<awygle>
the thing about darpa contracts is that most of the time they're sold and _then_ put out for bid, so i never put a lot of stock in that whole thing, but clearly the symbiotic eda people did
<cr1901_modern>
I was never told the the contract process is rigged during a class on project management taught by someone who works for lockheed :P
<awygle>
lol
<balrog>
from what I'm reading here it seems that they want to go well beyond FPGA
<balrog>
(and digital)
<awygle>
that seems to be true but there's no reason yosys couldn't be compatible with such a move
<balrog>
of course.
<balrog>
marcan: dunno if you noticed, Microchip is shipping a license-limited version of avr-gcc as part of XC8 (disabled optimization levels above O1 if you don't have a license), and hasn't provided source code
<marcan>
lololol they're still doing that shit?
<balrog>
oh they've been doing that, but now they took it a notch further, bringing it to AVR
<marcan>
le sigh
<balrog>
(also unrelated but shipped in the same package, their new C99 PIC18 compiler frontend is a hacked up Clang that generates their own PIC18 IR)
<balrog>
yeah, I brought it up in #fsf and someone there already reached out to them
<balrog>
as far as I'm aware avr-gcc isn't even their work -- the community contributed it way back
<balrog>
my guess is that as soon as they're done with atmel studio to mplab conversion tools they'll be phasing out atmel studio
<balrog>
and trying to force everyone to use xc8 for avr
<balrog>
(where you need an expensive or subscription license for any reasonable optimization)
<balrog>
I know that for XC16/XC32 (gcc derivative) they will not assign copyright to FSF / contribute upstream
<felix_>
not assigning copyright to another entity seems reasonable to me (i don't contribute to projects in my spare time that require that sort of stuff), but if they use gpl software they'll have to supply the source code if they ship the binary to customers and can't stop them from upstreaming that code...
<balrog>
felix_: FSF will not accept upstream contributions if copyright isn't assigned
<balrog>
for projects under their jurisdiction (GNU)
<balrog>
GCC, glibc, and others fall under this
wpwrak has quit [Ping timeout: 276 seconds]
wpwrak has joined ##openfpga
<felix_>
yep, i know. still their loss ;P
<felix_>
oh, that also makes it impossible to upstream things for anyone who's not the original author, right? o_O
<sorear>
see: openrisc
wpwrak has quit [Ping timeout: 276 seconds]
<Prf_Jakob>
felix_: Makes it impossible to upstream patches from anyone who hasn't signed the copyright assignment agreement.
<felix_>
uh oh :/
<Prf_Jakob>
IANAL, so I'm not sure if I work for company A and they have signed, I can contribute by virtue of my employment contract probably having a assignment clause as well.
Miyu has joined ##openfpga
ayjay_t has quit [Remote host closed the connection]
rohitksingh has joined ##openfpga
iximeow_ is now known as iximeow
<marcan>
yeah this is the FSF's loss, really
balrog has quit [Ping timeout: 256 seconds]
<marcan>
CLAs are evil
balrog has joined ##openfpga
azonenberg_work has joined ##openfpga
<digshadow>
marcan: if it stays open source, what is your main objection?
ayjay_t_ has joined ##openfpga
<marcan>
digshadow: it stifles contributions, then you're left with the choice between dealing with the CLA or forking and fragmenting the community
<shapr>
I see both sides of the issue, I don't know the answer.
<digshadow>
is CLA difficult to submit? or are people just morally opposed to it?
<marcan>
depends on the CLA
<marcan>
I just don't like the friction
<marcan>
and it gives the owner carte blanche to violate the terms of the nominal open source license
<marcan>
(for assignment CLAs anyway, which most are, otherwise why need a CLA?)
