|
Izzy Muerte
@
slurpsmadrips
Oakland, CA
|
|
Occult Programmer and C++ Bruja. void* made manifest. ✨🏴✨ ⚢ | ⚧ | xe/xyr
|
|
|
23.008
Tweetovi
|
664
Pratim
|
2.306
Osobe koje vas prate
|
| Tweetovi |
|
Izzy Muerte
@slurpsmadrips
|
8 h |
|
Also: I think I'm gonna remove the dict() type from IXM. I've been using it less and less and less. It's been a fun experiment, and I might add it back in the future. But using directory/global properties has been easier for tracking project state 😬
|
||
|
|
||
|
Izzy Muerte
@slurpsmadrips
|
8 h |
|
Yet, here I am trying to not only make it easy and data driven, but also permit translations in the near future.
|
||
|
|
||
|
Izzy Muerte
@slurpsmadrips
|
8 h |
|
After that, I'm back to working on the documentation. Hugo has changed a bit, and my design has as well. Documenting CMake calls is such a PITA, even their sphinx plugin can't be reused :(
|
||
|
|
||
|
Izzy Muerte
@slurpsmadrips
|
8 h |
|
It's needed though, otherwise folks have to specify the git repo, and I'm trying to keep the FetchContent_Declare call simple: FetchContent_Declare(ixm URL <url here>)
Next up is to get a decent github action setup.
|
||
|
|
||
|
Izzy Muerte
@slurpsmadrips
|
8 h |
|
Currently converting the cloudflare workers router for IXM to rust. Helps the bootstrap situation.
Sorry in advance however. I'm generating a zip file on the fly per request. Is it a good idea? probably not. Is anyone really gonna care? Probably... not???
|
||
|
|
||
|
Izzy Muerte
@slurpsmadrips
|
15 h |
|
No joke, I was said to myself "I want to make a Gibson reference. Time to throw the word 'crystal' in here somewhere"
|
||
|
|
||
|
Izzy Muerte
@slurpsmadrips
|
15 h |
|
I sit pensively next to the jack port in the cyber cafe. The tech barista hands me a macchiato and a crystal uplink. I thank them as I grimace from the heat. It reminds me of human touch.
I jack into the cyber grid. Just another Tuesday morning in Neo New San Bay Oaklandcisco. twitter.com/hexadecim8/sta…
|
||
|
|
||
| Izzy Muerte proslijedio/la je tweet | ||
|
Hacking Our H E X A G R I D
@hexadecim8
|
4. velj |
|
|
||
|
Izzy Muerte
@slurpsmadrips
|
16 h |
|
Happy to help! I’ve got tons more stuff planned too! :)
|
||
|
|
||
|
Izzy Muerte
@slurpsmadrips
|
3. velj |
|
Implicit int *is* dead in a standards conforming compiler :P
|
||
|
|
||
|
Izzy Muerte
@slurpsmadrips
|
3. velj |
|
Always a good sign. 😬 pic.twitter.com/gBsRyl5AqE
|
||
|
|
||
|
Izzy Muerte
@slurpsmadrips
|
3. velj |
|
I hope to make it. Options paper is gonna be a fun argument 😬
|
||
|
|
||
|
Izzy Muerte
@slurpsmadrips
|
3. velj |
|
easier than generating one :v
|
||
|
|
||
|
Izzy Muerte
@slurpsmadrips
|
3. velj |
|
For the target tuplets, it's still in draft form because I need to clean up some wording. Options file is gonna be "fun", because I'll be proposing both a legacy format (what folks currently support) and a modern format (literally just a JSON string/file)
|
||
|
|
||
|
Izzy Muerte
@slurpsmadrips
|
3. velj |
|
I used to know this all off the top of my head but we've made such huge strides that LL(k), LL(*), LL(1), GLR, IELR, I just can't figure out which one is "needed" these days.
Such is life.
|
||
|
|
||
|
Izzy Muerte
@slurpsmadrips
|
3. velj |
|
It's less that, I reckon, and more that parsing tech has come a long way, but we still teach LALR to kids. IIRC every C++ parser at this point requires a GLR capable parser, but it's easier to write a recursive descent.
|
||
|
|
||
|
Izzy Muerte
@slurpsmadrips
|
3. velj |
|
Removing K&R decls means we're one step closer. But I'm still blocking out edge cases for parsing and there are quite a few. Vendors might say no because it's more work for them but we won't know until we try :v
|
||
|
|
||
|
Izzy Muerte
@slurpsmadrips
|
3. velj |
|
It becomes optional, not removed. So `struct X { }` will compile, as will `struct X { };` some look ahead would be required. C89 code, though, won't be compatible at some point since 1) implicit int is banned in C99 2) K&R declarations are being removed :D
|
||
|
|
||
|
Izzy Muerte
@slurpsmadrips
|
3. velj |
|
That said, requiring it if you do `struct X {} y;` and not requiring it for `struct X{}` is feasible, I reckon :v
|
||
|
|
||
|
Izzy Muerte
@slurpsmadrips
|
3. velj |
|
It also implies a compiler will have to view the context of an incorrect declaration and ascertain the correct behavior. Even though it's been banned since C++98 to have implicit int, and banned since C99, the big three don't really treat them as hard errors :/
|
||
|
|
||