Don't wanna be here? Send us removal request.
Text
RBMK, MCAS and control theory
These topics are of course form the proverbial Horn of Plenty itself, with their unending supply of details, complexity, second- and higher order effects/interactions, and so on.
However, what I find most interesting is pinpointing the actual fatal design flaw. Because - and this is why this is fascinating - these situations were not like contacting a deadly disease, no, here there was always a point of return, right before the catastrophe. Basically right until the very last decision to activate the emergency control rod reinsertion procedure (AZ-5, EPS, scram!), or in case of the runaway anti-stall system (MCAS), before going below a certain altitude (which was a few seconds before crashing, so again at the very end).
After all, if the operators insert the control rods one by one and ramp up the water flow, things might break due to the enormous thermal load, but the explosion would have been probably avoided: https://www.youtube.com/watch?v=q3d3rzFTrLg&t=7m15s
And, in case of the Boeing 737 there's a standard procedure to stop electronics from interfering with the rear stabilizer angle (runaway horizontal stabilizer trim): https://www.youtube.com/watch?v=xixM_cwSLcQ&t=17m20s
Naturally, we can simply claim that it was human error. Though in case of Chernobyl the information wasn't even known, yet that was the official story. (And of course the chain of events was of course initiated by humans, and again the decision to withold information was again human-made, just as the flawed design, so they got it right, heh fukken tankies.)
And Boeing (even the CEO) blamed humans initially. (Bonus points for the similarity of infomration omission: MCAS was not documented in the 737MAX flight crew manual.)
Okay, so what were the flaws?
In case of the nuclear reactor, it was that the control rods weren't long enough. Basically the transition from ÜBER power to lower power went through a regime of ÜBERHYPER-to-the-11. (Because as the graphite moderator part of the control rods descended it increased reactivity at the bottom, which boiled even more water, etc.)
(Though the IAEA report from 1992 states that the reactor was so unstable that the explosion could have happened even just because of the water pump rundown itself.)
And in case of the Boeing 737MAX, the MCAS relied on systems with lower reliability than the system it influenced. Again, control theory, you cannot have reliable control if some unreliable system sometimes interferes. So claiming that the 737MAX flies the same as the NG because we solved it with some gadgets was rather thick bullshit.
2 notes
·
View notes
Text
basics of futurelock
Text. Yeah, just simple text please. It'll do. Or will it?
Who knows? Who knows anything anymore nowadays? Right? Yet people seem so certain.
Hence we come to this impasse.
It seems progress is blocked, the future simply pours down on us, washes over us, crushes us as its crests come down, muting and thrusting us back into our place. Into a primal state, locking us into an even more fierce race against each other over resources, possibilities, over our own future.
The dilemma is trivial, the solution is of course transcendental. Who has a say about what? What is right? What is just?
I used to (I still am, but you know how it is) get into people's faces and advertise Rawls' stuff - via a Pinker book, and I decided if it's good for him, it's probably kosher and it works.
But as I hinted at in the intro, we don't know anything nowadays, so it turns out that there are, of course, problems with Reflective Equilibrium as a constructive guide of ethics, mainly there's no guarantee of convergence, and even if we somehow make it work on a per mind basis, we have trouble with convergence toward a common value target. And I'm not saying it's doomed or anything, but let's just say it lost its relevance a bit as an argumentation tool. People are very skeptical when it benefits their biases.
So it's a kind of writer's block. I mean I haven't written anything for ~2 years now. Maybe because my head is dull and not because it was too full. Maybe not.
But maybe because it's just trite to whine, especially in circles. (Not that I have anything new, substantial, fresh, or somehow interesting to offer suddenly. No, but blogging was always just a crutch, the good old set of training wheels. Fearless experimentation with writing words, one after the other, just to see the cursor move, to feel the flow of unfunny little lines unrolling on the screen. And back to the basics is as good a tool as any, if it helps. If it works, it ain't stupid.)
The dilemma is trivial, because anyone who has taken the time to consider it like any complex process in a phase space knows that there are bound to be pathological states, attractors and other great dangers lurking in there. And "it" can be as simple as the economy or as big as our whole contemporary culture with all its history. (Path dependence, if you want to get technical.) Hence we can conclude that simple solutions are unlikely to work, as we will never know the initial conditions, our model will never be sophisticated enough anyway, and we don't even have the current state mapped out. (And to add a bit more difficulty, we are thinking hard about how to stay in the shadows in certain topics, as we are starting to see the problems with too much centralized observability.)
If simple solutions are doomed to fail, how do we make people to understand this? People seem to like details, faux sophistry, presenting themselves as witty, yet scorning others for eliding key bits and being intentionally cryptic. Of courrse, at the same time. Anyway, the answer seems simple, brainwashing, indoctrination, goal alignment, or you can call it simply education, but it doesn't matter. Intelligence without a utility function is meaningless. (I'd wager it's simply impossible to be a perfectly unbiased agent, because then that means uniform prior distribution and thus a complete lack of comprehension to interpret the world. Bootstrapping is the hard problem of consciousness after all. Evolution had billions of years and all the samples that worked are still just imperfect and smug copy-machines.)
So, how do we "align" others? How do we persuade them? Of course, if people are just again complex systems some might be simply too far gone that we can't ever reach them. What do we do with them? Are they a threat to us? (Yes. But do we want a comissariat to keep watch of them? That usually goes so well.)
It's hard to accept that you are wrong, it's even harder to accept that you are in a minority and you have to seriously hide your views. I already accepted the first part, but since I'm blogging I don't plan on accepting the second. I've experienced a very-very-very tiny bit of what that would entail, what that would mean in practice, and it's fucking bloodcurling, it's enraging, it's uprising (it's so revolting, I use it as an adjective), you could even say wake-fucking-uping.
