Daily (tech related) news diet

[TonyMac32]

4.14?

 

As for Rockchip, if any help is given @Myy from Rockchip for VPU, 4.4 will be irrelevant.

[chwe]

one from the category: "Scientific Mob"

 

https://quillette.com/2018/09/07/academic-activists-send-a-published-paper-down-the-memory-hole/

 

Disclaimer,  I didn't read the paper fully, not my field of interest...  Only the blog post. But this one shows clearly what's IMO going wrong in the scientific community.  It's not a problem that they fight against papers they don't like/ or can't agree for whatever reasons. Scientific dispute is healthy for science. The way they choose is IMO wrong. You don't stand for your opinions if you do it behind the scenes.  You don't stand for them if you force half of a board to resign in case a publication gets not rejected. 

As soon as this gets public, no matter how valid your arguments are, as soon as it's becoming public that you used your standing and your power behind the scenes, your arguments are negligible. You loose IMO all your credibility for your arguments by choosing the wrong way. The other way will work for sure, but it harms credibility on more than one field. Not only for yours, but also for future disputes. It harms the scientific community when they use the same methods which are responsible for a decrease in trustiness of business, media or politics. If you  belief that the authors arguments and theories are so wrong it shouldn't be hard to fight with proper arguments right? We should be better than trash credibility only to get a paper not published which might be questionable. Not my field to properly sort out his nor his opponents arguments. A 'censorship'  of a paper which 'survived' the peer review process of a Journal is IMO wrong, or at least it needs a discussion with public statements how this could happen. 

[chwe]

Just somehow a follow up from the last one:

 

https://www.chronicle.com/article/Should-We-Still-Cite-the/244450

 

Quote

This is a bind that we have yet to account for — how the process of building on academic work itself burnishes the reputations of people whose scholarship is good and sometimes even foundational, but whose characters are awful.

 

Similar as we do with 'good researchers'... Cite them where it's important to cite them.  A citation is not some sort of highlighting people you like, it's about link to previous work done of this field (when we talk about citations in the introduction).  Just avoid citing important papers done by assholes or sexists or whatever they did/do wrong is wrong too. IMO it's a fact that a bunch of well known researchers in my field aren't the polite people you can imagine...  Or to be more clear, the only reason you still want to deal with them is due to their huge knowledge and reputation on their field... Otherwise, you would never deal with them.

But stop citing them for this reason is IMO wrong.. It doesn't represent the research done on this field. Especially if they did important research on this field. There are IMO better ways to deal with assholes, sexists etc in research - cut their funding...  e.g. https://www.nsf.gov/od/odi/harassment.jsp

If an researcher realizes that funding his/her research isn't that easy anymore due to his/her behavior he/she might change the behavior or drop fully from research due to not possible to get it funded anymore. Cutting their power.. E.g. the MPG is known to have serious issues due to bad power balance between PhDs and their responsible 'directors'.  As a PhD you'll always be in the less powerful position, that's how research on universities works, but your university should do whatever is needed to balance this as much as possible. I assume that more PhDs will step up in the future if the PI was a pig during the thesis, which might harm the universities reputation in the long term. Some of those PIs might change when the university makes clear that they don't accept such a behavior, others not. But this seems to be a much smarter than stop citing them. Even when they're know pigs their research was important and cause it was important it has to be cited.  But cause citations are crippled anyway, it probably doesn't matter if we cripple it more (might be something for another post)... 

[TonyMac32]

This is a far too commonplace issue these days, unfortunately people want to take action on how they *feel* about something rather than by any objective facts.  Being an asshole does not mean you don't do good science.  The position of feeling over fact is antithetical to science, but here we are.

[chwe]

https://bugs.python.org/issue34605 @Igor shall we rename armbians master branch too?

 

I don't really understand what this 'clean ups' try to achieve? Does someone really felt offended by master/slave terminology in programming languages? Maybe I miss something here..

 

we could name it Gru:

 

well, maybe @zador.blood.stained might not think that GRU is appropriate..   

 

Spoiler

when I started as a lab technician we called them 'condoms for cats' - or 'condoms for x' (replace x with someone you want to piss off)... 

 

[TonyMac32]

I see too many people worrying about the words they are saying, and not worrying enough about the words they're being told not to say...  </political>

 

That was actually entered into the bug tracker?  ffs...

 

So, how about that Article 11?  Or article 13?  I think that counts as tech related, in a bad way.

 

On a lighter note, I work with a Transylvanian who sounds exactly like Gru.  Makes me smile every time.  He has a much easier life than the 1.55 m tall Scottish fellow we kept trying to get to say the Lucky Charms slogan..

