Squeak-dev March 2024

squeak-dev@lists.squeakfoundation.org

27 participants
58 discussions

Discuss & Review | Faster ProcessExplorer through AllProcesses list
by Taeumel, Marcel 18 May '24

18 May '24

...instead of "Process allSubInstances". The caveat is this: process creation gets a little bit slower while accessing the list of runnable *and* waiting processes gets about 40x faster: [ [] newProcess terminate ] bench. AFTER: '140,000 per second. 7.13 microseconds per run. 35.02 % GC time.' BEFORE: '224,000 per second. 4.47 microseconds per run. 4.38 % GC time.' [ aProcessBrowser updateProcessList ] bench. -- allow quick updates AFTER: '4,410 per second. 227 microseconds per run. 77.34453 % GC time.' BEFORE: '118 per second. 8.47 milliseconds per run. 2.59792 % GC time.' Why do we need this? If you have an auto-updating ProcessBrowser running in the background while doing time-sensitive benchmarks, you want a fast ProcessBrowser :-) Please see the attached change set. Thanks to Leon and Eliot for the hint! Best, Marcel

7 10

The Trunk: Kernel-cmm.1563.mcz
by commits＠source.squeak.org 06 May '24

06 May '24

Chris Muller uploaded a new version of Kernel to project The Trunk: http://source.squeak.org/trunk/Kernel-cmm.1563.mcz ==================== Summary ==================== Name: Kernel-cmm.1563 Author: cmm Time: 25 March 2024, 7:38:47.237222 pm UUID: 6986b006-23f1-4c46-b47a-7417539e6902 Ancestors: Kernel-cmm.1562 - Preserve "Map CONTROL keys to COMMAND keys" preference across image restarts. - With Tim's permission, removed LargePositiveInteger>>#asByteArray. =============== Diff against Kernel-cmm.1562 =============== Item was changed: ----- Method: EventSensor class>>startUp: (in category 'system startup') ----- startUp: resuming + resuming ifTrue: + [ Smalltalk platformName ~= Smalltalk previousPlatformName + ifTrue: + [Smalltalk platformName = 'Mac OS' + ifTrue: [ + self mapAltKeysToOptionKeys: false. + self mapControlKeysToCommandKeys: false] + ifFalse: [ + self mapAltKeysToOptionKeys: true. + self mapControlKeysToCommandKeys: true]. + self installVirtualKeyTable] ]. - resuming ifTrue: [ - Smalltalk platformName = 'Mac OS' - ifTrue: [ - self mapAltKeysToOptionKeys: false. - self mapControlKeysToCommandKeys: false] - ifFalse: [ - self mapAltKeysToOptionKeys: true. - self mapControlKeysToCommandKeys: true]. - self installVirtualKeyTable ]. + self default startUp! - self default startUp.! Item was removed: - ----- Method: LargePositiveInteger>>asByteArray (in category 'converting') ----- - asByteArray - - ^self as: ByteArray!

5 4

trunk debugger unusably slow...
by Eliot Miranda 20 Apr '24

20 Apr '24

Hi All, I'm trying to debug an inlining edge case in the VM for the threaded VM that Leon is doing. The debugger is unusably slow (steps taking 10 seconds, etc). I've raised this issue before but was brushed off. I suspect somewhere in the new (last two years?) highlighting code is taking a lot of time. Would the author(s) please provide me with some snippets of code i can use to try and measure performance? And/or would anybody be willing to pair (I'm on PST) in trying to get to the bottom of this? It's extremely frustrating to try and work when the debugger is unusable. _,,,^..^,,,_ best, Eliot

5 9

Questions Loading Maui from SqueakMap
by Edwin Ancaer 03 Apr '24

03 Apr '24

Hello, I wanted to load Maui from SqueakMap but as I'm running a 6.0 image, there is no corresponding Maui verson in squeakmap. I first tried to load version 1.53, but that is giving following error: [image: mabase1.png] Then I tried the Head version, but that proposes met to merge code. Not being a specialist, I don't want to rush into this. Is it safe for the other applications in the image to start changing existing code? Which of the proposed options can I take best?.... [image: merging.png] Is there any chance for a Maui version for the 6.0 image, and if not, what is the best option for the proposed merges? Please don't see this as a complaint. Free time is scarce for everybody, so "We just don't have the time to keep up with all this" is a valid answer for me. Kind regards, Edwin Ancaer

4 7

The Trunk: Kernel-cmm.1562.mcz
by commits＠source.squeak.org 02 Apr '24

02 Apr '24

Chris Muller uploaded a new version of Kernel to project The Trunk: http://source.squeak.org/trunk/Kernel-cmm.1562.mcz ==================== Summary ==================== Name: Kernel-cmm.1562 Author: cmm Time: 24 March 2024, 6:27:34.572036 pm UUID: 886bf6b9-a56b-4131-accf-3867a3836658 Ancestors: Kernel-ct.1561 Allow processPreemptionYields: to persist across image saves, as purported by its comment. =============== Diff against Kernel-ct.1561 =============== Item was changed: ----- Method: ProcessorScheduler class>>startUp: (in category 'background process') ----- startUp: resuming - Smalltalk processPreemptionYields ifTrue: [ - Smalltalk processPreemptionYields: false]. - resuming ifTrue: [ Smalltalk installLowSpaceWatcher. self installIdleProcess].!

10 36

drawing on morphs, damage recording, and canvas origin
by Lauren Pullen 30 Mar '24

30 Mar '24

Hi List, Easy question first. #refreshWorld warns not to use it often because of damage recording... but it doesn't say what to use instead. What is the correct method to mark a morph damaged? (I'm rendering a static 3D scene with a movable camera, so I only need to redraw it on demand by the camera.) Whenever I draw on a morph I always forget I need to offset coordinates by the morph's origin, so things end up getting drawn in the top left. I tend to open morphs in hand, so it draws correctly until I drop it back into the world. Why does the canvas in #drawOn: have a different origin depending on if you're holding the morph or not?

