When basing hiring decisions on emotional intelligence quotient (EQ), what reasonable accommodation would you make for candidates with Asperger syndrome, a disability that causes social awkwardness? Or is lack of autism considered a "bona fide occupational qualification" (BFOQ, BFOR, or GOQ) for most jobs nowadays?
I would be happy with Xubuntu, as it's the same OS that I used on my last netbook from fourth quarter 2011 to mid-2017. But does "reflash[ing] the firmware with something like MrChromebox's Firmware Utility Script" cause me to lose eligibility for warranty repairs on the hinge or power jack? I had to have my last netbook's power jack repaired under warranty once.
I visit your web page and stay for more than ten seconds, you get a penny.
How would the website know whether I viewed it for more than ten seconds if I've turned off JS?
I'm be totally for this rather than ads or site-specific paywalls or being data-mined.
And how would the micropayment processor assure readers of their privacy? Because the main problem I have with Google's "Contributor" micropayment system is that it shares a parent company with AdWords and DoubleClick and therefore likely shares Contributor users' browsing history as well.
Yes, there is something that ARM hardware cannot do. It cannot directly run the existing x86_64 or i386 software.
Neither can RISC-V. Both ARM and RISC-V must use interpretive emulation or static or dynamic recompilation to run an executable compiled for x86, x86-64, Z80, 65816, MC68000, SuperH, PowerPC, CLR, WebAssembly, or whatever other ISA.
(In context, the question appears to have been intended as "What does RISC-V do that ARM does not?".)
What proportion of the OEM cost of an ARM chip is licensing fees ?
Billions and billions...
That's not a "proportion". Let me rephrase Alain Williams' question: What fraction is "billions and billions" of royalties payable to ARM out of tens of billions and tens of billions in the overall cost of manufacturing hundreds of millions of processors?
Most (all?) chromebooks can be repurposed to run a full blown linux if you want to, or you can run chromeos in developer mode which is basically linux anyway.
As I wrote in this journal entry, a Chromebook in developer mode will wipe its storage if someone else turns it on and looks at it funny. This loses all installed software and all commits that have not yet been pushed to a remote repository. How would one go about repurposing a Chromebook to run GNU/Linux without running the risk of it being wiped?
Other than your $5 figure vastly inflating the value of ad impressions these days
The $5 figure is based on the minimum buy-in for a subscription to ad-free use of a website, which in turn is based on fees per transaction charged by payment processors as well as the opportunity cost of serving a paywall notice without ads to visitors instead of an article with ads. Some sites will offer access for, say, $5 per month or $20 per year (buy 4 months up front and get 8 free).
I would be perfectly fine with the option paying money to not be bombarded with ads and tracking scripts. It’s why I’m a subscriber at sites like Ars Technica.
You mentioned "sites", plural. To how many such sites do you subscribe? This becomes important if an article on a site to which you subscribe cites an article on a different site to which you do not, and you want to follow the citation. It also becomes important when searching the web, as Google Search ended its First Click Free policy six months ago, and it would become frustrating when most of the results are from sites other than those to which you subscribe.
If your website can’t survive without treating your visitors as a product then the website doesn’t deserve to exist.
By this measure, would you conclude that Slashdot "doesn't deserve to exist"? If so, why do you continue to use such a site? (You didn't post with Karma Bonus, so it's hard for me to tell whether you're offered the Disable Advertising checkbox.)
i'd rather sites that offer nothing of value just died
The availability of your home broadband Internet connection is subsidized in part by economies of scale from serving other subscribers in your city who enjoy viewing what you call "sites that offer nothing of value". Let's say hypothetically that most ad-supported websites close their doors a month from now, causing other subscribers to the ISP for your city to cancel home Internet service because their favorite sites had closed. Then the ISP gives you a termination notice on grounds that it is no longer profitable to offer home Internet in your city. How would you proceed? Would you instead access the Internet at a public library?
Couldn't you use your mobile phone as a Wi-Fi access point?
Not in my case. I have programming jobs for two different companies, one in an office and one from home. I work on projects for the latter to pass the time while riding the city bus to and from the former. Neither provides me "a company mobile". And with many of these being graphical and interactive (yet lightweight in CPU use), I would need to tunnel X11 or VNC over SSH, which would run up the latency and data usage even if I do manage to install some sort of X server or VNC viewer.
