<beneroth>
any significant differences to debian beside systemD ?
<aw->
beneroth: that's all
<aw->
it actually performs apt-get updates from the debian repos haha
<beneroth>
haha, nice
<beneroth>
good enough
<Regenaxer>
T
<beneroth>
systemd is like uefi
<beneroth>
aw-, nice board
<beneroth>
mobile server
<aw->
yeah for $50, should be good
<beneroth>
the pinebook looks also not bad
<beneroth>
for programming
<aw->
yes indeed, but I have a laptop, don't need another one
<beneroth>
I have a aging thinkpad
<aw->
thinkpad? indestructible
<beneroth>
T
<Regenaxer>
Indeed! I have one from 2005 still running fine
<beneroth>
the series got worse though, afaik
<aw->
Regenaxer: great purchase
orivej has quit [Ping timeout: 268 seconds]
<aw->
i deeply regret buying a Mac, it's really shit and overpriced
<beneroth>
yeah that was the joke: close the thinkpad, hit the other guy with it, open it and go back to work
<aw->
haha beneroth, like my keyboard, we're ready for zombie apocalypse
<beneroth>
I thought mac is the best hardware next to old thinkpads?
<beneroth>
aw-, hahah, good so
<aw->
beneroth: no, by far no
<aw->
the selling point is the "It just works"
<cess11>
Mac was good in the nineties and a short while after leaving PPC for PC architectures, or so I've been told.
<aw->
as long as you play by Apple's rules, but within a few years the Mac falls apart
<beneroth>
well I don't like the software. it's crippled (and often outdated). it just works because it doesn't do much. good enough for many people.
<aw->
i've already had to change my battery, power adapter, keyboard/trackpad, camera, HDD
<beneroth>
but I believed their hardware is better fitted together and stock stuff
<beneroth>
ok, so like everything beside the mainboard
<aw->
well you can't change anything yourself, except the HDD
<beneroth>
yeah that is BS
<aw->
so every out of warranty repair is crazy expensive
<aw->
they're not designed for long-term
<aw->
and even if you keep it for 5 years, then you're **forced** to follow their software updates
<beneroth>
especially this newer trend of glueing the battery into the devices (also with mobiles). planned obsolescence
<aw->
yes we've discussed this before
<beneroth>
T. forgot.
<aw->
it's a big issue, particularly when you pay $1500 for the device
<beneroth>
ok, so you say mac hardware (running other OS) is not feasible?
<aw->
hehe i've heard horror stories, i wouldn't try
<cess11>
I have a ZenBook that has been running almost like a server since -11-13 sometime. ~800 euros.
<aw->
a lot of people run MacOS and then Linux in a full-screen VM
<beneroth>
I've heard it's the best hardware to run windows xD. well in hacker community I usually only see thinkpads and macs.
<aw->
which is shitty because you can't use 100% of the RAM or CPU
<beneroth>
I never had mac hardware. worked with macs in art school 10 years ago and from time to time when helping friends.
<beneroth>
T
<beneroth>
well apparently the mainstream devs are on mac.. because they're not brave enough for linux but want some compatibility to the linux servers they have to work with. that is apparently also one of the main reasons for the windows linux subsystem.
<aw->
yeah that's fine
<beneroth>
I find it stupid. but yeah.
<aw->
well the Mac lets you focus on other stuff, so it's good if you don't care about actually controlling your hardware/software
<beneroth>
T. thought the gnu tools are all outdated. writting to ntfs is still a pain/not possible.
<aw->
cess11: awesome
<beneroth>
anyway. you make/get any casing for your rock64? or initially just to play around?
<aw->
beneroth: gonna try to fit it in in raspberry pi case
<beneroth>
I would like to build a arm64 DB server with a couple of disks
<beneroth>
I see
<aw->
same form factor but I think the power port is different
<aw->
beneroth: regarding that Pinebook, if I were to buy a new laptop, that would probably be my choice
<beneroth>
yeah, should be good enough for work & internet. but I also don't need a new laptop :)
<cess11>
Are there arm64 servers with support for plenty of RAM yet?
<aw->
although apparently it can't be used to watch youtube
<beneroth>
what do you call plenty, cess11 ?