2 2

The Trunk: System-mt.1452.mcz
by commits＠source.squeak.org 28 Mar '24

28 Mar '24

Marcel Taeumel uploaded a new version of System to project The Trunk: http://source.squeak.org/trunk/System-mt.1452.mcz ==================== Summary ==================== Name: System-mt.1452 Author: mt Time: 28 March 2024, 9:59:08.460064 am UUID: c4dd53f6-cf69-7b42-b554-f4e6320216cf Ancestors: System-cmm.1451 When using high-DPI screens, make enough room for the emergency evaluator. =============== Diff against System-cmm.1451 =============== Item was changed: ----- Method: Transcripter class>>emergencyEvaluator (in category 'utilities') ----- emergencyEvaluator + (Transcripter newInFrame: (0@0 corner: 320 px @ 200 px)) - (Transcripter newInFrame: (0@0 corner: 320@200)) show: 'Type ''revert'' to revert your last method change. Type ''exit'' to exit the emergency evaluator.'; readEvalPrint!

1 0

Re: [Vm-dev] Re: MIDI on Mac VM
by Vanessa Freudenberg 26 Mar '24

26 Mar '24

Of course I couldn't let this sit there. As of just now, SqueakJS supports actual MIDI 🥳 At least in Chrome: https://squeak.js.org/ You can run an image, drop in a MIDI file, and play it. I don't have a hardware synth so I tested MIDI output via GarageBand. The Steinway Grand sounds great playing the Moonlight sonata. Here's a video with sound: https://freudenbergs.de/vanessa/squeakjs/SqueakJS-MIDI.mp4 Should work in pretty much any Squeak version (I tried in 6.2, too). I tested MIDI input using the MidiKeys app since I don't have a real keyboard. It's implemented and the data coming into the image looks right, but the MidiSynth class's playback does not sound quite right. Also, it's not entirely clear what epoch/units the midi clock is supposed to be in – possibly because Squeak changed millisecondClockValue to not start at 0 anymore, whereas older images (or at least the MIDI code in it) expects that. If anyone wants to hack it, my WebMIDI plugin code is in: https://github.com/codefrau/SqueakJS/blob/main/plugins/MIDIPlugin.js Have fun 💜 Vanessa On Thu, Mar 21, 2024 at 9:44 AM Tobias Pape via Vm-dev < vm-dev(a)lists.squeakfoundation.org> wrote: > > Hi > > > On 21. Mar 2024, at 17:37, Vanessa Freudenberg <vanessa(a)codefrau.net> > wrote: > > > > A friend is trying to get this to work – is it supposed to work? > > > > > "SimpleMIDIPort primPortCount" evaluates to 0, even when I have a MIDI > controller plugged in. > > And "SimpleMIDIPort openDefault" gives me a primitive failed error. > > It could™ work, yet, I noticed that the API used is outdated since 10 > years, so there's that. > So, it might depend on the macOS used. > I wanted to fix that since forever > > Best regards > -Tobias

5 4

The Trunk: ReleaseBuilder-mt.247.mcz
by commits＠source.squeak.org 26 Mar '24

26 Mar '24

Marcel Taeumel uploaded a new version of ReleaseBuilder to project The Trunk: http://source.squeak.org/trunk/ReleaseBuilder-mt.247.mcz ==================== Summary ==================== Name: ReleaseBuilder-mt.247 Author: mt Time: 26 March 2024, 10:15:40.484051 am UUID: 6b05e554-d6ae-1046-adab-318f230b240a Ancestors: ReleaseBuilder-mt.246 Fix comment. Both mt.247 and mt.246 complement Kernel-cmm.1562 =============== Diff against ReleaseBuilder-mt.246 =============== Item was changed: ----- Method: ReleaseBuilder class>>prepareProcesses (in category 'preparing') ----- prepareProcesses "Re-start running processes if necessary. Call this at late as possible in the release process. Only required to processes that are not reset at image #startUp." Delay startTimerEventLoop. + Smalltalk processPreemptionYields: false. "avoid accidental round-robin; make users' Morphic-related workspace scripts/code more robust if forked at UI priority"! - Smalltalk processPreemptionYields: false. "avoid accidental round-robin; make users' Morphic-related workspace scripts/code more robust if forked at lower priority"!

1 0

The Trunk: ReleaseBuilder-mt.246.mcz
by commits＠source.squeak.org 26 Mar '24

26 Mar '24

Marcel Taeumel uploaded a new version of ReleaseBuilder to project The Trunk: http://source.squeak.org/trunk/ReleaseBuilder-mt.246.mcz ==================== Summary ==================== Name: ReleaseBuilder-mt.246 Author: mt Time: 26 March 2024, 10:11:12.181433 am UUID: 0b3b4f34-11f3-f34f-a0a4-72ec011dbfe0 Ancestors: ReleaseBuilder-dtl.245 For new images, by default, ensure inadvertent yield outside a process' own control. =============== Diff against ReleaseBuilder-dtl.245 =============== Item was changed: ----- Method: ReleaseBuilder class>>prepareProcesses (in category 'preparing') ----- prepareProcesses "Re-start running processes if necessary. Call this at late as possible in the release process. Only required to processes that are not reset at image #startUp." + Delay startTimerEventLoop. + Smalltalk processPreemptionYields: false. "avoid accidental round-robin; make users' Morphic-related workspace scripts/code more robust if forked at lower priority"! - Delay startTimerEventLoop.!

1 0