This is simply exposed as a table from tag -> true. While this is less
natural than an array, it allows for easy esting of whether a tag is
present.
Closes#461
- Use texture over texture2D - the latter was deprecated in GLSL 1.30.
- Cache the tbo buffer - this saves an allocation when monitors update.
Closes#455. While the rest of the PR has some nice changes, it
performs signlificantly worse on my system.
This moves monitor networking into its own packet, rather than serialising
using NBT. This allows us to be more flexible with how monitors are
serialised.
We now compress terminal data using gzip. This reduces the packet size
of a max-sized-monitor from ~25kb to as little as 100b.
On my test set of images (what I would consider to be the extreme end of
the "reasonable" case), we have packets from 1.4kb bytes up to 12kb,
with a mean of 6kb. Even in the worst case, this is a 2x reduction in
packet size.
While this is a fantastic win for the common case, it is not abuse-proof.
One can create a terminal with high entropy (and so uncompressible). This
will still be close to the original packet size.
In order to prevent any other abuse, we also limit the amount of monitor
data a client can possibly receive to 1MB (configurable).
timetout, max_upload, max_download and max_websocket_message may now be
configured on a domain-by-domain basis. This uses the same system that
we use for the block/allow-list from before:
Example:
[[http.rules]]
host = "*"
action = "allow"
max_upload = 4194304
max_download = 16777216
timeout = 30000
This registers IPeripheral as a capability. As a result, all (Minecraft
facing) functionality operates using LazyOptional<_>s instead.
Peripheral providers should now return a LazyOptional<IPeripheral> too.
Hopefully this will allow custom peripherals to mark themselves as
invalid (say, because a dependency has changed).
While peripheral providers are somewhat redundant, they still have their
usages. If a peripheral is applied to a large number of blocks (for
instance, all inventories) then using capabilities does incur some
memory overhead.
We also make the following changes based on the above:
- Remove the default implementation for IWiredElement, migrating the
definition to a common "Capabilities" class.
- Remove IPeripheralTile - we'll exclusively use capabilities now.
Absurdly this is the most complex change, as all TEs needed to be
migrated too.
I'm not 100% sure of the correctness of this changes so far - I've
tested it pretty well, but blocks with more complex peripheral logic
(wired/wireless modems and turtles) are still a little messy.
- Remove the "command block" peripheral provider, attaching a
capability instead.
When creating a peripheral or custom Lua object, one must implement two
methods:
- getMethodNames(): String[] - Returns the name of the methods
- callMethod(int, ...): Object[] - Invokes the method using an index in
the above array.
This has a couple of problems:
- It's somewhat unwieldy to use - you need to keep track of array
indices, which leads to ugly code.
- Functions which yield (for instance, those which run on the main
thread) are blocking. This means we need to spawn new threads for
each CC-side yield.
We replace this system with a few changes:
- @LuaFunction annotation: One may annotate a public instance method
with this annotation. This then exposes a peripheral/lua object
method.
Furthermore, this method can accept and return a variety of types,
which often makes functions cleaner (e.g. can return an int rather
than an Object[], and specify and int argument rather than
Object[]).
- MethodResult: Instead of returning an Object[] and having blocking
yields, functions return a MethodResult. This either contains an
immediate return, or an instruction to yield with some continuation
to resume with.
MethodResult is then interpreted by the Lua runtime (i.e. Cobalt),
rather than our weird bodgey hacks before. This means we no longer
spawn new threads when yielding within CC.
- Methods accept IArguments instead of a raw Object array. This has a
few benefits:
- Consistent argument handling - people no longer need to use
ArgumentHelper (as it doesn't exist!), or even be aware of its
existence - you're rather forced into using it.
- More efficient code in some cases. We provide a Cobalt-specific
implementation of IArguments, which avoids the boxing/unboxing when
handling numbers and binary strings.
- cc.pretty.pretty now accepts two additional options:
- function_args: Show function arguments
- function_source: Show where functions are defined.
- Expose the two options as lua.* settings (defaulting function_args to
true, and function_source to false).
These are then used in the Lua REPL.
Closes#361
- Use jacoco for Java-side coverage. Our Java coverage is /terrible
(~10%), as we only really test the core libraries. Still a good thing
to track for regressions though.
- mcfly now tracks Lua side coverage. This works in several stages:
- Replace loadfile to include the whole path
- Add a debug hook which just tracks filename->(lines->count). This
is then submitted to the Java test runner.
