Who the fuck thought it would be a good idea to use opening and closing tags as DA STANDARD for SHITML? Are the people at W3 dumb?
For those who don't know what I'm talking about, look how ugly and unnecessarily long HTML is.
<html> <head>MY WEBSITE</head> <title>MY WEBSITE</title> <body> <p>THIS KIND OF MARKUP IS <b><i><u>SHIT QUALITY</u></i></b></p> </body> </html>
But compared to the S-expressions
(html (head "MY WEBSITE") (title "MY WEBSITE") (body (p "THIS MARKUP IS" (b.i.u "SATORI QUALITY"))))
It looks far cleaner and makes your eyes feel right at home. It's easier to parse and won't take up as a few kilobytes / megabytes (depending on the page) less Internet bandwidth. Sure it has a lot of parenthesis but it's more attractive than the repeating shitstorm of </tags>.
I know there are a lot of implementations to this but there should be one STANDARD to rule them all. The next version of HTML in the standard should be something like this.
Opening and closing tags aren't [i]that[/i] bad, SGML-likes lend themselves to markup well when the amount of markup isn't under a few percent of the entire data.
What's really horrible is that tags may be implicitly closed. Whoever thought of that should be shot, it defeats the entire purpose of using a SGML-like language.
Are the people at W3 dumb?
Of course they are, it's the same team that defends DRM hooks in HTML with the "argument" that it makes the web more open.
SGML-likes lend themselves to markup well when the amount of markup [b]isn't[/b] under a few percent of the entire data.
I really hope you meant to say is, not isn't.
Name:
man(7) && mdoc(7)2013-09-17 14:54
>>1 Tim Berners-Lee, Yes, they can't do even proper \fIstyling\fR correct, and decide on things. We did, it was called [t]roff. Hipster Physicist didn't like it, so we are stuck in this mess. .PP >>2 We know, it is \fBMarkup\fR .PP >>3 Because Javascript was introduced. Should have kept it as Scheme. >>4 Meh. Other markups are better: org-mode, Markdown, Asciidoc, wikicreole, heck even wml. .PP >>5 tablecat's broken regex. .PP >>6 You an replace them with .RS if <like == true> <statement_limiter = '<>' > else <statement_limiter = default >
>>9 We know it is. But not enough to self reproduce.
We should really make a petition, and make a UI support for those markups. It's sad the Chrome has 46.02% of the browser market. Its all Javascript, sigh. Why can't they do one thing correct.
>>1 I agree with you, but if you think beyond just that, you realize you don't even need an html or a body object. Using S-exprs is one step towards a saner web, the other one is redefining the shit standard that has stupid things like tag arguments, <p> | <br> | <blockquote> | <q> | <em> | <i>, other redundant tags and even the le whoops forgot my closing tags sorry xDDD leniency of most shitML parsers.
Name:
Anonymous2013-09-18 7:47
But it's also a repeating shitstorm of PARENS
Name:
Anonymous2013-09-18 15:35
>>13 I'd rather have a ))) than a </div></div></div?.
I agree that HTML's syntax isn't that great, but if you want to try bootstrapping a new ML onto the internet, why constrain it to be ``HTML, but with parens instead of brackets and slashes''? Why not make a self-sustaining system like Gopher or the original WWW? There's plenty wrong with the web these days that could be fixed by completely changing the underlying ML, so try and grab as much of it as you can. It's not like you're going to do anything anyway.
>>15 That's what I said in >>12, except for the "it's not like you even know how to do it" part.
Name:
Anonymous2013-09-19 0:17
>>1,15, etc So given the impossibility of replacing HTML, how about this stupid solution? Write ()ML then compile it to HTML. You could even do it using a .js library on demand, allowing for improvements to the size of the content to be propagated at the cost of an increase in client-side processing time.
>>17 It has been done already, but I refuse to use any ML that retains the cruft of the HTML standard. (p text) is only slightly better than <p>text</p>, but the fact that p is stupid crap remains.
>>17 welcome to the 1990's where documentation generators will be the way of the future. No more need for nasty old databases talking to a script in the Common Gateway Interface to deploy output. And we will definitely not go with the live output while code is changing plan as we go. We will have none of that.
-CMS [sub]Agile based
Name:
Anonymous2013-09-19 2:05
>>18 I should have known it had been done. I agree with you about the tags as well. Div vs span vs p seems odd, and my God, css. Inheriting float from the previous sibling etc. is just silly. It simply isn't easy to do something that should be. The whole thing needs to be thrown out. That's sort of my thought with compiling to HTML: you wouldn't have a collection of HTML tags any more than C is a one to one map of assembly. Closer to the subject at hand, one could do for HTML what jQuery did for JavaScript, as limited as that may be.