I was wrong about optimism, and I know I'm wrong about using sticks and stones to fight for values, as every dream where folks take justice into their own sadistic little hands, but it also follows from the aforementioned that simple idealism would have very likely never worked anyway. And I don't mean don't be joyful, relaxed, coolheaded and hopeful about humanity's future, but at the same time accept the cold reality of the hard choices, face the possibilities that the calculus of rationality shows us, accept that it's a regular battle. Uphill, if you like. Detours, roadwork, ahead. Hard work. Work. Work ahead.
I gave a chance to a lot of things, because that's the lazy thing to do if you have rose tinted glasses - or you live in some kind of semi-translucent shell of the approrpriate biases.
It's 16:30, I have about 4 missed calls, probably 40 blinking chat windows, and about as many "important" unread emails. But those are just the symptomps of the other kinds of shit that I haven't dealt with yet. People I haven't talked to because it's easier to wait, enjoy the silence, why bring down the storm if it comes to you just the same eventually? Why fight the future if our bearing seems to be locked?
If we seem to be cast from stone? At least I seem to be the kind of black as night stubborn granite. Hard to mold after it cooled off, but viciously hot otherwise. And not in the pun way. Hah. (Though that's another post, another bag of laundry to go thought. Another staring contest with some kind of truth. Classic monogamy seems incredibly unfit for me when it comes to dressing for life from day to day, but in the grand scheme of things I'm not a Tinder troubadour, and never were. Nor were I unfaithful, nor do I want the meaningless [fishing for] fucks. Nor do I want to fight the everpresent discrepancy between said mental clothing and my occupying state of mind.)
It's getting dark, the storm is still a bit late though, but let me see about that.
2 notes
·
View notes
Text
Code Architecture
I watched a few Uncle Bob rants, or lectures, (rantures?) and naturally wanted to know more.
So here's a blog post from May 2012 talking about use-case centric architecture, and mentions that models are the hard part (because that's where most of the sharing happens in software stuff, because under this whole mess of code you usually have One True Source of Truth, the Database, and the structure of data in that is kind of fixed, therefore shared among all use-cases, hence the problem of context-dependendt modeling. Which is something we tried to do last year, and I think it's a pretty good idea. (Model the different aspects of the end-user depending on the context. In the sales module the end-user is a potential new contract, so maybe a Business; in the administration area, he-she is just a User like anyone else, in the shopping flow she is a Customer.)
The whole show with the files and directory structure is a bit laughable, because no one said that a directory layout is the right analogy for a building's blueprint. A building is (at best) a 3D structure. People or raw material come in at points and they can only go to a few places in permitted by the laws of physics, and due to gravity those 3 dimensions are 2 plus some points of connection between floors. And in a 2D plane with fixed amount of stuff you need long hallways and a lot of doors, a reception area and a lot of infrastructure to route the people and-or material to the place of processing. That kind of explains why frameworks are used in software.
The better analogy would be the user-visible portion of the application. If it has none, then maybe it's not as simple as a church? (Where the main and only function [feature in software speak] is that you can go into it and be there with a lot of other people.)
Also, Churches were built in many-many years, they are sometimes just as messy from an architectural standpoint as software. (Sure, the simple loadbearing structure is simple, but that goes for most software too.)
Anyway, Robert's points are not without merit, software architecture is important, and making code nicer and organizing it according to some kind of general system is even more a pressing concern, or more frankly a real pain point nowadays (because our software creativity is much further along than our programming languages), than it was before every application became a multi-multilayered mess of many components and aspects/dimensions.
So, again, it's not that it's okay to just mash a hundred files into a git repository add two lines into a README and call it a day, but maintaining documentation (blueprints for a many-dimensional structure!) manually simply takes too much time. (We're not building a Church, we don't have 40+ years.)
I think we need better tools! (Duh! :))
The STEPS research project (which became famous the moment it started, well, because it was the latest endeavor of Alan Kay, though I bet not many people were aware of it - I had no idea for example) is an interesting approach to combat this global complexity, by introducing new local DSL-s. The paper is well worth a read. HN thread (The result is from 2012, so coincidentally from the same time the Uncle Bob rant.)
Bret Victor initially advocated a more graphical approach see (as usual), but his DBX talk is more about the end of the 50-year* large hype-cycle peak, a sobering confession about the need for discovery, for fighting the momentum of J2EE (of enterprise programming, of building the large monsters that then take forever to die, but during the process bleed everywhere, trash and spasm and destroy lives - I'm thinking of security vulnerabilities in large systems that "will happen" in the coming years, see this video).
* - sure we should talk about which particular hype do we mean, but let's just say that IT moves so fast, yet at the same time has the same problems since the 80s (or 90s, if we mean the now ubiquitous nature of distributed systems thanks to the Network)
1 note
·
View note
Photo
Okay, okay, it's just techno, but yeah, it's pretty brutal, if you are not into "that kind of stuff" :)
Which reminds me, home come I actually am!? Maybe I'm just lazy and don't want to get sweaty in moshpits any more, yet still want crushingly powerful music to dance to!?
Who knows!
But this shit is amazing.

Antigone - Night Adrift [TOKEN59]
I mean just a few minutes ago I clicked on a recommended blog on the dashboard, and just as the linked track ended aaand since someone thought auto starting background music is awesome, Bastille started to play No Angels. Which sounded awesome, because the regular version has a sample from a movie (here), and as the thumping wubwubs subdued and a stringed instrument came alive, and then Dan started to sing, well, it was nice. Warm, human, cuddling, mesmerizing, reminding, reminding me of the hearth we used to share and now not, and the hardship that's moved in instead.
Hard times maybe call for hard techno, huh?
Ah, just kidding, I still listen to any shit that piques my interest enough!
On a related note, I can't wait for Rosetta this year. Until then, we have Ufomamut.

