Squeak-dev May 2017

squeak-dev@lists.squeakfoundation.org

40 participants
111 discussions

Graphing weather data
by tim Rowledge 30 Jun '17

30 Jun '17

Now that I have my Pi based weather station[1] in operation and mounted up on the peak of my garage roof ... … I need to do something with the data it provides. Sadly the code reading the sensors is in Python (for now!) but it sends it as MQTT packets to a mac-mini running the broker. I’m working with Craig on an MQTT client (very early state on http://www.squeaksource.com/@gWFeIvLx-rTHKE2R/Gtrt3pje) to read it but then what? Are there any current projects that will work in squeak5.1/spur that can draw nice graphs? What kind of graphs are good for weather data? Should there be chocolate? I guess I ought to store some amount of collected data so maybe it’s time I got to grips with Magma. I’m interested in all ideas that might lead somewhere fun with this. With those ESP8266 wi-fi enabled sorta-arduino things so cheap and easy to set up there must be a gazillion interesting things that could be monitored and analyzed and graphed and indeed actuated in response; let’s not forget that MQTT works both ways. How about a rain sensor sending a message that its raining (duh) which gets collected and turned into a message to your car to close the sunroof before the velvet seat covers (with painted portrait of Elvis, natch) get wet? Or a strain gauge on your security fence alerting you that something heavier than 20 pounds has climbed over so the remote gatling needs hotting up? tim [1] Pi model A+ with gpio attached interface board for rain bucket, wind speed/direction, outdoor temp/humidity, indoor temp, RTC and USB attached Ethernet interface to POE adaptor. WiFi wasn’t practical due to metal roof. Besides, POE means I can hard-reboot by pulling a plug in my server room :-) -- tim Rowledge; tim(a)rowledge.org; http://www.rowledge.org/tim Strange OpCodes: IIB: Ignore Inquiry and Branch anyway

12 37

Solarized dark theme: text selection colour a little hard to see
by Tony Garnock-Jones 30 Jun '17

30 Jun '17

In my image, I've replaced SolarizedTheme class >> #darkBackgroundHighlights with the following, since with the default I find it very difficult to spot which list item is selected, where the text selection is, etc etc: darkBackgroundHighlights ^ self base02 lighter (The original had just "^ self base02".) Does anyone else suffer similarly with the default? Should something like my change make it into the master image? Tony

7 8

Mouse button tester question
by H. Hirzel 29 Jun '17

29 Jun '17

Hello, http://wiki.squeak.org/squeak/2332 'Mouse button tester' has m := Morph new. m on: #mouseDown send: #value: to:[:evt| evt redButtonPressed ifTrue:[m color: Color red]. evt yellowButtonPressed ifTrue:[m color: Color yellow]. evt blueButtonPressed ifTrue:[m color: Color blue] ]. m openInWorld. red and yellow button work, but the blue button brings up the halo as it is noted on the wiki page (tested with a three button mouse). I wonder if there is a way to make the change of the button to blue possible? --Hannes

4 8

Rotary dial morphs
by tim Rowledge 29 Jun '17

29 Jun '17

I’ve factored out the barometer etc code from my weather station project and added http://www.squeaksource.com/RotaryDialMorphs for more general use. It seems to me there has to be some nice factoring of the style of the dial markings that could be applied, perhaps something like text styling and line styling that we see in drawing programs. Perhaps a student project even? Don’t forget that partial-rotary displays are also useful - think analogue VU meters (there’s an example included for that) and old car-dashboard ammeters etc. tim -- tim Rowledge; tim(a)rowledge.org; http://www.rowledge.org/tim Oxymorons: Good grief

