/dev/chat

« first day (2393 days earlier) ← previous day next day → last day (2569 days later) »

12:59 PM

@Gilles fair enough. Comment(s) deleted then.

1:22 PM

> od -x maps to od -t x2 which reads two bytes at a time, and on little-endian systems outputs the bytes in reverse order.

OK, @StephenKitt, WTF? That makes sense?

Why in reverse order? Is it only to confuse us poor idiots who don't get indianness?

Stephen Kitt

@terdon heh heh, welcome to the world of reading bytes

little-endian means that the lower-order byte is stored first

so when you read a two-byte word, you flip the two bytes to get the representation

i.e. if you stored the actual value 0AEB (as a 16-bit value), from a little-endian system, that would end up in memory as EB followed by 0A

so when you read a file containing EB 0A as 16-bit words, which is what od -x does, then print the values, you get 0AEB

lots of fun

there are worse systems than that, with mixed endianness

the Wikipedia page gives nice examples

terdon

That seems. . . masochistic.

But I guess there are good reasons for it.

Stephen Kitt

@terdon yes, I can’t remember what right now, but there are reasons

Endianness

Endianness refers to the sequential order used to numerically interpret a range of bytes in computer memory as a larger, composed word value. It also describes the order of byte transmission over a digital link. Words may be represented in big-endian or little-endian format, depending on whether bits or bytes or other components are numbered from the big end (most significant bit) or the little end (least significant bit). When addressing memory or sending/storing words bytewise, in big-endian format the most significant byte, which is the byte containing the most significant bit, is sent first...

says

> The Datapoint 2200 uses simple bit-serial logic with little-endian to facilitate carry propagation. When Intel developed the 8008 microprocessor for Datapoint, they used little-endian for compatibility.

terdon

Yeah. Blah blah simple blah blah propagation blah blah compatibility :)

Stephen Kitt

@terdon that’s a very good summary

terdon

1:32 PM

I'll file that under "some complex low level stuff". That's a pretty big folder for me.

Stephen Kitt

;-)

terdon

Thanks for the answer, by the way.

Stephen Kitt

You’re welcome! BTW I‘m 5 points away from 200 today, if you accept that answer you make me legendary ;-).

terdon

YES

Already did!

Stephen Kitt

woohoo

terdon

1:34 PM

caching though

@StephenKitt I'm still unclear on one thing though. Say I know the encoding and know od but it doesn't give me output I can read (od -c gave this 353) thing.

Stephen Kitt

yeah now we need to wait for the badge processing

terdon

Does that mean that I need to find the character table for that encoding and look up 353?

And can I safely assume that if it's in od -c then it is a decimal number?

Stephen Kitt

what do you mean by “doesn’t give me output I can read”?

no, od -c is octal too

for non-printable characters

iconv should always be able to figure things out if you know the encoding

terdon

@StephenKitt \n I can read, it's a representation designed for puny humans. 353 is not. That's what I mean.

Stephen Kitt

ah right

terdon

1:38 PM

$ od -c file
0000000 353  \n
0000002

So given the above and :

$ file file
file: ISO-8859 text

How can I go straight from 353 to ë?

Oh. By using the very first command you gave in your answer.

Stephen Kitt

yeah that ;-)

terdon

<-- idiot

Stephen Kitt

file can’t say much about the charset though

terdon

No no no no

Stop talking. I feel I understand now. Any more and you'll just confuse me again :P

Stephen Kitt

Excellent!

Kusalananda

1:50 PM

:-)

1 hour later…

terdon

3:13 PM

@tripleee am I missing something here? Your answer seems to be the same as mine.

Thomas Ward

3:27 PM

@terdon you has a summons

tripleee

@terdon we seem to have typed our sat pretty much the same time, and there were a couple of points I didn't see in yours, so I left it around. Feel free to merge in text from mine if you like.

I'm leaving for the extended weekend any minute FWIW...

Mobile keyboard, sorry...

terdon

Nah, just curious why you'd post essentially the same answer. "Cause I had already started writing it when you posted yours" is a perfectly fine answer.

mattdm

4:31 PM

Q: How can I replace a Ubuntu partition with a Fedora Workstation 25 one?

I am running a dual boot machine with Windows 7 and Ubuntu 16.04. I want to leave the Windows partition as is, and completely replace Ubuntu with Fedora Workstation 25. I have created a USB to boot Fedora from and started working through the "install to hard drive process". That part is all fine....

^ This got downvoted and closed as "too broad".

I really don't understand why. it describes a specific and common situation, and partitioning (particularly, removing existing partitions while keeping others) is scary. And there's a pretty straightforward answer.

What am I missing?

1 hour later…