Skip to main content

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - ersi

1
With Galaxy S8, Samsung has added a "desktop experience" to the external monitor, so the phone can replace a netbook now.
You have to buy a $150 dock.
Thanks for doing it so we can learn from your experience :)

I'm not super happy with my S8+. The wrap around screen is useless, I've turned all the related features off, as they are way too easy to accidentally activate. Perhaps of note also, I cracked the corner of mine rather quickly. Wrap around screen means don't drop it.
Does "wrap around screen" mean something like "edge" popularised by some earlier Galaxies? The first time I saw it, I understood it would be a problem to hold such a thing in hand when you want to just hold it, not touch anything. This would imply imbalanced grip and this in turn would imply dropping the thing, so investing in some good cover is compulsory.

My Note 4 survived much worse... I'm tempted to go back to it. With a custom rom i'm almost sure I can make up for the performance difference. Actually using the S8+ to its full potential tanks battery life.
I heard that S8 (and probably + too) at first has some features turned off and when you turn them on, it warns about battery. That's as good as scaring users away. Who doesn't know that actually using the phone consumes the battery?
2
But such words like common or general do tend to reflect a honed definition of the "regular" meaning -- the biggest catch being that it might be the 19th century meaning while the language has otherwise moved on.
Language has moved on or life in general has moved on. Or the guy who authoritatively fixed the meaning in the 19th century did not make it relevant to life in the first place.

Science may be international, but that doesn't mean it's entirely homogenized. In the English-speaking world those two generally just aren't grouped together like that. You've got the social behavioral science of economics and then you've got this separate commerce and business administration thing of which I'm not really sure how it's classified. Perhaps it's more vocational than the arts & sciences, even if it's university-level.
Yes, the obvious problem with building ontologies from umbrella concepts is the failure to reflect cross-disciplinary stuff. A true ontology goes by distinctions that cannot be crossed.
3
The word "common" basically just signifies that which we normally mean when we say x, which can be distinguished from the broader concept or category of super-x. It does make translation harder because that which Dutch or English speakers commonly mean by "economics" is not the same as German speakers. Too bad for the machines. :P
That's not how it works in science. In science you build terminology, concepts that relate to each other in fixed ways, specifically in the way scientists think reality works.

In this case, the distinction is economics broadly considered and then a subset of it. It's like in biology they give names to species. Names of species always have two parts. Bigger groups, such as families, have one part. Often a name of the species consists of "common" + family but the function of the compound is exactly the same no matter what stood there instead of "common".

To clarify the terms for the whole science (to other scientists), some people build ontologies. Biologists believe their taxonomy describes real-life ontology. Economists also arrange their terminology so as to make it an into an ontology. Yes, I have seen them do it. Funny when they try it with financial terms.

In that sort of system, something like "common economics" would not be that which people commonly mean when they speak about it. It would be about a specific spot in the ontology. When they fail to put "common" before it, it would be elsewhere in the ontology and the scientist would go nuts.
4
DnD Central / Re: Infrastructure
Schematic metro lines are considered a major invention, made by a guy I don't remember the name, that realized that when using the subway people don't need to know the geographical location but only the stop's succesion. Therefore, a virtual diagram can be much more useful than a real map.
With such insight very complex maps could be turned into easy to interpretate straight lines.
It's the way I drew maps when I was four.
5
I suspect that both you and @krake are basically confused by English here, not by German.
I'd say that English manages to be even more confusing in this case than German.

German compounding is sane enough in this specific case, even though "Volks-" is the wrong word for the purpose for which it's being employed and direct translation from German to Estonian (which happens to be how this term got into Estonian universities) makes it worse, but be that as it may, "political economy" is worse than that and "common economics" is just as good as "economics" with nothing to specify it.
6
Quote
Die Volkswirtschaftslehre (auch Nationalökonomie, Wirtschaftliche Staatswissenschaften, oder Sozialökonomie, kurz VWL) ist ein Teilgebiet der Wirtschaftswissenschaft.
Isn't it directly deducible here that

- Wirtschaft = economy
- Wirtschaftswissenschaft = economics = science about economy
- Volkswirtschaftslehre is something more specified compared to Wirtschaftswissenschaft because the former is (said to be) the subset of the latter, not the other way round

???

