Posts by HereIAmJH 178 posts • joined 24 Aug 2010
it became apparent quite quickly he wasn't so much interested in solutions, more "how far can I push this before it gets silly". A true engineer.
Not an engineer, a YouTuber. Taking sponsorship money to build over the top projects for entertainment. Not that entertainment is bad.
It's not like any of the Pi limitations that he found were unknown. And SSDs for speed? Can you even do 10g NICs or link aggregation with a Pi? Fast access on a NAS does no good if you can't serve it.
systemd
I don't remember why I gave up on Devuan, whether it was a hardware problem, or not being able to get newer packages for software.
I ended up trying Artix (based on Arch) a couple years ago and have been happy with it, mostly. Plasma with OpenRC and rolling updates. (no versions)
I recently rebuilt one of my 'servers' using it. I just need to get the X2Go server working so that I can re-deploy it to it's tiny cubby where it will be mostly forgotten again for the next 5 years. (monitors my CCD cameras on a remote property) A problem that still exists after decades, apps install perfectly on one distro and need lots of tinkering to get running on another.
I do kind of miss dpkg, but I'm learning new things with pacman.
It's all theater, it's not an actual threat. No phone company will hesitate even for a minute because of this. If they receive a lawful request for information, they will give it. Because it's the law. If McCarthy doesn't like that then he needs to get enough of his fellow members to change the law that they passed.
And as someone else pointed out, this isn't even a demand for information yet. It's simply a preservation request. Which means the Committee is saying, if you have info related to these people, do not purge any of it from your systems. Congress could ask for a preservation request for every person in Washington DC, and it wouldn't result in a single byte of data being released to the Committee.
Hmm, I bought most of my appliances in 2000. Durable goods and all that. Recently I just bought an old stove on Craigslist to replace the cooktop (broken ceramic) and clock (minute up button failed) on mine.
Out of washer, dryer, stove, and fridge, all are still functional with a few issues. Replaced the defrost heater on the fridge, and it's icemaker requires patience sometimes. Cooktop and clock on the stove. Replaced the drum belt (under warranty) once, and drum belt and tensioner pulley once (not warranty) on the dryer. And I'm considering just rebuilding the dryer with new heat elements, drum belt and pulley if I can find the parts. Maybe a redesign on the drum belt, it has been an ongoing problem.
Haven't had to repair the Washing machine in 21 years, so I have probably jinxed that.
OTOH, I have a John Deere commercial mower that I just scrapped because of a broken planetary gear in one of the hydrostat transaxles. Parts are NLA, and you can't even buy the whole transaxle any longer, even if you wanted to shell out $2k for it.
I wonder who DOES dump all of that plastic?
It comes from a lot of places. Most storm sewers drain straight to waterways. So when someone tosses that water bottle out of the car window it gets washed into a river that flows into an ocean. Plastic bags are a huge problem, because animals get trapped in them, or they break apart and add to the micro plastic problem. There is just trash that is dumped in the ocean. Little regulation outside of territorial waters and it would be extremely hard to enforce. And then there is the fishing gear (ropes, nets, etc) that others have mentioned.
If it's collecting in mats in the ocean on its own, so much the easier to dredge it.
It's not just collecting in mats or patches in the ocean. It's also washing up on beaches.
http://in-mocean.org/wp-content/uploads/2019/10/cropped-DJI_0174.jpg
Having just disassembled one of my old server cases, the plastic on the front is ABS. 99% of the rest being sheet steel. Supposedly very easy to recycle once separated.
Hopefully these shortages will drive demand for recycled plastics, encouraging more recycling. Last year my city considered stopping curbside recycling because demand for plastics and cardboard was so low that they were having to pay ever increasing prices to get rid of it. And even then, they couldn't be certain that it wasn't just ending up in a landfill.
I made the point that we had to start somewhere, and separating recyclables at the source was the correct place to do it. Even if we had to pay to dispose, and they still ended up in a landfill. Because if we didn't do it, it was guaranteed to be trash. Paying for principals. But it also means the supply is there, we just need companies to start demanding it.