Ufomamut - Ecate album
1 note
·
View note
Text
Specs
Seriously. Bah. 5 minutes ago I was fuming and silently raging about how inefficient our consumer reality is. I spend a lot of money on pieces of technology manifested by a Dell P2415Q monitor (4K, UHD or whatever, but the important thing is, that it can do 3840x2160 @ 60Hz, allegedly over HDMI - but for that you need HDMI 2.0 - but luckily my Gigabyte GV-N760OC-4GD rev 1.1 has DP output - it took a few days for me to notice it, but meh, I usually forget about extra connectors; it happened to me that I bought additional DDR2 RAM and upon installing it I discovered the DDR3 slots right next to the DDR2 ones).
So it should work, right? Nvidia has a very spartan description of the things the chip can do, Gigabyte has a bit more informative "tech specs" page, and according to our modern bible (wikipedia) mDP (mini DisplayPort, because Dell was kind enough to include a - shitty, but at least not separately sold - DP-mDP cable) has the necessary bandwidth (because it can do what DP can do, just with a smaller form factor connector).
And I get the nicely stylized error message from this black beveled 24" beast that ... the input timings are not appropriate and I should use one from the list of supported timings. Oh really? Then go fuck yourself with anything from your list, because that list is not really present anywhere on the web. (Because it's in the EDID, about a hundred bytes that can be read from the monitor's NVRAM (or EEPROM) via a VESA command.
But, even though Nvidia driver and/or Windows 10 was able to read that the monitor supports 60 Hz, selecting that resulted in the uselessly infuriating error message.
In the end I had to play with CRU and override the data presented by the monitor.
Maybe the monitor truly supports "LCD native" timings (176 pixel front porch, 88 pixel sync width, 296px back porch, 560px for blanking) and the card, or the driver, has problems with the necesary 594MHz pixel clock, but there was no way to figure this out. The NVIDIA Control Panel is buggy and slow and dumb as fuck, amusingly clunky at best. The Windows 10 settings are as useless as ever. However, with a bit of trial and error and setting (declaring, specifying!) the custom resolution to use "LCD standard" timings (3840x2160, front porch 48, sync width 32, back porch 80, blanking 160, and the corresponding vertical timings are 3 5 54 62, with a +/+ polarity ...) and the magical 60.000 Hz works!
Again, not just from an engineering perspectice, this is a big no-no, a usability nightmare, a proper fuckup. I know that most consumers are just slightly more intelligent than the cardboard box they get with their new shiny electronic device, but information [knowledge] is power. At least they could empower their friendly neighborhood techno witch doctor, so the abstract folk could help mere monitor mortals. But no, no way to get the necessary magic numbers from the device itself, no "show me the black magic" button/option anywhere.
And I know everybody loves when their stuff is white and "everything just works", but that is just a stupid simple QA step, a false illusion of choice and freedom. Especially it concerns the latter, because iDevices don't like non-iDevices, so they can guarantee the magical interoperability, by ... actually not interoperating with anything outside of their close knit of components. Vendor lock-in. Monopoly.
Oh, and just as a sidenote, I would still like to hang retailers and shoot manufacturers for not labeling, presenting and displaying detailed enough specs. It's not enough to say that it's a DisplayPort cable if there are 3 versions. (I know, I should just assume you are selling the crappy 1.0/1.1 version, but usually that's not the case. Usually manufacturers, like Dell, are lazy to give sufficient information for helping potential buyers navigate this forest of false choices.)
Okay, enough ranting for today. :)
0 notes
Text
systemd and D-Bus activated services?
So there's this very useful accountsservice package (program). (I haven't read up on why it's better than parsing /etc/passwd, and why there isn't a library already for that -- hint, there is!, -- but it's not really relevant now. The original idea in 2010 was to factor out user accounts (so login) related parts of the display manager, because they are used when presenting the switch user dialog from inside a desktop environment. Probably there was no logind back then and somehow this daemon stuck.)
Anyway, lightdm can probably work without it, but let's say I just want it gone when it's not needed. And since it's a dbus activated service, (systemctl cat accounts-daemon shows BusName=, and it's also registered as /usr/share/dbus-1/system-serivices/org.freedesktop.Accounts.service) we can just use a timer to stop it after it's done responding to a request.
# /etc/systemd/system/accounts-daemon-autostop.service [Unit] Description=fuck yes! accounts-daemon stopping. [Service] Type=oneshot ExecStart=/bin/systemctl stop accounts-daemon # /etc/systemd/system/accounts-daemon-autostop.timer [Unit] Description=fuck accounts-daemon After=accounts-daemon.service [Timer] OnActiveSec=10 [Install] WantedBy=accounts-daemon.service
And just systemctl enable accounts-daemon-autostop.timer.
This should be a core systemd feature perhaps? After all socket activation (and on-demand) implies that it's off when it's not in demand.
0 notes
Text
The year of the Windows X/Linu X/OS X desktop
So, first, Windows 10. Error C19001DF.
Thanks Microsoft. Yes, I know this stuff is hard. But fuck you! (Don't worry, be the end of this post everyone will get theirs!)
KDE/Kubuntu always fucks up. Now with "Plasma" (because calling it KDE 5 was too mainstream) the fancy notebook always asks for wifi passwords. But I found out that after removing them from the previously remembered connections, and connecting again, it works! Woo! It's a quirk maybe because those passwords were saved into a Gnome keyring, or whatever. And naturally the best part is, people can't write error messages. And error handling in general is completely stuck at the C19001DF level every fucking where.
I just spent a day fiddling with apt-get and aptitude later, trying to untangle this mess. Multiple desktop environments just can't live side-by-side. Nor multiple versions of whatevers. It's just sad. And I though maybe NixOS will save us, but then I saw what they think of package management... meh, at least there's ad-hoc install, but people still don't get how declarative configuration is the future, but the present should always be about small incremental imperative changes. I want to install a package, I click-clack here or there, or tak-tak-tak something on the keyboard, and using the declarative state it should work with me. The system should allow me to specify actions, and it can of course compile those to declarative constraints ("install blabla" -> "ensure package blabla is installed"). But it migh be at least getting there.