<cess11>
32 GB or more.
<aw->
hmmm cess11 i don't think so
<beneroth>
wehh the pine A64 board supports up to 128GB
<beneroth>
and how much is it? can't find it, neither on the rock
<beneroth>
much ram would be interesting for a strongly-caching pilDB
<aw->
2GB RAM on the pinebook
<aw->
4GB on the ROCK64
<beneroth>
2 is really a bit low. 4 on the rock is okay, good enough for running some dozen normal pil apps.
<aw->
cess11: apparently there are servers with 64-bit ARM with +32GB RAM.. but i dont know who makes them
<aw->
maybe Supermicro?
<aw->
beneroth: i could run hundreds of pil apps with 4GB
<cess11>
Great, I'll keep an eye out for them.
<beneroth>
aw-, T. just wanted to be on the saver side of the estimate, haha ^^
<cess11>
I want to do cache and parsing heavy stuff for single or few users, sort of. Disc I/O won't be worthwhile so system SSD and some slow, cheap RAID, and processor heavy lifting can be outsourced.
<cess11>
But I'm not in a hurry.
* beneroth
has similar applications in mind
rudha has joined #picolisp
<aw->
cess11: you can use intel cpu
<aw->
if you want lots of RAM
<cess11>
T
<cess11>
Purism and another company sells some towers that seem OK.
mtsd has joined #picolisp
<beneroth>
cess11, I guess you would want to attempt a pil vs elasticsearch et al ? question is how RAM-conservative those systems are. guesswork: pil is probably more RAM-saving, by having more fragmentation (cell by cell), so some cost to execution performance and bulk operations (?)
<beneroth>
difficult
<cess11>
I want pil as foundation since it is extremely programmable and almost no limits.
<beneroth>
same
<beneroth>
dev time matters more than execution time. took long until I grokked that.
<beneroth>
execution time is much much easier to scale :)
<cess11>
Yeah.
<cess11>
I want several things but one of them is a real big news parser looking for and aggregating anomalies as well as notes on themes currently important compared to a few recent measures.
<beneroth>
in german, but there is a english voice-over (was life translation, so quality might be suboptimal)
<beneroth>
a guy who donwloaded spiegel website (german newspaper), and than did some data mining on it for fun & illustrate how data mining works with this presentation
<cess11>
RAM heavy but processing is easy to plan and prioritise.
<cess11>
Nice, thanks.
<beneroth>
e.g. news article authors, publish time, if commenting was allowed or not or for how long, etc
<beneroth>
did articles change over time etc
<beneroth>
some interesting measurements
<beneroth>
relationships of topics/keywords
<cess11>
I've been doing similar things with pil and found it very, very efficient already with the built-in indexing and parsing functions.
<beneroth>
T
<beneroth>
I haven't made yet strong use of all the string indexing possibilites
<cess11>
Besides counting and comparing it would be easy to implement some supervised ML as well and since we're in a war ridden and tense time I think it would be a good idea to monitor the world rather than just read the news.
<beneroth>
T
<beneroth>
it is important to be consciousness of the filters. (there are always filters... in the media used, in the availability, in the wording, in your mind and background)
<beneroth>
having enough information is a problem of the past,mostly.
<cess11>
Yar.
<beneroth>
now the information is drowning in noise and disinformation.
<cess11>
What is and isn't talked about and by whom is more important than the details, usually.
<beneroth>
ah, metadata
<beneroth>
I also see motivation as a problem.
<beneroth>
the absence of easy solutions nor any fully-conceptualised proposed solutions strongly reinforces humans to stay in their comfort zone and largely ignore the issues and risks. "can't do anything meaningful about it, so what"
<cess11>
For first pass parsing, yes, automated classification and guessing can be based on it and constraints, with observation of 'silence' on issues on top of manual work with how mass communicated propaganda works in Sweden.
<cess11>
Yep.
<beneroth>
a majority seems to be willing to prove the pessimistic answer to the drake-equation to be true (self-extinction).
<cess11>
I think they're mostly passive rather than explicitly pessimistic.
<cess11>
It's hard to create some life and thought outside work for a lot of people/
<cess11>
*.
<beneroth>
true
<beneroth>
true, another problem.