- On test completion, we emit a luacov.report.out file.
As the debug hook is inserted by mcfly, this does not include any
computer startup (such as loading apis, or the root of bios.lua),
despite they're executed.
This would be possible to do (for instance, inject a custom header
into bios.lua). However, we're not actually testing any of the
behaviour of startup (aside from "does it not crash"), so I'm not
sure whether to include it or not. Something I'll most likely
re-evaluate.
`local varname = value` results in `varname` being inaccessible in
the next REPL input. This is often unintended and can lead to confusing
behaviour. We produce a warning when this occurs.
This uses the system described in #409, to render monitors in a more
efficient manner.
Each monitor is backed by a texture buffer object (TBO) which contains
a relatively compact encoding of the terminal state. This is then
rendered using a shader, which consumes the TBO and uses it to index
into main font texture.
As we're transmitting significantly less data to the GPU (only 3 bytes
per character), this effectively reduces any update lag to 0. FPS appears
to be up by a small fraction (10-15fps on my machine, to ~110), possibly
as we're now only drawing a single quad (though doing much more work in
the shader).
On my laptop, with its Intel integrated graphics card, I'm able to draw
120 full-sized monitors (with an effective resolution of 3972 x 2330) at
a consistent 60fps. Updates still cause a slight spike, but we always
remain above 30fps - a significant improvement over VBOs, where updates
would go off the chart.
Many thanks to @Lignum and @Lemmmy for devising this scheme, and helping
test and review it! ♥
Unknown blit colours, such as " " will be translated to black for the
background or white for the foreground. This restores the behaviour from
before #412.
- Write to a PacketBuffer instead of generating an NBT tag. This is
then converted to an NBT byte array when we send across the network.
- Pack background/foreground colours into a single byte.
This derives from some work I did back in 2017, and some of the changes
made/planned in #409. However, this patch does not change how terminals
are represented, it simply makes the transfer more compact.
This makes the patch incredibly small (100 lines!), but also limited in
what improvements it can make compared with #409. We send 26626 bytes
for a full-sized monitor. While a 2x improvement over the previous 58558
bytes, there's a lot of room for improvement.
This functions the same as shell.run, but does not tokenise the
arguments. This allows us to pass command line arguments through to
another program without having to re-quote them.
Closes#417
- Remove the parenthesis around the text (so it's now
"Computer ID: 12"), rather than "(Computer ID: 12").
- Show the tooltip if the computer has an ID and no label (as well as
when in advanced mode).
- Lint references to unknown fields of modules, excluding the keys and
colours modules. This caught several silly errors in our stub files,
but nothing else.
- Lint on using unknown globals. This highlighted a couple of really
silly mistakes. Fixes#427.
- Add documentation for fs.attributes, fs.getCapacity and pocket, as
they were not defined before.
Co-authored-by: JackMacWindows <jackmacwindowslinux@gmail.com>
- I'm excluding pocket computers, as they have such a tiny screen I'm
not sure the screen estate is worth it.
Pocket computers /generally/ aren't people's starter machine, so I
think this is fine.
- Prune the motd list, and try to make them a little shorter. I think
this list is more of the interesting ones. We can modify this list in
the future, as we get more feedback.[^1]
- Also fix paint/edit not adding an extension when they should. This
was caused by the settings rewrite, as the explicitly provided
default shadowed the one provided by bios.lua.
[^1]: ~5 months ago I asked for some feedback about enabling motds by
default. I only got something constructive back today >_>.
This replaces the allow/block lists with a series of rules. Each rule
takes the form
[[http.rules]]
host = "127.0.0.0/8"
action = "block"
This is pretty much the same as the previous config style, in that hosts
may be domains, wildcards or in CIDR notation. However, they may also be
mixed, so you could allow a specific IP, and then block all others.
- The store is now split into two sections:
- A list of possible options, with some metadata about them.
- A list of values which have been changed.
- settings.define can be used to register a new option. We have
migrated all existing options over to use it. This can be used to
define a default value, description, and a type the setting must have
(such as `string` or `boolean).
- settings.{set,unset,clear,load,store} operate using this value list.
This means that only values which have been changed are stored to
disk.
Furthermore, clearing/unsetting will reset to the /default/ value,
rather than removing entirely.
- The set program will now display descriptions.
- settings.{load,save} now default to `.settings` if no path is given.
