Posts by Charlie Clark 12166 publicly visible posts • joined 16 Apr 2007
I seem to remember that 10nm was considered to be a bad idea – you get problems at this scale without sufficient benefits in lower power – which is why other fabs went straigt to 7nm: half the size so lots more transistors on the die, lower power and still the same quantum / RF problems to deal with. This gave Intel an advantage for a few years because of its established 14nm process.
The laziness and hubris argument doesn't match the reality
Not regarding process technology. Intel was a bit guilty of assuming that it didn't need to do much to continue to succeed in the data centre space, where it made all its profit. And, let's face it, this is where it's still doing very good business. But it repeatedly fluffed mobile and wasted a lot of time trying to compete with nVidia for graphics and waited to long to get into the custom chip business.
The company deserves a lot of criticism but looking at the numbers, I wouldn't mind failing that bady.
Moore's law really refers to the number of transistors on the chip. But it was only an observation of improvements in process technology. Clock speeds hit upper limits a long time ago which is why we started seeing multiple cores as a consolation. Intel used to be top of the process game but lost the crown a few years ago as it lost focus chasing after the mobile, machine learning, etc.
Firms that build physical things will always be more vulnerable until full automation
But robots don't buy things. Anyway, any modern electronics assembly line is close to clean room standards. Though this is more for the benefit of the components than the workers. But a supply-side analysis is flawed anyway, because the problem is going to be weaker demand as more and more people go on short time or lose their jobs. In such situations a new I-Phone might not be top of their priorities.
Exposing a Salt master to the internet is not best practice and firewall security should be implemented.
Anything that can be remotely configured using Salt, Ansible, etc. should never publicly allow root login without secure authentication. Generally, however, the scripts are limited to setting new instances up.
My Jabra Sport headset still has great battery life and has done sterling service for years, mainly giving me navigation instructions while cycling. Unfortunately, it has recently developed a problem with sound generally being lost of on one side. This is probably a dodgy connection but I haven't worked out to fix it. The replacement is a Sony headset (SP500) with external drivers. While the battery life isn't as good, you can get better cheaper, the headset can safely be worn in both ears while allowing me to hear what's going on: traffic, announcements, etc. Great when cycling though less great on a plane…
When I'm in the right environment the sound is also better than the Jabra (but I don't want a sports headset to listen to Ludwig van) and it also seems slightly less sensitive to interference.
Users can expect to get five hours of playtime between charges, or 2.5 hours of talk-time. That's roughly on-par with Apple's AirPods Pro.
At that price? For something almost designed to get lost? No chance! Also, how do they handle background noise for the wearer when they're out running? I'm pretty impressed with my Sony's handling of this when out cycling. Yes, it means that I don't get the highest fidelity but I'm much more concerned about hearing traffic and announcements around me!
EXPLAIN ANALYZE is your friend but only if you've set up your database correctly. How many queries rely on stupid "optimisations" because data is nor normalised properly or indexed reasonably?
That said, SQL doesn't necessarily always make writing efficient queries easy, though Common Table Expressions do help take some of the pain out of using subqueries.
Postgres has seen continued uptake since MySQL was sold to Sun and particularly since Sun was sold to Oracle. Though this is somewhat ironic because Oracle is definitely taking better care of MySQL than MySQL AB ever did. The uptake has allowed companies like Enterprise DB, 2nd Quadrant and others to invest in Postgres which has helped make releases better and more predictable and added very important functions for larger installs. It now also has good support on the various cloud platforms for those companies who don't want to invest in DBAs and aren't worried about losing control of their data.
But I guess it's always going to be hard to find developers who are prepared to try and understand how relatiional database systems are supposed to work. Note, this doesn't necessary mean learning all the underlying relational algebra, but being able to define relationships logically and understand the flexibility of making projections from normalised data.
Ranked number four in the database market by DBEngines, PostgreSQL is often left in the shade of fellow open-source stalwarts MySQL and MariaDB.
Only by journos who don't know the difference: Postgres is streets ahead of MySQL in terms or functionality and reliability and the only open source choice for those who want to get off the Oracle bus. And Postgres does this all with a single codebase, ie. no need for different table types for "real" data, no tricy licensing.
Swarm 64 is really impressive for companies that need this kind of throughput – let's face it any FGPA-based system isn't going to be cheap – so banks, etc.