..which will be progressively blinded by the clouds of Internet satellite constellations shot up there, Starlink being the first one.
Simple, make it a requirement that Starlink(SpaceX) create a replacement for Hubble and launch it on their dime. Then we'd be good for the next 30 years.
You could call it a good faith gesture to astronomers, or a cost of doing business like subdivision developers that have to do initial public roads and utilities.
I occasionally get the 'pleasure' of doing technical interviews for contractors that my team is looking to hire. I'm sure I was an awful person in a previous life, and karma is paying me back.
Recently we reviewed a couple dozen CVs for an upcoming position. We interviewed a handful. And only found one that had the experience reflected on his CV. Interviewing applicants is like online dating; you find the perfect person, only to find out when you meet them that their profile (CV) was 90% bullshit. Not sure I can continue to live with the disappointment.
If you find the perfect position and you want to pad your CV with tech that you don't have work experience with, use the Internet. There is probably a YouTube video for everything. Being able to intelligently discuss that tech may not get you the job, but it's better than the stupid look on your face while you try to figure out a response to a question that you know nothing about.
The point is, for various legitimate or not reasons, there are plenty of people who are not responsible for the electricity supply they use, and therefore have a theoretical incentive to mine until their cards produce the magic smoke, even if it does cost £2 of electricity to mine £1 of ₿ (or, more likely, ETH).
My first thought was what happens when smart guy in the IT department at my employer decides to turn this feature on for all the Work From Home laptops? The company could do lots of mining at the employee's expense. And many would never know why their power bill suddenly went up. Thanks to group policies, the only option I'd have is turn off the laptop every evening. Basically meaning more unpaid overtime because I'd have to wait for long-running tasks to complete before I could log off. Or find a new job.
Breach of contract was my thought too. Which makes me wonder if Woz was compensated by Reilly. If there was an exclusivity clause. And if there was a timeline for performance. Surely Reilly didn't expect to use Woz's name forever.
The article hints at copyright on marketing materials. So I wonder if it didn't just have the wrong focus, and the lawsuit is not actually about a competing boot camp using his name. In the end apparently both failed.
My guess would be that they hardened their network; turning off protocols, crypto suites, and ports that are vulnerable or not needed. And as a result their distributed software couldn't make the connections to various services it needed.
As an example, I've seen .NET 4.7.x applications running on Win2016 that would insist on using TLS10 on a system that only has TLS12 enabled. So the apps fail when you start locking down.
Since they neglected security until it was too late, now they will have lots of heartburn while they clean and secure their systems.
This is the major failing of understanding to the entire OSS / Linux community:
It's not a failing of the communities, it's a matter of perspective. If you have the code, you have the option of fixing it yourself or PAYING someone else to do it. You don't have to depend on the maintainers to address your concern. OSS isn't a guarantee of a free ride.
And people who contribute to OSS projects do so in many ways, not just coding. You can code, write documentation, test, give $$$, help manage the community.....
The market will find it's level. As the throughput drops off the heavy users will move on. The article argues that providers will use Starlink as an excuse to not build out infrastructure that they are already not building out. And the argument being used against Starlink could also be used against 5g wireless building into rural areas. Small wireless providers that currently support rural areas are going to get hurt regardless, if they can't keep up with technology. From the customer's perspective it all improves, or they stay with their current options.
I tried Hyper-V, but none of the Linux distros I needed to run would install. I would be willing to try WSLg, but they require running preview Windows builds to do it currently. I'm willing to risk a beta for the one KDE/Linux app I really need. But I don't want to beta the base Windows OS too. I guess I need to stop procrastinating about getting QEMM installed. Would have been nice to have my one KDE app as a shortcut on my Windows task bar.
Other than the 1400 jobs that probably won't materialize, it sounds like they just resized everything to sunk cost and both sides are basically walking away. But someone has to have a win by 'saving' the taxpayers some money. (that was never going to get paid out)
The lawsuits against IBM and Microsoft weren't about being a monopoly. If it was, they would have been broken up like AT&T. It was about anti-competitive behavior. Using their dominant positions to block competition. Example, tying software and service contracts to hardware purchases. One of the biggest reasons Microsoft exists is IBM was afraid that if they created and supplied the OS for the PC market they would get sued for tying them together.