This is a backport of 1.15's terminal rendering code with some further
improvements. This duplicates a fair bit of code, and is much more
efficient.
I expect the work done in #409 will supersede this, but that's unlikely
to make its way into the next release so it's worth getting this in for
now.
- Refactor a lot of common terminal code into
`FixedWithFontRenderer`. This shouldn't change any behaviour, but
makes a lot of our terminal renderers (printed pages, terminals,
monitors) a lot cleaner.
- Terminal rendering is done using a single mode/vertex format. Rather
than drawing an untextured quad for the background colours, we use an
entirely white piece of the terminal font. This allows us to batch
draws together more elegantly.
- Some minor optimisations:
- Skip rendering `"\0"` and `" "` characters. These characters occur
pretty often, especially on blank monitors and, as the font is empty
here, it is safe to skip them.
- Batch together adjacent background cells of the same colour. Again,
most terminals will have large runs of the same colour, so this is a
worthwhile optimisation.
These optimisations do mean that terminal performance is no longer
consistent as "noisy" terminals will have worse performance. This is
annoying, but still worthwhile.
- Switch monitor rendering over to use VBOs.
We also add a config option to switch between rendering backends. By
default we'll choose the best one compatible with your GPU, but there
is a config option to switch between VBOS (reasonable performance) and
display lists (bad).
When benchmarking 30 full-sized monitors rendering a static image, this
improves my FPS[^1] from 7 to 95. This is obviously an extreme case -
monitor updates are still slow, and so more frequently updating screens
will still be less than stellar.
[^1]: My graphics card is an Intel HD Graphics 520. Obviously numbers
will vary.
This would return true for any block with a fluid in it, including
waterlogged blocks. This resulted in much broken behaviour
- Turtles cannot place blocks when waterlogged (fixedd #385)
- Turtles could move into waterlogged blocks (such as fences),
replacing them.
This is relatively unoptimised right now, but should be efficient enough
for most practical applications.
- Add textutils.json_null. This will be serialized into a literal
`null`. When deserializing, and parse_null is true, this will be
returned instead of a nil.
- Add textutils.unserializeJSON (and textutils.unserializeJSON). This
is a standard compliant JSON parser (hopefully).
- Passing in nbt_style to textutils.unserializeJSON will handle
stringified NBT (no quotes around object keys, numeric suffices). We
don't currently support byte/long/int arrays - something to add in
a future commit.
- Remove stub for table.pack/table.unpack.
- Remove Lua 5.3 bitlib stub. We're not on 5.3, there's no
point emulating it.
- Change peripheral.call to correctly adjust the error level. This is a
terrible hack, but I believe the only good option.
It'd be good to remove load as well, but it's a little more complex due
to our injecting of _ENV.
Closes#363
It hasn't been http_enable for yonks - slightly worried I didn't notice
this earlier.
Also don't refer to ComputerCraft.cfg - the name has changed several
times across versions, so let's leave it ambiguous.
- Return EPOCH if a zip entry's creation/modification/access time is
missing.
- If a BasicFileAttributes.*Time method returns null, use 0 as our
time. This shouldn't happen, but is a good sanity check.
Fixes#371
This adds documentation comments to many of CC's Lua APIs, and
a couple of the Java ones, through the use of stubs. We then
export these to HTML using illuaminate [1] and upload them to our
documentation site [2].
Uploads currently occur on pushes to master and any release/tag. The
site is entirely static - there is no way to switch between versions,
etc... but hopefully we can improve this in the future.
[1]: github.com/SquidDev/illuaminate/
[2]: https://tweaked.cc/
This means we are already provided with file attributes, which halfs[^1]
the time taken to scan folders.
[^1]: This dropped the fastest scan time from ~1.3s to ~0.6. It's by no
means a perfect benchmark, but this is still an obvious improvement.
- fs.getCapacity just returns the capacity of the current drive, if
available. This will be nil on rom mounts.
- fs.attributes returns an lfs like table of various file attributes.
Currently, this contains:
- access, modification, created: When this file was last accessed,
modified and created. Time is measured in milliseconds since the
epoch, same as os.epoch("utc") and what is accepted by os.date.
- size: Same as fs.getSize
- isDir: same as fs.isDir
Closes#262
- Move timer handling to IAPIEnvironment, rather than OSAPI. This means
the environment is reset on startup/shutdown, much like normal APIs.