"They're magically delicious!"

 

we got cussed at a lot, which was just as funny. (I'm half Scot and half German, I am triggered by this cereal box, and Budweiser. )

 

This cereal is made of whole Monsanto grain!  Extra GMO!  And corn syrup marshmallows.    I'm depressing myself, back to sipping whisky and eating Spätzle  <--- if that isn't comfort food, I don't know what is.   @chwe we should start a "local foods" thread.  Mostly because I'm hungry and want to learn some recipes.

[chwe]

https://lkml.org/lkml/2018/9/16/167

 

Linus takes a break (as it sounds at least for the rest of the 4.19 release) to fix his own behavior.. 

[chwe]

Yesterday was about linus don't understand that some people might felt offended by his communications style. Today Coraline Ada's blog post about her year at github. 

 

https://where.coraline.codes/blog/my-year-at-github/

It's somehow interesting... IMO they face both the same problem. Their communication style isn't  fully compatible with the community they are in..

 

Coraline Ada:

Quote

In speaking up like this, I felt like I was simply doing my job.

Linus Torvalds:

Quote

The fact is, people need to know what my position on things are. And I can't just say "please don't do that", because people won't listen.

 

Coraline Ada:

Quote

Starting in December, in my weekly one-on-one meetings with my manager, we would review all of my written communication (issues, pull requests, code reviews, and Slack messages) to talk about how I could improve. It felt ridiculous but I went along with it, and did my best to address my manager's feedback and concerns. I even got a book on constructive communication and effective collaboration and reported in regularly on what I was learning from the reading. My manager seemed satisfied with my progress.

 

Linus Torvalds:

Quote

This is more like the time I got out of kernel development for a while because I needed to write a little tool called "git". I need to take a break to get help on how to behave differently and fix some issues in my tooling and workflow.

 

For me this is somehow ironic.  Both stand hard for their position which might not be a mainstream position which brings them quite often in trouble and they have to defend them-self.  IMO, being 'inclusive' means that you accept both behaviors as long as they don't cross some lines of common sense. We've to accept that not everyone has the same opinions on several topics and what people assume as 'normal behavior' . Both, Linus and Coraline aren't fully what I would call average but in an inclusive environment both should find their place and felt somehow welcomed as long as they accept that maybe not everyone shares their opinion on a *random topic*.  Cause actually, some of the features she and her team developed for github are really useful e.g. as she describes:

 

Quote

One small feature that I shipped got a lot of attention: the first-time contributor badge. Basically, this was a little badge that a project maintainer would see next to the name of a new contributor to that project.

I think it's beneficial for bigger projects to spot new contributors either to have a closer look at the patch or just to congrats for the first contribution..  

 

and the other one:

Quote

A couple of months later, I finished up a feature that I was very excited about: repository invitations. With repository invitations, no one could add someone else to a repository without their consent. Being invited to contribute to a repository resulted in an email notification, from which the recipient could accept or decline to join and even report and block the inviter.

IMO I would call it a bug that the invitation is only visible via email and not reminds you friendly as soon as you login with your github account but well.. Nothing is perfect..  But I also like that I could not be randomly added to groups I don't know or I don't want to be part of..

 

And my last two cents on this one:

Quote

GitHub touts its values, but has consistently failed to live up to them. Values that are expressed but that don't change behavior are not really values, they are lies that you tell yourself.

Well IMO those 'values' are written by HR and Lawyers (to ensure they can't be sued for them..  ) and that's why a bunch of people inside a company don't give a fuck whats written there.. They don't grow organic it's more some sort of GMO, the majority accepts them cause they're cheap... e.g. 

Quote

"Positive Impact. We believe in making the world a better place through our work. Make GitHub a role model for the industry. Be a great neighbor and member of the community. Build inclusive culture."

-- From "Values" by Chris Wanstrath

smells like the marketing intern together with the law intern played a round of buzzword bingo to make an HR senior happy who had to make management happy cause they pushed him to write down some sort of 'what we are'.. If this stuff doesn't come from the people who do the work.. They don't feel responsible to follow these rules/values.. That's why IMO the companies HR department should have different 'values' than the development unit and they may have different values than the production department.. 