<marcan>
pick your license well and you don't need a CLA
<digshadow>
think in that case they decided not to accept outside contributors to avoid CLA
<felix_>
only reason i see for a CLA is that the entity wants to be able to change the license without having to ask the contributors for permission and that's something i strongly dislike
<azonenberg_work>
felix_: yeah thats the only reason i see for them
<digshadow>
in the case of m-stack, its because development is also supported by selling commercial licenses. At least personally I'm not fundamentally opposed to that if that funds it being open
<iximeow>
felix_: that's generally the reason i've seen given when i ask, yeah
<felix_>
hmm, i wonder if a project having a dual license (e.g. gpl and paid non-gpl) causes problems with external contributors
<digshadow>
well in m-stacks case, it forcibly did, but I'm sure it does in general as well
<digshadow>
thinking about FSF, the main reason why I would see them CLA is to change to a later GPL version, although generally things say "GPL version X or later"
<marcan>
yeah, the funny thing is the FSF *already* has relincensing power over not just their software, but 90+% of the world's GPLed software
<marcan>
*relicensing
rohitksingh has quit [Quit: Leaving.]
<marcan>
(which IMO is ridiculous, which is why I try to avoid the "or any later version" wording)
<azonenberg_work>
digshadow: They don't want "or any later version"
<azonenberg_work>
The FSF wants the ability to move things to be GPLv4+ only
<felix_>
.oO(register gpl4.org, put the 2 clause bsd or apache 2 text on it and call it gpl4) *noise of lawyers imploding* ;P
<awygle>
Exactly, CLAs are solely so you can keep selling commercial licenses, still take other people's work, and prevent them from selling commercial licenses.
<azonenberg_work>
and deny the ability to use it under (say) GPLv2 rules
<awygle>
In the one case where they could be used for good, legally risky code, they... Aren't.
<awygle>
GPL is almost certainly trademarked, no?
<marcan>
azonenberg_work: but they can *already* do that
<marcan>
without any CLAs
<azonenberg_work>
marcan: can they?
<cr1901_modern>
How does a CLA prevent someone from making a fork, NOT contributing changes back, and the fork selling commercial licenses?
<marcan>
you can arbitrarily further restrict licensing on a multiply licensed work after a subsequent contribution
<azonenberg_work>
Hmm, i guess?
<marcan>
and they can't retroactively change licenses for old versions, CLA or no CLA
<azonenberg_work>
cr1901_modern: Because the fork doesnt own the copyright
<awygle>
cr1901_modern: the fork won't hold copyright, so they can't relicense, so they can't make a commercial license
<azonenberg_work>
so they can't release it under non-GPL terms
<azonenberg_work>
They can sell commercial support services etc
<cr1901_modern>
I meant assuming the fork keeps the source open under the same license
<azonenberg_work>
But they cant authorize a third party to use it as a binary blob
* cr1901_modern
INAL
<azonenberg_work>
Which is really the point of commercial licenses
<azonenberg_work>
You're buying a "don't release my proprietary source" option
<digshadow>
azonenberg_work: m-stack has apache license, which seems like it would be okay for most commercial projects
<cr1901_modern>
Oh b/c GPL is invasive?
<awygle>
digshadow: what I said above about CLAs for risky code applies directly to Project X-Ray. using a DCO is transparently a move to protect The Patron. a DCO might as well say "exhibit A in Xilinx V. Everybody But The Patron".
<awygle>
even if that's not the actual intent, that's the vibe
<digshadow>
awygle: I was not involved in licensing / DCO, all of my projects are BSD licensed
<marcan>
DCO isn't really a CLA
<marcan>
it's just some CYA legalese
<marcan>
where I say CLA I mean copyright assignment or near-equivalent CLAs
<awygle>
marcan: that's my point, this is a case where a CLA would be a good thing, at least arguably
<azonenberg_work>
cr1901_modern: exactly
<marcan>
Linux requires a DCO
<marcan>
it's literally a line in the commit message
<marcan>
that's fine
<marcan>
it doesn't grant any extra rights to anyone special
<marcan>
it just says "I'm not violating any licenses or copyright by submitting this"
<awygle>
digshadow: I am not accusing you personally of anything, just conveying my feelings to you as a representative of the group.
<awygle>
marcan: it's saying "I did this". which implicitly says "$COMPANY had nothing to do with this". So if the code is challenged in court, $COMPANY is safe.
<awygle>
Often that's not a big deal. Sometimes it is.
<marcan>
which CLA/DCO are we talking about exactly?