- Websocket.receive now accepts an optional timetout (much like
rednet.receive). This starts a timer, and waits for it to complete.
Closes#201
- contains now performs a case-insensitive comparison. While this is a
little dubious, it's required for systems like Windows, where foo and
Foo are the same folder.
- Impose a depth limit on copyRecursive. If there are any other cases
where we may try to copy a folder into itself, this should prevent
the computer entirely crashing.
See #354
- Remove Lua script to generate recipes/advancements for coloured
disks, turtle upgrades and pocket upgrades. Replacing them with Lua
ones.
- Generate most block drops via the data generator system. Aside from
cables, they all follow one of two templates.
- Remove *Stream methods on IMount/IWritableMount, and make the channel
ones the primary.
- Fix location of AbstractTurtleUpgrade
- Make IComputerAccess.getAvailablePeripheral and .getMainThreadMonitor
mandatory.
- IComputerAccess throws a specialised NotAttachedException
Most of the port is pretty simple. The main problems are regarding
changes to Minecraft's rendering system.
- Remove several rendering tweaks until Forge's compatibility it
brought up-to-date
- Map rendering for pocket computers and printouts
- Item frame rendering for printouts
- Custom block outlines for monitors and cables/wired modems
- Custom breaking progress for cables/wired modems
- Turtle "Dinnerbone" rendering is currently broken, as normals are not
correctly transformed.
- Rewrite FixedWidthFontRenderer to to the buffer in a single sweep.
In order to do this, the term_font now also bundles a "background"
section, which is just a blank region of the screen.
- Render monitors using a VBO instead of a call list. I haven't
compared performance yet, but it manages to render a 6x5 array of
_static_ monitors at almost 60fps, which seems pretty reasonable.
- Add a cc.pretty module, which provides a Wadler style pretty printer [1].
- The cc.pretty.pretty function converts an arbitrary object into a
pretty-printed document. This can then be printed to the screen with
cc.pretty.{write, print} or converted to a string with cc.pretty.render.
- Convert the Lua REPL to use the pretty printer rather than
textutils.serialise.
[1]: http://homepages.inf.ed.ac.uk/wadler/papers/prettier/prettier.pdf
GPS requests are now sent and received on CHANNEL_GPS by default
instead. This means it should not be possible to distinguish
computers (and thus locate them) via their GPS requests.
"exit" now has a custom __tostring method, which prints an explanation
message. This is very similar to how Python achives the same
functionality:
lua> exit
Call exit() to exit
lua> exit()
> Actually leaves the REPL
Unfortunately we can't apply the config changes due to backwards
compatibility. This'll be something we may need to PR into Forge.
CraftTweaker support still needs to be added.
It appears that WB opens containers manually, and thus all of our stubs
network stubs are entirely ignored. Thus the only solution here is to
stub out the whole network handler code.
Thankfully this is simple enough - we do the same for Plethora and 1.14.
Fixes#328
This provides the following methods:
- dan200.computercraft.turtle.removeUpgrade(id: String)
- dan200.computercraft.turtle.removeUpgrade(stack: IItemStack)
- dan200.computercraft.turtle.addTool(id: String, craftItem: IItemStack[, toolItem: IItemStack][, kind: string])
While it's pretty minimal, it should allow for a reasonable amount of
functionality.
Closes#327 and #97.
We now use illuaminate[1]'s linting facilities to check the rom and
bios.lua for a couple of common bugs and other problems.
Right now this doesn't detect any especially important bugs, though it
has caught lots of small things (unused variables, some noisy code). In
the future, the linter will grow in scope and features, which should
allow us to be stricter and catch most issues.
As a fun aside, we started off with ~150 bugs, and illuaminate was able
to fix all but 30 of them, which is pretty neat.
[1]: https://github.com/SquidDev/illuaminate
Before it would remain the same across world reloads, and thus would be
out-of-date after leaving the first world. This architecture technically
allows for running multiple servers at once, though that's not going to
matter that soon.
This fixes them not rendering particles when broken. Particle rendering
is a little janky right now, as it uses the whole texture - we should
probably split up the texture into smaller images. Fixes#315
This should fix several issues (see #304, etc...). I'll try to get round
to PRing this into Forge at some point, though on the other hand this is
/super/ ugly.
This shouldn't matter either way - we don't expose it in the creative
menu, and there's no recipes for it. This should shut up a log message
though. Fixes#305.