[chwe]

Todays special: What could probably go wrong? (It's in german, sorry for that.. )

https://www.heise.de/newsticker/meldung/Schweizer-Steuer-App-speicherte-alle-Daten-oeffentlich-in-der-Cloud-4167240.html

Random company decides that people may want a app which may help you to fill your yearly tax form (that's the latest point where I probably would be skeptical but well, seems that people want to share all their tax data with *random company*). To save money, they decide to outsource the development for such a app to an contractor from india, contractor horribly fails to develop the app which is based on an AWS Bucket with public read/write rights for more or less everyone (as far as I understand the text correctly)..  *random security researcher* tells the company that they do wrong, they think it's more appropriate to just ignore him than to see if there's a real issue until media covers up the story. One of the databases he found stored all the user credentials in plain text and all data uploaded to the app could be seen by *random security researcher*..  Just WTF?!??

If you don't have the resources to develop such an app on your own just don't do it! There's absolutely no excuse to bring such a shitshow to your end-users.. At least you need in-house capabilities to review the garbage you bought from your contractor.. How can you think it's somehow okay that user credentials for an app storing tax data of them in plain text? If you don't even get this point right you shouldn't be allowed to offer such a service.  And the best part of it, the app is called 'Steuern59.ch' (tax59) because this app cost you 59CHF (~60$)..  

reminds me to this one:

 

Edit:

just for fun, from the steuer59.ch website:

Quote

Sicherheit

Ihre Daten werden sorgfältig aufbewahrt und nicht an Dritte weitergereicht. Nachrichten werden verschlüsselt und Ihr Login basiert auf einer 3-Faktoren-Authentifizierung, wie Sie es aus dem E-Banking kennen.

'free translation'

Quote

Security

Your data will be stored carefully and will not be passed on to third parties. Messages are encrypted and your login is based on 3-factor authentication as you know it from e-banking.

So I guess 

1 factor: the guy who wants to see you taxdata needs Internet access

2 factor: he probably needs java script activated to brows through AWS services

3 factor: he needs a free AWS account cause those open AWS Buckets are only open to people having such an AWS account

 

Sorry for being sarcastic but that's the only way to deal with such news.. 

 

BTW: there is a java based tax app provided by the govt here in switzerland since years. You can load the form from the last year so that all your data is there and you just have to adjust it to the new year.. And it's not only Windows.. They provide an OS X and even a Linux version for it.. It's not that we've to submit those forms on tablets of stone...

[chwe]

There's an interesting follow up of Linuxs CoC story going on since days:

https://www.newyorker.com/science/elements/after-years-of-abusive-e-mails-the-creator-of-linux-steps-aside

 

Some parts I want to comment here:

Quote

Torvalds’s response was conveyed by the Linux Foundation, which supports Linux and other open-source programming projects and paid Torvalds $1.6 million in annual compensation as of 2016. The foundation said that it supported his decision and has encouraged women to participate but that it has little control over how Torvalds runs the coding process. “We are able to have varying degrees of impact on these outcomes in newer projects,” the statement said. “Older more established efforts like the Linux kernel are much more challenging to influence.”

and I hope this doesn't change.. Otherwise I hope that they kick out half of the 'Board of Directors' and replace them by community members to ensure they're good represented... 

 

Quote

Intel, Google, IBM, Samsung, and other companies assign programmers to help improve the code. Of the eighty thousand fixes and improvements to Linux made in the past year, more than ninety per cent were produced by paid programmers, the foundation reported in 2017; Intel employees alone were responsible for thirteen per cent of them. These same companies, and hundreds of others, covered the foundation’s roughly fifty-million-dollar annual budget.

Even when this claim is true.. All those companies could close their company without the kernel. It is in intels interest that their super shiny new server CPUs are well supported in Linux cause they'll run quite often Linux when they sell them.. 

 

Quote

There are very few women among the most prolific contributors, though the foundation and researchers estimate that roughly ten per cent of all Linux coders are women. “Everyone in tech knows about it, but Linus gets a pass,” 

out of curiosity cause I don't do in tech, I do in science... Compared to where you work, is 10% for 'tech' few, average.. etc.? In chemistry I would assume 10% would be on the lower end.. I think (without having statistics at hand) ~20% would be average. In biology it should be more ~40% maybe (just a guess)

 

Quote

For a research project, Squire used e-mails from Torvalds to train a computer to recognize insults. According to Squire’s tabulations, more than a thousand of the twenty-one thousand e-mails Torvalds sent in a four-year period used the word “crap.” “Slut,” “bitch,” and “bastard” were employed much less frequently during that period.

since when is 'grep "crap" linus.torvalds@*'  called train a computer?  Well, if we have now a look into the research paper:

Quote

There are other mild profanities referenced in [17] that were not used in significant enough numbers to be included, for example, 'bitch', 'cock', 'slut', 'bastard'.

well.. the statement is IMO misleading.. Much less vs. not used in significant... And as far as I understand, the paper analyzed not only Linus and slut, bitch and bastard where searched for everyone writing there not only Linus (and even then it wasn't there in a significant amount). Cause the "crap" comes from table 3 where they highlight Linus... 