Although, this is a bit ridiculous:
To install KDE 5, add this entry to your in your /etc/nixos/configuration.nix: services.xserver.desktopManager.kde5.enable = true;
Installing KDE 4 and KDE 5 at the same time is not recommended because some files have overlapping names.
Why? What the fuck? Why the fuck does it want to know about what desktop manager I use? It should manage packages (and package address spaces, as it's their "killer app", their reason d'etre) and fuck off.
So it's good that we have Linux "distributions", and countless volunteers working on making them safe and sound, and innovative, and clever and cross-platform, but when just renaming a package can effectively brake hundreds of others, I get the feeling that we are seriously lacking efforts on the sanity front. This whole computing endavour is not picking a few critically rotten low hanging fruits. (It would be pretty easy to check that a rename does not break a few important packages, such as kubuntu-desktop, xubuntu-desktop and ubuntu-desktop can all coexist at all times would be a great slogan.)
OS X auto wifi hopping. Just fuck you. Seriously, how is it a feature that it automatically connects to an Access Point with a stronger signal? What if that, you know, lacks the stuff I need, like Internet (or Intranet - because it usually does that when it cannot find Internet) access? But of course, the topping on the iCake is, that you can't disable this as far as I know.
0 notes
Text
Atomic fences
Restartable Sequences, and also by Paul Turner
Also, on a related note we started working on a little program, that allows building a kind of distributed machinery for managing ... stuff. All in Rust. Of course.
The title refers to memory fencing (read and write barriers) that are required for making certaing operations atomic (which means some degree of local consistency, usually total local consistency in the full virtual address space).
I'm no way a pro-programmer churning out multi-threaded code, but finally I think I grasped the whole towering stack required to reason about what happens when more than one executing unit is active.
It took a while, but. Now I want to properly do asynchronous stuff. Event driven, with one thread, no thread pools hiding under the carpet, and other magical machinery obscuring the subtle details.
Also, finally we're approaching Python 3.5 with gradual typing. And I want to jump into that a bit too. (With pyenv and mypy it's as easy as installing 3.5.0b3 and annotating things! Right? Oh and putting mypy into the continuous integration process, so it statically checks the code before running tests.
I should blog more. (It's a recurring theme. I should do more stuff, instead of just reading about cool stuff. Or at least write more about cool stuff. Meh. This is a suboptimal world.)
0 notes
Text
EU VAT MOSS WTF!!
So, let's say you are a polite Canadian company and provide software-as-a-service over the Internet to other companies, some of whom are EU-based. Uh, okay, I might have to pay some taxes because I heard "the EU" is herp-derp and have turn't down for VAT.
Yes, value added tax is a consumption tax, quite fair if you think about it, it but since requiring consumers to report their consumption would be neigh impossibly impractical developed states require sellers to add the amount of VAT to prices. (Interestingly, VAT is applied throughout of the supply chain, whereas sales tax, which is mostly a USA thing, only applies at the last leg.)
Good. Let's pay taxes to "the EU". The problem is, if the EU can't fucking communicate. see this and this
You somehow figure out that there is a "mini one stop shop" (don't worry there is no non-mini "one stop shop", or ... well, there is, but it's useless, it's just a list of links for generic single point of contact sites for each Member State - also I think someone mucked up that translation, because "points of single contact" doesn't make much sense).
Okay, so by default you would have to register with each tax authority of every MS (member state), but naturally ain't nobody got time fo' that. Maybe this MOSS thing will help.
Somehow you find their practical guide, it's 31 pages, with a fucking table of contents and mothersaddening orphanmaking useless text. Or the 90 pages long explanatory notes to the directives. A 10 page long overview maybe finally shine some light on what the heck is going on. (Some fine folks even made a report about .. something, but I can only guess that its production required most of the leftover props from Breaking Bad.)
No, I mean, I'm sure they are fine, but... the thing about people who run software-as-a-service things, they don't want to deal with this bullshit.
It is amazing, how bad the help pages are. They contain no direct links to any page defining what commerce activity is covered under "electronically provided services", also it turns out, you need to basically negotiate with a tax authority and hope for the best. (Because every MS might define these differently. for example)
What's up with the Irish Tax Authority's IT guys? huh and wat Less Jameson, more takatakakatka!
If you only search for "eu vat", you can find some sort of mini salvation on a blog.
Finally, a half bad PDF, just a bit less than 2 pages containing only the relevant information.
They should make a shiny infographics about this. I sincerely hope some day it'll be illegal to put Word documents online as information in PDFs, and people will be punished by publicly force feeding them printed copies of the hideous info-monsters until they choke on the fucking disclaimers.
Next time we might try to look at this completely futuristic and never-before-heard concept of transferring money from an account at very prestigious US bank to a wild and unexplored land, called abroad!
1 note
·
View note
Text
oh, the workarounds of life
charles stross on bitcoin
so instead of hitting humanity until it evolves/educates itself into seeing the benefits of those things enumerated in the post, we should just opt for the easy-to-abuse tyranny, that'll just breed the same abusers that we want to get rid of via the tyranny to make the nice things work reasonably well. :|
why do people forget game theory before lunch? and if I'm thinking this right the above doesn't even need second order quantifiers, it's just the self-evident problem with top-down management/complexity.
bah!