This fixes monitor rendering underwater (closes#314). By default,
isSolid returns true if the render layer is SOLID. As we use CUTOUT due
to our funky TE rendering, we need to override this to return true
anyway.
This will cause some side effects, as monitors now blocking light
propagation, but I don't think that's the end of the world.
This is sufficiently useful a class, that it's worthwhile exposing it.
Hopefully we can slowly encourage other mods to migrate to it (well, at
least in 1.14), and so make error messages more consistent.
Also:
- Add Javadoc for all public methods
- Clarify the method names a little (getNumber -> getDouble,
getReal -> getFiniteDouble).
- Make the *Table methods return a Map<?,?> instead of
Map<Object, Object>.
Previously we were just returning the current tile. However, if someone
was holding a reference to this inventory (such as a GUI), then it'd be
outdated and invalid once the turtle had moved.
This caused a couple of issues:
- turtle_inventory events would not be fired when moving items in the
turtle GUI.
- As of 75e2845c01, turtles would no
longer share their inventory state after moving. Thus, removing items
from a GUI using an invalid inventory would move them from an old
tile, duplicating the items.
Fixes#298, fixes#300
Not quite sure when this changed, but I'm fairly sure isMouseOver wasn't
a thing when I wrote this. Or I'm a plonker. Both are possible.
Also fixes mouse dragging not being handled in turtles.
Fixes#299
Closes#293. Doesn't really solve anything there aside from exposing the
number, but sadly there's not really anything obvious I can do on my end
- the command API just doesn't expose anything else.
- Adds cc.completions module, with a couple of helper functions for
working with the more general completion functionality (i.e. that
provided by read).
- Adds cc.shell.completions module, which provides shell-specific
completion functions.
- Add a "program completion builder", which allows you to write stuff
like this:
shell.setCompletionFunction( "rom/programs/redstone.lua",
completion.build(
{ completion.choice, { "probe", "set ", "pulse " } },
completion.side) )
Closes#232
It appears several mods inject their own drops on the LOWEST priority,
meaning that we capture the existing drops, and the other mod will clear
the (now empty) drop list and add its own, resulting in dupe bugs.
While I'd argue it's somewhat dubious doing this on the LOWEST priority,
it's not a battle I'm prepared to fight. For now, we just remove the
block/entity drop handlers, and handle all drop logic when entities are
spawned.
Fixes#288
This is the behaviour on 1.14 already, so it makes sense to backport to
1.12.
Any mod may now insert files into assets/computercraft/lua/rom, and
they'll be automatically added to the default ROM mount. This allows
other mods to easily register new programs or autorun files.
See #242
Lua 5.2+ uses loadfile(filename, mode, env), not loadfile(filename,
env). While this is a minor incompatibility, it'd be nice to be
consistent as much as possible.
We try to handle the incorrect case too, as obviously we don't want to
break existing programs.
This moves expect from the bios into a new craftos.expect module,
removing the internal _G["~expect"] definition. Apparently people were
using this irrespective of the "don't use this" comment, so we need to
find another solution.
While this does introduce some ugliness (having to load the module in
weird ways for programs, duplicating the expect function in memory), it
does allow people to use the function in a supported way, and removes
the global ugliness.
I'm not entirely sure how useful all of these will be yet - still
trying to work out what/when to test things, but hopefully this'll
be a useful datapoint.
This is equally an ugly hack, but means we're at least not constructing
entities with null worlds any more.
Ideally we could always use the turtle entity, but this will require a
bit more of a refactor.
Fixes#265
- Add Forge's "name" field to the loot tables. This doesn't resolve all
our missing loot providers, but it's a start.
- Add back GUIs for pocket computers, printouts, view computer, etc...
Equivalent to `pastebin run`, but allows running arbitrary URLs
instead.
Is this a little questionable? Yes - people shouldn't be downloading
and running code from the internet. But hey, people do that already,
so we might as well make it convenient.
So very little works, but it compiles and runs.
Things to resolve over the next few days:
- Horrible mappings (should largely be resolved by tomorrow).
- Cannot send extra data over containers - we'll have to see what Forge
does here.
- Turtle models are broken
- No block drops yet - this will largely be cherry-picking whatever I
did on Fabric.
- Weird inventory desyncs (items don't show up initially when
interacting with a CC inventory).
- Probably lots of other things.
- Adds a CheckStyle configuration which is pretty similar to CC's
existing one.