 

Just from the paper to:

Quote

Limitations to our method were that we did not search for obscured words (e.g. f*ck, sh*t),

well.. grep could solve that for you... 

 

Quote

Many women who contribute to Linux point to another open-source project, Python, as a guide for Linux as its faces its #MeToo moment. Guido van Rossum, a white, male programmer from the Netherlands, ...

Ahhh, got it we've a #MeToo moment in linux now (I thought it was about Linus language, not that he's the harvey weinstein in tech. Well, since Avital Ronell got her title IX for abusing a man the hashtag got probably democratized for everyone and everything). I thought we don't discriminate people towards their color, color etc... So why is Guido van Rossum decriben as white male programmer.. At least from all the contacts I had to python I had no idea whenever he's 'white' or not and I honestly never cared if he's white or not. Besides those few members here with photos of them-self as 'Profile Photo'  it's also not clear here.. And honestly too, I don't care even.. 

Quote

van Rossum told me, adding that if the leaders communicate abusively “it will attract people who either share that attitude, or at least don’t see a problem with it.”

IMO this could also be related to the field you work.. e.g. the tone in physical chemistry is also often hard, people doing in physical chemistry must have a special 'mind set' (others would simply call them crazy cause they think that 100 dimensional rooms are for beginners ) to get a clue what's going on..  Those people thinking in 100 dimensions mostly don't waste much time with 'being nice' (you didn't get that you've to transpose this matrix to ensure blabla does what it should.. How dump you are - something they told me probably once a week, btw. I assume that most physical chemistry departments would see it as a success when 10% of them where female but also the females there calling you dumb when you don't get it - are they know 'badly influenced' by the men working there or is it that just the crazy people doing in PC... not sure honestly).. 

Quote

He joked that perhaps he could find a technical fix for his bad behavior—“Maybe I can get an e-mail filter in place so at[sic] when I send e-mail with curse-words, they just won’t go out.”

Well, I don't think that was a joke...  More or less every bigger company I worked for had such filters... 

Quote

It announced that the “Code of Conflict” had been replaced by a “Code of Conduct” that forbids “insulting/derogatory comments” and behavior “considered inappropriate in a professional setting.” Complaints will be heard by the foundation’s technical-advisory board, which has ten members, all men.

Spoiler

Board members:

• Chris Mason (Facebook’)

• Dan Williams (Intel)

• Greg Kroah-Hartman (Linux Foundation Fellow)

• H. Peter Anvin (Intel)

• Jonathan Corbet (LWN.net)

• Olof Johansson (Tesla)

• Rik Van Riel (no idea, not mentioned in the summary)

• Steven Rostedt (VMware)

• Steven Rostedt (Google)

• Tim Bird (Sony)

 

I assume they get mostly elected by merit so if only 10% of the submitters are women and the article complains that no women is under the top maintainers (it this true? Never checked who's responsible for *random subsystem*) chances are high that a 'technical board' will have only men.. You could argue that the TAB is maybe the wrong board decide whenever such complaints occur. 

 

For me there are two parts on this.. Calling someones work crap or calling him crap. I yell to 60-70% of the reactions I perform in the lab (sometimes with one of the 'not used in significant enough numbers'-words  | grep -v  ['bitch', 'cock', 'slut']). I don't call often other peoples research crap , there are those 'more polite forms' like 'not interesting' or 'may not scale well' or 'might need some adjustments'  (depending on what I want to criticize)..  Is this better than calling it crap.. I'm not sure.. I'm quite sure it's in general 'more accepted'. My work has sometimes been told crap (and its polite forms) since I started in a lab, a bunch  of it was crap and bunch of it will be crap in the future. Sometimes you've bad days and you produce crap (and you know it), sometimes you're just wrong and therefor you produce crap and sometimes the one calling your work crap was/is wrong. If you really believe that it isn't crap you often find a way to explain the other one that he is wrong (and if he doesn't get it you can tell him politely that he should "go *have sex with you alone*")... 

