Indie game storeFree gamesFun gamesHorror games
Game developmentAssetsComics
SalesBundles
Jobs
Tags

bug report's Sticky

A topic by chess created Mar 17, 2022 Views: 338 Replies: 19
Viewing posts 1 to 14
Developer

hey! post bug's here please. i lose track of them too easily if they get posted in the discord and then i Forget Forever. thanks!

(1 edit)

coming in hot with a bug fresh off the presses! if you type like "test test" and then a newline at the end and click play it only plays one "test".
i think its probably when the input is split into tokens, the newline character is not trimmed and becomes part of the last token and obviously "test\n" is not a valid word
hope this helps!


chess edit: right on the money! cleaned \n out of entry strings before passing em to vox. will prolly have to do an input sanitization pass at some point but for now this'll do. thanks!

i still see something similar. the following (with a newline at the end) seems to skip the last note
!tc vox ^s n17 +* +3-*

i just hear the first two notes. on v1.2.0.2

but i dont think it'll be a common issue for live voxes, since i dont think you can enter a newline on a twitch message

Developer

oh weird... so, i cant replicate this by simply pressing enter, but i CAN replicate it by hitting shift+enter...??? i wonder if shift+enter is doing a \r\n or something, since i Only clipped \n out. i'll give it a look tomorrow, thanks for the headsup

(2 edits)

it looks like normal doubling isn't working if the '*' (repeat) character is used as the first note in the pair.

^s n17 ^d *+3 *+5

just plays the first note, but this works:

^s n17 ^d n17+3 *+5

on 3/17/22 __ ver 1.2.0.1


chess edit: fixed! ditto was foolishly trying to ditto commands. i didn't have to account for this with the previous ditto functionality and forgot about it. oops!

(1 edit)

1:
^s n19 * ^l=16 * ^l=4 * ^l=8 +* +0* +* -* +* -* ^l=1 +0*
only plays the first two notes

2:

-10buzwarn +* +* +*

plays a low pitched noise, a slightly higher pitched low pitch noise, and then two expected pitch noises

on 3/17/22 __ ver 1.2.0.1


chess edit: fixed n fixed! bug 2 was due to, if the previous clip had a +3 and the ditto had a -1, they would add together (to create a +2) rather than overwrite. another quirk of the way i changed ditto handling...

v1.2.0.2

this seems like a related ditto bug:

!tc vox ^s ^d n17 ^d +3-* +5* ^d * ^d +3-* +5*

this is using double chaining to play 3 notes (c major chord). note how the second chain is missing the base c note when played. in this case im using * for notes except the first one, since the note/sfx never changed. but if you change this to

!tc vox ^s ^d n17 ^d +3-* +5* ^d n17 ^d +3-* +5*

it plays correctly. in this case the first note of the second chain was replaced with n17 instead of a ditto

Developer

so, that's not actually a bug, but a syntax error on your part:

dittos copy both the clip and pitch of the previous clip. if you want to return to a cnote, you can do +0 or -0 on that ditto to return to neutral. this is actually detailed on the vox page, which surprised me bc it feels like something i'd forget to write down anywhere lol

oh! thanks for the explainer. and 'rtfm' for me hehheh

ah ok i think i have a better test case to trigger what i was talking about:

^s ^d -n17 +4* ^d -* +4*

this plays two doubles of a low b (-1) and fsharp (+4). the second double is missing the b when ditto is used. it seems like going negative is tricking something. using n17 on the second double works.

if you switch the notes, the b isn't played at all!

^s ^d +4n17 -*

but if you do:

^s ^d +4n17 -* -*

(ie. playing the low b individually after the double) even that is wrong. it sounds like a +4-*

Developer

LOL i looked at what was going on and it was bc i... must have hit undo too many times when fixing the -/+ compounding handler for ditto, so it was just checking for pluses on the ditto Twice rather than checking for plus And minus. i tested it and it sounds right (??? i think) to me now, i'll upload the fix

(8 edits)

v1.2.0.3

possible vox leaks. play:

^s n1 ^d n14

notice the intentional syntax error (a double with no second). and you only hear the n1 which is probably correct (ie. not playing invalid syntax)

then play

^s ^n2

and notice the missing morshunote is found!

im not sure if this will be a problem for live rook voxes though

On build v1.2.0.3, inputting:

^s ^bpm=150 hl_hev>.28 +0* *

...will output audio as if though a plus is on the last ditto, or: 

^s ^bpm=150 hl_hev>.28 +0* +*

Note: the same behaviors happens even if the +0* is -0*, but with any number other than 0 it works as expected.

bug with the previewer textbox.  ctrl + c and ctrl + v work as expected, but ctrl + x adds nothing to the clipboard while still erasing the text.

ver 1.2.0.3

Play Highlighted doesn't work as intended after ^d is used in grant mode. Simple example;

^g ^d yeah huh here * *

Full preview plays correctly, Play Highlighted on "here * *" uses normal vox sounds instead

As of 1.2.0.3, an unmodified ditto does not properly copy the previous pitch if + and - were used in combination; it effectively ignores the - portion unless you only used -.

^s +-n1 * . +2-* * . -5* *

The above results in the notes playing at 105%, 110% (when it should be 105% again), 115%, 120% (when it should be 115% again), 75%, 75% (correct).

Also, less important, but "+2-3n1" (combining + and - with a number after the -) doesn't work, though I'm not sure whether it ever did. "+2---n1" does work, however.

This is extremely niche and not a big deal but I thought it was kinda funny: If a lot of bass notes are doubled up with other notes in quick  succession, the bass notes start fading out:

^s ^l=16 ^d n10 kk14 ^d n10 kk14 ^d n10 kk14 ^d n10 kk14 ^d n10 kk14 ^d n10 kk14 ^d n10 kk14 ^d n10 kk14 ^d n10 kk14

As of 1.2.0.3, the Play Highlighted code that checks preceding state changes seems to not proess persona switching properly if you use multiple persona codes, as opposed to just using one as a toggle.

^g a. ^m a. ^g a.

If you enter this and play just the last "a" (without highlighting the preceding ^g), it plays as Morshu, presumably because it sees an even number of ^g's and figures that means they canceled out. (In fact, if you add extra ^g's to the start such that there's an odd number overall, highlight-playing the last "a" will properly use Grant.) This seems to happen with any persona codes arranged in this way.

On 1.2.0.3:

^s ^d dingdingding one ^d dingdingding two three

Seems to skip the second dingdingding.  However, the following works as expected:

^s ^d dingding one ^d dingding two three

1.2.0.3, again a very niche one: Notes are skipped if you enter song mode with `^song` instead of `^s`. Timing changes are still applied, though:

^s ^l=4 animal n1 n2 n3 ^s o k.
^song ^l=4 animal n1 n2 n3 ^s o k.