<marcan>
it's literally just legalese saying "I'm not breaking any copyright laws"
<marcan>
you're allowed to submit stuff on behalf of a company
<marcan>
people do that all the time
<marcan>
you're allowed to submit stuff other people wrote
<marcan>
as long as the licensing checks out
m_w has joined ##openfpga
<awygle>
"The contribution was created in whole or in part by me"
<marcan>
;or
<awygle>
yes okay
<awygle>
that's not the point. the point is it's disingenuous for a company to support a project in secret and make contributors attest to their participation simultaneously
<awygle>
*publicly attest
<azonenberg_work>
awygle: i think that is the whole rationale behind it being a "black project" @ $SPONSOR
<marcan>
I don't see how the DCO for that project makes a difference to anything
<marcan>
the license is permissive to begin with
<awygle>
it's not about permissive licensing, it's about being sued by Xilinx
<azonenberg_work>
Plausible deniability if Xilinx ever tries to sue somebody over it
<awygle>
azonenberg_work: right, which is scummy
<awygle>
imo
<azonenberg_work>
My guess is that they agreed to sponsor the work as long as they werent publicly attached to it
<azonenberg_work>
at which point you can either take the money or leave it
<awygle>
you can plausibly argue that the work existing is a higher social good which overrides the scumminess
<azonenberg_work>
And each contributor has the option to either join the project at their own risk, or not
<azonenberg_work>
I chose not to
<awygle>
not sure I agree with that view, but I'm fundamentally not involved in any of this so oh erll
<awygle>
*well
<azonenberg_work>
Incidentally, that kind of lines up with my general thoughts on licensing etc
* cr1901_modern
can't muster outrage for this and isn't quite sure why
<azonenberg_work>
The reason I go non-copyleft is because I am totally OK with a commercial fork of my code
<awygle>
azonenberg_work: for *paid* contributors, I'm with you. for soliciting unpaid contributions, no.
<azonenberg_work>
And I'm totally OK with it competing with my free, open version
<marcan>
I'm honestly still not entirely sure what happened that I should be outraged about
<marcan>
all I see is a repo with a DCO :P
<awygle>
yeah we went through this. you want a scope, rqou wants to be famous for making a scope, I want to live in a world where scopes exist
<marcan>
I'm missing backsotyr
<azonenberg_work>
Users have the ability to choose between extra features or seeing the source code
<marcan>
*backstory
<azonenberg_work>
i.e. in general i care more about freedom of developers
<cr1901_modern>
"a scope"?
<azonenberg_work>
whereas gnu seems to care more about freedom of users
<awygle>
cr1901_modern: starshipraider
<cr1901_modern>
oh lol
<awygle>
we were discussing our different motivations for such a project last time I saw them both
<awygle>
I may be slightly slandering rqou here, in which case I apologize
<awygle>
Basically azonenberg_work wants maximum liberal license, rqou wants attribution, and I want copyleft
<digshadow>
I want maximum impact
<awygle>
ENODEFINITION. what do you mean by impact?
<cr1901_modern>
B/c I'm an idealist who never learns, I still believe in GPL. Of course, since software ate the world, rolling your own when you hate the GPL is what ppl do now.
<cr1901_modern>
Also, I can legitimately think of use cases (emulators) where you genuinely don't want commercial companies using your work
<digshadow>
ultimately I want better EDA tools because the commercial ones aren't very good. Open source seems a pretty good way to get there because it will allow people to play around with things. Primarily I care about what will grow the community the most, licensing is derived from that
<awygle>
Okay, that sounds similar to the motivations I ascribed to azonenberg_work earlier. Although I'd argue that GPL without CLA would prevent a company from taking your work and closing it again, which seems useful to me.
<awygle>
I find trying to understand everyone's motivations for licensing endlessly fascinating.
<digshadow>
VPRs license allowed it to essentially become quartus. I view that as a success
<azonenberg_work>
awygle: yeah, whereas I say that if they take the work and close it, people will just continue to use the open version unless the closed one has significantly better feature sets
<azonenberg_work>
In which case, people can choose based on merit
<awygle>
yep and I disagree with both of you, which is fine
<awygle>
for the record, quartus isn't a win for me because they've captured the value created by the VPR devs and profited from it without passing that profit along to the VPR devs.