 

For the first part of the article, citing from research papers in newspaper.. That goes often horrible wrong. A research paper is somehow the 'concentrate' of your work, you're often limited to x pages (depends a bit on format and publication you choose), so you don't waste much words (e.g. explaining that 'bitch', 'cock', 'slut', 'bastard' belongs to all emails not only the one from Linus - sometimes this is also used to hide shitty results, but different story). You just expect that readers of such journals do understand 'how such stuff is written'. Scientific journalist in newspapers get this often but when journalists which write mostly not science related stuff write about papers and stuff inside those papers, they often fail. Or came to a wrong conclusion..  Don't get me wrong, paragraph describing that 'bitch', 'cock', 'slut', 'bastard' wasn't used significantly is IMO questionable written and a peer reviewer IMO should've made a comment to clarify if it belongs to Linus or LKML at all (IMO it should be in the paragraph above if it belongs to LKML but well.. both tables are somehow questionable - but 'Proceedings of the 2015 48th Hawaii International Conference on System Sciences' isn't something I regularly  follow to understand how tables in their communications are properly written). 

[chwe]

https://www.consumerreports.org/car-safety/over-the-air-update-left-tesla-model-3-without-key-safety-features/

 

Seems that not only Armbian is sometimes affected by 'software update gone wrong'  After update no more automatic emergency braking and Autopilot...  I'm a bin concerned towards all those OTA update mechanisms. If this becomes a general trend I'm quite sure there will be companies messing up so hard that this becomes a security issue in the future. Don't know how much this is regulated yet, but I think it should be. It's not only you as driver might be harmed when things go wrong.. 

[TonyMac32]

6 hours ago, chwe said:

Seems that not only Armbian is sometimes affected by 'software update gone wrong'  After update no more automatic emergency braking and Autopilot...  I'm a bin concerned towards all those OTA update mechanisms.

 

Well, if a real company was doing it, perhaps it wouldn't be such a joke.  auto emergency braking being randomly disabled can kill.  If anyone looks back 15 years from now and wonders why self-driving cars are still not mainstream, they will only need to look up "Tesla" on Wikipedia.

 

6 hours ago, chwe said:

Don't know how much this is regulated yet,

None.

 

[chwe]

5 hours ago, TonyMac32 said:

 

12 hours ago, chwe said:

Don't know how much this is regulated yet,

None.

well, as expected..   At least VW could fix mostly their add blue issues.. Not the one that their 'purified pee'-tank is to small (as we know software can't fix faulty hardware), but at least that a reasonable amount of pee is injected. Next hotness here will quite sure be gasoline engine producing to much particulate matter...

At least govt. should be somehow informed what they change here.. As with everything else, we might need some scandals here to get the attention which is needed before someone cares. 

 

7 hours ago, TonyMac32 said:

If anyone looks back 15 years from now and wonders why self-driving cars are still not mainstream, they will only need to look up "Tesla" on Wikipedia.

or uber.. 

 

Urea/Pd/NOx chemistry is actually quite interesting for those interested in heterogeneous catalysis there's a lot of research involved to get those SCR cats working. I may have a look if I find a recent paper explaining it a bit in detail. 

[TonyMac32]

Well, Uber isn't selling their cars, let alone with an (in my opinion) criminally misnamed semi-autonomous under-engineered system. "Autopilot" is anything but an autopilot. And it seems to suck.

Sent from my Pixel using Tapatalk

[chwe]

2 hours ago, TonyMac32 said:

"Autopilot" is anything but an autopilot.

well, if you're clear on your opinion towards a Musk product:

 

he might not be that happy... 

 

He probably get sued for it..  

https://www.thesun.co.uk/news/6792661/who-vernon-unsworth-thai-cave-rescue-pedo-guy-elon-musk-sue/

[TonyMac32]

https://lkml.org/lkml/2018/9/23/212

[chwe]

For those laugh at me that I still use Opera as my preferred browser (e.g. @TonyMac32 and @Igor ):

 

https://blog.cryptographyengineering.com/2018/09/23/why-im-leaving-chrome/

 

Quote

A few weeks ago Google shipped an update to Chrome that fundamentally changes the sign-in experience. From now on, every time you log into a Google property (for example, Gmail), Chrome will automatically sign the browser into your Google account for you. It’ll do this without asking, or even explicitly notifying you. (However, and this is important: Google developers claim this will not actually start synchronizing your data to Google....

 

Just no.. I'm aware that no matter what I do, chances are high that google will track me around the world as soon as I open a random webpage due to more or less every page uses some sort of google services for statistics etc. But they start to connect and link those data everywhere, and it's not that they do it in public, it's everytime that you get those informations later with some sort of 'we're sorry, we should annotate such a change, believe us it's not that bad and we'll do it better the next time' followed by the next 'we're sorry...' cause they 'forgot' it once again to annotate their new shitty approach to collect data. I'm somehow locked-in for google services (e.g. my phone would simply refuse to work without a google account) and out of laziness I use this gmail account for a bunch of other stuff too.  But happily chrome doesn't offer something unique I need.

