English Language & Usage: Multi-Layered Discourse Room

12:02 AM

@MετάEd We can but that would be confusing.

@Cerberus what I doubt is what you can infer from the fossil record (and this sort of also gives doubts to the other question about the linguistic record)

Cerberus

How so?

Mitch

so you find one set of bones in one spot. and suppose you can date them exactly.

then you find another set of pots at another location.

Cerberus

Pots?

Mitch

OK pots -and- bones.

and then maybe some of the same pots at a third location that are somewhat similar.

what bothers me is that archeology will then infer a whole family tree out of that (yes, I exaggerate)

Like they used a pinhole camera, saw london, and then saw los angeles and then from that drew a map of the world.

there is so much ancient data that may very well be out there that could entirely change every inference.

Cerberus

I am sometimes suspicious about what I read from archaeology.

Mitch

12:10 AM

we've been lucky with thdinosaur bones that have been found in mine shafts and on outcroppings in mongolia that by accident were buried in a mudslide 100M years ago.

Cerberus

But palaeontology is something else.

Mitch

that's bones too.

Cerberus

And I know how suspicious people are about certain things in my field.

Some things I agree are unfounded.

Mitch

and we're just hearing about these ideas from scholarship now.

Cerberus

But other things that seem far-fetched to outsiders are quite reasonable once you truly know how they do it. Things that you think are impossible to know about the past.

Mitch

12:12 AM

not seeing all the doubt and appeal to (unfounded) authority of the past scholarship.

Cerberus

What do you mean?

Scholars have never been infallible.

Mitch

lots of things 'figured out' in the past were not science based.

Cerberus

Nor will they ever be so.

Mitch

effing Aristotle is the prime example.

Cerberus

That is essential to science...

Mitch

12:13 AM

he just made stuff up.

Cerberus

Sure.

Mitch

and he was a super genius.

Cerberus

But how does that help us?

Discard all science?

Mitch

think of all the lesser geniuses over the years that made stuff up but we trust them because it is written down.

@Cerberus physics and chemistry work pretty well, so I'm not doubting that.

I think of etymonline. how the hell do they 'know' some of those things?

the OED just wrote them in an authoritative manner, but still made up out of thin air as much as anything here at ELU or the Straight Dope or Urban dictionary.

"Word X sounds a lot like word X'. I bet they have the same history" -> "X and X' are cognate" just made to sound more authoritative.

The OED is pretty good, but some instances (I can't think of one now) seem very questionable.

I'm having a post-modern crisis.

where nothing means anything anymore except what you feel like.

Cerberus

@Mitch I'm sure many mistakes are made in those fields as well.

@Mitch That's not how a good etymologist works.

I'm sure there are errors and leaps in logic.

But he normally doesn't just make up conexions.

Mitch

12:20 AM

those are much easier to find errors. build a machine and if it doesn't work then the science wasn't right

Cerberus

He studies all the forms of a word in countless sources between certain dates. You just don't see the fact that he has done that in the dictionary.

Mitch

@Cerberus sometimes it looks like that.

Cerberus

@Mitch I'm sure tons of things in physics and chemistry are not (yet) used in actual machines...

MετάEd

The plural of Full English Breakfast is obviously Fulls English Breakfast. The English are always pulling stuff like Attorneys General.

Cerberus

@Mitch Maybe you should read the notes of an etymologist to one of his entries.

Mitch

12:22 AM

@Cerberus some words and spellings are rare. so hard to find alternate instances to corroborate

Cerberus

The dictionary tells you nothing about his method or his evidence.

Mitch

@MετάEd The proper plural is 'Breakfast Buffet with Omelette station'

@Cerberus sure there's a lot of theoretical physics that is just messing with mathematics. The mathematics is derived and supportable logically, but may not have any expression in the real world at all.

MετάEd

@Mitch or Warren Buffet, if there's hare on the menu.

Cerberus

Certainty decreases when complexity increases.

You could say mathematics is the simplest science.

Mitch

@MετάEd I didn't know he's a vegetarian.

Cerberus

12:25 AM

Where you can have the greatest certainty.

MετάEd

@Mitch Vegetarian?

Cerberus

Then physics.

Etc.

Mitch

@Cerberus biology is the simplest. mostly no explanations, just pointing and naming.

Cerberus

Umm.

Nobody claims to know with certainty exactly how evolution works.

