3>
awygle: I just picked it on principle
<qu1j0t3>
thousands*
<awygle>
i know it's not a huge conference or anything, i just happen to have family in that area, but still thought i'd mention
<awygle>
qu1j0t3: actually i picked MIT on principle all through college and for a while afterwards on a knee-jerk rejection of copyleft :)
<awygle>
eventually i turned on my brain and realized that maybe i should think a bit more
<qu1j0t3>
yeah, i honestly don't understand that position, but I know it's popular.
<awygle>
i could go into my best guess at the philosophies behind each position but... i need to go buy groceries lol
<qu1j0t3>
on another topic, i had a lot of fun and success breadboarding today... http://imgur.com/TRiZIPy (ignore the Whovian cookbook, I don't watch the new Who)
<qu1j0t3>
awygle: yeah, i don't expect to persuade anyone, but i have those data points from my own experience
eduardo__ has joined ##openfpga
X-Scale has quit [Ping timeout: 255 seconds]
eduardo_ has quit [Ping timeout: 260 seconds]
<pie__>
<cr1901_modern> w/ BSD license, corporations will just take
<pie__>
<qu1j0t3> cr1901_modern: yeah.
<pie__>
<cr1901_modern> w/ GPL, they'll contribute back, pay for a different license, or do nothing at all :P
<pie__>
qu1j0t3, to be fair that was mostly implied in what you said and not very explicit
<pie__>
just my 2c
<pie__>
<awygle> in the specific FPGA scenario, there are many incumbent closed source products, so if the options are "pay $$$ for GPL'd code" or "pay $$$ for commercial code" then it comes down to how many $s
<pie__>
also support?
<qu1j0t3>
pie__: I tend to assume this is better known :-)
<qu1j0t3>
yeah, support is also a potential revenue generator for GPL code, and my dual licenses include that as a quoted item (though it wasn't really taken up)
<pie__>
qu1j0t3, i mean it was obvious to me but thats might be either because i know you better or something else, i just wouldnt have necessarily decided to assume mysefl, but now im makin unnecessary meta noise so eh. :D
<qu1j0t3>
to be fair, paid support is also a possibility for BSD code
<pie__>
my main problem with licensing so far on all the code i havent written is that id go probably GPL as a matter of principle but i want to make some money xD
<qu1j0t3>
re-read the above :-)
<balrog>
there's a reason azonenberg picked LGPL
<balrog>
:)
<balrog>
I'd probably go GPL too
<balrog>
I just find it rude to contribute code that sneaks in a license change
promach has quit [Remote host closed the connection]
<pie__>
more money for more lab equipment for more free things :D
<pie__>
free beer as in obtaining a crippling moral obliation to contribute to society in return >;P
X-Scale has joined ##openfpga
<awygle>
LGPL is an interesting case as well. i tend to round it off as "library license" but obviously that's lazy thinking.
<balrog>
awygle: it was renamed for that reason
<balrog>
it used to be called Library GPL
<balrog>
because RMS/GNU decided that GPL is more suitable for many libraries