(I wouldn't say opera is much better here, but at least they don't have the budget and dominance to act as bad as google can..  I try to set 'good' privacy settings but honestly I don't trust my browser enough no matter which one I use, they'll always find a way to collect my data... )

Edit:

 

19 hours ago, TonyMac32 said:

https://lkml.org/lkml/2018/9/23/212

btw. @wtarreau answered as well in this ongoing 'war' https://lkml.org/lkml/2018/9/20/257 ... seems that:

Quote

Everybody has been gossiping around the water cooler instead,..

(https://lkml.org/lkml/2013/7/18/426 I know, that's an old one...  )

 

Curios if this has an impact...  At least the new CoC has bigger echo than some people thought..  Discussion culture is always interesting as long as everybody agrees that we don't have to agree on everything to have a good conversation.. I would say, the best discussions I've had were often with people I disagree on certain topics. 

[lxde-OSIREN]

that sure is a lot of cache mb chwe, I always clear all but saved passwords, if you want to be the middle man of your data to see and control all network activity in and out of your computer why don't you get yourself an esp32

 

for example: lets say you know what the data form of a google update looks like coming in from your esp32, maybe it will say like "update=ui,version," you can set up your esp32 to trim out that update data so nothing gets changed.

 

someone go comment on my post:

 

[chwe]

19 hours ago, helper-dialog said:

that sure is a lot of cache mb chwe,

just to be clear here, that's not my twitter account..  I don't use twitter at all (I grew up when an SMS was 0.25$... so a tool limited to 150 chars wasn't something I thought is exciting.. IMO twitter here in Switzerland is for politicians stalking journalists and journalist stalking politicians but well.. that's their problem not mine).

 

everyone who wants to have the same 'browser-history' in all his used chrome browsers should be free to opt-in for such a feature.. People might think it's useful and they're free to  do so but I don't like the 'google decides what's best for you' approach is healthy. Similar to google decided that www.example.com was shortened to example.com.. Unluckily there are pages where example.com isn't the same than www.example.com - have fun if you are one of the lucky guys who needs such a website. For me a browser is something not really personalized.. I have a few bookmarked pages (~10) and my 20 open tabs, if a page gets closed out of those 20 tabs it was either not interesting anymore or the browser crashed..  I've a few autologin cookies. E.g. the mailaccount I don't use for serious stuff, this forum & github (only on one machine which only I use) everywhere else I need my keepass otherwise I would be lost cause I can't remember passwords..  For those like me google shouldn't decide that our browsers are now automatically linked to our gmail accounts without rational reasons why this is needed and if they do.. well, I simply avoid to use their browser anymore.. They're free to do so, I'm free to use other browsers, just reminds me once more that I shouldn't build up my 'infrastructure' only on google products. Their 'free beer' slowly turns out to be a bud light.

 

19 hours ago, helper-dialog said:

someone go comment on my post:

It might need some explanation..  

btw. you can also post interesting findings in this thread.. I don't own it and it was never my intention to be 'my personal' thread, actually I hoped that others share interesting stuff too.  

 

 

On 9/24/2018 at 1:13 PM, chwe said:

Urea/Pd/NOx chemistry is actually quite interesting for those interested in heterogeneous catalysis there's a lot of research involved to get those SCR cats working. I may have a look if I find a recent paper explaining it a bit in detail. 

https://www.tandfonline.com/doi/pdf/10.1080/2055074X.2015.1105615

 

Received 26 june 2015; accepted 12 september 2015 --> I think that's 'just before VW messed up' (or at least before it was publicly known.. ). It's a 'short' (~18 pages or so - for a review article it's rather a short one) review summarizing catalytic reactions in diesel engine aftertreatment.  Tl;tr it's not that easy.. e.g a perfect material must be active under 'dirty conditions' , has to work under various conditions (cold start engine), shouldn't harm the environment (you don't gain much when it reduces NOx fully but releases ammonia instead)..  also 'catalyst bleeding' - means that besides of exhausts, degenerates of your catalyst comes out of the pipe and to keep the engine economically it has to be 'cheap' and shouldn't increase fuel consumption.  Research in heterogeneous catalysis is a way 'harder' than homogeneous catalysis cause you've a way more variables which must be taken into account (e.g. surface, gas flow etc.) also the fact that a car engine doesn't run in a steady state the whole time. In marine ships (where the fuel is even dirtier than diesel) this is mostly solved with scrubbers where you just 'wash' the exhausts and let it (partly) go into the seawater (gets less attention, cause we mostly don't care) such a ship engine runs quite often for longer times in steady states which makes exhaust aftertreatment a way easier. 