MετάEd

@Cerberus Yes, in mathematics one can be completely certain of something that doesn't exist at all.

Mitch

12:26 AM

@MετάEd Hardly, I barely knew her!

Cerberus

@MετάEd See? That makes it easy.

terdon

@Mitch Tosh! Depends on what you're studying. There are some really hard concepts in biology. Also some very serious math.

MετάEd

Whereas in psychology, one can have an opinion about something life-threatening.

Cerberus

shrugs

We all die anyway.

And then you get to visit me.

MετάEd

@Cerberus I'd rather take my chances with St Peter. Though I don't believe in either one of you.

terdon

12:34 AM

@Cerberus So if we die before you, we're screwed. Gonna have to wait another twenty years or however long it takes for your non-smoking self to come and open the gate.

Or, since you won't be guarding it, I suppose I could always slip in.

Mitch

@Cerberus heaven is nearby?

Cerberus

@MετάEd Guess where he is.

@terdon Hah.

I suppose I can't really "die" in the conventional sense.

@Mitch If that's how you experience it, sure...

terdon

Was Cerberus ever actually alive? I don't remember his origins.

Cerberus

Life and death don't mean much to him, I'd say.

He just is.

Mitch

1:16 AM

@Cerberus That's pretty dismissive for someone who isn't even badass enough to get Into hell.

Kit Z. Fox

2:35 AM

@Cerberus Rob also pinged you from his new room, in case you missed it, which you probably did because I'm pretty sure you can't ping someone from a room unless they have visited it.

Johan Larsson

8:40 AM

´.

user174558

10:22 AM

@Cerberus Evolution may not even exist.

Anonymous

10:36 AM

anyone here?

I have a quick question about the English languate.

*language.

Mitch

12:25 PM

@JasperLoy they're pretty sure evolution 'exists'. Practical consequences: take the full prescription of antibiotics or you risk helping groom antibiotic-resistant bacteria.

It's sorta like the theory of gravity. You may doubt it or even not believe it, but it helps make sure bridges stay up.

user174558

12:39 PM

@Mitch Well, I am not sure they are so sure, lol.

user174558

@Mitch Well, gravity is nothing but a model to explain things dropping.

user174558

Now that I have gotten 200 I may retire.

Mitch

1:25 PM

@JasperLoy Oh, they are pretty sure. That's something you can count on. The scientists are very sure. Sure of, not just of the phenomenon (very sure that evolution happens), but also very sure of the mechanisms (DNA and genes). There's still lots to figure out of the details, but the general process is pretty well established.

@JasperLoy Woo hoo! That's science!

terdon

1:40 PM

@JasperLoy They are 100% sure. We can observe it in real time. There is absolutely no question about that. The only thing that is questioned are the details of how exactly the forces driving it work. That mutations occur and are then spread across the population is not debatable.

Not among actual scientists working in the field, anyway.

And no, having a PhD from the "Center for Creationist Study" or whatever, does not qualify you as a scientist working in the field.

Mitch

1:54 PM

@terdon Rather they may be out standing in the field.

terdon

Indeed.

Mitch

Actually there is always doubt. (in my opinion) there is more doubt about evolution/genetics than there is about newtonian mechanics because there are many fewer phenomena to check with mechanics. also, newtonian mechanics is approximately correct, with Einstein giving better approximations. Also, lots more supplementary theories needed in evolution/genetics (geology and radiation for dating bones, optics for microscopes, germ/cell theory, DNA)

It's all very well supported (new data all fit the theory) so sure 99.999... percent. But there are still things to doubt like where does the weird shape of the folds in the outer ears come from (what genes do that?).

Mr. Shiny and New 安宇

@Mitch The thing is, you're mixing up sub-fields here.

Knowing that evolution works is different from knowing which particular gene does what.

For one thing, it's almost never one particular gene doing one particular thing.

Mitch

Also 'creationism' isn't used seriously by people you and I would call creationists anymore. I think Intelligent design ma even be going out of fashion. here's some new euphemism I'm sure.

@Mr.ShinyandNew安宇 I intentionally said 'genes' instead of 'gene'.

Mr. Shiny and New 安宇

@Mitch Still. It's akin to having doubts about chemistry because you're not sure what chemical process gives rise to, say, the specific material that horseshoe crab shells are made of. or something.

terdon

2:07 PM

@Mitch Most of those are irrelevant to the core theory. Mutations are all you need and those we can observe in droves. No less than anything in mechanics.