That said, >>19 has a point as well: if you look at the HTML CMSs generate it's pretty terrifying. One would need to write an optimizing compiler, so to speak.
Your assignment: Write a S-expression -> XML parser in Haskell or any functional programming language of your choice. Extra credit for included unit tests!!!
>>17 How about not? That way, you would have to deal with the quirks of HTML and the quirks of your replacement. That sort of thing is why we have Automake and its family (the Parcae of programming), as well as all those goddamn X-to-javascript projects.
Name:
Anonymous2013-09-19 21:23
>>22 Well I've thought about that. JavaScript is difficult to write a converter for, because it's genuine code. HTML on the other hand is, as has been exhaustively explored in this thread, not a real programming language. Its sole output is visual. So it's possible that in compiling your ()ML you might receive build errors if a rendering engine (webkit, gecko, etc.) put something in the wrong place, or hid it when you wanted it to display, etc. It's probably possible to automate that process. So if you set out clear specifications for your language, with no confusion as to where things are or what's on top, you could be automatically notified if something failed to land where it should.
>>27 You had a user base! You're on your way, son! Just think - if your little cousins played with sticks and snot, how many parents would pay to keep their little ones occupied? Mister, you're on to something!
It's your 10th birthday, and your favourite uncle has just turned up.
Giddy with excitement, you run to the door to see what he's brought for you.
'Yay-' you start to say, as he unloads a roundhouse kick in your face.
Happy birthday, kid! Have fun eating through a straw the rest of your life.
...Fuck you, [Tim Berners-Lee].
Name:
Anonymous2013-10-04 21:58
>>32 I'm being cynical, but it was about time that the current web dies anyway. This is our chance to fix what is fundamentally wrong with its standards.
Name:
Anonymous2013-10-04 22:30
>>32,34 LOL, I love this guy's comment: #9 By: TheGlitchEcliptic, October 3rd, 2013 01:52
We need an xkcd-style super villain to create an implausible scenario highlighting the potential abuses of this.
"The terrorists have planted thirteen bombs across the city and their locations are hidden in this website's source code!"
"I can't access the source code, sir!"
"Damnit! If we don't get those locations we might as well call off the bomb squads and call in the coroners!" The timers are already counting down!"
"Hang on, I just googled a web forum where an amateur programmer offers a Firefox plugin that exploits a workaround!"
"Thank god!"
"The link's no good anymore! It was hit with a DMCA takedown!"
"What?! Who filed the complaint?"
"....the terrorists, sir."
"Damnit! How are they always one step ahead?!"
Name:
Anonymous2013-10-04 22:39
>>32 I feel physically disgusted by reading this. What the fuck? This is way too much for me.
Name:
Anonymous2013-10-04 22:45
>>34 Your not being cynical, you are actually being an optimist, thinking we can actually fix it, and making an implementation in time. I hated the web anyways. USENET was and should have been the norm. I just hope the secushare guys can make an translators for must protocols out there, even a http2psyc gateway: http://about.psyc.eu/Software
>>39 D fock? Any message you send is encoded in proper MIME, which any client can use to interpret however specification required. And hyperlinking was as easy as saying:
Since it is distributed in nature, and you can edit your message as to not have identifying information on the nodes. It is still the best federated P2P pseudonym network. high and low latency capabilities, and great caching. Even better when entire channels were a community hosted DHTs, of many different things.
Don't blame the protocol for your perceived lack of UI support for multimedia presentations and hyperlinked pages. Most common ones could and by default do: wikipedia.org/wiki/List_of_Usenet_newsreaders#Free.2Fopen_source_software wikipedia.org/wiki/Comparison_of_Usenet_newsreaders
Name:
Anonymous2013-10-06 20:37
>>40 In practise, do most people instruct their clients to render Usenet posts in the same manner as a web page? If not, Usenet is only as good as plaintext email and is no good for presentations and hyperlined pages.
>>41 I just said the moderns do. Browse some and try 'em. You have also used email with HTML documents and other attachments, right? So much so there are now webmail clients. The stupidity of that is beyond me, but it proves the point: UI for clients can easily adapt the MIME files however they wish, in whatever medium of choice. Heck, even in hotmail and gmail you can view ppt, pdfs, and video!
>>42 I know right[i]![/i] We just force them anyways:
To view this page properly, you need to have an SexpML render on your browser. You might want to consider installing one of these:
Name:
Anonymous2013-10-06 21:32
>>43 What if Scheme were the default scripting language of the web, and SexpML were the default markup language? Imagine all the money Mozilla would have thrown at it. Imagine the faster-than-C JIT compilers. Imagine the amount of libraries it would have. Imagine all the people using it for their epic Reddit clones!