<beneroth>
democracy cannot work if common people haven't enough resources to think and enact their options
rudha has quit [Quit: Leaving]
<cess11>
Very true.
<cess11>
Besides parsing it is fairly trivial to query open REST API:s as well so guess-based automated aggregation from diverse sources as complement and perhaps explanation of what news and other mass media streams say seems interesting too.
<cess11>
The object system is efficient for this since linking is fast and takes away the need for keeping several copies of information to push up performance in lots of situations.
<beneroth>
another idea I had (which is less likely that I will implement it in a long while) is to compare news sources, calculating their shannon values (informative value), too see which sources produce insights/original work vs. content farms and press agency copycats
<beneroth>
cess11, T, the big advantage of graph databases over relational databases is JOINS
<beneroth>
joins are cheap in graph.
<cess11>
Also to compare how a source reports on an issue compared to how they usually perform in informativeness. If an outlet is usually informative but not on an issue currently all the rage, that's anomalous and interesting.
<beneroth>
T. additionally, sources are probably specialised, meaning they don't have the same quality on all topics, so you want to find out on which they shine. might be interesting, even bill-able information for their own business decisions.
<cess11>
The flexibility to develop this as a running system isn't easy in basically any other language, except perhaps logic programming but that's included, as far as I can tell.
<cess11>
Most likely.
<beneroth>
it is with pilog
<beneroth>
it is even more awesome as you expect it, even when you think it is awesome
<cess11>
T, since it goes in 'all directions' through backtracking it can easily describe as well as construct data model spaces.
<beneroth>
haha, oh yeah. just you wait, I might likely attempt to automate data modelling (ERD design) eventually :)
<beneroth>
just have to finish automating all the involved processes first
<beneroth>
how nice you payed for the non-optional component which lets the NSA pwn you
<beneroth>
(and others, in case you trust them)
<cess11>
It'll be harder to clean up than Mirai/"IoT" botnets once it really catches on, I think.
<beneroth>
how to even detect it?
<cess11>
Catching packets with Burp or firewall filtering, or not at all, I suspect.
<beneroth>
I think so too.
<cess11>
And then it means the physical unit needs to be replaced which few will do.
<beneroth>
well even this patches by intel and OEMs (firmware update), you can't be sure if you haven't already caught something.
<beneroth>
and I bet most vulnerable machines will not be patched
<cess11>
Yep.
<cess11>
And yep.
<beneroth>
reasons for going away from Intel (not that the others are safe, but diversity is costs for attacker). and reason for stuff like pilOS.
<cess11>
Public service IT in general must be considered as compromised now, since it is so extremely attractive for security services and criminals and data or information oriented companies.
<cess11>
T
<beneroth>
what do see as "public service IT" ? cloud infrastructure? or IT of gov/public services?
<cess11>
The latter.
<beneroth>
ah ok
<cess11>
Medical data, tax data, and so on.
<cess11>
The state sucks at keeping up to date with anything unless it is a dictatorship or forced to by an angry public.
alexshendi has joined #picolisp
<beneroth>
well my stance is that a nation should follow a strict non-IT-warfare, strict pro-defence policy. doing research and publicly publish vulnerability and patches. also enforce anti-hacking laws within its jurisdiction, including against domestic groups. only that way a country gains plausible deniability in "cyber warfare".
<beneroth>
faking wrong attacker attributions is so easy. MAD doctrine doesn't apply to internet.
<cess11>
Mine is very similar.
<cess11>
The swedish parliament is expected to give police hacking capabilities soon, without regulating state trojan procurement or other method developments first.
<beneroth>
problem is I see currently no chance for getting this into law. not even here in switzerland. mainstream, even the subgroup of mainstream politicians, are too uninformed and get disinformed. its a topic difficult to explain. and secret services and the military industrial complex has to much incentives to take part in this circus.
orivej has quit [Ping timeout: 255 seconds]
<cess11>
A few weeks ago someone pulled out lots of police data in a breach.
<cess11>
So I'm quite worried.
<cess11>
Yeah.
<beneroth>
well in switzerland the population recently voted yes to extend existing data retention laws, also agreed that secret service might spy on all connections which are not strictly within switzerland (so, haha, all internet connections), and hack peoples devices (if judge stamped)
<cess11>
It will be interesting to watch cops find out about honey pot hackback offensive security the hard way, though.