Mitch

@Mr.ShinyandNew安宇 there is the phenomenon of evolution (over time there is change; evidence fossil record) and there is the mechanism (genes/DNA; evidence:biochem)

@Mr.ShinyandNew安宇 yes, the 'doubts' are mostly inside-baseball controversies over .. well.. inside-baseball issues.

terdon

@Mitch That is not an issue of evolutionary theory, as such. Sure, there are things whose evolutionary history we can't immediately infer by looking at their current form. That, however, is no more a problem for evolution than the fact that we can't see a galaxy that is on the other side of the universe for the theory of gravity.

@Mitch The DNA evidence is by far the most easy to obtain and most conclusive.

Mitch

@terdon I'm being pedantic. Of course, all the observations fit. (considering them as corroboration is boring)

@terdon I think I get it but am not sure, can you elaborate?

terdon

I know. I'm just saying there is probably less doubt for evolution than for mechanics. Since biology is, perforce, limited to a single planet, we can be far more comprehensive with it than we can with sciences that attempt to describe the entire universe.

Mr. Shiny and New 安宇

@Mitch You're right, of course, that the fossil record is so woefully incomplete that it's like trying to restore a compact disk's recording when all you have is a handful of fragments of the disc. And some of the fragments may be from other albums.

Mitch

2:11 PM

@terdon I can throw a ball and see it (mechanics) easier than I can see DNA in action (biochem)

Mr. Shiny and New 安宇

@Mitch You don't need to know what genes do what to, say, do statistical analysis on genes to quantify the changes over time.

terdon

@Mitch Marginally so, yes. I mean that sequencing a cell's DNA is really simple and fast these days. Yes, you can't do it at home yet but it is also no big deal.

Mitch

@terdon I say doubt about evolution is more than for mechanics because there is just a lot more intermediary theory needed to show that eyes evolved than to show that the bomb will hit a certain point.

terdon

@Mitch Ah, but that isn't really what the theory of evolution is all about. its basic tenet is that selective pressure leads to change. That's all you need to show in order for the theory to be accepted and that is trivially demonstrated.

You just take two sequences from two different species and observe how they are different and yet so similar.

Mr. Shiny and New 安宇

@terdon There's also drift that occurs when there's no pressure

Mitch

2:15 PM

@terdon it's not as trivial as mechanics

terdon

@Mr.ShinyandNew安宇 Yes, and all sorts of other things.

@Mitch No, not quite, but not that far either.

Mitch

@terdon too much extra machinery (both thoughts and actual machines) needed to say 'not that far'

terdon

@Mitch OK, fair enough, but it's not like you need a particle accelerator or anything. Plus, the groundbreaking work on all this was done by a monk who was planting peas in his garden. Hardly very advanced :)

Mitch

you need extra theory to justify gravity (optics of telescopes to follow moons of jupiter)

@terdon particle accelerators aren't that complicated. they just need to be big and have lots of power. lots more intermediary operations needed to sequence genes.

terdon

You need very little extra theory to support evolution. Again, Darwin and Wallace were only armed with a notebook and a crude microscope.

@Mitch Not really actually. It's just a small machine these days, you pop in your test tube, switch it on and go for a coffee.

Sequencing has truly become trivial.

Mitch

2:19 PM

How do they know, when sequencing genes, that you didn't get mitochondrial DNA mixed in? How about bacteria living in your mouth (form cheek swab). From another person (whose food you just stole from their plate, you moocher)?

@terdon and a lot of travel

@terdon lots more theory has gone into making that possible

terdon

@Mitch You don't. You have to filter those out. It's easy today since we know the mitochondrial sequences and can recognize the bacterial ones. Also, since you often target specific genes, this is done using the polymerase chain reaction (PCR) whose great advantage is that it amplifies your target sequence. So you end up with several orders of msagnitude more of sequenceX than the rest. That makes it easy to filter out the noise.

@Mitch Mostly chemistry rather than theory. It is a combination of clever biochemical tools. The theory is the same as it was in the 70s, it's the tools that have improved.

Mitch

chemistry is more theory

ancillary

but still necessary support

terdon

True. I gues I sort of take all that as read.