Wait a minute, that last thing sounds awful. Writing a browser from scratch is a better idea. Hopefully Hacker Jews won't hear about it.
Name:
Anonymous2013-10-06 21:40
>>44 With EME being pushed by TBL himself, people actually started removing the dust from gopher. It's a ridiculous amount of optimism, but this might be the chance Scheme needed.
Blame the Netscape Managers for that big blunt that plagues the world.
>>46 I'd built it so that you have a choice. Plus, Scheme was great to run interpreted, like JS is today.
>>47,48 No, he just write the following in his features.conf: CONFIG_SCRIPTING_SPIDERMONKEY=no
Name:
Anonymous2013-10-07 3:19
>>49 A lot of JS gets ``accelerated'' today and it's a big source of security holes. What I want is a decent, very simple bytecode specification that can be interpreted easily and safely, crafted in a way that you would really have to really go out of your way to fuck up implementing an interpreter for it. Also very little to none interfacing with native code; none of that DOM shit with hundreds or thousands of C procedures interacting with javashit code.
Give me a couple of days, and I can write a standard. Unless you want to write one first. Be sure to make the headers secure enough for clients to reject server requests, esp. for does that do not want to yield identifying info.
I can't understand for the life of me why you people insist that SGML-like languages are ``fine''.
"Oh, HTML sure is utter shit, but SGML is good". "The W3C are a bunch of retards. Instead of HTML-shitbucket, they should have stayed with SGML-turd".
Can someone explain this reasoning? I've seen it quite frequently here. I just don't find any reason to think <tag></> is not something a retarded nigger kike would come up with.
>>57 Am I missing something, or is this just SQLite?
Name:
Anonymous2013-10-08 1:33
>>58 Its a API standard that was dropped by W3C for database creation and management on a web browser, for making other datasets, like cookies, JSON files, flash balls, java bytecode, and the kind. The main reasoning behind the dropped is because the only implementation that supported it was SQLite at the time. No one else pitch in others, like SmallSQL , H2 DBMS, LevelDB, Kyoto Cabinet, LevelDB, LMDB, or Berkeley DB. But it was a great idea they should have kept, like they have kept Javascript, CSS, and HTML!
Name:
Anonymous2013-10-08 17:16
>>59 I honestly don't see why storing absolutely all your data (presentation format and content) in an obfuscated format you have to access through an API is a good idea. Also, it was made by the W3C, and that's not exactly a signal of good design.
SQLite for the data that needs to be in a database and SexpML for the rest is fine.
Name:
Anonymous2013-10-09 6:38
>>60 I'd use it for housekeeping and data (cookie management?). The rest of the document files would be sent on regular HTTPS, like any document. But with the client side scripting, it is a greate way to have the client self-organize the response, with a well formatted database.
Optionally, you can extend the database to keep all types of data, so that the user has only one or 2 files to keep per site/page/domain/etc.. At the least we agree. Maybe make a server option for clients to ask the type of formatted response they want.
Polish notation is harder for plebians to parse than simple tags.
Name:
Anonymous2015-11-03 3:20
Something I've noticed among most (all?) popular languages is that they are easy to format with dumb text editors. My first experience doing anything resembling programming was creating a homepage in Microsoft Notepad. At the time, the thought of indenting my HTML for readability had never occurred to me, and I wrote my HTML with 1 tag per line (with the exception of formatting tags like <b>. But had I been aware, it would have been very easy for me to format, even in Notepad: just hit Tab a few times before you type.
Which I ended up doing in one of my first brushes with "real" programming, C++. Writing "guess the number" games and other trivial examples out of a book (before it went waaay too far over my 10~ year old head with C++ arcana) in Notepad, Tab'ing away at every new line. Of course, when I discovered "real" text editors and their ability to auto-indent newlines, I was overjoyed, but nevertheless I had been undeterred by the tedium of formatting HTML and C++ by hand for a few years.
Several years later I discovered SICP through /prog/, and with it, Lisp. I didn't quite "get it" the first time, thinking of the book as any other introductory programming book but with needlessly complicated and math-heavy examples, and though of Lisp as just another programming language, but with strange formatting. It certainly didn't help that I didn't have a text editor capable of formatting sexps at the time. Mashing space to line up sexps correctly got old fast, so I used an Algol-like formatting until I gave up on the whole ordeal.
