[ / / / / / / / / / / / / / ] [ dir / bestemma / general / imouto ][Options][ watchlist ]

/tech/ - Technology

You can now write text to your AI-generated image at https://aiproto.com It is currently free to use for Proto members.
Name
Email
Subject
Comment *
File
Select/drop/paste files here
Password (Randomized for file and post deletion; you may also set your own.)
* = required field[▶ Show post options & limits]
Confused? See the FAQ.
Expand all images

[–]

 No.912377>>912397 >>912445 >>912463 >>912464 >>912466 >>912496 >>912513 >>912529 >>912561 >>912743 >>913225 >>921332 >>923369 >>923776 >>923787 [Watch Thread][Show All Posts]

Why is the left still the format of choice?

 No.912390>>912396 >>912556 >>923776

I would say it's far easier to parse than json.


 No.912394>>912447

Why did XML and JSON become popular instead of s-expressions, which is perfect for data formatting?


 No.912396>>912556

>>912390

it's easier to look at, not easier to parse. Look at all that text bloat.


 No.912397

>>912377 (OP)

That json whitespacing really triggers my jimmies


 No.912404>>912443

XML is a markup language.

JSON is a shit. But people love shit if it works and they don'y have to come up with something better.


 No.912443

>>912404

>JSON is a shit

You have to at least justify this annoying syntax with a reason


 No.912445

>>912377 (OP)

Because it's not as fucking hideous as the right.


 No.912447

>>912394

How would you unambiguously express a mapping data structure as a S-expression?

That is, if you encode and then decode a mapping data structure you always get a mapping data structure out, and if you encode and then decode a sequence you always get a sequence out.

A mapping data structure would be something like Python's dictionaries, and a sequence would be something like Python's lists.


 No.912454>>912891 >>923787

File (hide): c211e8a64d66449⋯.png (9.68 KB, 357x349, 357:349, ws.png) (h) (u)

Rate my language


 No.912463>>912524

>>912377 (OP)

Ironically android 'solved' xml by not putting content in the tags, rather the attributes of the tags define what actually is the content, the only time you would have content within tags is to use more tags.


 No.912464

>>912377 (OP)

>Why is the left still the format of choice?

XML has been dead for 15 years man


 No.912466>>912467

>>912377 (OP)

I prefer XML over JSON that hideous spacing makes me want to die.


 No.912467

>>912466

anon you can format xml exactly the same way. its a text format


 No.912468>>912475

json and xml suck ass. Edn is the best format i've seen


 No.912475>>912476

>>912468

csv dude


 No.912476

>>912475

>I love a format that is impossible to consistently parse

GJ mate


 No.912496>>912501 >>912507 >>912754

>>912377 (OP)

https://8ch.net/tech/0.json

Look at that complete puke. Now imagine how much easier it would be to read if it was xml


 No.912501

>>912496

Neither of them are intended to be human readable or editable, your argument is invalid.


 No.912507>>912532

File (hide): f64186bc5c49352⋯.png (184.36 KB, 882x950, 441:475, Screenshot from 2018-05-10….png) (h) (u)

>>912496

You know XML looks like dogshit too when you strip out all of the whitespace.


 No.912513>>912514 >>912519 >>912754 >>913153

>>912377 (OP)

Because the one on the right is Reddit spacing


 No.912514>>912557

>>912513

Its literally not tho


 No.912519>>912557 >>912565 >>912782 >>912938

>>912513

What people call "reddit-spacing" Is really just programmer-spacing. You can't use tabs to indent paragraphs (otherwise the comments are misaligned from the code), so you have to use a double-enter to denote them.


 No.912524

OP, XML is being used since the late 90's. There's way more content, tools and systems developed with it in mind.

If it was the other way around, probably XML wouldn't be largely adopted.

I just wish you never, EVER, have to deal with SOAP.

