| Age | Commit message (Collapse) | Author |
|---|
|
In both the engine and SST it's used in more places than just console
printing, so it makes more sense to give it a more appropriate nanme and
location.
|
|
|
|
|
|
I need to write a pre-commit hook to check this or something.
|
|
This has been a known issue for a long time. Let's finally fix it!
|
|
These were never actually used before or this would have been caught
sooner, but as they're being used in some in-progress code this
obviously needs fixed now.
The virtual table indices in use were wrong for some engines due to the
funny thing Windows does with reverse-ordered overloads. The code for
building the plugin's own virtual tables explicitly accounts for this
but calling into existing cvars failed to do so, due to a skill issue.
This fix is kind of inelegant and stupid but making it elegant and
non-stupid is less of a priority at the moment. It should happen
eventually with future improvements to the gamedata system, for some
definition of "should," "eventually," and "future."
|
|
Removes the mem_store functions because they're never used anywhere.
|
|
Note: fixes.c was changed this year, so correct the copyright year too.
|
|
Nobody seems to have been affected by this in the wild yet, thank
goodness. Will probably be a while before there's an actual release, so
hopefully people will continue not to be affected.
I suppose it's only an issue for paths longer than 128 characters, so
it's not too likely to matter under normal circumstances...
|
|
|
|
|
|
Thanks Bill for figuring out what the problem was.
|
|
This was an accidental regression in an earlier refactoring.
|
|
|
|
|
|
Thanks Aciidz for pointing out the strings with no newline in them, and
thanks Clang for warning me I wasn't using a function for ages while I
was too lazy to delete it.
|
|
Another big one. Here's a list of things:
- Since the upcoming C23 standardises typeof(), use it as an extension
for the time being in order to allow passing arbitrary types as
macro/codegen parameters. It wouldn't have been a big leap to do this
even without standardisation since it's apparently an easy extension
to implement - and also, to be honest, this project is essentially glued
to Clang anyway so who cares.
- Likewise, bool, true and false are becoming pre-defined, so
pre-pre-define them now in order to get the benefit of not having to
remember one header everywhere.
- Really ungodly/amazing vcall macro stuff now allows us to call C++
virtual functions like regular C functions. It's pretty cool!
- Events can now take arbitrary parameters and come in two types:
regular events and predicates.
All this makes the base code even uglier but makes the feature
implementation nicer. In other words, it places more of the cognitive
burden on myself and less on other people who might want to contribute.
This is a good tradeoff, because I'm a genius.
|
|
Shoutouts to Lokki for accidentally reporting this bug.
|
|
|
|
It seems like the nice thing to do.
|
|
This is for somewhat later. I'd always planned to use it - it existed
already in earlier private repos, in fact. I just didn't bother to
import it here in case it wouldn't actually be needed, but with the way
current plans are going, it's definitely going to be needed, so here it
is.
|
|
Spotted by Bill the other day.
|
|
|
|
|
|
Only if the code page in question is that of the system, obviously;
Source itself will fail to start otherwise anyway.
This fixes issues for at least one Russian user who'd named a "Games"
folder "Игры," which I am told roughly translates to "Games" (and also
exactly translates to "Games").
|
|
Copyright note: the stuff Bill wrote is all still in the other file.
|
|
Committer's note: this is somewhat adapted from Bill's original code,
written a while back, but he gets full credit for actually doing the
hard part.
|
|
Also centralise NEXT_INSN macro, into its own header at least for now.
|
|
|
|
|
|
|
|
|
|
Some of the alias stuff was kind of stolen from earlier figuring-out
Bill did. More Bill code is also on the way. :^)
|
|
|
|
It'll be less confusing not to have development builds versioned the
same as existing releases. That's probably obvious, but I'm still new to
this releasing software business!
|
|
|
|
|
|
Preparing for release, I spose!
|
|
You can widen the text field to read the whole thing, but that's not
immediately obvious. Let's just shorten the text and make it nice and
concise so people can figure out what to do more easily.
|
|
|
|
Currently only works in 3420 and 5135 and uses hardcoded offsets with a
byte pattern sanity check. Future work includes making it more widely
compatible, and also doing the crazy thing I wanted to do but gave up on
wherein the actual textures and stuff get patched in memory to sync up
all the colours.
Oh also, a couple of vtables were erroneously made executable, so I went
ahead and fixed that while I was at it.
|
|
|
|
|
|
Meh, whatever.
|
|
A stupid, stupid bug I'd probably never have caught for years and years.
Thanks Evan Lin in the Portal Discord for pointing it out!
|
|
|
|
People want to be able to write a scheduled-release passphrase to demos
to prove that said demos were recorded during a timeboxed event. Ideally
this is something that'd be part of the custom demo data and general RTA
run management stuff that's planned, but of course that doesn't exist
yet. When it does exist, we'd probably want rid of this nonsense, so
it's marked hidden and I don't plan to document it anywhere "official."
It's just here for people who really need it in the short term.
|
|
|
|
This is how it should've been to begin with, but I was lazy.
|
|
I still haven't bothered to flesh out all the possible indices for
GetEngineBuildNumber but we're back to being able to use it to detect
The Last Stand (as a result of some old version testing Aciidz did -
thanks again!). This means we can do away with the terrible map file-
based hotfix. Also, turns out sst_l4d_testwarp doesn't work if the
plugin gets loaded early via VDF because certain SendTable offsets start
out negated, so we work around that too now.
|