2 2

color names for crayons
by Ralph Boland 29 Jun '17

29 Jun '17

I am using Squeak 5.1. Sorry if this has already been fixed. It's nice to have the additional named colors (the crayon color names) but unfortunately if you do for example: "Color banana name" you don't get #banana but instead get the unnamed color specifics. I wrote the following code to fix this. Feel free to use it or even add it to the next version of squeak under the squeak license. Ralph Boland initializeCrayonNames "Name crayon colors. The following colors were removed as they are duplicates: maraschino (red) lemon (yellow) snow (white) licorice (black) cantaloupe (lightOrange)" "Color initializeCrayonNames" "example: Color blueberry name" | crayonColors | crayonColors := #(aluminum aqua asparagus banana blueberry bubblegum cantaloupe carnation cayenne clover eggplant fern flora grape honeydew ice iron lavender lead licorice lime magnesium maroon mercury midnight mocha moss nickel ocean orchid plum salmon seaFoam silver sky snow spindrift spring steel strawberry tangerine teal tin tungsten turquoise). crayonColors do: [:color | self named: color put: (Color perform: color)].

2 2

How should one implement a custom window color for a custom window?
by Eliot Miranda 29 Jun '17

29 Jun '17

Hi Marcel, Hi All, at one point I had implemented custom window colors for the VM simulator window thusly: InterpreterPrimitives class>>initialize "Initialize the window color. This is probably Squeak-specific, so..." [Preferences setWindowColorFor: #cogVirtualMachineSimulatorWindowColor to: (Color colorFrom: InterpreterPrimitives windowColorSpecification brightColor)] on: Error do: [:ex| Transcript show: 'Failed to set window color preference in ', thisContext method methodReference asString] SimulatorMorphicModel class>>registerWindowColor (Preferences windowColorFor: self name) = Color white ifTrue: [ Preferences setWindowColorFor: self name to: (Color colorFrom: self windowColorSpecification brightColor) ]. This approach is now deprecated, causing the load script to protest. But the deprecation message isn't helpful. The setter setWindowColorFor:to: just says the method is deprecated, not suggesting any alternative. The getter Preferences class>>#windowColorFor: just says 'Use Model >> #customWindowColor', but the method isn't implemented in any class. Where's an example of the right way to do it? Could the deprecation warnings be more informative? thanks _,,,^..^,,,_ best, Eliot

2 1

Updated and finished removal of ExternalForm (and related classes)
by Chris Cunningham 29 Jun '17

29 Jun '17

Hi. The task I assigned myself a couple of weeks ago is complete. ExternalForm is no longer in EToys, and (as much as possible) support for it has been factored out of Form. The classes are now in http://www.squeaksource.com/Balloon3D, in the package GraphicsExternal*. These classes should still work if just filed into a current image with all of the packages applied. Further, the test associated with this class have been removed from FFI and placed into the same package noted above. To finalize this, these two packages should be moved to Trunk after you are satisfied that work as advertized: http://source.squeak.org/inbox/EToys-cbc.299.mcz http://source.squeak.org/inbox/Graphics-cbc.375.mcz and this package to FFI: http://source.squeak.org/inbox/FFI-Tests-cbc.9.mcz Also, could the following packages be purged from the InBox? They are not ancestors of the packages needed (and I didn't realize I have to save these as myself to handle this task myself. Packages: http://source.squeak.org/inbox/EToys-cbc.293.mcz http://source.squeak.org/inbox/EToys-cbc.294.mcz http://source.squeak.org/inbox/EToys-cbc.296.mcz http://source.squeak.org/inbox/GraphicsExternal-cbc.1.mcz http://source.squeak.org/inbox/GraphicsExternal-cbc.2.mcz (the last 2 are now correctly in the Balloon3D repository) (*I looked into packaging these are Graphics-External as Andreas did, but the significant difficulty of not messing up the host windows packages in the Ffenestri sub-package made me toss the towel in and go with the simpler packaging) Thanks, cbc

2 1

Re: [squeak-dev] missed commit mails
by Jakob Reschke 20 Jun '17

20 Jun '17

Hi, The Daily Commit Log mail has not appeared any more since Oct 20. Is this an intentional change or a regression? Best, Jakob