<beneroth>
the problem with all this is, it is actually not such a big issue, quantitatively. only highly increases the risk of some singular people (e.g. high-profile journalists, activists, etc). and throws A LOT useful money into bullshit. but when the systems become a danger for the majority, it will be powerful enough to not be changed. the whole thing of opposition forming against it, demonstrations, movement etc. is the one thing it is best at detecting and pr
<beneroth>
eventing.
<beneroth>
hm
<beneroth>
haha, right cess11 :)
<beneroth>
well the canton of zurich buyed hacked team (from italy) equipment for millions CHF just weeks before hacked team got totally pwned, rendering the equipment useless.
<beneroth>
also, at the time they bought that system the use would have been illegal by law.
<cess11>
Heh.
<beneroth>
but yeah.. apparently one trick is that swiss police asks german police to plant a trojan on a swiss suspects devices, and probably vic versa. that way, they don't violate the laws as the laws usually don't care about foreigners, or even if they do, its much less likely to be challenged.
<cess11>
Yeah, I also expect it to turn into social and political control once it becomes too obvious it doesn't catch criminals beyond the stupid ones, that could be handled in other ways.
<cess11>
I think that is common, but largely unknown and hard to confirm.
<beneroth>
well it's just to easy to be used for that. first lets target the nazis and child molesters, then lets move on to dissidents (russian fake news), etc.
<cess11>
Yar.
<beneroth>
it is also just kinda scaling up the societies capability to keep consent and culture.
<beneroth>
problem is you can easily end up with a vulnerable monoculture.
<cess11>
Came a study of chinese dissent control recently that showed it employs a million or two persons that post nice and happy stuff everywhere, suffocating discussion. It will be here too eventually.
<beneroth>
hehe, yeah
<beneroth>
and they don't argue against dissidents, they just flood it with distractions and noise.
<cess11>
I expect that refugees might be the first to be exploited for it in the West.
<beneroth>
already done.
<cess11>
Soft stalinism.
<cess11>
Really?
<beneroth>
facebook / google content checking is done by low-payed (and low-educated) personal in morocco and germany afaik
<cess11>
Ah, right.
<cess11>
It is similar, but commercial rather than driven by state interests.
<beneroth>
"Derkaoui had spent three weeks working in Morocco for oDesk, one of the outsourcing companies used by Facebook. His job, for which he claimed he was paid around $1 an hour"
<cess11>
Horrible work, I'm sure. Done a fair bit of 'social media moderation' on a service I had some years ago, some men seem to have nothing to do besides finding and passing on despicable stuff for no reason.
<beneroth>
you think this is a gender/sex thing?
<beneroth>
also, this can be automated to DOS a community
<cess11>
It was real people, only men and fairly constant, so yes, I think it is gender related. I see some of it on Facebook and elsewhere too but obviously not to the extent as when I did digging in my own database when complaints came in.
orivej has joined #picolisp
<cess11>
Some came back over and over and went to a fair bit of work to keep on sexual or other kinds of harassment. Which they didn't see it as.
<beneroth>
I question your sample size and I suspect your userbase was mainly male for which you probably didn't correct
<beneroth>
well I think it would be useful to differ between intended and unintended harassment
<cess11>
When I began we had ~100k active users. I've seen thousands of cases.
<beneroth>
in my view, intended harassment is a behaviour problem, while truly unintended harassment is a cultural difference / perspective thing, not bad behaviour per se
<cess11>
Those I have in mind didn't see it as harassment, they had all sorts of excuses, some even came up because they got blocked and complained about it.
<beneroth>
the biggest groups I moderated was some hundred people, not thousands, so you have more experience
<beneroth>
well, if they were sincere than I would say some communication was suboptimal.
<cess11>
Helgon.net/.se, once a huge community but reduced by Facebook like many others.
<cess11>
Well, quite unacceptable to send snuff or porn to a stranger you've just recently started talking about music or interests with.
<beneroth>
well, depends on the context and culture. some people apparently exchange cooking receipts on pornhub, so I guess in that context a porn link would not be inappropriate
<cess11>
I agree with most of that description of Slack but it isn't because of it, the interface just makes things obvious that regular speech sort of hides.