TIFKAM also had a thought-out standard and has some good aspects to it but it went too far on the flatness, making non-touch UX really awful. It's interesting, and constructive, to compare the design manuals since then with especially Material Design emphasising the importance of visual feedback in elements to support the interface, as has been standard in design manuals for years. I like Material Design not least because the manual includes some "design defence" and examples of what does and doesn't work, so trying to educate and not just dictate.
The use of big, bold tiles based on signage has its place but quickly becomes a problem when you have a lot of them competing for attention: it's okay for passive consumption but poor for highlighting actions. I find the UI of Office 2016 for Mac a refreshing change from the coloured crayon of Office 2010 but I am really struggling with Office 365 for Windows with its stickmen minimalism: the lack of depth is downright confusing.
Both Google and Dropbox made them. Google's was famously known as "unladen swallow" but I think they then shifted to Go for the stuff they were going to use it for. Dropbox pushed MyPy which ended up force-feeding type hints into Python. Some people have backported some stuff to Python 2 but I don't think there's anyone now seriously interested in actively developing a fork. It took a long time but Python 3 is now better for most situations.
IIRC I think there are bits of Numpy that offload the calculations to Fortran libraries, which make C look positively slow. But it could be another part of the Python stack. It's certainly not uncommon for Fortran (and C and C++) libaries be wrapped and exposed for use in Python, I know that CERN does this for some the LHC work. Python really excels at this kind of task meaning that the mathematical geniuses write the code that normal scientists without Comp Sci or Maths PhDs get to call from Python. Meaning that stuff is both fast and safe.
My understanding is that some of the Trio stuff has been included. But I think this is also par for the course for making such substantive changes to the language. You start with an implementation as specification and work with the feedback. This has been the case for several important features in Python.
I remember that when yield was introduced as a keyword that I really struggled with the generators and would have preferred a different keyword for generators, which is sort of where we're at at the moment. At some point I think the runtime will be good enough to decide when some code is to run asynchronously or not but in the meantime a parallel stack is probably unavoidable.
It's not all about strings and bytes. The work for some C extensions is not non-trivial because they're no good, because they'te non-trivial. At least so I've been told by a Python core developer… ;-)
The switch would have been a real problem if the original Python 3 schedule had been maintained which placed the entire burden of migrating code on the developer: all downside with, at the time, no upside (slower, needs more memory). Fortunately, the PSF sponsored the migration of some popular libraries and frameworks and in the meantime, for the few projects that haven't made the change alternatives have become available: the large number of libraries as open source helping here. One notable example was pillow, the fork of PIL which Fredrik had abandoned.
Most code that runs on Python 2.7 will run on Python 3 without too much work. However, if you have C-extensions and other low-level stuff the amount of work required to migrate can be considerable. I know of a couple of libraries that won't make the change because of this. Though at the same time it should be noted that this also means that their user base has shrunk with time as other solutions become available.
Then again, in many environments Python 2.7 can be expected to run reliably for many years, just as all the Cobol and Fortran code continues to do. Anything with public networking code is likely to struggle over time.
I can not see what python3 does that perl does not do.
For system administration, probably nothing. But there are now many problem domains for which there are outstanding Python libraries: biology, statistics, engineering, etc. And of course, "machine learning" and "AI".
And then, of course, the killer reason for many switching to Python: maintainability because readable. Sure, some people love their own perl code but hate others'…
YMMV but I was just trying to answer the question.
Typically, 5G handsets have commanded something of a premium, thanks to the cost of the modems, plus the esteem of having perceptibly faster mobile internet.
The marketing departments demanded the new name. In fact, in many situations people won't notice the difference. Good 4G networks won't have contention problems and shitty 5G ones will. On a good 4G network you can already stream video even in reasonably crowded places and on a shitty one you'll be twiddling your thumbs all the time as the connection drops.
He says a lot of things. Getting people to overreact to his mutterings on Twitter is one of his main aims. Always worth looking at what any actual Executive Order says and whether it stands the inevitable judicial review.
Meanwhile, one of the results of the anti-immigration policy is, surprisingly not an uptick in the amount of people, particularly from the underrepresented demographic groups, training as IT professionals, but the offshoring of more development to places where these can be found in abundance.