I assume that some 'catch and release' approach could be interesting - materials which adsorb NOx at lower temperatures temporarily and release it at higher temperatures followed by an SCR which is then (hopefully) warm enough for a sufficient catalytic degeneration of NOx gases. Probably not that easy to develop.. I'm sure, VW has some spare cars to test such systems... 

 

 

And for those who got confused by:

On 9/21/2018 at 12:50 AM, chwe said:

e.g. the tone in physical chemistry is also often hard, people doing in physical chemistry must have a special 'mind set' (others would simply call them crazy cause they think that 100 dimensional rooms are for beginners )

 

Here's an example why I thingk these guys are crazy:

https://www.nature.com/articles/nature14566 (unfortunately, nature isn't open access by default, but google shows you at least the draft from it http://www.kroto.info/wp-content/uploads/2015/05/C60-DIB-orig-manuscript-1.pdf)

It was assumed  that the two bands at 9,632 ångströms and 9,577 ångströms which were reported in 1994 are related to a fullerene (C60) molecule. So those bands were found when they looked into spectra of "light' from diffuse interstellar clouds.

(https://en.wikipedia.org/wiki/Fullerene)

and some ideas are there that PAHs (polyaromatic hydrocarbons) could be 'gods molecule' synthesize RNA (simplified RNA is a single stranded DNA).  Some RNA strands are known to be able to replicate themselves (http://science.sciencemag.org/content/323/5918/1229) and if those RNA strands now somehow find a liposome ('lipid bilayer ball' - basically similar to a cell) and by 'luck'  they synthesize other RNA enzymes which now can build up and functionalize those liposomes, we would have an explanation how the first cells were produced. 

Compared to fullerene PAHs are flat and due to pi-stacking they can build layers (that's what your graphite pencil does ). The distance of such a stack is roughly the one needed for RNA synthesis:

(https://en.wikipedia.org/wiki/PAH_world_hypothesis)

 

so confirm that those bands belong to C60+ find possible PAHs in the universe and you might be as famous as Darwin cause you solved the fundamental question why (Edit: it's not really a why, it's more a how could it happen) we're here..  Unfortunately the bands found in the universe didn't match (good enough) with the one determined on earth for a long time due to a bunch of small 'errors' which can have an impact on the measured spectra (e.g. temperature, pressure etc.). 

So some physical chemists spent a bunch of time to 'simulate'  an environment similar to the one you will find in those interstellar clouds just to confirm that this small mismatch is due to environmental reasons...  You must have a special mind-set to go motivated to work day for day just to shift a spectra a few ångströms so that int matches for a molecule 99% on earth don't even know (and care).  Not by faking results - which may be possible, cause there aren't many research groups on earth which are able to confirm or deny your results.  Years of work, to confirm that 2 bands found in 1994 belongs to a molecule which looks like an adidas football from 1970's  (IMO adidas should be part of their funding... )

(https://en.wikipedia.org/wiki/Adidas_Telstar)

 

I admire their endurance in doing this kind of research but it's obviously not the stuff of work the majority of people is capable to do - somehow similar to kernel development where the majority on earth also doesn't recognize the hard work which was done - until some newspapers report one of Linus 'famous' rants out of thousands mails he wrote. 

 

 

Edited September 26, 2018 by chwe

[lxde-OSIREN]

18 minutes ago, chwe said:

Similar to google decided that www.example.com was shortened to example.com

When I host with Microsoft IIS and I setup the aaa dns records, I can type in www.orangepi.org for users to visit if I own the name, if I don't also make another record that says orangepi.org which is a very simple task, than the one without www wont work.

[chwe]

a short follow up to chrome:

https://www-blog-google.cdn.ampproject.org/c/s/www.blog.google/products/chrome/product-updates-based-your-feedback/amp/

 

seems that google accepted that this was a shitty move:

Quote

We’re also going to change the way we handle the clearing of auth cookies. In the current version of Chrome, we keep the Google auth cookies to allow you to stay signed in after cookies are cleared. We will change this behavior that so all cookies are deleted and you will be signed out.

 

Quote

We’re updating our UIs to better communicate a user’s sync state. We want to be clearer about your sign-in state and whether or not you’re syncing data to your Google Account.