<beneroth>
you used/using slack?
<cess11>
Sure.
<beneroth>
I haven't
<beneroth>
ok
<cess11>
Yeah, at work.
<cess11>
It is IRC with a base RAM consumption of roughly 2 GB.
<cess11>
But you can edit your messages.
<beneroth>
the slack thing: I'm not sure this female vs male culture are global. mabye some tendencies, but I suspect that e.g. female culture in US is different to female culture in switzerland. though nations are not good borders for this.
<cess11>
It is occidental, post-christian/-victorian.
<cess11>
I think.
<beneroth>
which thing? I mostly think the thing they call "hedging" in the article
<beneroth>
may main stance is that I find separation by sex-role/gender to group people not very useful. I think grouping humans by other aspects than gender is more useful to reflect individual preferences and needs
<cess11>
That sexism on Slack and elsewhere. Muslims and jews tend to supress women in sligtly or clearly different ways.
rudha has quit [Remote host closed the connection]
rudha has joined #picolisp
<cess11>
It is just one component but the moderation I did and had in mind cut through income and class groups.
<beneroth>
well I also take the stance that most suppression is happening within the gender groups. in-group behaviour and pressure.
<cess11>
Of course.
<cess11>
It's passed on by role models mostly, I believe.
<beneroth>
T, though those are constantly in flux
<beneroth>
so I would say one questions on the topic of harassment/displeasure is, for a community, how well are the values and expected/tolerable behaviours signalled? is there even a consent on this within the group, or is the group indeed made up of multiple groups. etc.
<clacke[m]>
> It is IRC with a base RAM consumption of roughly 2 GB.
<clacke[m]>
cess11: I need to steal that quote
<beneroth>
hehe, yo great summary
<clacke[m]>
do you want attribution?
<cess11>
Nah, I'm all for copyleft.
<clacke[m]>
you mean CC0 :-)
<clacke[m]>
The number of unsolicited dick pics any one of my female acquaintances have to endure as soon as they do anything to become remotely minimally famous is ridiculous.
<clacke[m]>
A male friend of mine used his last name as IRC handle back in the 90s. It could easily be mistaken for a female first name. He got quite a lot of private messages presupposing he would be interested in meeting total strangers.
<clacke[m]>
The number of men harrassed on Japanese or Chinese trains, or the number of Iraqi and Syrian men abducted for sex slaves are insignificant compared to the nimber of women. There is definitely a gender pattern, and it is cross-cultural.
<beneroth>
maybe, but does this give us any vector to attack the problem short of getting rid of men?
<beneroth>
I don't see the usability of this angle
<beneroth>
also, high under-reporting of violence/sexual abuse of men is probable
<cess11>
Not getting rid of male bodies, but we might need to get rid of manliness.
<beneroth>
I disagree.
<beneroth>
manliness is a fuzzy and moving concept. contains also a lot you surely don't want to get rid of. so why even work with that concept and not more detailed ones.
<beneroth>
we want to get rid of abuse.
<clacke[m]>
it means men are the ones that need special attention in teaching empathy. this is not useless data.
<clacke[m]>
getting rid of manliness ... I would prefer to see it as redefining or refining manliness
<beneroth>
that is a continous and ongoing process
<clacke[m]>
Lazarus Long manliness is compatible with existing manliness but our society would probably be better off with more of it
<beneroth>
well why not abandon gender-roles at all ?
<clacke[m]>
yes, there are no fire-and-forget measures in shaping social relations
<clacke[m]>
and no delegation to higher authorities, every parent is responsible -- and every adult that shapes childreb
<beneroth>
whatever the definition of the current in-group "manliness" or "femaleness" is, there will be people of the related sex which will not fall into this definition
<beneroth>
so why not just get rid of this discriminations (in the literal sense) entirely?
<clacke[m]>
I think we have tried abandoning gender roles, and it is part of the current problem where young men don't know how to behave.
<beneroth>
just as a though provocation
<beneroth>
well we have also young woman who don't know how to behave
<clacke[m]>
the binary model is too strict and not everybody fits. but a lot of people do fit. of course we need to be sensitive to, and accepting of, people who need to mix and match to feel at home.