<awygle>
anyway. back to work.
pie__ has quit [Ping timeout: 240 seconds]
<digshadow>
thats not completely accurate. They've maintained a reasonably close relationship and theres been some cross pollination between VPR and quartus over the years
<awygle>
it's not a contractual relationship though, at best it's charity. and it sounds like you're talking about code, while I'm mostly talking about money.
<digshadow>
gotcha
<awygle>
👍
ayjay_t_ has quit [Read error: Connection reset by peer]
ayjay_t has joined ##openfpga
<rqou>
having _been through_ and having helped perform a CLA/relicensing, I'm never going to do that again
<rqou>
(going from "herp derp" straight to CLAs with the associated political bullshit)
<balrog>
marcan: the FSF abusing "or later" would not end well for them imo
<rqou>
except if you ask certain people they have already done that
<balrog>
FSF claims that they want copyright assignment because it makes it easier for them to enforce
pie__ has joined ##openfpga
<shapr>
I think that's true
<shapr>
oh wait, my opinion is uninformed, I don't have any law training
<shapr>
I guess that sounds reasonable
<shapr>
ok, that's closer to reality
pie__ has quit [Remote host closed the connection]
<rqou>
from my own experiences (including with certain people in this channel), i absolutely will not link any of the projects i care about and control with "business-y" things like CLAs or a sponsor
<rqou>
these things never end well
<rqou>
whether it's unnecessary friction or developers acting maliciously
<balrog>
I'd take a reasonably-permissive license over CLA
<balrog>
however when you're working with an upstream like GNU (for GCC), you may have to deal with CLA
<rqou>
BSD-style permissive licenses with warranty clauses give me the attribution and protection i want/feel i need
<balrog>
personally I'd probably use GPL, because "GPL without CLA would prevent a company from taking your work and closing it again"
<rqou>
yeah, fortunately i haven't had to deal with gcc yet
<rqou>
i don't care about that scenario
<daveshah>
Personally I'm very happy with ISC
<balrog>
I do because it happens all too often
<daveshah>
Nice and simple without going all the way to PD type stuff
<rqou>
as long as i get attribution I don't care very much
<rqou>
anyways, again, i _helped supervise_ going from "herp derp" to CLAs
<rqou>
it __sucked__
nrossi[m] has quit [*.net *.split]
Wallbraker[m] has quit [*.net *.split]
eddyb has quit [*.net *.split]
_florent_ has quit [*.net *.split]
Zorix has quit [*.net *.split]
feuerrot has quit [*.net *.split]
stefanct has quit [*.net *.split]
kmehall has quit [*.net *.split]
lb5tr has quit [*.net *.split]
<balrog>
rqou: what was the "herp derp", and what was the motivation?
eddyb has joined ##openfpga
lb5tr has joined ##openfpga
_florent_ has joined ##openfpga
stefanct__ has joined ##openfpga
<rqou>
herp derp ~= it should be (as in, everybody wanted it to be) open-source but nobody checked all the boxes to actually make it such
<balrog>
oh, it wasn't open source?
nrossi[m] has joined ##openfpga
<rqou>
as for motivation, I don't want to discuss that in public right now
Wallbraker[m] has joined ##openfpga
<rqou>
it was "a group of people put some code together intending it to become open source"
<rqou>
but it didn't have license headers or proper attribution for code we borrowed
Zorix has joined ##openfpga
<balrog>
ah :/
<balrog>
yeah that is a pain to deal with
eddyb has quit [Changing host]
eddyb has joined ##openfpga
eddyb has joined ##openfpga
kmehall has joined ##openfpga
jfng has quit [*.net *.split]
nurelin has quit [*.net *.split]
shapr has quit [*.net *.split]
shapr_ has joined ##openfpga
feuerrot has joined ##openfpga
jfng has joined ##openfpga
nurelin has joined ##openfpga
pie_ has joined ##openfpga
<rqou>
honestly part of the reason this experience sucked was how much certain people involved wanted the group to be very "corporate"
<rqou>
wow, doesn't that sound familiar?