Mitch

looking for a single known gene then is easier much than 'the' sequence, correct? both the chemical ops and the sequencing part?

terdon

@Mitch Yes. Quicker, certainly. Getting the entiore sequence of a genome is not that hard these days either, though.

We have cool new tricks, collectively called "next generation sequencing", which basically are different ways of cutting the DNA into little bits, making many copies of these, sequencing them en masse and reassembling the sequence to get the entire genome.

Mitch

2:27 PM

more questions: in mammals all cells have the same gene sequence? except for minor sporadic variants (transcription errors over multiple cell replications) and except for sex cells (eggs and sperm)?

terdon

Mostly, yes. And not only in mammals, all living creatures share the same DNA across all their cells.

Mitch

@terdon A good transporter mechanisms will have to replicate all these sequences (I'm working up a business plan)

terdon

The mostly is because mutations happen all the time. You will have a few cells that have tiny differences.

@Mitch Heh, yes, but that will be the least of your worries. Presumably, you'll be able to take an image at the atomic scale.

Mitch

@terdon how do reptiles and amphibs change their (gross morphological) sex then under environmental pressure?

Matt E. Эллен

English spelling - a bit mad, but perhaps the best system around

►

terdon

2:30 PM

I'd have to look up the details but what probably happens is some form of suppression. There are mechanisms that can render a chromosome inactive. If these are linked to temperature in reptiles, I can imagine a system where, say, XX is female and X is male so one X is inactivated by the temperature.

Mitch

@terdon Well, there a re different techniques obviously. If you can compute the positions of all atomic particles, surely that is a more accurate replication method. but the funding model is way different (need much higher computing power and power power). That's pretty expensive. Our funding model is the restriction here not the science. When you're restricted to trillions of dollars, you have limitations on the science you can implement.

terdon

OK, but getting the sequences won't help. That's only part of the story. If I were to just clone you at the other end of the transporter chain, all your memories would have been lost. Not to mention that scar you got when you ran into the kitchen door at age 5.

Mitch

@MattE.Эллен and you missed the one on penannular brooches?

@terdon 1) re brain state...hmm yes, you'll probably need to transfer the electrical distriubtion over the brain too to maintain consciousness state.

2) re the scar, you have to maintain morphology so the scar would stay. scar removal would be the bonus you get on the paid app.

(unfortunately, regulations would enforce removing communicable diseases and carcinomas. goddam you regulations!)

and as long as the replication is far enough away, you avoid the uncomfortable situation of having to battle your doppelganger to death. of course investors get to watch such contests for free.

Mr. Shiny and New 安宇

2:53 PM

@MattE.Эллен Not a bad video but he gets some details of the Chinese wrong.

Mitch

@Mr.ShinyandNew安宇 at a rough approximation, english orthography, for one based on an alphabet, is much more in the direction of the ideograph. makes the 'whole word' reading-teaching method very plausible.

Mitch

3:14 PM

Holy Jesus effing Christ. The ODO did not choose 'emoji' as the word of the year but rather an emoji, namely the crying laughing one? Kudos on being postmodern. Now STFU Donny.

Matt E. Эллен

@Mitch his video?

I don't even know how I found that one. I'm too tired to remember

@Mr.ShinyandNew安宇 what about?

Mitch

@MattE.Эллен Yeah his. I watched your link, then looked at the 'suggested' videos on the right column. You can spend hours traveling down that path.

Matt E. Эллен

@Mitch :D

Mr. Shiny and New 安宇

He over simplifies how Chinese works. The written language isn't automatically understandable to people who use other dialects. Essentially: the written language has dialects too, and your reader needs to know the one you're using. In the past the written dialect was Classical Chinese.

Now it's standard Mandarin.

If you are not trained in classical Chinese you can't read old texts.

And classical Chinese is to modern Chinese as Latin is to modern European languages.

Today, everyone in china is taught Mandarin and standard writing. In the old days, when the local dialects were more important, nobody could read except the elite, who were reading the equivalent of Latin.

A lot like how Europe used to be, really.

Heck, classical Chinese was used in Japan and Korea for writing. They'd converse in their local language and write in Chinese.

Matt E. Эллен

I see. interesting

Mr. Shiny and New 安宇

3:30 PM

So while it doesn't undermine his point in the video, the thing is that Chinese is even less optimal than he makes it out to be. It's not nearly as universal as people assume it is.

Anonymous

3:48 PM

@Mitch It might seem plausible, but it's very harmful for the minority of students who actually need instruction.