0 notes
Link
nem, ez nem. viszont még jó ideig eléggé véges erőforrásokért megy a játék. még a legcsudajobb országoknak is szüksége van külkereskedelemre, és senki se szeretné, ha a honos cégeinek a leányvállalatait mindenféle helyi Vida Ildikók kopasztanák meg és lokál CBA-klónok mutatnának csúnyákat. ez már önmagában egy elég jó érv, hogy miért kell kicsit foglalkozni a kérdéssel, akárhova is megy a halandó. (materialista utilitariánus?)
meg annak is örülni kell, hogy még nem kell kötelezően turulmintás csadorban járkálni, és sajnos úgy látszik az ilyeneket nem lehet csak úgy alapjognak venni, mert pár százalékért bármilyen őrültségre képes egy kvázi-szövetséges. kicsit aktívabbak lettek volna a népek, kicsit empatikusabbak, kicsit proaktívabbak a csendes többségesek, valószínűleg most nem őrült kalasnyikovos kaftános kurafik krosszoznának a közelkeleten. (szentimentális?)
dehát "why not both"? még szavazhatsz a távol kényelméből. (pragmatikus?)
Meg nyitott hatarok vannak. Ez nem a 19. szazadi amerika ahol ha valaki kuzdott valamiert akkor tudta hogy a vilag elvonalat viszi elore. Miert kuzdjon valakiis olyan dolgokert amit mashol 50-100-150 eve elertek? Egyszerubb kivandorolni es 21. szazadi problemakkal foglalkozni.
58 notes
·
View notes
Text
Ugyanezt, annyira ugyanezt mondtam a kis Coldplay-fan ismerőseimnek, hogy mindjárt rákeresek a Skype historyban. (nem találtam, gyász! a pre-2014 előzmények a bitbarbárság martelékává lettek.) De ez eléggé igaz a MUSE-ra is. Hát meg úgy alapvetően ez a pop (lásd a Pitchfork 2014 50 ajánlott albumának a fele ilyen).
Tehát május 8-án a Roham "bárban", Black Cobra. Garantáltan 0% intro:
https://www.youtube.com/watch?v=-FH2cJ8eZ20
https://www.youtube.com/watch?v=aSR6RZ-5-yI
https://www.youtube.com/watch?v=-eyGSixo9hM
https://www.youtube.com/watch?v=brUqO6EJGd4&list=PLBD746ABFB523781E
;DD
Amugy most nagyon messzirol asszocialok, de beszelgessunk a Coldplayrol
engem rohadtul idegesitett, amikor igy megprobaltam meghallgatni oket, hogy folyamatosan olyan elmeny volt az osszes szamuk, mintha egy-egy kurvahosszu introt hallgattam volna, es mindig vartam, hogy most mar mindjart ELKEZDODIK de aztan vege lett, es jott a kovetkezo szam.
4 notes
·
View notes
Photo
Az emberek vakon bármit be "endorse"-olnak. Csak el kell kezdeni nyomatni a Fireball, Frost Orb, (Chain) Lightning, Blink, Swordfighting, stb. skilleket. A fele amúgy is létező JS-fasz. A másik fele meg az lesz nemsoká!
Már csak azt kéne kitalálni, hogyan lehetne NPC-nek tagelni a kedves-de-foglamatlan rekrútereket!? :o
Mikor lesz ennek vége?
Még mindig él az ötletem, hogy a LinkedIn API-jára építsünk egy szolgáltatást, ahol tényleg csak ilyen neveken fut majd mindenki. Regisztrációkor választhatsz magadnak karakterosztályt, hogy mondjuk wizard, warrior, monk, ninja, priest vagy druid leszel, és a LinkedInben levő adataid alapján bepontoz a rendszer. Pontokért cserébe szintek járnak, a cégek meg megmondhatják, hogy lvl 10-es JavaScript warriort, vagy lvl 80-as HR druidot keresnek.
10 notes
·
View notes
Text
[miniblog] systemd
upstream bugs
no way to run standalone
neither --system nor --user ("Trying to run as user instance, but the system has not been booted with systemd.")
so you must go for systemd-sysv and reboot. (kexec-tools comes handy)
error message about missing /run/systemd/private on 208, and then on 215 some even less straightforward error.
Ctrl-C (SIGINT) is too mainstream.
localectl just gives up if systemd-localed is not running without mentioning that it wants the *d ("connection timed out")
systemd-localed just exits if it cant find /etc/vconsole.conf and /etc/default/keyboard [sure just touch empty files, it's a retardeD after all, "Failed to read locale data: No such file or directory"]
there's no way to simply add a service unit to a target (you have to use ln like peasants of old [times]!)
systemd-modules-load.service just fails if it can't load a module that has no .ko files on the system
debian packaging bugs
vconsole.conf and default/keyboard not packaged
it might not be a bad thing to install the locales package (systemd-localed does not depend on it :o)
oh, and systemd-215-5-blablabla.deb fails to properly depend on libgpg-error0 systemctl: /lib/x86_64-linux-gnu/libgpg-error.so.0: no version information available (required by /lib/x86_64-linux-gnu/libgcrypt.so.20)
y u wontfix this
also, when someone just does an innocent apt-get install systemd and then an upgrade, all hell breaks loose, because dpkg wants to use systemctl to restart services, but systemd is inoperable pending reboot, and if you don't install -sysv, then .. you're stuck with a non-backward compatible systemctl/systemd setup. boring bugreport
1 note
·
View note
Note
kár. kár. varjak. gondolták sokan.
az okozott kár, egy lelkes csoportvezető, családapa, férj és barát pszichéjén elég jól tetten érthető, ha megnézzük a sírós videóját. (nem néztem meg, anélkül is elég egyszerű átérezni, hogy milyen szar lehet neki - pedig csak 30 másodperc!)
mi van, ha bordélyházban dolgozol? vagy a pornóiparban? vagy otthon távmunkázol, és az irodádba azt teszel, amit akarsz? meg amúgy is, olyan, de mégsem az. hmm, nem is pucérak. tök lehangoló. erős, ügyes, független nők, pisztollyal a kezükben. hogy mer ilyet munkahelyen viselni bárki.
hogy függ össze a sajtónyilvánosság a társadalmi igazságossággal ebben az esetben? a külső alapján ítélni nem ősi román szokás? csak nem azt akarod mondani, hogy magára vessen, amiért úgy öltözködött ahogy?