Plus it so happens that the terminology in Estonian is as crappy as in German because German was the role model for the terminology. When talking about a country's economy (i.e. aspects of GDP), the scientists say a word that translates like "people's householding" which is plain stupid. It can, with some stretch, be understood when talking about one's own country, but not when talking about some other country.
7
To be fair, that is essentially what it says even if it is stylistically awkward. Volkswirtschaftslehre means something like "common economics" or "common science [or teachings] of economics" and the rest is nothing but synonyms...
When you say plain "economics", you are not saying much, not distinguishing anything, such as, say, microeconomics (economics of a single entreprise) from macroeconomics. In contrast, Volkswirtschaftslehre, Nationalökonomie etc. is rather specific. It's taking stock of the economy of a country, i.e. aiming at a sensible definition and calculation of GDP. "Volks" or "National" etc. may look like poor words for "of a country" but that's the meaning.
9
The Lounge / Re: What is your weather now?
...the greenhouse model describes climate like a pot of water: On constant added fuel, water will reach a boiling temperature. When the boiling temperature has been reached, water does not become hotter anymore. Instead, it becomes turbulent and it will blow away the lid of the pot.
Record cold for May last week. Record heat for May this week. Confirmation bias.
10
Sent.
Queued up on the e-reader. Thanks :up:

Edit: What are you guys talking about? Have you considered the word snacka? And it doesn't mean to have a snack, even though "ha litet snack" is definitely a thing.
11
Are we now to accept any Photo-Shopped pic as "news"? :) Have we all become Democrat Party members?
You mean cover illustrations are Liberal Mainstream Media Lies whereas Trump's twittering is "without the filter of the fake news"? Way to go!
12
It's a big round number of years since the Reformation this year. Not seeing much celebration, except some odd attempts to swipe the old differences under the rug, so it's safe to determine that people have forgotten what the Reformation was all about.

For one, Luther said that churches should not be overloaded with distracting art. Even this simple well-founded point has been forgotten.


13
Mine had to do with normative guidelines for orthography, not with historical development. Basically, my point there was that I can do better than some of the details of the current norms are. Particularly boring.
14
PS If you're interested I could mail you my bachelor's thesis. I like to think I wrote a pretty decent overview of the inner workings of machine translation.
Hey, that would be great! It might marginally touch on the thesis that I am planning. Unfortunately I didn't get it going this year, even though I am doing my best to catch up with the science.

My own batchelor's thingie is most likely uninteresting to you. It's contrasting Finnish and Estonian orthography, not very interesting even for myself these days.

Hopefully my vivaldi.net address works. If not, I will make it work.
15
The basic error with this as grounds for firing Comey is that it vaguely resembles juridico-political reasoning. Trump doesn't do that sort of thing.
What you meant to say is that you don't understand how our system of government works...You seem to think it's all about the personality of who's in power.
Not at all. I'm only saying that with Trump in power, that's what it observably is. It wasn't like this before, but with no chance to remove him, he may make this a permanent feature of your political culture.

This point is perhaps too specific and nuanced for you so that it never gets through to your skål.
16
Of course you're quite right that the machine doesn't actually know what things mean. But at the same time it's capable of some very impressive tricks related to noticing patterns. It doesn't know what pets or farm animals are, for example, but the algorithm can actually notice that words like cat and dog are often used in similar contexts. As such it has been demonstrated to be able to make a very educated guess at translating a word like dog between two languages even if there is not a singe human-made translation of the word available to teach it.
Yes, I know that "noticing patterns" is what machines do and it looks often outright amazing. However, I also know that machines know nothing about the thing called context unless it's pre-defined and fed in (by humans, duh). When some of the semantic or grammatical categories happen to realign due to cultural progress (or regress, fads, new inventions), the machine cannot update itself.