Mitch

@snailboat oh sure. but a student of chinese isn't really learning from that video.

Anonymous

@Mr.ShinyandNew安宇 Or something very Chinese-like.

Anonymous

It's kind of crazy jumping through that many hoops to encode a language so entirely unlike Chinese in a way that makes it look superficially similar

Mr. Shiny and New 安宇

@snailboat yeah. My understanding is that it started as classical chinese and over time they adapted it to the local language. But both countries have since introduced native writing systems, and neither are logographic.

Anonymous

Japanese does still use a lot of Chinese characters

Anonymous

3:53 PM

Hangul is pretty neat :-)

Mr. Shiny and New 安宇

yeah

Anonymous

Kana evolved from kanji.

Anonymous

@Mr.ShinyandNew安宇 Technically, languages rather than dialects. A lot of Chinese languages aren't mutually intelligible with Pǔtōnghuà.

Anonymous

So it's strange to call them dialects.

Anonymous

Victor Mair has been trying (without a lot of success) to get people to use topolect for 方言 (fāngyán).

Anonymous

3:58 PM

It's supposed to be more or less a calque, meaning literally "the speech of a place".

Anonymous

It's kind of a neat coinage :-)

Anonymous

@JasperLoy We can actually observe some examples of evolution as they happen.

Anonymous

→ Peppered moth evolution on Wikipedia

Mitch

4:17 PM

@snailboat ooh... spicy.

wait 'moth' not 'mouth'. Yechhh

Mr. Shiny and New 安宇

@snailboat yeah I'm being imprecise, I know

Some of the Chinese "dialects" are barely related at all.

user174558

@snailboat OK, but does that mean that we evolved from apes? I don't know.

Mr. Shiny and New 安宇

@JasperLoy There are multiple lines of evidence to support this.

Mitch

@JasperLoy that's a gross oversimplification. The analogy is more like we are cousins of the apes. The current apes (chimps, gorillas, orangutans, humans) all have a common ancestor, say 5M years ago.

user174558

OK. I have not done much reading into evolution, so I am just speaking my mind.

Mr. Shiny and New 安宇

4:25 PM

The videos on this site are very easy to understand an informative on the topic of evolution.

Especially the "What is the evidence" video

Cerberus

4:49 PM

Greetings.

What's all this evolution talk?

Everyone knows all species are immutable.

Mitch

@Cerberus Trolls are immutable

Cerberus

They are an existing species, of course.

Mitch

Of course.

Cerberus

And immutable.

For have you ever seen a troll mute?

Mitch

No of course not. I've never seen them immute either.

Cerberus

5:05 PM

They may not immute, but they are immutable. One can immute them, even if one does not.

Mitch

All this talk is making me hungry

Cerberus

So sexy.

terdon

@MattE.Эллен That only defends alphabets, not the insanity that is English spelling. Pretty much all languages carry their history in their spelling, yet most of them manage to do it in a way that makes at least a modicum of sense.

Cerberus

Heh.

I would rather say English spelling has been updated less than that of most other languages.

Back to where we started? That's Windows 1 on the upper left.

Johan Larsson

@Cerberus 1.0 is nicest

Cerberus

5:21 PM

The logo or the OS?

Are you on Windows 1.0? That would explain a lot of things...

Anonymous

Also, current evidence suggests that birds are a kind of theropod dinosaur.

Johan Larsson

@Cerberus the logo

Anonymous

And honestly, it's fun to refer to birds as dinosaurs. "Oh, what a cute little dinosaur you are!"

Anonymous

ucmp.berkeley.edu/diapsids/avians.html

Mitch

"There are 20 dinosaurs making a nest right outside my front door"

Cerberus

5:26 PM

@JohanLarsson Hmm too bad.

I'm not sure I understand the captcha joke...

@snailboat Absolutely!

Johan Larsson

@Cerberus double precision

Cerberus

What does that mean?

Johan Larsson

floats are not represented exactly

JohnB

@Cerberus 0.30000000000000004.com

terdon

$ perl -le 'print 0.1000000000000000000002 + 0.2000000000000000000002'
0.3

Anonymous

5:34 PM

crystal@wings:~$ perl -e 'for (0.1, 0.2, 0.1 + 0.2) { printf "%0.99f\n", $_ }'
0.100000000000000005551115123125782702118158340454101562500000000000000000000000000000000000000000000
0.200000000000000011102230246251565404236316680908203125000000000000000000000000000000000000000000000
0.300000000000000044408920985006261616945266723632812500000000000000000000000000000000000000000000000

Cerberus

@JohnB I see.