<clacke[m]>
but I think it helps non-binary people too, to have a language to express I'm a bit of this and a bit of that
<beneroth>
I don't think this "feel at home" and "gender-group" is the same. huge overlaps for most people, but not all.
<beneroth>
grouping by e.g. hobbies (readers), or e.g. personality traits look more accurate to me than grouping by gender
<clacke[m]>
excuse me, I need to help a misbehaving boy brush his teeth :-)
<clacke[m]>
Bedtime protocol in force. But basically what I want to say is that we seem to have two bell curves, and we need role models and ideals that work for many people.
<beneroth>
I basically agree, but I want to note that we can't throw all people into a single bin. different groups with different ideals need to be able to co-exist. and social movement between these groups should be possible.
<beneroth>
which consequently makes this a non-solvable problem I think, there is no ideal solution but it is a red-queen race environment
<beneroth>
meaning the parameters and context is constantly moving, and so have the adjustments, too.
<beneroth>
and this whole act is playing in the meta-environment of society and survival
<cess11>
Gender roles are quite slow moving, binary and heavily commercialised in the West today.
<beneroth>
well I would say we had some pretty big and significant changes in the last 100 years. even in the last 30 years.
<cess11>
Sure. To some extent the european neofa
<cess11>
scism is a reaction.
<beneroth>
this are short timespans for social change I believe. at least when we look at historic data. though the speed of it seems to have increased over time, same as the speed of technological development though not so much
<beneroth>
well
<beneroth>
I think the neofascism is more of a symptom of economical development than of true counter-movement on the cultural level.
<cess11>
How come?
mtsd has left #picolisp ["Leaving"]
<beneroth>
when looking at it in terms of Maslows pyramid of needs, I think quick changes on a upper level are more likely caused by an underlying level then having a cause on that level itself
<clacke[m]>
It is my impression that sexes probably work well as groups to screen for certain behavioral issues, like we screen women for breast cancer, men for prostate cancer, certain African ethnicities for sickle-cell anemia and certain European ethnicites for malign melanoma.
<cess11>
But it is mostly middle class men expecting disaster or seeing it in immigrants that they have nothing to do with that join it, right? It isn't mainly a movement among the poor.
<beneroth>
clacke[m], well thee you don't screen for gender but for sex/body differences
<beneroth>
s/thee/there
<beneroth>
cess11, afaik the nazis in germany and the alt-right in USA cluster with the poorer geographical areas, no?
<cess11>
They also argue irrationally, have the resources to make lots of noise and spend time doing politics through jokes, at least in the broader neofascist movement, which is bigger than the outrigt nazi/alt-right part.
<beneroth>
the upper end of that spectrum is probably more visible
<beneroth>
yeah, but that doesn't mean the others don't exist, they are just too far removed from your view (geographical, or medial e.g. no internet), or too busy to survive to voice their opinions
<beneroth>
but when voting happens or other, more disrupting events, there are suddenly groups popping out in bigger numbers out of nowhere
<beneroth>
lack of measurement
<cess11>
In the US voter suppression, not turnout, gave the far right success.
<beneroth>
they have a broken system anyway
<cess11>
We all do, I'd say.
<beneroth>
well suppression including self-suppression?
<beneroth>
well I happen to like the swiss system quite well.
<cess11>
Rather intense anti-Clinton propaganda and her campaign didn't even think they needed to cater to those catholic and dark skinned groups at that point.
<beneroth>
and internal fights and sabotage of sanders.
<cess11>
It works but it is seriously risky since bigotry and oligarchy has the upper hand in many issues due to the amount of voting and the somewhat low turnout in them. In Sweden we get other kinds of corruption instead, of course.
<cess11>
Yes, but that was kind of early in the election, the effects were limited I think.
<beneroth>
my point about the broken system was that a two party system plus some other factors make it pretty hard to make adjustments. in a working democratic systems ideal outcomes are also not guaranteed, but there are fewer parameters to take care of
<cess11>
It was a very close election, decided in the last week or so.
<beneroth>
though I think I read about this earlier this or even last year, so I wonder how this is a new revelation now
<beneroth>
google promises to better its behavior (haha)
<beneroth>
reason for it is probably not technical optimizing (as google spokeperson told) but tracking if an advertisement target entered a store or not (one of the newer features for google ad customers)