An Atom CPU is no worse in performance than a similarly clocked Pentium 4 CPU.* Thus an Atom laptop can still hold its own running Xubuntu, especially for things like lightweight hobby or contract programming work to pass the time on the bus commute to and from one's day job.
* Yes, this is telling about how inefficient NetBurst was, but bear with me.
One difference is that Google Chrome, the pack-in browser on a Chromebook, is more capable (in support for web platform features) than Safari, the pack-in browser on a Mac or iOS device. And any third-party web browser on an iPad will have exactly the same deficiencies in support for web platform features as Safari due to their shared Apple WebKit engine.
The real version of AdBlock Plus has been malware since they started deciding some ads were acceptable for the end user.
If you oppose all web advertisements, would you prefer having to pay $5 for each distinct domain that you visit in a month?
How about we gets less intrusive and trespassing ads?
Personally, I agree. And I admire Daring Fireball's print-like model, also seen on Read the Docs, where the advertiser sends the ad image to the publisher and the publisher hosts it. Firefox Tracking Protection blocks ads that track me but allows publisher-hosted ads, such as those on Daring Fireball and Read the Docs. But I imagine that fibonacci8 would disagree because "deciding some ads were acceptable for the end user" would amount to "malware".
It even suits my needs while traveling because my travel device only needs to SSH to my main machines, and provide a web browser.
Good luck SSHing from a moving city bus. It won't stay near one Wi-Fi access point long enough for your Chromebook to associate. If you're buying cellular Internet service just to use SSH from your Chromebook, you end up needing to include the price of a cellular subscription over the course of your Chromebook's useful life in its effective price.
The real version of AdBlock Plus has been malware since they started deciding some ads were acceptable for the end user.
If you oppose all web advertisements, would you prefer having to pay $5 for each distinct domain that you visit in a month? That'd make web search engines a lot less convenient. If you have a third option in mind other than ads or paywalls, I'd be interested to read it.
Email isn't Twitter. There is no reason to not use the full link
Other than that RFC 2821's definition of "text line" limits SMTP line length to 998 characters. How should someone who needs to send a longer URL to your clients do so?
I mean email is OK in the beginning but then allow us to switch to proper 2FA (oauth, gauth type stuff).
The advantage of email authentication is that SMTP is federated in a manner that doesn't require a preexisting business relationship between each identity provider (IDP) and each relying party (RP). OAuth-based authentication protocols, such as OpenID Connect, require each RP to register with each IDP to receive a client ID that uniquely identifies the RP to that IDP. Because this client ID cannot be reused with other IDPs, this is an n^2 problem, meaning doubling both the number of RPs and the number of IDPs means quadruple the overall effort to issue and track client IDs. The OAuth 2 specification describes an optional Dynamic Client Registration (dyn-reg) feature that allows an RP to automatically obtain a client ID from the IDP on the RP's first connection to that IDP, but to my knowledge, none of the popular IDPs offer dyn-reg.
Moreover, a lot of IDPs offering OAuth-based authentication expect RPs to keep part of the client ID secret. This is especially true of IDPs that use the HMAC-based OAuth 1.0a, such as Twitter, instead of the bearer token-based OAuth 2. This makes authentication from a desktop application, mobile application, or single-page web application impossible because the client ID cannot be kept secret from someone with a copy of the executable and a debugger. I've explained this before, as has Okta's OAuth guide, but Twitter has threatened to revoke and refuse to reissue client IDs that leak.
Slashdot Mirror
When basing hiring decisions on emotional intelligence quotient (EQ), what reasonable accommodation would you make for candidates with Asperger syndrome, a disability that causes social awkwardness? Or is lack of autism considered a "bona fide occupational qualification" (BFOQ, BFOR, or GOQ) for most jobs nowadays?
If you seek freelance gigs as a DJ, that might help.
I would be happy with Xubuntu, as it's the same OS that I used on my last netbook from fourth quarter 2011 to mid-2017. But does "reflash[ing] the firmware with something like MrChromebox's Firmware Utility Script" cause me to lose eligibility for warranty repairs on the hinge or power jack? I had to have my last netbook's power jack repaired under warranty once.
I expect that reading terms of service would go a long way to letting you know if your "likely" is "actual."