I am surprised those leftovers are so big.

terdon

This one is even worse:

 $ perl -le 'print 0.3 - 0.2 - 0.1'
-2.77555756156289e-17

Cerberus

Weird.

Johan Larsson

@Cerberus that is why learning pi to 17 digits is a good idea

terdon

Yup. Happily, that doesn't affect most every day applications.

Cerberus

5:36 PM

Is there no way for the software to filter that out after the calculation is finished?

@JohanLarsson Fun!

Anonymous

@Cerberus It's efficient to store numbers that way. You don't have to, though. There are lots of other ways to store numbers.

Cerberus

Such as by using more decimals?

Anonymous

If it matters you can certainly avoid introducing that sort of error.

Cerberus

Right.

And that is really so costly that it isn't done normally?

Mitch

@JohanLarsson Only if you're computing the travel distance of an electron

Anonymous

5:40 PM

@Cerberus People use whatever is appropriate to the task at hand. (Except when they don't :-)

Anonymous

If you're drawing a 3D scene for a person playing a video game, you might not care that much about precision so much as doing lots of calculations really fast.

Anonymous

If you're writing software that handles money, you might want to use a different representation.

Anonymous

There isn't a single best way to store numbers that's used all the time.

Cerberus

Naturally.

I'm just surprised that it already shows up after so few decimals in our daily software.

I always assume computers are infinitely fast at simple math.

Anonymous

crystal@wings:~$ bc
bc 1.06.95
Copyright 1991-1994, 1997, 1998, 2000, 2004, 2006 Free Software Foundation, Inc.
This is free software with ABSOLUTELY NO WARRANTY.
For details type `warranty'.
.1000000000000000000000000000000000000000000000000000000000000000000\
00000000000000000000000000000000000000000000000000000000000000000000\
00000000000000000000000000000000000000000000000000000000000000000000\
00000000000000000000000000000000000000000000000000000000000000000000\
00000000000000000000000000000000000000000000000000000000000000000000\

Anonymous

5:50 PM

No errors :-)

Anonymous

Well, no error.

Cerberus

That's more like what I expected.

Johan Larsson

@Mitch No.

Cerberus

6:12 PM

@snailboat Thinking of you.

From an article about how snails are becoming popular as pets.

And in beauty parlours.

Mr. Shiny and New 安宇

6:46 PM

@Cerberus Computers are really good at integer math, as long as your integers fit in the representation, i.e. they fit in 32 bits or 64 bits or whatever.

Computers are really good at math that has decimal points if you don't care about precision.

The reasons are that it is easy to build math processors that can do that sort of arithmetic.

MετάEd

@JohanLarsson Oh. That's precious.

Mr. Shiny and New 安宇

If you want arbitrary precision for very large numbers (or very small numbers) then you need a special software library that will do things for you. It won't be nearly as fast as doing it using the built-in hardware features.

MετάEd

@JohanLarsson That's from SMBC.

Johan Larsson

what is smbc?

MετάEd

@JohanLarsson It's an online comic. Saturday Morning Breakfast Cereal.

Johan Larsson

6:49 PM

@Cerberus that is what happens if you have globals yes.

Matt E. Эллен

@terdon I was trying to tell a Greek just the other day that English is hard, but she dismissed me as foolish

Kit Z. Fox

English is hard. Let's build rockets!

MετάEd

@KitZ.Fox Great idea. No intelligent life here.

Kit Z. Fox

I was riffing on the Barbie thing, but "English is hard. Let's go mathing!" made less sense.

MετάEd

I know Barbie, and I knew Ken. Ken would have been much more excited about building rockets.

terdon

7:08 PM

@MattE.Эллен Well, granted, Greek spelling is very hard but there are rules. At the very least, you always know how a word is pronounced. And yet, its etymology and history are still right there. You might have trouble spelling a word you hear, but you can always pronounce a word you read.

Kit Z. Fox

Soo I'm relatively new to coordinating UAT, and there is one user who apparently thinks that test scripts are useless and conversational-style stream of consciousness email is much more what I need. So she has kindly done that.

Matt E. Эллен

@terdon true, true.