<prpplague>
rqou: did you get the right cover sheet for your TPS report?
<rqou>
ok, not that corporate; we're still in the bay area
wpwrak has joined ##openfpga
stefanct__ has quit [Quit: quit]
stefanct has joined ##openfpga
shapr_ has quit [Changing host]
shapr_ has joined ##openfpga
shapr_ is now known as shapr
<pie_>
i missed most of the conversation but azonenberg_work also has a well paying side-job :p
* pie_
still hung up on whether he will be able to make enough moneez on open sauce to not starve
<pie_>
*part time
<pie_>
whats a DCO?
<jn__>
developer's certificate of origin, i suppose
<felix_>
only complaint i heard on those was that some people were unhappy that using a pseudonym instead of a "real" name isn't allowed
<pie_>
oh damn :P well I suppose thats understandable
<pie_>
hard to sue a pseudonym :p
[X-Scale] has joined ##openfpga
X-Scale has quit [Ping timeout: 256 seconds]
gnufan1 has joined ##openfpga
[X-Scale] is now known as X-Scale
<rqou>
i don't see the need for a DCO, but i suppose it's fine as long as nobody tries to police "real names"
zino has quit [*.net *.split]
mietek has quit [*.net *.split]
marex-cloud has quit [*.net *.split]
pakesson has quit [*.net *.split]
JSharp has quit [*.net *.split]
mietek has joined ##openfpga
pakesson has joined ##openfpga
JSharp has joined ##openfpga
zino has joined ##openfpga
mietek has quit [Changing host]
mietek has joined ##openfpga
marex-cloud has joined ##openfpga
JSharp has quit [Changing host]
JSharp has joined ##openfpga
<awygle>
I agree with rqou on DCOs most of the time
<awygle>
the argument up thread is based on fairly unique circumstances
<rqou>
yeah i agree that a certain person here has been acting really scummy
dfgg has quit [*.net *.split]
kuldeep has quit [*.net *.split]
Ultrasauce has quit [*.net *.split]
mIKEjONES has quit [*.net *.split]
asy has quit [*.net *.split]
mIKEjONES has joined ##openfpga
asy_ has joined ##openfpga
<awygle>
I have no comment on any individual person, just the corporate sponsor. whatever I may privately think, individuals can do what they want and I can choose to work with them or not
pie_ has quit [Ping timeout: 265 seconds]
dfgg has joined ##openfpga
gnufan1 has quit [Ping timeout: 268 seconds]
gnufan1 has joined ##openfpga
<shapr>
I LIKE OPEN SOURCE CODE
<shapr>
that's my strong opinion
<shapr>
plz don't make me fail at installing the Xilinx tools again
<shapr>
so much frustration
rqou_ has joined ##openfpga
zkms has quit [Ping timeout: 256 seconds]
kuldeep has joined ##openfpga
Ultrasauce has joined ##openfpga
rqou has quit [Ping timeout: 256 seconds]
rqou_ is now known as rqou
<felix_>
if you're running fedora, make sure to have ncurses-compat-libs installed; otherwise the installer just hangs withour an error message. well at least last year that was a problem; haven't installed a much newer version yet
zkms has joined ##openfpga
pie_ has joined ##openfpga
uovo has joined ##openfpga
oeuf has quit [Ping timeout: 272 seconds]
zino has quit [Quit: Leaving]
zino has joined ##openfpga
Hamilton has joined ##openfpga
GenTooMan has joined ##openfpga
Hamilton has quit [Quit: Leaving]
azonenberg_work has quit [Read error: Connection reset by peer]
<awygle>
i hate tweets like this. it relies on you already holding the same opinion as the author. _why_ does this press release thing mean CLAs are bad? it doesn't mention a CLA at all! (i assume redis has one and that's why they were able to do this but nothing in the tweet or blog says that)
<iximeow>
(it does mention a CLA in like hte last sentence, though i broadly agree with you - i don't understand what's bad about their CLA in that context. though i didn't read it)
<awygle>
okay i take it back "Contributor License Agreement" is the last 3 words of the press release
<iximeow>
:P
<awygle>
read before tweeting^WIRCing, wygle.