For example, it may be able to do lexical innovation based on regular derivation patterns, but it cannot tell which innovations actually take root with the language community. This must be updated manually. The machine doesn't do educated guesses. It does mechanical trial and error - it might get it right, but it's fundamentally still pure chance.
17
If anyone is still interested: The commission of an "independent" counselor egregiously muddies the waters...
The original investigation was of Russian interference in the 2016 U.S. election. By making such a criminal investigation, the Justice Dept. has rejected an intelligence investigation. (Lots of special safeguards apply in criminal investigations that are moot in a counter-intelligence investigation...) But maybe a "counselor" is not a prosecutor? :) Of course, the average Democrat won't know that!
The basic error with this as grounds for firing Comey is that it vaguely resembles juridico-political reasoning. Trump doesn't do that sort of thing.

For example, as you well know from Trump's interviews with O'Reilly, when asked how he would respond to military threats from Putin, his answer was, "I would call him and I would say, "Don't do it. Just stop it. Don't do it.""

Do you have any examples of him thinking any deeper than this? Or are you like him in that when asked about factual data, you can reply, "Many people have come out and said I'm right," and when pressed further, "Forget all that. We're gonna do it."?
18
I remember reading somewhere that Google Translate has two core elements:

- synthesising available human translations
- machine translation

What may appear as a passable work of the latter is normally the former. When a translation at Google Translate is surprisingly good, it means you happened to hit on an oft-translated and well-synthesised topic. Machine translation in its proper sense continues to be a still-born idea.
19
In case you didn't know, that's a Russian minaret. :)
Your fav news source covers it best https://www.rt.com/viral/388848-cnn-time-russian-minarets/
20
Comey did technically usurp -with the help of President Obama and Atty General Loretta Lynch- the AG's role...
And Trump sees this the same way? Hardly. Trump has no clue of anything called a technicality and in this case you are simply imagining it up a la Breitbart.

ersi, why post a link to a story you haven't read? :) (Or is it merely a case of having a deficient sense of humor?)
Of course I have read it. You evidently have no idea why it's funny.

First, stories appear accusing Trump of being an intelligence leak. This fits the overall picture of Trump's unstable character.

Then Putin rushes in Trump's defence. This is funny because, when you think it through, he really doesn't. He knows he is making the scandal worse and this is exactly why he is doing it.
21
American law can do nothing against the president that decides to behave like a total scumbag.
What -exactly- has he done that so aggravates you?
Currently the relevant question is: What exactly did Comey do to aggravate Trump so that he ended up fired?

The documentation of Mr. Trump's request [to Comey to shut down the investigation concerning Flynn's ties to Russia] is the clearest evidence that the president has tried to directly influence the Justice Department and F.B.I. investigation into links between Mr. Trump's associates and Russia.

And what exactly does Trump find so cute about Putin so as to be overgenerous with apparently classified information to him? Putin enjoys embarrassing the clown further http://6abc.com/news/putin-jokes-he-would-release-record-of-trumps-russian-meeting-to-congress/2003397/
22
Britain is an island. Scotland is another island. Almost. It's a fake island.
23
American law can do nothing against the president that decides to behave like a total scumbag. Even before Trump became the president, everybody knew he was a total scumbag (plus consistently acting like it).

Chew on that one for a while. Four years at least.
24
You can pin packages...
This is a gross technical term. You "pin" only in the graphical package manager. It means "to exclude from the common update cycle".

On my distro, the graphical package manager (even though pretty and popular and eagerly developed) is not quite safe. It happens often enough when the update notification explicitly recommends command-line.