A few years later, I'd come to understand emacs just well enough to make writing Lisp possible, and with that, I was able to truly appreciate the beauty of SICP and Lisp. But how many others like me never made that leap? And how many others like me understood emacs but hated using it so much that they never touched a piece of Lisp code outside SICP ever again? I'd love to write more Lisp, but I think I'd have to write my own text editor and Scheme before I could justify it, so instead I stick with my comfort zone of C.
Likewise, I made a pass at Haskell around that time (again because of /prog/'s memes). Likewise, the strange syntax, formatting, and dependence on emacs and vim turned me off (though in Haskell's case the semantics of a purely functional language and the community's obsession with faux-intellectual formalizations of simple concepts turned me off harder).
Call my an idiot all you like for having bad taste in text editors and programming languages, but my point is that if you need a special text editor to write in a language, it will probably never ever take off amongst the masses of "simpler-minded" programmers that make or break platforms, let alone the non-programmers that once wrote their own homepages in Notepad. So while today we can sit around and bemoan the comparative ugliness of HTML and Javascript, the web as we know it might not exist without their relative ugliness.
Name:
Anonymous2015-11-03 3:34
>>64 What are you talking about? It's incredibly easy to format S-expressions, even in the dumbest of the text editors, like nano, ed or notepad. This is not a valid argument either, since it would be easier to write an hypothetical Lisp-like language on notepad than writing Javashit in Atom.
Name:
Anonymous2015-11-03 4:32
>>65 Formatting SGML in a shitty text editor: Most people just write it all left-justified anyway, so there's nothing to format. Including the tag name in the closing tag makes it easy to figure out where you are in the tree, even if you lack indentation to convey the structure.
Formatting C and most Algol-like languages in a shitty text editor: New scopes and scope-like constructions require an indentation level. Hit tab 1-4 times before you type. If you need more than 4 indentation levels or so, you're probably doing something wrong.
Formatting sexps in a shitty text editor: Small sexps are inline: (* 2 4) More complicated sexps that are split across multiple lines have each parameter indented to the same amount. (+ (* 3 (+ (* 2 4) (+ 3 5))) (+ (- 10 7) 6)) Spaces are required to properly indent this code. It's a lot easier to indent 1 tab per indentation level than it is to spam the space bar twenty times in the kind of deeply-nested code that Lisp encourages. If you change a function name, every sub-expression is now improperly aligned, and you will have to realign them manually: (plus (* 3 (+ (* 2 4) (+ 3 5))) (+ (- 10 7) 6)) And if you don't follow these rules, it quickly becomes extremely difficult to follow your code. More importantly, shitty editors lack paren matching, so you'll be wasting tons of time manually counting parens and running into subtle bugs when you make mistakes.
This never ever would have flown with the tech-illiterates copying and pasting their way to their first homepage that drove the early web.
Name:
Anonymous2015-11-03 5:07
>>66 Not wanting to use emacs doesn't mean using a shitty editor. The point is that s-exprs are super easy to edit in any basic non-shitty editor.
Also, emacs is shitty.
Name:
Anonymous2015-11-03 5:20
>>67 And my point is that the web became what it is because people could easily make complete pages in fucking garbage like Notepad (or even worse, directly in an HTML form, which come to think of it was what I really did as a kid), which was usually all they had available to them. If you had to find and install anything (which almost certainly involved spending upwards of $100 in those days) and spend more time learning how to format the sexps on top of the document markup already blowing your neophyte mind, it wasn't going to happen for 99% of user/creators.
This argument is pointless because you can easily make a DSL that outputs the HTML the way you want to without even having to deal with the mark-up yourself.
Name:
Anonymous2015-11-03 14:19
>>68 Don't be a retard, nobody makes webpages in notepad anymore. Why do you think shit like ``Zen Coding'' and many HTML/CSS/JS preprocessors exist? It can't be because they're beautiful and easy to use.
Why do you insist on HTML being easier to format, if you want form an injection from HTML to S-expressions? Replace <tag> with (tag and </tag> with ). Then your notepad webpage becomes
(b (i (o text) ) )
which looks absolutely retarded but is easy to follow even in fucking ed. It's not like your left-aligned HTML looks better anyway.
>>71 I never said HTML or Javascript is superior (believe me, I hate the web just as much as you fags). But there are people ITT bemoaning Tim Berners-Lee, Eich, and Netscape for not giving us the pure Lisp web from the start, and I'm just saying that the web probably would have never taken off if they had designed it that way. It probably would have been supplanted by some ugly corporate Java AIDS that ex-C++ programmers and Bobby making his first webpage would find more intuitive.
And yeah, I know no one writes straight HTML anymore, but they sure as hell still write Javascript. And even today the web appers would probably be scared away by the sexps.