- Add the Gradle license plugin.
- Ensure the existing source code is compatible with these additional
checks.
See #239
- Make mcfly's stubbing system a little more fault-tolerant.
- Add a small utility function which redirects print, printError and
write to capture their output, rather than printing to the terminal.
This can then be matched against in order to determine a program's
output.
It's a little flakey - you can't use it multiple times in an it
block, etc... but it's a nice feature.
- Add a small couple of tests to delete as a proof-of-concept.
- Convert existing changelog over to use Markdown. This mostly involves
wrapping code in backticks, and marking things as headers where
appropriate.
- Copy all of CC:T's release notes over to the changelog. This is
somewhat more verbose than Dan's notes, but keeping them in sync
seems reasonable (and allows for automation!).
As 'require' operates relative to the current program's directory,
rather than the current directory, it meant we were trying to load files
from /rom/programs.
This is never a good idea, so we add the current directory to the
package path, allowing you to use require as one'd expect.
- Define an expect(index, actual_value, types...) helper function which
takes an argument index, value and list of permissable types and
ensures the value is of one of those types.
If not, it will produce an error message with the expected and actual
type, as well as the argument number and (if available) the function
name.
- Expose expect in the global scope as _G["~expect"], hopefully making
it clear it is internal.
- Replace most manual type checks with this helper method.
- Write tests to ensure this argument validation works as expected
Also fix a couple of bugs exposed by this refactor and the subsequent
tests:
- Make rednet checks a little more strict - rednet.close(false) is no
longer valid.
- Error when attempting to redirect the terminal to itself
(term.redirect(term)).
This only renders the bounding box on non-screen edges of the monitor,
meaning you have an uninterrupted view of the screen when hovering
hover.
Closes#219
Rendering an item worked in principle, but had several caveats:
- The terminal did not fit well within the item's texture, so we had a
rather large border.
- The "correctness" of this was very tied to Minecraft's item rendering
code. This changed a little in 1.13, causing problems like #208.
Instead we effectively reuse the computer GUI rendering code, though
also handling coloured pocket computers and rendering the modem light.
This fixes#208, and hopefully fixes#212.
Actually, many *globs*. It additionally prints the glob if no files
matched it, since that's clearer.
Also move the ComputerTestDelegate's filesystem to be disk-based. This
is what actual computers use, and the MemoryMount is a little broken.
- os.time, when given a table, will act the same as PUC Lua - returning
the seconds since the epoch. We preserve the previous string/nil
behaviour though - os.epoch("local") is equivalent to PUC's
os.time().
- os.date will now act accept a string and (optional) time, returning
an appropriate table.
Somewhat resolves the madness which was dan200/ComputerCraft#183, and
hopefully (though probably not) makes @Vexatos happy.
This changes the previous behaviour a little, but hopefully is more
sane:
- Only require the socket to be open when first calling receive. This
means if it closes while receving, you won't get an error.
This behaviour is still not perfect - the socket could have closed,
but the event not reached the user yet, but it's better.
- Listen to websocket_close events while receiving, and return null
should it match ours.
See #201
We were using += instead of =, meaning the budget always grew,
rather than growing while there was still space. As a result, computers
were never correctly rate limited.
Further more, if a computer went into a deficit, we would continue to
increase the budget by a negative amount, exponentially decreasing until
overflowing!
Yes, this is a very embarrassing mistake. I'd been aware that rate
limiting wasn't working as expected for a while, I hadn't realised
the problem would be this stupid.
This uses the same behaviour that repeaters and comparators do for
determining their input, meaning that redstone directly connected to the
computer is read (as you would expect).
It's worth noting that this is a shift from previous behaviour.
Therefore, it runs the (small) risk of breaking existing builds.
However, I believe it is more consistent with the expected behaviour.
Originally introduced in 173ea72001, but
the underlying cause has been happening for much longer.
- Fix order of method name parameters. Looks like this has been broken
since 1.11. Woops.
- Catch RuntimeExceptions too, as Forge converts checked into unchecked
ones.
Fixes#179
I promise! The joys of using -SNAPSHOT I guess...
This will now correctly cause orphaned threads to be cleaned up,
reducing the risk of thread saturation.
If mod loading fails, we'll continue to load colour handlers. As
blocks/items have not been registered, then we'll throw an NPE.
See MinecraftForge/MinecraftForge#5682. Somewhat fixes#168.
SquidDev