So let's do that. The Contributor TOS cites the general Google TOS and Google Payments TOS, which in turn cite the Google Privacy Policy and the Google Payments Privacy Notice. The Google Privacy Policy states in footnotes on "advertising services" and "linked with information about visits to multiple sites" that Google routinely uses Google Analytics data to "improve relevance" of advertising by building an anonymized interest profile about each viewer, aka the "TiVo thinks I'm gay" phenomenon. The latter footnote explicitly mentions "remarketing", a common adtech method that stalks viewers around the Web and whose common failure mode involves showing viewers things they'd already bought. Contributor is part of the Funding Choices service, which Google advertises as providing analytics to publishers about viewers who use third-party content blocking tools, though the Funding Choices TOS requires participating publishers not to correlate anonymized analytics data with actual PII.
My suggestion is that when a website — any website — offers terms of service you have to agree to, you actually read them
How long would you expect a reasonable person to spend carefully reading dozens of pages of terms of service before giving up?
Micropayments.
I visit your web page and stay for more than ten seconds, you get a penny.
How would the website know whether I viewed it for more than ten seconds if I've turned off JS?
I'm be totally for this rather than ads or site-specific paywalls or being data-mined.
And how would the micropayment processor assure readers of their privacy? Because the main problem I have with Google's "Contributor" micropayment system is that it shares a parent company with AdWords and DoubleClick and therefore likely shares Contributor users' browsing history as well.
Unless I'm misreading rv8's description, it looks like rv8 works the other way around, to run RISC-V code on x86-64, not to run x86-64 on RISC-V.
Yes, there is something that ARM hardware cannot do. It cannot directly run the existing x86_64 or i386 software.
Neither can RISC-V. Both ARM and RISC-V must use interpretive emulation or static or dynamic recompilation to run an executable compiled for x86, x86-64, Z80, 65816, MC68000, SuperH, PowerPC, CLR, WebAssembly, or whatever other ISA.
(In context, the question appears to have been intended as "What does RISC-V do that ARM does not?".)
What proportion of the OEM cost of an ARM chip is licensing fees ?
Billions and billions...
That's not a "proportion". Let me rephrase Alain Williams' question: What fraction is "billions and billions" of royalties payable to ARM out of tens of billions and tens of billions in the overall cost of manufacturing hundreds of millions of processors?
Most (all?) chromebooks can be repurposed to run a full blown linux if you want to, or you can run chromeos in developer mode which is basically linux anyway.
As I wrote in this journal entry, a Chromebook in developer mode will wipe its storage if someone else turns it on and looks at it funny. This loses all installed software and all commits that have not yet been pushed to a remote repository. How would one go about repurposing a Chromebook to run GNU/Linux without running the risk of it being wiped?
Other than your $5 figure vastly inflating the value of ad impressions these days
The $5 figure is based on the minimum buy-in for a subscription to ad-free use of a website, which in turn is based on fees per transaction charged by payment processors as well as the opportunity cost of serving a paywall notice without ads to visitors instead of an article with ads. Some sites will offer access for, say, $5 per month or $20 per year (buy 4 months up front and get 8 free).
I would be perfectly fine with the option paying money to not be bombarded with ads and tracking scripts. It’s why I’m a subscriber at sites like Ars Technica.
You mentioned "sites", plural. To how many such sites do you subscribe? This becomes important if an article on a site to which you subscribe cites an article on a different site to which you do not, and you want to follow the citation. It also becomes important when searching the web, as Google Search ended its First Click Free policy six months ago, and it would become frustrating when most of the results are from sites other than those to which you subscribe.
If your website can’t survive without treating your visitors as a product then the website doesn’t deserve to exist.
By this measure, would you conclude that Slashdot "doesn't deserve to exist"? If so, why do you continue to use such a site? (You didn't post with Karma Bonus, so it's hard for me to tell whether you're offered the Disable Advertising checkbox.)
i'd rather sites that offer nothing of value just died
The availability of your home broadband Internet connection is subsidized in part by economies of scale from serving other subscribers in your city who enjoy viewing what you call "sites that offer nothing of value". Let's say hypothetically that most ad-supported websites close their doors a month from now, causing other subscribers to the ISP for your city to cancel home Internet service because their favorite sites had closed. Then the ISP gives you a termination notice on grounds that it is no longer profitable to offer home Internet in your city. How would you proceed? Would you instead access the Internet at a public library?
Couldn't you use your mobile phone as a Wi-Fi access point?
Not in my case. I have programming jobs for two different companies, one in an office and one from home. I work on projects for the latter to pass the time while riding the city bus to and from the former. Neither provides me "a company mobile". And with many of these being graphical and interactive (yet lightweight in CPU use), I would need to tunnel X11 or VNC over SSH, which would run up the latency and data usage even if I do manage to install some sort of X server or VNC viewer.