<azonenberg_work>
awygle: whats funny is that this doesnt actually change anything
<awygle>
but still, no explanation just "LOOK AT THE BAD" really pisses me off
<azonenberg_work>
What it means is, you can't make money selling a pre-provision redis server
<azonenberg_work>
you have to sell a blank linux server and make the end user install redis instead
<azonenberg_work>
this benefits the community how?
<awygle>
"including without limitation fees for hosting or consulting/support services"
<azonenberg_work>
it benefits the redis devs how?
<awygle>
is the real money shot probably
<awygle>
"wanna use redis commercially? fuck you pay us"
<q3k>
CLAs on permissive copyleft are acceptable if a bit superfluous
<azonenberg_work>
awygle: my reading of their goal is to say "redis as a service" is prohibited
<q3k>
CLAs on GPL is evil
<q3k>
oh shi,t I was scrolledup. neverming.
<awygle>
lol
<azonenberg_work>
But "redis plus your code as a complete system" is legal
<q3k>
*nevermind.
<q3k>
ah, no, we're still on CLAs
<azonenberg_work>
i.e. they STILL dont take a profit
<q3k>
MY POINT STANDS
<awygle>
still directly relevant to the conversation, sadly
<azonenberg_work>
awygle: IMO, such licensing is how you get your code forked/reimplemented
<pie_>
can you agree to a CLA without actually signing any papers?
<awygle>
if you find yourself wanting to put a CLA on your GPL code, just don't accept any contributions
<azonenberg_work>
if google/amazon decides they dont like software X's new license they're going to build a replacement
<azonenberg_work>
They're not going to pay license fees
<awygle>
azonenberg_work: google and amazon are not ever customers
<azonenberg_work>
Then google open sources their version and redis dies :p
<q3k>
google is likely to fork your code even if you pay them to use yours though
<awygle>
but there are thousands of smaller businesses that could use redis
<q3k>
because of how different their usecases usually are compared to the rest of the world
<awygle>
and they're unlikely to invent it themselves because ^
* awygle
wonders how today became Mad About Licenses Day
<iximeow>
if it's not Mad About Licenses Day it's Mad About Underfunded OSS Ecosystems Day, take your pick
<q3k>
that's everyday
<awygle>
that's my secret - every day is both of those days
<q3k>
awygle: hi5
<rqou>
pie_: in the specific case i was involved in it DID involve actually signing some dead tree
<rqou>
yes, it sucked as much as you might imagine
<awygle>
i suspect the general answer to pie_'s question is "depends on the judge"
<awygle>
as is often the case with legal stuff
<pie_>
do any of you know someone that has actually managed to sell open source software?
<pie_>
im constantly hung up on this but im still coding my thing
<jn__>
pie_: redhat?
<rqou>
openssl successfully sold fips compliance consulting, does that count?
<pie_>
either i OSS it or buy a 500$/month qt license, but i mean, i *want* to open source it anyway
<jn__>
pie_: or rather: define "sell"
<q3k>
pie_: depends on how you phrase it
<pie_>
rqou, i dont want to do consulting lol...
Bike has joined ##openfpga
<q3k>
pie_: hashicorp sell commercial/extended versions of the oss offering
<q3k>
pie_: gitlab does the same and on-prem
<pie_>
im writing a desktop app, so end user
<pie_>
well i guess most of those wouldnt really care about source or no source either way
<q3k>
pie_: a lot of devops/sre sort of products are marketed this way - get the f/loss version for free and deploy it and integrate it yourself, or pay us to do it aaS and give you LDAP integration
<q3k>
pie_: krita does that
<q3k>
pie_: although apple fucked them over
<jn__>
oh, Ubuntu and Ardour (a DAW) tell you to pay money when you download their stuff
<pie_>
wait i thought krita was just totally free
<q3k>
pie_: ie you could pay for krita on the app store even though it was free/gpl
<pie_>
apple fucked krita over?
<pie_>
ah
<q3k>
pie_: but then apple insisted they can't link to the free version / source / homepage because that's against their policy
<pie_>
i mean i know i can technically sell download access even if it was gPL
<pie_>
q3k, huh.