Not sure about about a second wave given how low infections are in some states, it will just continue. Even so, the total deaths are likely to be close to that a of a heavy flu pandemic (we have these regularly) but they're heavily skewed with the elderly and especially the poor over-represented.
Given the appalling turnout of US presidential elections, whoever is best able to get the vote out will win. Will be harder for the orange one this year. But anyone who really thinks they know this far out, is taking the wrong pills. A lot of things can and will happen between now and November and any of them could decide the election.
Your source is interesting, and vastly disagrees with this:
Totals are the same on both. SVT is the national broadcaster. Figures are updated in near real time.
Infection rate rate of increase is linear, which is probably the most important number, though no guarantee this won't pick up again. Count for the weekend is in and last week they were moaning that other countries weren't counting the dead in retirement homes properly.
I just wish broadcasters would cover this say 1/10 th as much as they do the horror stories. The antibody tests will drive policy in Scandinavia.
Not as far as I know. The Norwegians apparently introduced the restrictions to reduce call outs by the emergency services to people who have accidents (usually when drunk). But it's worth noting that the Scandinavians generally do what the governments recommend, which can't necessarily be said of other countries.
In the UK "infectious" generally denotes airborne (keep your distance, probably viral) and "contagious" generally denotes physical contact (wash your hands, probably bacterial). But, of course, many people do use them interchangeably. With respect to covid-19 plenty of fresh air and sunshine is as good as we can do, though I look alarmingly around me at the increasing number of people who are busy trying to convert themselves from potential sources of viral infection into definite sources of bacterial infection by wearing cotton masks for sustained periods in public. Of course, they're just responding to what they're coming across in the media…
The whole tracing idea is really a red herring and about wanting to be seeing doing something. We are dealing with a virus that is extremely infectious (though apparently not contagious) with quite a long incubation period where people are infectious without displaying symptoms. This makes testing for the presence of the virus useful for studies only, because by the time you have the results, the damage has been done.
More importantly, we have to assume that the disease is endemic. This means that, until we have reliable treatment and a vaccine that it will continue to spread and the only thing you can do is control the rate of infection. But, unfortunately, it also means that those particularly susceptible to the disease will, at some point, succumb and probably die.
Reliable testing for antibodies should help us determine whether sufficient "herd immunity" has been built up in specific areas. The results from Sweden, which continues to have the most relaxed approach to the pandemic, will be key on this. At the moment the presence of antibodies does not guarantee immunity, because we don't know enough about possible reinfection rates, nor whether it will sufficiently suppress the virus so that people who are immune can be considered as no longer infectious. The hope* is that Sweden has naturally passed peak infection.
This is why both Norway and Denmark have decided to reopen schools and nurseries. Though, it's worth pointing out that people in Scandinavia generally keep their distance from each other anyway and a great many have summer houses they can escape to.
"standards compliance" is a a set of moving goal posts. Featue detection has been the preferred approach since Google started using the UA string to tell IE users that they were going to lose support for YouTube. But the downside of feature detection is that it is granular enough to be usable for fingerprinting.
My biggest beef with any mechanical charging port is that it will suffer from wear and tair. The magnetic adapters on MacBooks up until a few years ago were the way to go for anything that is likely to plugged and unplugged a lot. Maybe we'll get there with USB-D… or wireless charging will become ubiquitous.
But for manufacturers they're a dream because they can remove other ports from the machine and expect the user to user a breakout box, which provides what they need. It's just a pity that the a lot of the peripherals are not suited to all devices: I cannot use the HDMI adapter of my Planet Gemini with my other phone or a laptop and the breakout box doesn't have HDMI and has itself only one USB-C connection which means it can't be used for charging and video, though Planet "solved" this problem by bricking its ports allowing only one to be used for charging and one to be used for data.
If this phone had made its appearance in November last year at the same time as the 11, I would've pounced on it. Instead, I bit the bullet to replace my 6s with an 8 for around £450 (plus some accessories).
I think that might be the case for many who held off. Sure, a lot of people with I-Phones will be loathe to pick up an Android. Though, if they ever do, they might be surprised at how little difference there is between a lot of the phones nowadays but the timing is likely to be an issue and they might be worried about their investments in Apple's walled garden (or anyone else's for that matter).
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