igazából nem tudom megítélni, hogy szexista-e a képi világa az ingnek, mert nem tudom mi a pontos definíció, de nem látom, hogy tárgyiasítaná a nőket úgy általában, és talán ez egy elég alap feltétel szokott lenni.
kár. kár. kár, hogy erről kell itt írogatni, mikor itt van ez a 100% tisztaságú magas-oktánszámú cukiságcsomag, amiről senki se beszél, pedig hát ezek kurvajók (a kurváktól elnézést, hogy egy pozitív-asszociációs jelzős szókapcsolatban fordulnak elő):
ez meg ez, és ugye ezek.
persze attól még van itt egy elég komoly gond. nem újkeletű, de hála az internetnek elkezdtünk róla beszélni, mert .. feltűnik, amikor egy férfi felvesz egy pólót, és ettől "nők milliói" megalázva érzik magukat.
konkrétan, mindenkinek vannak érzékeny pontjai; például genetikai eredetű, mondjuk csípőficam, tök jelentéktelen, mégis vannak a bunkók akik kinevetik azt, akinek nem volt szerencséje. ejnye-bejnye bunkók, nyomás vissza az óvodába empátiát tanulni. ettől persze, ha bicebóca gyerekkori pszichés traumáját kezeletlenül hagyja a társadalom, és 20 év múlva mindenre túlérzékeny féldiplomás lesz, attól azért meg lehet neki mondani, hogy jó, most már kurvára álljál le. (megint csak elnézést a kurváktól, hogy elindulatszavasodtak), és amikor kurva sok ilyen túlérzékeny ember van, megtalálják egymást a globális hipersztrádán, akkor csúnya dolgokat tudnak csinálni, és senki nem fogja helyretenni őket, ahogy a bunkókat se lehet csak úgy visszazavarni oviba. (meg ugye nem mindenkin segít még pár év építőkockázás.)
de az nagyon nehezen lehet társadalmi probléma, hogy a férfiak öltözködésükkel megalázzák a nőket, az viszont lehet, hogy már túl sokan túlpörögnek, mint ugye a Dallaszban, egy műbalhén.
komoly gond, hogy milliók élnek apátiában a facebookot görgetve, és a 2000es évek AIM korszaka óta nem lett jobb a helyzet. sokan érzékenyek, sokan túlérzékenyek, sokan meg akaratukon kívül érzéketlenek és így bántanak másokat, és míg 50 éve legrosszabb esetben családostul átköltöztek az áldozatok egy másik városba, manapság csak egy lassú google keresésnyire van az előző élete minden pórul járt embernek, legyen akár ártatlanul megvádolt férfi, vagy az igazáért kiálló nő.
és mivel a 2015 (is) az Ad Hominem éve lesz, had kérdezzem meg, hogy a kedves mama hogy van? - najó, nem. ennyire nem vagyok ambivalens, de annyira vagyok [túl]érzékeny, hogy az egész estémet elnyomja ez a kis cirkusz, és itt üljek hajnal 1kor egyedül egy laptopot nyomkodva, azon mélázva, hogy vajon mégis, mi lesz ebből? vajon simán kihordja lábon? (mármint a pólós bácsi) vagy jön majd a herp-derp viszont-reblog? (mármint erre a posztra)
hogy lehet egy pólón felháborodni? talán ezen esetleg.
de kár, de kár, még tobb varjú. mert ahogy lehet tényleg ízléstelen vagy gyűlöletkifejező pólókat viselni (pl. egy "kill all gays" felirat elég egyértelmű és a társadalom kb. 0.0%-a számára vicces), úgy lehet erről cikkezni, posztolni, kommentálni, twiitelni és sipákolni. ezekre reagálni. viszont-reagálni és így tovább.
finom részletek. ahol laknak az ördögök. lehet írni öltözködési tippekkel karöltve utalásokat, jelezni, hogy zavart hogy a TV-ben láttam, hogy csöcsös csajos pólóban pózoltál, de ebből a tényből levezetni azt, hogy az esa (ill. a Rosetta csapat, s mégpontosabban maga a bácsi) lenézi a nőket, az logikai hiba.
és aki logikai hibákat követ el, azokat falhoz kell állítani!
de majd kiderül, hogy mi lesz. valószínűleg adható erre egy morális kalkulus, egy számítás, egy önmeghatározó formula, amit elég ideig futtatva kiderülhet, hogy mi a jobb, ha mindenki kicsit érzékenyebb lesz, vagy ha mindenki kicsit elfogadóbb. vagy ezek kombinációja. (elvégre az is ironikus, hogy a gonosz pólós T-lenyomó-aylor Megegalázó Matt kezdett pityeregni azon, hogy pár fruska fricskának vette a frenetikus fesöniszta felsőjét!) viszont majd azt is bele kell faktorálni az egészbe, hogy lesznek mindig is emberi hibák, hibás emberek és egyéb körülmények, amikor tévedésből esnek bántódások.
és szerintem kurva nagy kár, hogy elveszítettünk egy színes egyéniséget a TV-k képernyőjéről.
Astrophysicist Katie Mack: "I don't care what scientists wear. But a shirt featuring women in lingerie isn't appropriate for a broadcast if you care about women in science.” twtr
oh. azon kívül, hogy tévedett, mert nem fehérneműben rohangálnak, legalább releváns a véleménye, hiszen egy közeli szakterületen dolgozik.
mondjuk arra is kíváncsi lennék, hogy az általa elvárt konstans autoauditnak milyen következményei lennének. (úgy járunk, mint pistabácsi a kommunizmusban, amikor kapálni menet megnézte, hogy párttagsági megvan? lenin élete megvan? kiváló dolgozó kitűző felbiggyesztve? oké. fúbazdmeg a kapát otthonhagytam!) meh
na mindegy. lényegében lényegtelen, hogy mit gondol egy szociológiában-nem-oly-jártas arról, hogy miért nem áldoz több nő a tudomány oltárán, ilyenformán csak egy fals adatpont (hiszen ő már pont tudós, azt kéne megkérdezni, aki még nem az, hogy most ettől szívesebben jönne-e vagy menne-e).