<q3k>
yeah, and they made quite a bit of money off of that
<jn__>
Ardour's model is: If you get the binaries, you pay; if you compile from source or use $DISTRO, you don't pay
<q3k>
because turns out people are okay paying you a few dollars if they get it updated and easily
<rqou>
in totally unrelated news, some coworkers are/were struggling with the perennial windoze issue of "UTF-16?!"
<q3k>
especially rich mac users
<pie_>
q3k, right...guess i should buy a mac and use cocoa or whatsit for them rich mac users...
<rqou>
imagine how much developer effort Microsoft continually wastes everywhere downstream
<pie_>
q3k, i was thinking a phone version miiiight be viable
<rqou>
by going the opposite direction of the rest of the world all the time
<q3k>
rqou: rumor says the A versions of WinAPI is scheduled to become UTF-8 compatible soon
<q3k>
*are scheduled
<iximeow>
q3k: no kidding? that would be nice
<pie_>
q3k, i should probably do some more market research
<rqou>
i thought msft repeatedly said that was impossible?
<awygle>
lots of people "sell OSS" by selling a hosted version
<awygle>
oops q3k said that
<awygle>
i should just let q3k handle these discussions, we are clearly simpatico
<q3k>
awygle: despactito 2: sympatico
<q3k>
awygle: (sorry, I'm tired)
<awygle>
alexa, play sympatico
<awygle>
i'm tired too, i think it's the smoke. i slept fine and it's not quite 4pm here.
<rqou>
q3k: I can't find the reference right now, but i thought msft insisted that there was no way *A apis could become utf-8 because of assumptions that one "character" is at most two bytes?
<rqou>
iirc they basically explicitly said "too late, deal with it"
<q3k>
rqou: yeah, rumour. also can't find the source right now anymore.
<q3k>
awygle: i'm on a 1500kcal diet and also tried bouldering. almost crashed and fell asleep at the gym
<rqou>
interesting rumour for sure
<rqou>
i guess too many people decided to "deal with it" by ignoring it and ignoring windows :P
<awygle>
q3k: damn, good luck. 1500 kcal is way below my TDEE without bouldering
<awygle>
(bouldering is fun tho)
<rqou>
e.g. i thought i could just ignore the problem and just do "opaque bytes from argv go into fopen" but apparently whitequark told me this doesn't actually work
<q3k>
awygle: thank you. and bouldering is super fun.
<q3k>
awygle: especially now that I make myself go at least twice a week again, i'm back to making progress
<awygle>
nice :) i should do that now that my foot is better
<awygle>
$$ tho :( which is weird cuz it's just a big rock
<q3k>
so for indoor bouldering here I pay 8eur/entrance
<q3k>
which is acceptable
<q3k>
(assuming you buy a bundle of 10 entrances)
<awygle>
I think the local place is like 80$/mo?
<q3k>
similar pricing then, assuming you go twice a week
<q3k>
the problem with the place I go to it's that it's somewhat far away and the public transport here sucks
<q3k>
50 minutes by bus usually :/
<rqou>
that's called "sucks?"
<awygle>
yeah the problem is they don't offer a non-monthly thing
<awygle>
and my visits tend to be... sporadic
<rqou>
I'd call that "perfectly average" bus speed here :P
<awygle>
especially since they're outside biking range
<q3k>
rqou: well, sucks compared to most european public transport
<rqou>
:P
<q3k>
rqou: being not-worse-than-US-west-coast is the lowest bar in the world
<rqou>
lol
<awygle>
on the plus side i bike past two regular gyms on my way home so that's pretty easy. can swim, lift weights, etc.
<rqou>
just for reference: my home<->work is 2 hours by public transportation
<q3k>
rqou: lolwut
<q3k>
2hrs one way?
<rqou>
yeah
<q3k>
how far away is it?