Microsoft was using their dominant OS platform to block competitors by illegally tying non-OS products. Office had, and possibly still does, undocumented APIs to improve it's performance and integrate it with Windows. Internet Explorer, nuff said. MS kept modifying Win32S until it reached a point that OS/2 could not run large parts of the Windows software market.
As far as Twitter buying competitors, they would have to pass a Federal review to do so. And would likely fail. Consider the processes that derailed the AT&T purchase of T-Mobile and then approved T-Mobile and Sprint merger. Canadian Pacific merging with KC Southern railways is a current example. Although to be clear, none of them are monopolies in the legal definition.
Monopoly is defined by barriers to entry. Such as your local phone company being the only one getting permission to run phone lines. Or extreme cost for another provider to run duplicate infrastructure, like gas and water. But even extreme cost doesn't make a monopoly by itself. Cell phone providers are regulated because they are an oligopoly. It costs billions to build a new nationwide network.
A competitor to Twitter has neither of those. Parler and Frank are both examples of how it is not a monopoly. You could launch a new platform on a hosted website, and then grow out of it. Register a domain, spend an afternoon coding, and a few $$ a month. The barrier to entry is getting users (customers) to switch platforms. Which means that the new sites have to be competitive. You know, the pure capitalism that they claim to want.
Everyone used MySpace before Facebook came out. AltaVista used to be THE search tool before Yahoo and Google. I haven't seen an AOL CD for a very long time. I'm sure there is a whole list of Internet companies that used to be the leader that no one could compete against. Until someone did.
Chinese wall goes back a long time, and probably didn't originate with IBM. The term likely referenced the Great Wall of China and is used in business to denote a segregation of information with-in the company. Often for regulatory purposes.
The case you are talking about, at least the second half, I have always heard called a Clean Room Implementation. Although I guess they could work in parallel. You get the requirements (headers, API definition) and write your own implementation. There have been a number of JVMs built this way. Clean Room's purpose is to negate lawsuits, but doesn't always work.
For me, the contract screams "walk away". Because I have been around long enough to recognize the scam. It's just another variation (although a little better for the individual) of the tech schools with no certifications that advertised on late night TV to train you in IT or nursing and then get you a job in the industry. They never did.
It also seems like the salaries are pretty low. What the developer was getting is Code Monkey pay around here. $15/hour is what they are wanting for minimum wage in the US, and that is just short of £23k, without discounting for the £7k/yr training costs.
It doesn't matter how experienced you are, there is going to be OJT before you are up to speed with a new company. Without details of what the training was, there is no way to tell if it had any value to the employee. If it is business specific, for example; here is our development process and this is our tool stack, then it's probably a sunk cost for the business and provides little value to the employee.
And like someone else mentioned, I don't see how the Director's time is training. That would have been a red flag for me on the contract. It sounds more like they have high turnover and they use new employees to cover the Director's paycheck.
These guys did get some very beneficial training though, Read, and understand, the contract BEFORE you sign it!
You either didn't read, or grasp the complexity of the calculations. It's not operations per second. I'm assuming you're talking instructions, or MIPS. Calculating a device location requires multiple data points for single device within a designated timeframe (< sec). Then looking at the signal strength to triangulate. You have to take into account topology and any obstructions that will affect signal strength. Consider this, you have 3 towers surrounding a park. Two have open field between them and the device. Tower 3 has woods. Tower 3 is going to have a weaker signal during the summer and about the same strength once the leaves drop. You also have to take into account hills or buildings. All these factors will affect the accuracy of your location calculation. If your use case is that you want to monitor traffic on a street, your accuracy would need to be within a radius of a hundred meters or so.
And throwing more cores at it likely won't solve your problem. With today's architectures you're more likely to be I/O bound than CPU bound. While you can determine a location with 1 or 2 towers, accuracy really suffers. 3 towers is really the minimum for it to be reliable, but you could have more in areas where towers are close together. A payload is going to be a minimum 25 characters. So every 10 seconds you're going to need transfer ~75 megabytes of raw data. (25 bytes x 3 towers x 1m devices, fixed width, no delimiters) And a million devices is a pretty small number.
