« first day (253 days earlier)      last day (407 days later) » 

El'endia Starman
El'endia Starman
12:09 AM
O.o Uhhh...I blame the poltergeists.
 
El'endia Starman
El'endia Starman
12:24 AM
@quartata I got the same error, and after googling, it seems like the fix is trying repeatedly until it works. Took me three or four tries altogether, I think.
 
quartata
quartata
12:39 AM
@frontend
func:foo() {
  \alert("Retrieving something");
  \return(\div[style="some css](\backend()))
}

@backend
func:backend() {
  \return("2 + 2 = 4");
}
Ugh Markdown pls 
<body>
  <foo />
 </body>
This is difficult on mobile... anyways
The functions with the frontend annotation are compiled to JS. Any crossing between the two ends is done with AJAX magic
And then you can call functions like HTML tags from the view
 
quartata
quartata
12:56 AM
@El'endiaStarman
 
El'endia Starman
El'endia Starman
This could be really cool.
 
quartata
quartata
Oh just pinged you as soon as you edited that
 
El'endia Starman
El'endia Starman
Good timing, I guess. :P
 
ASCII-only
ASCII-only
1:11 AM
What do the annotations indicate
 
quartata
quartata
Where the code is placed when compiled
 
ASCII-only
ASCII-only
Oh just realized you can have multiple of each location if I understand correctly
 
quartata
quartata
As JS in the view or as a CGI script
 
ASCII-only
ASCII-only
So basically here backend clientside would get compiled into a POST request to /backend (some location derived from the function name?)
 
quartata
quartata
Function calls or global variable access that exist on the backend get relayed from the frontend to the backend with a POST request yes. I might do something with websockets instead for bidirectional
 
ASCII-only
ASCII-only
1:47 AM
IMO only use websockets for realtime connections/long-term connections
 
quartata
quartata
2:07 AM
@El'endiaStarman Did you fix the parse bug?
@ASCII-only That's the only time the backend would be making a call to the frontend anyways
 
El'endia Starman
El'endia Starman
@quartata >_> No...
Non-work things happened.
I'm fairly sure I'll just check to see if the current node is AST_root and if it is (at EOF), then just finish it.
 
quartata
quartata
Sometime tomorrow I'll make lists not suck. Then it's really just generators and we're good to go. I'll make @doc be a nop so I can write docs and leave them in for later when the docgen works
 
El'endia Starman
El'endia Starman
I think I'll try using WebSockets to implement a REPL.
 
quartata
quartata
@includec needs dlopen so that'll have to wait until we start cythonizing
Oh hey, we never did floats...
 
El'endia Starman
El'endia Starman
Haha, nope.
I have another 6 am shift tomorrow morning, so I'mma head to bed now. G'night!
 
quartata
quartata
2:18 AM
g'night
 
 
13 hours later…
quartata
quartata
3:41 PM
This is super spooky. 
  generator.state = DONE;
  swapcontext(&generator.generator, &generator.me);
  //puts("how did I get here");
  throw std::out_of_range("");
This segfaults, but commenting out the exception doesn't
But uncommenting the puts doesn't change the output
so that's never reached and yet it's somehow causing a segfault
Maybe trying to make generators in C++ without a library was a mistake
 

« first day (253 days earlier)      last day (407 days later) »