>>912463

I didn't know that. Here's one example I found after a quick search:


<!-- RelativeLayout in which we set green color for the background -->
<RelativeLayout xmlns:android="http://schemas.android.com/apk/res/android"
xmlns:tools="http://schemas.android.com/tools"
android:layout_width="match_parent"
android:layout_height="match_parent"
android:background="@color/greenColor"
tools:context=".MainActivity">

<TextView
android:id="@+id/firstTextView"
android:layout_width="wrap_content"
android:layout_height="wrap_content"
android:layout_centerHorizontal="true"
android:layout_margin="20dp"
android:padding="10dp"
android:text="First Text View"
android:textColor="@color/white"
android:textSize="20sp"
android:textStyle="bold" />
<!-- second TextView -->
<TextView
android:id="@+id/secondTextView"
android:layout_width="wrap_content"
android:layout_height="wrap_content"
android:layout_below="@+id/firstTextView"
android:layout_centerHorizontal="true"
android:layout_margin="20dp"
android:padding="10dp"
android:text="Second Text View"
android:textColor="@color/white"
android:textSize="20sp"
android:textStyle="bold" />

</RelativeLayout>

It doesn't look bad at all.


 No.912529

>>912377 (OP)

DTDs

< Yes, that's a pun where I'm calling you a retard and answering your question.

XML has a defined mechanism for metadata about itself.


 No.912532

>>912507

>VSCode

enjoy your telemetery, dumbass


 No.912551

>anything designed for javascript


 No.912556>>912615 >>912753

>>912390

>>912396

XML is easier to parse because everything is in descriptive tags

JSON is a mess to parse because everything is nested in brackets


 No.912557>>912570 >>912586 >>912896

File (hide): bb5537ac6b936e3⋯.gif (698.63 KB, 1221x768, 407:256, REDDIT.gif) (h) (u)

>>912519

>>912514

reddit spacing you fucking niggers


 No.912561>>912571

>>912377 (OP)

Lol OP, come back when JSON has transformation capabilities as mature as XML, like xslt (pro tip you can't without named elements). Don't show me some edgy library or pseudo standard, because I know they are all shit.

After that, they might as well introduce namespacing support, it's there for a good reason in XML and that shitty JSON language just rapes you without it when you try to recognize by it's structure (pro tip you can't without a namespace).

JSON is a fucking joke, it should only be used for your soyboi projects and your's alone.


 No.912565>>913087 >>921955

>>912519

What people call "reddit spacing" is actually markdown.


 No.912570>>912589

File (hide): 6a6c04211b35c59⋯.jpg (26.16 KB, 500x324, 125:81, DcZRUDPVQAAtNfC.jpg) (h) (u)

>>912557

please leave


 No.912571>>912813

>>912561

> Don't show me some edgy library or pseudo standard

Yeah unless its literally in the official JSON RFC its fake shit even if trivial solutions have been around for 10 years! You tell 'em!


 No.912586

>>912557

>formatting code is reddit spacing

this board gets worse and worse every day


 No.912589

File (hide): 81d50f70a9372b0⋯.jpg (40.21 KB, 500x324, 125:81, nou.jpg) (h) (u)


 No.912615>>912753

>>912556

Do you even syntax analysis ?


 No.912743

>>912377 (OP)

A more important question: EBML vs. JSON?


 No.912753

>>912556

This.

>>912615

Write a fast json parser in c or c++ that shits out values, string or otherwise, with object,array,value as a header in single lines.

example:

["movies",378,"info","original_title"] "Flammen & Citronen"["movies",378,"info","actor_roles","Mads Mikkelsen"] "Citron"["movies",378,"info","actor_roles","Thure Lindhardt"] "Flame"

["movies",378,"info","titles",4] "Flame And Citron"

And most importantly, it has to work for non-standard implementations of JSON.


 No.912754

>>912496

><p class=\"body-line ltr \">Welcome to \/tech\/ -

It makes you puke because of the XML inside the JSON not because of the JSON itself.

>>912513

>reddit spacing

>zero empty lines

NO


 No.912755>>912756

Whenever I see XML, I smell Pajeet.


 No.912756>>912757

>>912755

Same here. It's degree of complexity is simply unnecessary in all situations.


 No.912757

>>912756

It's also heavily associated with SQL, Java, .NET, and "Enterprise" frameworks.


 No.912768

maybe they think they need the supports, sometimes its necessary


 No.912782

>>912519

>people

It's literally just one really dedicated aspie. That's the entire 'reddit spacing' "crowd"


 No.912813>>912871

>>912571

Name me the fucking standard for it then.