2 2

Interpreter VM on Debian testing/unstable 64 bits?
by Bert Freudenberg 07 Jun '17

07 Jun '17

Does anyone have access to a Debian testing/unstable system? Could you check if Etoys really gets stuck, and maybe figure out why? I suspect it might be because they're running on x86_64. The VM is 4.10.2-2614. - Bert - ---------- Forwarded message ---------- From: Víctor Bettachini <victorb(a)gmx.net> Date: Sun, May 14, 2017 at 9:50 PM Subject: Bug#862576: etoys: Doesn't get beyond Squeak security key generation To: Debian Bug Tracking System <submit(a)bugs.debian.org> Package: etoys Version: 5.0.2408-1 Severity: grave Justification: renders package unusable Dear Maintainer, *** Reporter, please consider answering these questions, where appropriate *** * What led up to the situation? Just running etoys&. * What exactly did you do (or not do) that was effective (or ineffective)? Run etoys& * What was the outcome of this action? It gets stuck at 99% of the Squeak security key generation procedure. * What outcome did you expect instead? The same procedure in a Jessie installation in the same system just took a few seconds and led to eToys main screen. *** End of the template - remove these template lines *** -- System Information: Debian Release: 9.0 APT prefers testing APT policy: (500, 'testing') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 4.9.0-2-amd64 (SMP w/2 CPU cores) Locale: LANG=es_AR.utf8, LC_CTYPE=es_AR.utf8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) Versions of packages etoys depends on: ii squeak-vm 1:4.10.2.2614-4+b1 Versions of packages etoys recommends: ii etoys-doc 5.0.2408-1 etoys suggests no packages. -- no debconf information

4 6

The Trunk: KernelTests-eem.329.mcz
by commits＠source.squeak.org 31 May '17

31 May '17

Eliot Miranda uploaded a new version of KernelTests to project The Trunk: http://source.squeak.org/trunk/KernelTests-eem.329.mcz ==================== Summary ==================== Name: KernelTests-eem.329 Author: eem Time: 31 May 2017, 10:43:19.068997 am UUID: 62f89132-ccc4-4461-8c04-596678fe9686 Ancestors: KernelTests-ul.328 Kernel tests harvested from Terf. =============== Diff against KernelTests-ul.328 =============== Item was added: + ----- Method: NumberParsingTest>>testFloatReadWithImplicitLeadingZero (in category 'tests - Float') ----- + testFloatReadWithImplicitLeadingZero + "Test support for implicit leading zeroes when reading numbers from Strings." + self should: [-0.22 = '-.22' asNumber]. + self should: [0.22 = '.22' asNumber]. + ! Item was added: + ----- Method: SemaphoreTest>>testSemaAfterCriticalWait (in category 'testing') ----- + testSemaAfterCriticalWait "self run: #testSemaAfterCriticalWait" + "This tests whether a semaphore that has just left the wait in Semaphore>>critical: + leaves it with signaling the associated semaphore." + | s p | + s := Semaphore new. + p := [s critical:[]] forkAt: Processor activePriority-1. + "wait until p entered the critical section" + [p suspendingList == s] whileFalse:[(Delay forMilliseconds: 10) wait]. + "Now that p entered it, signal the semaphore. p now 'owns' the semaphore + but since we are running at higher priority than p it will not get to do + anything." + s signal. + p terminate. + self assert: ((s instVarNamed: #excessSignals) = 1)! Item was added: + ----- Method: SemaphoreTest>>testSemaInCriticalWait (in category 'testing') ----- + testSemaInCriticalWait "self run: #testSemaInCriticalWait" + "This tests whether a semaphore that has entered the wait in Semaphore>>critical: + leaves it without signaling the associated semaphore." + | s p | + s := Semaphore new. + p := [s critical:[]] fork. + Processor yield. + self assert:(p suspendingList == s). + p terminate. + self assert:((s instVarNamed: #excessSignals) = 0)!

1 0