 

if you now also implement a turn off sync.. that would be great.. (people might want to go from synced back to not synced without killing all their cookies or figure out which one is responsible for it.. just as an idea.. ). I hope at least that they understand that such changes should be annotated in public.. not that you've to figure it out on your own or rely on passionate testers.. I'll give your browser still a break in the penalty box..  

[Igor_K]

 

[chwe]

 If you've roommates like that... well.. you might search for new ones..  But hey.. he got 15$ and a nanopi..  

and it doesn't look like armbian..

 

[Igor_K]

There are another benchmarks of "something" https://www.phoronix.com/scan.php?page=article&amp;item=16-armlinux-sep2018&amp;num=1

[Jens Bauer]

Did you read about this security hack ?

(Bloomberg)

[chwe]

1 hour ago, Jens Bauer said:

... Probably not, but I just wanted to ask if you've read about this security hack.

 

well, realistically they put it in one of the delivered blobs than using their hight-tech microchips.. ATF (when only available as blob) is a great opportunity to serve such a feature..  or rootmydevice (just to avoid panic here: rootmydevice is fixed in armbian long time ago). 

 

BTW. the register wrote a nice cover up where they analyzed the denials of the affected companies and some points where the authors of the blooberg story may can be wrong:

https://www.theregister.co.uk/2018/10/04/supermicro_bloomberg/

one aspect I was really surprised here:

Quote

Bloomberg reporters receive bonuses based indirectly on how much they shift markets with their reporting.

hell yeah! I get more money if I mess up the market.  I really hope this doesn't end like the Subprime Credits where everyone got a house cause *someone else's bonus* was higher if he gave you a credit even when he knew that you'll probably never pay it back.. 

[Jens Bauer]

3 minutes ago, chwe said:

 

hell yeah! I get more money if I mess up the market.  I really hope this doesn't end like the Subprime Credits where everyone got a house cause *someone else's bonus* was higher if he gave you a credit even when he knew that you'll probably never pay it back.. 

 

Heh, we can't really know for sure exactly what happened, but there's a pretty good chance that it's not all fantasy.

We know it's definitely technically possible to do this - and I'd say it's likely that it happened (if a large company says 'minor security breach', it's likely not just minor...)

 

I kept thinking through reading the article ... why didn't they hide between the PCB layers ... and 3/4 through the articles that was actually mentioned as well.

 

But the next thing would likely be to implement such junk inside CPUs or other chips. Intel-based devices are just very interesting, because they're mainstream and they're everywhere. ARM is also widely spread by now - this makes the ARM architecture a target. Maybe this won't happen through the CPU itself, but "bad RAM" would likely be the most likely way it would be implemented.

-So now CPU vendors need a way to verify that the RAM is not modifying its contents on its own...

 

Security by obscurity might not be so bad after all... Using a good old Z80, which noone expects in a modern computer, could protect against such attacks.

[Jens Bauer]

... Besides, it's probably just those people behind bot-nets that are responsible ... make sure you use SSH-keys for logins and disable password authentication.

[chwe]

merged it with daily tech news, I think it fits well here not? 

42 minutes ago, Jens Bauer said:

But the next thing would likely be to implement such junk inside CPUs or other chips. Intel-based devices are just very interesting, because they're mainstream and they're everywhere.

I thought that was already done, they call this intel management engine..  

 

Well in arm world they push SPI flash for bootloaders.

I assume there's plenty stuff left if you decide to hijack them..  Compared to the tiny one bloomberg claims to be there on the Supermicro boards.. Also most (all?) arm SoCs have some sort of a secondary microcontroller which could be compromised (e.g.  Allwinners AR100)

 

Let's face it. If they want, it's not much a problem to compromise *whatever device* they want. From network infrastructure, IoT garbage, mobile phones, tablets, etc. The only protection we have is an economical one. If it gets public, some govts. may decide that it's no longer allowed to fab critical infrastructure in china which will harm their industry (we would probably lose 5-10 years updates for such a critical infrastructure cause nobody here knows how to produce such stuff for an acceptable price anymore.. ). So is it worth to harm your electronics industry? Not my decision, I would assume it's not worth. Fact is nobody cares about the Snowden documents anymore and they had some enlightening SOPs for "how to be a good spy" in it.. 

 

A not govt. related one:

 

smart engineers they have in the card payment industry.  Angular momentum determination won't be the first thing which comes to my mind to determine if a card reader is compromised or not. 

[lxde-OSIREN]

looks like an add on eeprom