And sure, that is doable with a dedicated pipe going to servers dedicated to the task. But cost/benefit isn't there. You pay for the data. You pay for the dedicated lines to each telco. Then you pay to process the data to get it into a form that you can overlay on your city map. Or you go out to the location and install a dedicated traffic monitoring device that you can move around town as needed.
The data is there, but I doubt that your average city has the competence to handle that volume of data. The cell market is much larger than it was 10 years ago, and growing all the time.
You need to understand what is being collected: every time your device communicates with a tower there is an entry created in a database. The data was originally collected so that cell carriers could diagnose and tune their networks. It contains data like tower id, handset identifier, signal strength, and antenna sector. Note that it does not contain an actual handset location. But your phone will be talking to more than one tower at a time, if there are more available, so that it can choose the best tower based on strength and available bandwidth.
For a city to use this data they need location, so they can determine movement. That means matching device records from multiple towers and triangulating every device on every tower within their jurisdiction, while taking into account obstructions and topography. Then they have to overlay that on maps to get useful traffic patterns.
Imagine you have only a million handsets active in a given area, with each one communicating with towers every few seconds. You need 3 towers to get a good location. Even if you only checked every 10 seconds (which might not be often enough for traffic, depending on speed limits), you have to make 6 million calculations a minute on 18 million datapoints. Just to get the 'raw' location data. Then throw in the need to get the data from ALL carriers in the region so that you have the full picture.
It's not impossible. And I suppose the Telcos could sell anonymized data that has been pre-processed to determine locations. But its a huge task that would be pretty expensive for local governments on a budget.
The user needs to have the ability to turn it off. It needs to be off by default. And there needs to be a visible indicator that it is turned on.
From the article:
"Wi-Fi will cease to be a communication-only standard and will legitimately become a full-fledged sensing paradigm,"
No, Wi-Fi will cease to be a communication standard, and will become a surveillance standard.
I didn't sign a lease agreement, and I haven't signed up for any service. Granted, my machine is still in the box, but none of the pre-purchase literature said a word about needing to pay for a service to use the machine.
Consider this, you buy a new printer. On top of the supplies, the manufacturer now wants you to pay them a fee to use their cloud service to print anything. Would you buy that printer?
There is a lot of comparison to 3d printing, but the device I have is a simple 2d device not unlike a plotter that scores the media. The most common use is vinyl decals, but they are also used to cut gaskets that are hard to find. (that is why I bought it)
If I'm going to have to pay for a cloud account, I'll look at the competition. Basically they made a bad business decision trying save money by using cloud instead of a locally installed app. Now they need to find a way to pay the hosting fees.
Gold, silver, diamonds all have industrial uses. So while they may not be worth the 'ooh, shiny' value, there is a floor on their value as long as that use exists.
You can't say the same for crypto currencies. They have value for 3 reasons; bypassing government control, the latest new thing, and lots of suckers to fleece. Governments aren't going to allow crypto currencies to remain unregulated. They can't, because currency regulation is how government policies are enforced. IRS can't have people hiding assets in BTC. DEA can't have drug dealers converting their cash to launder it. Governments can't enforce sanctions if they can't control the money.
My experience has been that when you have a weak signal, your battery usage increases. If I'm going to be in an area with poor data coverage I'll turn 4g off on my phone to keep it from killing my battery in just a few hours. Verizon has very little 5g coverage. If you want to use it, you should probably figure out where that is and only turn it on when you are in those areas. For most Verizon customers, that will mean leaving it turned off for the next year while they build out their network. But that isn't the message they want to give their customers, which is probably why the tweet was deleted.
So it is difficult to see what this addition to a crowded market has to offer - apart from the "Pi" branding.
The sole extent of my knowledge on the Pico is this article, so far. But it appears to be a mid-point between rPi and NodeMCU type boards like ESP32. I still have some questions about code deploy, supporting libraries, and communication options. But I can see it falling in an area where ESP32 doesn't have enough programming/processing capability but running a full blown OS (Linux) is too fat.