During a recent Wall Street Journal Facebook Q&A, Taylor was asked about whether tattooed men can achieve greatness and said, "The people I work with don't judge me by my looks but only by the work I have done and can do. Simple."
oh. tévedett.
twtr
oh.
kemény, kemény probléma ez. és kiváló állatorvosi lova az etikatudománynak. fogalmam sincs, hogy mi a helyes rawlsi megoldás, de nagyon érdekel, viszont szeretem a nevetséges pólókat is.
ignore this tmblr haxx, needed for correct rendering
Shirtstorm?
vicces, ahogy csak egy szót ide bedobsz az ask-ba, kicsit automata-szerű érzésem lett tőle :DDD
nekem az az álláspontom, hogy kár, hogy Matt Taylor azt az inget viselte, amit. önmagában már ez is egy probléma (munkahelyen ilyesmit viselni kábé olyan, mintha az irodád falára meztelennős plakátot aggatnál - diszkriminatív és nőket lealacsonyító; sajtónyilvánosság előtt duplán elfogadhatatlan), emellett valószínűleg egy tünet is azzal kapcsolatban, hogy 1) hogyan gondolkodik ez a konkrét férfi a nőkről, 2) az eddigiekben milyen visszajelzéseket kapott a környezetéből (pl. a munkahelyén is) ezzel az attitűdjével kapcsolatban.
az viszont pozitív, hogy reagált és bocsánatot kért (mondjuk egy halovány fokkal jobb lett volna, ha kerek perec kimondja esetleg még azt is, hogy igen, ez bizony szexista dolog volt).
ezen túlmenően: igazán csúcs lenne, ha nem kellene olyan letaglózóan unalmas dolgokról beszélnünk, mint amilyen a férfiak helyenként felszínre bukkanó szexizmusa, és ehelyett olyan hallatlanul izgalmas dolgokról lehetne inkább szó, mint amilyen az, hogy hogyan lehet egy űrszondával landolni egy üstökösön. (tessék, itt egy érv, hogy miért lenne klassz leszokni a nők számára megalázó gesztusokról - ha már a nők emberi méltósága önmagában úgysem elég meggyőző szempont.)
8 notes
·
View notes
Text
Opening the lid
This laptop I look at almost every day, now rather sleepy, just got woken up, because I was reading on my phone, and finally decided that I'm tired but can't just hit the pillows yet, so I opened the lid and it unwind itself from digital suspension, welcomed me with the royal colors of something like this and promptly went back to sleep again.
Huh. Fuck, it's doing it again. I was no stranger to this, I just have to push the power button to make it properly wake up, but it's frustrating, especially because no one else is having this problem, otherwise it'd be fixed by now. And so it's probably a race condition when I close the lid at least 2 commands start to execute the suspend procedure - at least that's my guess, and when it resumes the unfinished suspend finally reaches its goal. Yaay. Great. :|
And I don't care. I just don't care much.
It's a fundamental failing of system design. It's hard to see what goes on. People solve problems by describing them to other people, I could and probably even ought to open a ticket on Ubuntu's issue tracker over at launchpad, but I won't. I'm not really that kind of a person. I want understanding, and I want it all. Full spectrum, top-to-bottom, vertical and horizontal, yadda-yadda. I find it unacceptable that user sessions on the Linux Desktop are in this completely chaotic state for more than a decade. There are services, implemented by a group of smaller programs, utilizing other services and depending on various other components, and session managers and different levels of the desktop stack living and breathing underneath this fucking glitter.
And it's just wrapped in one complex system after another. The hardware, the kernel, the various drivers in the kernel and their configuration driven by whatever programs are usurping control from each other.
And then when people want to cut through this mess some other folks get angry (see the systemd "debate", also this).
But it's still not enough.
I've recently talked to one of my acquaintances, he seems like a great guy and wants to switch to IT, because ... well, whatever he does nowadays is just not really a career-sector. And I can't recommend a good enough material to him, honestly, I don't even know where to start.
He is 46 and mourns his passion, his dream. He was 20 and wanted to change the world. People loved him and he loved people. He was creative and spontaneous, risk-taking and great with people.
I'm sure folks love me, I even try to reciprocate, but I don't feel fuck either way, I can't seem to find anyone to love. I've drowned my dreams in hard liquors, the things and the how I'm passionate are not what people usually call as such, and I don't even know what's being great with people. I've turned a bitter cynic over the years. And not without reason.
Because everything is just fucking connected, part of the same monstrously multitudinous system, and it goes without much saying, that information technology is eating it all, slowly codifying every inane part of life.
There is a game theoretic and understandable reason why things are not better at the moment, I know why it's not currently getting magically "better and better". Sure, it does slowly, and in certain aspects, but still, my soullessly imagined high standard of perfection is not even on the horizon. Because, why the fuck it would be? The world is not driven by any external benevolent pedantic force, it's just the good old greedy hand of economics, psychology and the whole social setting of the human condition.
I'm writing this not merely because there is a lack of ideal interactive knowledge-transfer tools when it comes to teaching and education, or software documentation or discoverability (and transparency) of complex systems, nor because I just had enough of the utter insophisticatedness of others, but for the sad reason that I've been noticing that I'm incrementally getting farther and farther away from said society.
I'm still a young idealist, but only because I would be sooooo soo bad at running from others, working alone, preparing for betray (mine and of course of others), yet it seems I'm doing nothing but these.
I wanted to find a picture for all this, but after ~15 minutes of imgur on random mode, I settled for this, though I have great friends, but they can't seem to help either. Great.