<rqou>
20 miles
<q3k>
(narrator: that's 30 kilometers)
<rqou>
it's normally about ~45 minutes by car with traffic
<q3k>
that is pretty terrible
<q3k>
although, I used to do 16km home/school in warsaw, and that took 40 minutes by bus
<q3k>
so not too far off
<awygle>
my home<->work used to be 17 miles by car, took 45 minutes in the morning and 90 minutes at night by bus
knielsen has quit [Read error: Connection reset by peer]
<awygle>
which is 27km
<awygle>
ish
knielsen has joined ##openfpga
<awygle>
so it's better here than the bay
<rqou>
don't forget that my home<-> work route passes through what fiora described as a nice transit wall of darkness
<awygle>
bay bridge et al?
<rqou>
east bay <-> silicon valley
<awygle>
oh yeah ok
<awygle>
well regardless i now live 1.5 miles from work and it's great :D
<rqou>
e.g. Fremont<->Berkeley is a less unreasonable 50 minutes for 40 miles
<awygle>
(rent? what rent? i don't pay rent, i just only get paid one time a month)
<rqou>
q3k: fwiw there is now finally a serious proposal (although much delayed) to fix this particular transit wall of darkness
<rqou>
although getting the bay area to actually operate sanely is basically an impossible effort
<q3k>
rqou: for a week i did sunnyvale -> goog mtv campus via 101 onto amphitheather pkwy
<awygle>
enjoy "coming in 2035" signs
<q3k>
rqou: that was living hell
<q3k>
rqou: fuck that shit
<rqou>
looooool
* awygle
waits anxiously for the train to reach redmond
<awygle>
ugh 2024. why do trains take so long to build
<rqou>
<also insert fiora cursed not-fallout-zones map of the bay area here>
<awygle>
also why can't they open the stations one at a time from the current transit instead of waiting til the whole thing is done
<awygle>
these are both serious questions, idk if there are good reasons or if it's just bad planning
<rqou>
why can't we just ask MTR corporation run US transit? :P
<awygle>
is that boston?
<awygle>
oh no it's hk duh
<rqou>
although hk has a housing crisis of their own but at least transit works
<iximeow>
a world where a 5 mile drive is 30+ minutes
<azonenberg_work>
iximeow: i live 15 miles from work
<azonenberg_work>
You know how long it takes me to get there/
<iximeow>
i feel like i don't want to know. <1hr?
<azonenberg_work>
A 10 minute bike ride, a ~30 min bus ride
<azonenberg_work>
transfer delay
<azonenberg_work>
35 minutes on a ferry
<azonenberg_work>
then another 25ish mins on the bike
<iximeow>
i was ready to say "that's not so bad!", naively assuming your first line was the whole thing
<azonenberg_work>
Lol
<azonenberg_work>
oh and that doesn't count the ~5 min drive from the hotel i sleep at
<azonenberg_work>
to the house my bike is parked at
<azonenberg_work>
So i have to leave the hotel maybe 7:25 or so to make the 7:50 bus
<azonenberg_work>
Which puts me on the 8:50 boat
<azonenberg_work>
Which puts me in the city about 9:20
<azonenberg_work>
And at the office shortly before 10
<iximeow>
so.. you work remote often? :D
<azonenberg_work>
Yes :p
<azonenberg_work>
The current apocalyptic weather doesnt help
<azonenberg_work>
I have to go to the office tomorrow and i'll be wearing my 3M 6300 with P100 + nuisance organic filters
<azonenberg_work>
plus my usual bike helmet and safety glasses
<awygle>
yeah i got an N95 filter and mask
<awygle>
huge help
<azonenberg_work>
awygle: yeah N95 is probably adequate for the smoke but I dont have any N95 cartridges and I have P100s
<azonenberg_work>
So why not
<azonenberg_work>
i gues the one plausible downside is more friction from the finer filter = slightly more inhalation effort
<awygle>
yeah i mean if i already owned P100 i'd have used that
<awygle>
but N95 was chaper so that's what i bought
<azonenberg_work>
Yeah
<azonenberg_work>
Either one will be a loooot better than nothing
<azonenberg_work>
The main thing is, cartridge respirators with silicone face pieces are vastly more pleasant to wear for a long time than the disposables
<awygle>
true. but disposables are cheaper :p and my commute is only 15m or so