 No.912818

>not JSX

shiggery


 No.912859

My eyes like reading JSON. My computer likes reading XML. JSON was made for my computer to read, not me. My computer is not me. JSON is not XML. Therefore:- my computer thinks JSON is shit.


 No.912862

they're both shit, and no, XML is not the format of choice in any "modern" web shit. only enterprise bros use it and even then not much anymore


 No.912871

>>912813

>If its not in the official JSON RFC it does not count


 No.912891

>>912454

I would.


 No.912896

File (hide): f85edc866b62e49⋯.gif (751.86 KB, 1221x768, 407:256, REDDIT_OVERLOAD.gif) (h) (u)


 No.912938>>912942 >>912947 >>921363

File (hide): bbfaf47ed518e8a⋯.png (53.08 KB, 866x475, 866:475, redditspacing.png) (h) (u)

>>912519

Holy shit how can you be THIS stupid. Educate yourself.


 No.912942

>>912938

What if you just use org-mode a lot


 No.912947>>913108

>>912938

You still don't understand that people who program for a living have reasons to type that way.


 No.913087

>>912565

why are people posting markdown to a website that obviously doesn't have markdown support then


 No.913108>>913109

>>912947

>programming for a living


 No.913109

>>913108

>having a job

>not living with your mother


 No.913119

XML unironically looks better unless you're a symbols autist


 No.913151>>913285

xml is more powerful and therefore still needed. you nigger


 No.913153

>>912513

nice one


 No.913225

File (hide): e80a7ae28cfc433⋯.jpg (7.47 KB, 299x168, 299:168, Agiri.jpg) (h) (u)

>>912377 (OP)

why not?


 No.913285

>>913151

no, it isn't, but neither is JSON


 No.921332>>921334

>>912377 (OP)

both of them use too many special symbols that require pressing two keys to get the symbol


 No.921334>>921337

>>921332

Fuck off YAML shill. Go jerk off over the whitespace in your python scripts.


 No.921337

>>921334

>bootyblasted over the superior format having whitespaces

heh


 No.921363>>921780

>>912938

Reddit uses markdown. It's markdown spacing. Markdown

follows a long line of email formats. Many email clients don't

softwrap, so if you want paragraphs you need to hard wrap.

HTML ignores hardwrapping, because it has it's own

softwrapping, so when you translate markdown to HTML, it

deletes your hardwraps. Where else is softwrapping ugly as

fuck? Oh right, when youre programming.


 No.921369>>921895

Why not just use S-expressions and load that?


 No.921372>>921393 >>921986 >>922012

(empinfo
(employees
((name "James Kirk")
(age 40))
((name "Jean-Luc Picard")
(age 45))
((name "Wesley Crusher")
(age 27))))
Fixed.


 No.921393>>921792

>>921372

>adhoc unstandardized sexpers

durr


 No.921531

XML and JSON were both disasters, but fail at different things so are used by different groups.


 No.921780

>>921363

its not that hard to not press enter twice


 No.921792

>>921393


<data> ::= (empinfo <empinfo>)
<empinfo> ::= (employees <employee>*)
<employee> ::= ((name <string>) (age <number>))

wow that was hard.


 No.921895

>>921369

My nigger.


 No.921955

>>912565

Markdown would be escaping linebreaks with \


 No.921986>>921987

>>921372

Looks like unreadable shit. You can tell where the words are, but if you fucked up the parentheses (for example remove one of the last 4) there's no way you would know at glance.


 No.921987>>921989

>>921986

>there's no way you would know at glance.

That's what computers are for.


 No.921989>>921992

>>921987

Computers are for binary data, not your gay expressions.


 No.921992>>921993 >>921998 >>923774

>>921989

Hello? These are formats for exchanging data between computers. Yes, we like if they are human readable, but of course the hierarchy will be ruined if you delete a character. You could make the same argument about json. You could make deleting a '/' from a closing tag in XML fucks it up too.


 No.921993

>>921992

*make a mistake... which fucks it up...


 No.921994

Anyone here like YAML?


 No.921998>>922005

>>921992

>computers can't understand file formats unless they're in some markup text language

Do you even computer?