How would widely used mail user agents know to recombine a URL sent on two lines?
An Atom CPU is no worse in performance than a similarly clocked Pentium 4 CPU.* Thus an Atom laptop can still hold its own running Xubuntu, especially for things like lightweight hobby or contract programming work to pass the time on the bus commute to and from one's day job.
* Yes, this is telling about how inefficient NetBurst was, but bear with me.
What percentage of the high-school students in the US are in this group [of students taking programming]?
100 percent, if the College Board gets its way. The College Board administers SAT and AP tests that high school students take to determine their eligibility to attend university.
One difference is that Google Chrome, the pack-in browser on a Chromebook, is more capable (in support for web platform features) than Safari, the pack-in browser on a Mac or iOS device. And any third-party web browser on an iPad will have exactly the same deficiencies in support for web platform features as Safari due to their shared Apple WebKit engine.
The real version of AdBlock Plus has been malware since they started deciding some ads were acceptable for the end user.
If you oppose all web advertisements, would you prefer having to pay $5 for each distinct domain that you visit in a month?
How about we gets less intrusive and trespassing ads?
Personally, I agree. And I admire Daring Fireball's print-like model, also seen on Read the Docs, where the advertiser sends the ad image to the publisher and the publisher hosts it. Firefox Tracking Protection blocks ads that track me but allows publisher-hosted ads, such as those on Daring Fireball and Read the Docs. But I imagine that fibonacci8 would disagree because "deciding some ads were acceptable for the end user" would amount to "malware".
I can see why one would purchase a cheap laptop with Chrome OS for their children in middle school or high school
Middle school maybe. But how would a high school student taking AP Computer Science complete his homework using Chrome OS?
It even suits my needs while traveling because my travel device only needs to SSH to my main machines, and provide a web browser.
Good luck SSHing from a moving city bus. It won't stay near one Wi-Fi access point long enough for your Chromebook to associate. If you're buying cellular Internet service just to use SSH from your Chromebook, you end up needing to include the price of a cellular subscription over the course of your Chromebook's useful life in its effective price.
And where are your "main machines"? If at home, many home ISPs use NAT that blocks incoming connections.
The real version of AdBlock Plus has been malware since they started deciding some ads were acceptable for the end user.
If you oppose all web advertisements, would you prefer having to pay $5 for each distinct domain that you visit in a month? That'd make web search engines a lot less convenient. If you have a third option in mind other than ads or paywalls, I'd be interested to read it.
The Chromebook isn't a full blown laptop that can run all sorts of high end software.
True, but it did crowd more versatile compact laptops out of the market. To what extent did the introduction of the Chromebook in third quarter 2011 cause inexpensive compact laptops to cease being a market segment at the end of 2012?
Email isn't Twitter. There is no reason to not use the full link
Other than that RFC 2821's definition of "text line" limits SMTP line length to 998 characters. How should someone who needs to send a longer URL to your clients do so?
I mean email is OK in the beginning but then allow us to switch to proper 2FA (oauth, gauth type stuff).
The advantage of email authentication is that SMTP is federated in a manner that doesn't require a preexisting business relationship between each identity provider (IDP) and each relying party (RP). OAuth-based authentication protocols, such as OpenID Connect, require each RP to register with each IDP to receive a client ID that uniquely identifies the RP to that IDP. Because this client ID cannot be reused with other IDPs, this is an n^2 problem, meaning doubling both the number of RPs and the number of IDPs means quadruple the overall effort to issue and track client IDs. The OAuth 2 specification describes an optional Dynamic Client Registration (dyn-reg) feature that allows an RP to automatically obtain a client ID from the IDP on the RP's first connection to that IDP, but to my knowledge, none of the popular IDPs offer dyn-reg.
Moreover, a lot of IDPs offering OAuth-based authentication expect RPs to keep part of the client ID secret. This is especially true of IDPs that use the HMAC-based OAuth 1.0a, such as Twitter, instead of the bearer token-based OAuth 2. This makes authentication from a desktop application, mobile application, or single-page web application impossible because the client ID cannot be kept secret from someone with a copy of the executable and a debugger. I've explained this before, as has Okta's OAuth guide, but Twitter has threatened to revoke and refuse to reissue client IDs that leak.
I live in the United States, and Slashdot is headquartered in the United States.
Queue up the guy that will say
Just one "guy"? I thought "queue" meant there would be more than one.