Third-party software in general (such as Opera or Vivaldi) often maintains its own repos that integrate with the system.
True mostly for Debian and Ubuntoids. There are Linux-friendly software producers, such as Vivaldi etc. that normally produce a Debian and Fedora package, sometimes also OpenSUSE and more. But if you have a different distro, you can't expect the Debian/buntu package to work just so. (Yup, a "Linux" package is guaranteed to work only on a single select distro. This is an unintended consequence of separate package management systems for every distro. A distro normally has its own separate repositories which in turn implies a separate package management system which in turn has brought about incompatibility of package formats across Linux distros.)

If the thing is not in the repository, you must grab the source or tarball and compile locally. Which is why I use Manjaro/Arch whose nicety is, on top of the official repository, an extraordinarily large bunch of user-built software packages and compiler scripts that are conveniently shared.

Just to keep it clear, "BSD" is not a Linux distro. It is another operating system - or better, another class of operating systems descendant from the old Unix - or something like that. :)
This is another gross technicality. Nothing gets clearer this way.
25
No. The main point is: It's kernel updates versus everything else. It's just that kernel updates for a single point version tend to be security updates and minor bugfixes only. If you want a kernel feature update, you generally upgrade to a different kernel version number (upgrade as distinguished from update). Linux makes it easy to keep several kernel versions on board and you can switch them by rebooting.

If I remember right, the last time I looked at a *Buntu, there were three levels of updates, one level for kernel (labelled Security, the way you like, but it was really more like for kernel only) and you could refuse that level.

Browsers and such also receive security updates, obviously, but they are not distinguished from ordinary app updates in any way. Generally you can click around and make selections in the thing called updates notifier, but there are a lot of interdependencies within Linux, so in reality you cannot tailor things too much there.

Linuxes fall roughly into two classes. In the more popular class it's generally assumed/recommended that you update everything. In those distros (including Ubuntoids) you don't go to the Firefox website to update your Firefox. You wait for the Firefox updates in the distro repository[1] and you can only update when the volunteers who work for the distro have completed administering all the packages in the repository so users can update their computers from there. You can refuse updates for some select packages/apps, but you must know what you are doing.

There's the thing called package manager (related to, but distinct from updates notifier) that is specific to each distro. You will need to learn to know the package manager for your distro. When things break and the package manager is a good one and you know it well enough, you can revert updates, change labels/tags on the packages etc. to fix things. When you end up not liking a distro and you choose another, you must learn to know the other package manager.

Generally it's the easiest and safest to refuse kernel updates. Other packages/apps tend to have more interdependencies that can break.

Then there's the following important catch. You may refuse an update, either for a select package or for the entire system. Then comes the next update cycle and you will receive the previous updates along with it anyway. It's just the way Linux generally works. It cannot work anyother way, because interdependencies are many and it takes a lot of work to keep a distro consistent. To completely bypass this, there's another class of Linuxes, the hypernerdy old-school tinkerer class.

In that other class of Linuxes, e.g. Slackware, perhaps even better - BSD, the distro repository (the server that hosts the updates for the distro) hosts just the bare essentials, because it's generally assumed that people update the apps/packages locally and individually. In this class, you see an update for Firefox on the internet, just like on Windows. Then you go download the Linux version for Firefox and compile it locally. Compile = build an installable package. It's a good skill because you can modify it to your heart's content. Then install it. If a dependency breaks, you hunt for an update for the dependent package too or you figure out the changed dependency requirements and recompile things from source, applying the changed requirements. In a similar way you can compile and recompile the kernel, applying or removing whatever parts you want/need.

In small ways, we can/must practise all this on the more popular Linuxes also, whenever we want to install something that is not hosted in the repository or when we want (and know how) to modify the kernel.
Repository is the server that hosts updates for the distro. It's just like Microsoft hosts Windows updates, but the difference is that Microsoft does not host updates for desktop apps like Firefox or Libreoffice, whereas Linux hosts updates for everything from kernel through system libraries up to desktop apps and it's recommended to update everything all at once, as much as is hosted and administered in the repository.