0 notes
Text
Ófaszom, so meta, wow, lámpaernyő-kiállítás
miért miért miért.
nyilván nem kötelező ezeket olvasni, nyilván aki ezeket komolyan veszi, az magára vessen, aki ezeket kommentálja - mindenekfelett pont a tumblren(!), azzal meg mivan?
főleg, ha csak azért, mert a facebookon látta, hogy egy ismerőse lájkolta egy nem-ismerőse posztjaként, tehát ahelyett, hogy helyben feküdne bele az ebolába, még cipeli is azt tovább máshova el. át. ideodaszétösszevissza.
DE.
a hatos pont mellett nem lehet elmenni szó nélkül!
6 KÉT MINTA EGYSZERRE: Miért? Ezek voltak tiszták.
meglepő, mert a listában valószínűleg ez az egyik legmélyebb igazságot magában hordozó pont. ártatlan indikátor arra, hogy az elijesztett nőszemély igazából egy tervezőgrafikust vagy bécsi dizájnert akar leszopni, nem egy szinte általános alanyú univerzálisan kvantifikált férfit.
jó, mondjuk a 26-os is vicces. (mármint, hogy a túlszedett a gond.) 36-os, ahahah.
a 39 persze érdekes, mert igazából dzsoni depp semmivel sem kiemelkedőbb "férfi" a felsorolt szempontok szerint, mint az átlag, viszont kívánatossá vált, mert státuszszimbólum, hiszen sztár. (mert sokkal erősebben asszociálja a psziché a sikeres kalózzal, borbéllyal és egyéb karakterekkel, akik megmentették a világot és miegyebet tettek, mint a színésszel, aki amúgy nyilván sikeresen összerakott pár [kb 350] millió dollárt.)
kiváló példa Mike Tyson, családi erőszak és 2003-ban még csődeljárást is kért maga ellen, pedig mekkora főnök volt a Másnaposokban a tigrissel. (igen, valószínűleg jót tett neki a 10 év 2003 óta.)
érdemes megjegyezni, hogy bulvárlapok teljes oldalakat szentelnek a "sztárok" rettenetes ruhaválasztásának. (ahol főleg nőket szapulnak; (konstruktív) kritikának nem nagyon nevezhető volt amit láttam, egyszerűen csak leírták, hogy "kalap cipővel? hát ez gáz!" és hasonló magasszintű mélyelemzéseket nyomtak le, stilisztikailag megkérdőjelezhető oldalelrendezésekkel sokkolva az analitikusabb (így talán többnyire az óvatlan férfi) olvasókat.)
41, műanyag szemüveg. infantilis. hah. pont a napokban lájkoltam egy facebook posztot/galériát/képet, amin egy kedves ismerősöm ugrabugrált egy rikító citromszerű sárga napszemüvegben a barátnőjével a világ másik felén (többek között pont Hollywoodban is), és nem úgy tűnt, hogy a barátnője ijedezne a képeken a szemüveg miatt! hah :D
46, jaok
49, na igen. ez is olyan, mint a másnaposság. nekem egy fürdés megoldja, de az ilyen listákon nem segít egy 90 fokos mosás sem.
szóval a listának vége, de.
Gondolom azóta sem a közoktatás része az illendő és stílusos öltözködés, így érthető ha van aki érdeklődés hiányában soha nem szembesül azzal, hogy bűn-e vagy nem amit épp visel.
ezzel kezdődik a "cikk" (jó, ha már nem csak lista, mert legalább egy kétmondatos lead is került hozzá, akkor hívhatjuk cikknek, oké). ez ugye egy gyorsan belátható racionális (ill logikai és erkölcsi) katasztrófa, de. (hiszen tök jó lenne, ha már az lenne a legnagyobb baj, hogy minden közoktatásból kikerülő leendő PhD-ssal csak annyi a baj, hogy csálén áll rajta a zakó vagy a smink! és hát elég sok orosz realista értekezett a bűnről, de ... valahogy biztosan nem jutottak a végére abban a hétszázezer oldalban, és a divat pont lemaradt.)
de amúgy fontos lenne azt a szociális problémát vizsgálni, hogy elég sokan tengetik-lengetik az életüket magányosan, vagy keserű kompromisszumokkal kapcsolatokba ragadva. (és nem azért, mert mind önként vállalt remete vagy mert a cellatársán kívül más nincs.) ennek valahol oka az, hogy az embereknek fingjuk nincs arról, hogy mit akarnak ők, mit akar(hat)nak mások, és hogy most akkor mivan. (többek között [még én is, bah!] teljesen meglepődtem ezen a PornHub insights poszton, már-már annyira, hogy eléggé könnyen elképzelhetőnek tartom, hogy módszertani hibák jelentősen torzíthattak az eredményeken, de amíg valaki nem mutatja ki ezeket, hajlamos vagyok készpénznek venni az írást)
tehát az 50es lista lehet teljesen baromság - a szerző jól megsejtetten egyből tiltakozásra buzdító sorokkal indítja a cikkét, sejtve valamiféle visszhangot - de attól még vannak alapvető dolgok, amikkel nem árt tisztában lenni a hogyan ne ijesszünk el nőket témában (don't be unattractive), persze nagyrészt nem a ruha teszi az embert, hanem .. a gének. há-há! (könyörtelen rögvalóság, kénköves pokoljárás, és túl sok ingyen pornó a magányos estékre!)
szóval az emberek nagyon erősen hagyatkoznak a szemükre, mégis vakok a kérdésben.
lámpabúra toposz, ha valakinek esetleg kéne.
na, most hogy már kidolgoztam és szétinterneteztem magamat az elmúlt két napban, lehet, hogy tényleg kéne alaposabban fürdeni, mert azért az tud ijesztő lenni amikor meg akar ölni a saját hajam ;o
2 notes
·
View notes