Name:
Anonymous2015-11-03 22:51
>>73 Do you not think that web appers are just comfortable with that shit because it's what they've been using all these years? As in, if javascript was a lisp dialect in the first place, nothing else would be different. In that universe, web appers might be scared away by braces and K&R indentation.
Name:
Anonymous2015-11-03 22:55
>>73,74 A proper web wouldn't have any appers working on it. A proper web is one made for EXPERT PROGRAMMERS by EXPERT PROGRAMMERS.
Name:
Anonymous2015-11-03 23:15
>>75 The fuck you talkin'bout nigger? The web was made so writers could publish documents, not all this fancy schmancy shit.
Name:
Anonymous2015-11-04 2:49
>>76 The web was made so manly military men could trade secrets.
Name:
Anonymous2015-11-04 8:39
>>73 How in the fuck do you faggots thing are people going to be scared away by (b foo), when every simpleton shitcoder accepted <b>foo</b> without complaint?
Get off my /prog/. It's supposed to be shit, but you're an insult to shit.
Name:
Anonymous2015-11-04 9:07
Stop wasting time discussing this subject. The ship has long sailed and HTML will not change to accomodate the inconsequential whining of some BBS. Gzip compression takes care of the worst of it.
If you want you can always use JADE when you're "programming" (upmarking) or write your own S-EXP => HTML preprocessor if your autism compells you to.
>>74 This is the lie functional programmers and Lisp programmers have been telling themselves, that their respective languages never hit the mainstream through historical accident, that C-like languages won through a "worse is better" approach, and that there's a self-reinforcing cycle of unfamiliarity.
Not so, the languages are shit. They at best let you write passable code if you manage to simulate some of the mutability of good languages, but any sufficiently large project will inevitably boil down to a sticky macro and/or operator soup.
Name:
Anonymous2015-11-05 1:55
>>81 Your "good languages" make the easy easier, and the difficult impossible.
When you get into massive projects, Lisp will keep you from pulling your hair out. It's enjoyable and productive in the large. Your languages are only "fun" in toy programs.
When you get into massive projects, Java will keep you from pulling your hair out. It's enjoyable and productive in the large. Your languages are only "fun" in toy programs.
Now that's more like it!
Name:
Anonymous2015-11-06 4:50
>>78 For the formatting reasons I explained above, and more importantly because 99.99% of them are going to give up the second they see (+ x y).
Never thought I'd say this on /prog/ but you are seriously overestimating the intelligence of the average computer user.
Name:
Anonymous2015-11-06 8:05
In an alternate dimension, we are using binary formats for our html. Web is not running in debug mode, everyone is happy
Name:
Anonymous2015-11-06 8:13
>>84 How would they see (+ x y) in HTM fucking L? There aren't any arithmetic operations allowed in current HTML version, so why should there be ones in LispHTML? Besides, the average computer user doesn't even need to look at HTML.
Name:
Anonymous2015-11-06 9:06
The most important feature is restricting the grammar of any new markup to be decidable by a regular language. An endless number of problems would be solved by simple, finite grammar checkers.
Name:
Anonymous2015-11-06 22:08
>>85 In an alternate dimension, secure sandboxes let everyone run whatever the fuck native code they want so there's no incentive to run everything through a web browser just because it's the only untrusted path that 99.99% of users will accept.
>>86 I thought the entire point of the Lisp web idea was to unify the page content language and the embedded scripting language.
Name:
Anonymous2015-11-07 0:55
>>88 That is a truly profound idea. The top level form in the page namespace should evaluate to the form that is used to render the page. Then markup and script become one. You can run arbitrary code in the middle of a ``HTML'' form.
>>1 SGML did not need closing tags Well, it did need them but like this: <head>...</>
Name:
Anonymous2015-12-31 3:52
Why not M expressions?
Name:
Anonymous2015-12-31 4:34
>>92 M expressions are Fortran-inspired shit that people rejected even back in the 1950s.
Name:
Anonymous2015-12-31 17:17
>>93 But do you know why? M-expressions make the code readable. Readability is the last thing you want if you are trying to convince people that Lisp has magical properties and is the only language that can be used for AI.
That's also why they killed LISP 2. Nobody will believe that AI can arise from something they understand.
Name:
Anonymous2015-12-31 19:57
>>94 You don't know the first thing about trolling. This shit is anything but "more readable":
: let var val [ m-expr; m-expr; m-expr ];
Name:
Anonymous2015-12-31 20:44
>>91 SGML apologists like you should set themselves on fire.
Name:
Anonymous2016-01-01 0:54
>>96 I don't even like SGML, I am simply expressing a truth. EAT A DICK