This was a verification of years of simulation and modelling that was supposed to prove the design before any construction even started. A sensor reading out of range shows there is an important detail missing from the model that makes the simulation invalid.
This is what bothers me. When I'm testing software I never worry about the things I know to test. I build my test cases and expect everything to pass. I don't make excuses why something 'almost worked'. If it's test parameters then you fix them just like a defect. What worries me is there are ALWAYS things that could fail that you don't know about. You either find them and fix them during testing, or it bites you in the ass in the real world.
Boeing isn't going to build a dozen test vehicles and 'fail early' like SpaceX does. With their recent track record, saying "no worries, it's just a bad test parameter" isn't good enough. It was just a 'bad test parameter' on a clock that caused their launch problem on their test capsule.
They had equipment failure and their test failed under perfect conditions, and the problem is the test parameters were too restrictive and everything will be all right when we put people on top of it and light the candle?
Well, failure to make orbit was a success previously. So maybe.
Since they had inserted themselves into the build process, hashes for the deployed code (integrity check) would have come from the already modified code. As a result, if a clean build was done it would actually trigger the alert as being suspect. You'd need to bump your release and issue a whole new set of hashes.
Maybe I missed it, but I didn't see anything in the article that linked the homes to employment. They are just homes being built in areas where Amazon is opening new offices.
If you have to be an Amazon employee to live in these new homes, then it's problematic. It's bad enough in America that you have to make career decisions based on employer 'supplied' health insurance. I can't imagine having to consider a job search and home search at the same time.
OTOH, if Amazon went in and said 'our analysis of the demographics surrounding our headquarters show a lack of affordable housing, and we can do something about that.' I don't see a problem, even if they earn a profit from rentals, if it's open market housing.
I would actually prefer a light weight core browser that supported features through extensions. With a proper sandbox and security model. There are so many default 'features' in many browsers that I will never use. And before long your browser is consuming a gig of ram. That was the original promise of Firefox, and look where we are now.
Technically, this would be range checking. Making sure the values you are manipulating are within an expected range. In this case, > 0. A pretty common occurrence is list handling. Users shouldn't ever see 'subscript out of range', for example. But coding practices have largely become let the shit fall where it may. and we'll address it in a future release.
Bounds checking, OTOH, or the lack of it, is the more dangerous incompetency. This is where no one is checking if the data will fit in the memory allocated. And leads to buffer overflows and exploits.
Delphi had both bounds checking and range checking since the 90s. Other languages don't. Either because someone sees them as syntactic sugar, or because it slows down their app. Seriously, there is enough other crap slowing down apps much more than bounds checking. It's time for some improved tools.
Technically, we do pay for GMail. There is value in our digital footprint. We just aren't paying in a common currency. It's more of a 'payment in kind'. But make no mistake, Google is providing us a service and we are giving them valuable info in return. GMail is no more a gift than Google search is.
Having said that, I'm depending more and more on my domain parked on GSuite and I'll probably look to move it onto another service provider, migrating away from Google mail servers.
The Internet isn't changing, you've just finally become aware of what is has been for over 2 decades. Probably since it moved from an academic only platform to the general public, it has been a commercial enterprise. Amazon launched in '94. eBay moved from 'sell stuff you don't need' to stores a long, long time ago. I rarely buy anything used there anymore.
If you want to create your own web space, you can still use a web host or dynamic DNS. Most people prefer the ease of Facebook or Instagram. The only business model for public web sites that seems to succeed is free content with advertising. As such, there is going to be a battle over what the appropriate amount of advertising is going to be.
Except Edge and Opera are now Chrome with a new wrapper. I've been reasonably happy with Opera, since I got tired of Firefox's bloat and bullshit. And if Google makes this the new core Chromium, then it's going to be hard to get away from. It would take a lot of work for open source developers to stand up a new browser engine. Sure, you could fork the code. But Google could just bury it deeper with cross dependencies. Think SystemD.
The Register - Independent news and views for the tech community. Part of Situation Publishing
Biting the hand that feeds IT © 1998–2022