Anonymous

@Cerberus I love snails, but I wouldn't want one anywhere near my eyes! Pictures like those are a little bit terrifying to me :-)

Anonymous

Snails try rasping at anything they crawl on.

Kit Z. Fox

Guess that would be one way to have your lenses corrected.

Anonymous

7:23 PM

Eep!

Anonymous

@Cerberus Snails are getting really popular as pets in some parts of the world. Particularly the UK, I think. But I think having pet snails in the US still makes you weird. Well, makes me weird :-)

Anonymous

@KitZ.Fox I'm probably going to have nightmares now :-)

Cerberus

7:41 PM

@Mr.ShinyandNew安宇 Right, but the software will work on top of integers, right? So in the end what the computer sees is integers.

But anyway, I understand the problem.

It's just different than my subconscious assumption.

@JohanLarsson Hmm globals? As in variables?

@snailboat Hmm the radulae.

But as long as you keep your eyes closed?

@snailboat But weird is good, no?

Mitch

@Cerberus sort of. they have limited precision.

Cerberus

I have to admit I don't know of anyone here with pet snails.

Mr. Shiny and New 安宇

@Cerberus well, the computer processor only has certain operations it supports. But adding two integers together pretty much the fastest thing the cpu can do. when you start doing complicated things to ensure that precision is never lost you are making your program much much slower.

Cerberus

@Mitch You mean they have limited precision when approximating non-binary fractions, right?

Just as we have limited precision with 0,33333333333.

Mitch

meaning, that even if you give a number 64 bits to work with, depending on how you use those 64 bits, you can lose precision during operations.

@Cerberus yes, that's one problem.

Cerberus

7:45 PM

@Mr.ShinyandNew安宇 But approximating fractions basically means you just use a very large number of integers and perform calculations upon them, doesn't it?

Mitch

(that could be the cause of the .1+.2 = .300...004 prblem

@Cerberus most computers don't use 'infinite' precision.

Cerberus

@Mitch Of course, there would have to be limits. I was just surprised that the limit already became apparent when you have so few decimals.

Mr. Shiny and New 安宇

@Cerberus well, yes, but using more than one memory location to store a single "value" is necessarily much much slower.

Mitch

floating point is what the use which means you get lets say ...

Cerberus

@Mr.ShinyandNew安宇 Right.

Mitch

7:46 PM

X bits for the significant digits, and Y bits for an exponent

Cerberus

Not sure what that means.

Mr. Shiny and New 安宇

IEEE floating point

The IEEE Standard for Floating-Point Arithmetic (IEEE 754) is a technical standard for floating-point computation established in 1985 by the Institute of Electrical and Electronics Engineers (IEEE). Many hardware floating point units use the IEEE 754 standard. The standard addressed many problems found in the diverse floating point implementations that made them difficult to use reliably and portably. The current version, IEEE 754-2008 published in August 2008, includes nearly all of the original IEEE 754-1985 standard and the IEEE Standard for Radix-Independent Floating-Point Arithmetic (IEEE...

Mitch

so when you're doing operations around 10^(-53) everything works out OK, but if one number is around around 10^25 and the other around 10^(-25) some digits will be lost because that's the best that can be done

There is 'infinite' precision, but even that will still lose precision because of irrationals

Cerberus

Ummm.

I don't get it: why are digits lost when you have fewer decimals?

Mr. Shiny and New 安宇

generally you have a certain number of bits/digits to represent your number, and you have a scale.

Mitch

7:51 PM

sorry, this is all telescoped

Mr. Shiny and New 安宇

so you can represent 123x10^53 and you can represent 123x10^-53 but if you add them together you lose precision.

Mitch

you keep a finite # of significant digits, at no matter what the scale. so 1 = 1.0 x 10^0 = 1.00...00 (let's say 50 digits) x 10^(0.00..00) (Let's say 10 digits for the exponent).

then ...oh... @Mr.ShinyandNew安宇 said it

Cerberus

@Mr.ShinyandNew安宇 OK that makes sense.

@Mitch Ah OK, and that is done even for numbers like 1?

Anonymous

@Cerberus I think weird is neutral with respect to goodness.

Anonymous

There's good weird, bad weird, and in between weird.

Cerberus

7:58 PM

Yes.

But, in practice, weird is often fun.

Transcript for

English Language & Usage: Multi-Layer…