>You could make the same argument about json

No you can't, the difference is that XML and JSON are much more human readable. If it's missing something then you can see it easily.


 No.922005>>922012

>>921998

Wrong. JSON also suffers from bracket hell at the end of complicated structures ]}]]]}. I cannot help you understand the first part, just count it as yet another thing you failed to understand.


 No.922012

>>922005

JSON isn't read by humans in a form with compressed brackets like LISPfags like doing with their shit. See >>921372 . No one would ever do that kind of madness with C.


 No.922135

Much much easier to parse XML than JSON I guess? Not that you'd know, as you've never programmed even a reverse polish calculator parser.


 No.922138>>923780

XML is much more powerful


 No.923369

>>912377 (OP)

Name spaces. Metadata. Validators.

>but who ever uses those?

Apparently enough.


 No.923744

XML is easier to read, but JSON has semantics built-in which make object composition more straightforward. You have lists, maps, and singe-values. When I need to serialize data in a program I'm writing, it's conceptually clear how the objects or data will be mapped into a JSON object. With XML, you have only two types to work with: tree and single-values. There is a wider conceptual gap between the data being serialized and the format itself.


 No.923751>>923790

JSON has no types and adding types makes it look like shit. It has no comments and adding comments makes it look like shit. It can't handle binary data at all. Almost no JSON parser actually parses JSON correctly. Don't believe me? Try adding a large floating point number and then parse it and re-serialize it with your library of choice. You'll get data loss with almost all of them. It's webdev-tier garbage.


 No.923752

JSON is a true choice for White People, you are literally using a nigger to do the job for you.


 No.923774

>>921992

Hey javanigger, if you want to transfer data between computers efficiently, you compress it into a raw datastream. XML is best used for configuration data that needs to be both human-readable and easy to parse for computers (and in many cases, also easy for humans to edit in a text editor).


 No.923776>>923778

>>912377 (OP)

They're both unreadable but at least you can have an IDE that will highlight what's XML and what is not, and like anon said >>912390 XML is easier to parse. So that makes XML better.

And if you use JSON you're getting into the java ecosystem which anyone who values their time, money and sanity wants to avoid like the plague.


 No.923778

>>923776

>IDE that will highlight what's XML and what is not

meant

>IDE that will highlight what's XML/JSON and what is not


 No.923780

>>922138

Oh yeah, I feel really powerful when I parse an XML document and my computer needs to connect to some remote server to finish parshing it.

XML is powerful in the same way Javascript is.


 No.923787

File (hide): 92d4e2a142cf2e5⋯.gif (231 B, 48x32, 3:2, valid_html.gif) (h) (u)

>>912377 (OP)

Crucial:

Schemas

Namespaces

Convenience:

Comments

Cleaner diffs (that fucking comma at list/object end)

Minor:

Can be used for markup (quite niche)

Tool & library support that comes with being standard for decades. XPath, XSLT, etc (JSON is catching up)

Controversial:

Attributes (cancerous & misused in 95% cases, but perfect fit for 5%)

Only problem with XML is that is was used for simple cases, where it's functionality was overkill.

>>912454

>keys without quoting

>let's double syntax complexity for minor convenience

Into the trash it goes. There is a reason why JSON syntax is one page and YAML is a fucking book


 No.923790>>923863

>>923751

>Try adding a large floating point number and then parse it and re-serialize it with your library of choice.

>json.loads('1.1', parse_float=decimal.Decimal)

>{0, 1123, -9} = jsonx:decode(<<"1.123e-6">>, [{number_format, decimal}]).

And?


 No.923863

>>923790

>json.loads('1.1', parse_float=decimal.Decimal)

Now try re-serializing it with dumps. While it can be done, it's ugly.

>{0, 1123, -9} = jsonx:decode(<<"1.123e-6">>, [{number_format, decimal}]).

Erlang is a meme. Try parsing a string with \ud83d in it.

And then try sending long floats to a browser and see what happens.




[Return][Go to top][Catalog][Screencap][Nerve Center][Cancer][Update] ( Scroll to new posts) ( Auto) 5
95 replies | 9 images | Page ?
[Post a Reply]
[ / / / / / / / / / / / / / ] [ dir / bestemma / general / imouto ][ watchlist ]