Planet Squeak Code

Simulaiton:
Symbolic names for addresses on the rump C stack.
More asserts tracking c/processor stack pointers in the MT VM.

by Eliot Miranda (eliot.miranda@gmail.com) at November 25, 2020 07:25 AM

An alloca version of stackStringValue: (stackEphemeralStringValue:) could be generally useful.

by Eliot Miranda (eliot.miranda@gmail.com) at November 25, 2020 03:08 AM

Avoid really ugly generated code for InterpreterPlugin>>sizeField:.

by Eliot Miranda (eliot.miranda@gmail.com) at November 24, 2020 02:15 AM

- fixed a typo: columnsDescriptions -> columnDescriptions

by Levente Uzonyi (leves@elte.hu) at November 23, 2020 03:54 PM

- added tests for uuid and uuid array parsing

by Levente Uzonyi (leves@elte.hu) at November 23, 2020 03:51 PM

- terminate released connections in #isInFailedTransaction state

by Levente Uzonyi (leves@elte.hu) at November 23, 2020 03:51 PM

- map Postgres's uuid to Squeak's UUID
- support reading uuid arrays as well

by Levente Uzonyi (leves@elte.hu) at November 23, 2020 03:50 PM

Fix the InstvarInterfaceExtractor hiwerarchy for full blocks. Avoid a crash in Squeak since Squeak raises an exception if one attempts "foo removeLast: 0". Reduce some unnecessary duplication in reloadStack. Improve performance a little: Nuke input and saveStacks accessors (lazy initialization isn't necessary). use #== for comparing type tag symbols.

by Eliot Miranda (eliot.miranda@gmail.com) at November 23, 2020 03:12 PM

Fix error when no data for selected country.

by Thiago da Silva Lino (thiagolino@yahoo.com.br) at November 23, 2020 12:19 PM

MTVM:
Separate from MultiProcessor. Have CogThreadManager>>releaseVM & tryLockVMOwnerTo: simulate thread switches, since these are effectively the only places they occur in this simple threading scheme. The full complexity of MultiProcessor isn't needed.

To make this work change all senders of tryLockVMOwnerTo: to invoke CogThreadManager's simulation-only wrapper around Cogit>>tryLockVMOwnerTo:.

Simplify and correct checks and initialization of the processor's thread-specific stack pointers.

Provide a multi-threaded version of primtRumpCStack that prints the individual stacks for active threads.

by Eliot Miranda (eliot.miranda@gmail.com) at November 23, 2020 05:46 AM

Separating from MultiProcessor.

by Eliot Miranda (eliot.miranda@gmail.com) at November 23, 2020 05:40 AM

Extend image version compatibility.
For Squeak 6 alpha (current trunk):
Remove use of deprecated hideVScrollBarIndefinitely: with possible cosmetic impact on older image versions but otherwise harmless to remove.
Provide a compatibility method to handle the renaming of includesSubString: to includesSubstring:
Handle the removal of CrLfFileStream.
For circa Squeak 3.8:
Use ifNotNilDo: instead of ifNotNil: when passing a block argument.
Use ifNil:ifNotNilDo: instead of ifNil:ifNotNil: when passing a block argument.
Note, early images require Pragmatizer to undo pragma usage

by David T. Lewis (lewis@mail.msen.com) at November 22, 2020 11:34 PM

Makes AnimatedImageMorph delays publicly accessible. In contrary to e.g. nextTime, delays are an essential part of an animated image, for instance when converting it into a different representation from an external client.

by Christoph Thiede (Christoph.Thiede@student.hpi.uni-potsdam.de) at November 22, 2020 04:55 PM

And splObj: is needed for the last case in cStringOrNullFor:.

by Eliot Miranda (eliot.miranda@gmail.com) at November 19, 2020 11:42 PM

More effort is required to delete the unnecessary read before Spur is:instanceOf:compactClassIndex:.

by Eliot Miranda (eliot.miranda@gmail.com) at November 19, 2020 11:27 PM

But we can do a little better than VMMaker.oscog-eem.2888. Better comment argAssignmentsFor:send:except:in: and make sure that side-effect-less longAt: is recognised as such, hcen inlining of SpurMemoryManager>>#isClassOfNonImm:equalTo:compactClassIndex: is much improved (less noise).

by Eliot Miranda (eliot.miranda@gmail.com) at November 19, 2020 10:32 PM

Fix a Slang regression in VMMaker.oscog-eem.2884. The elision of unused argument assignments also elided argument expressions that had side-effects. The fix is to retain epxressions that have side effects but generate assignments of those expressions to inlined variables only for variables that are actually used in the expansion.

by Eliot Miranda (eliot.miranda@gmail.com) at November 19, 2020 10:06 PM

Re-enables handling of #launchDrop events when the VM is configured as a singleton and has been invoked again, though currently only handled with a rudimentary fallback implementation.

See http://forum.world.st/Changeset-Enhanced-integration-of-drag-n-drop-from-host-tp5123857p5124332.html. Thanks to Jakob (jr) for the feedback!

by Christoph Thiede (Christoph.Thiede@student.hpi.uni-potsdam.de) at November 17, 2020 07:26 PM

Allow per-instance control over TextDiffBuilder showInsertBeforeRemove.

by Tony Garnock-Jones (tonyg@leastfixedpoint.com) at November 10, 2020 08:57 AM

Fixes recent regression of the COPY button in the version inspector.

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 10, 2020 08:21 AM

Makes the order of inserted/removed lines in text diffs configurable. Set the default to the unix-y way, which is remove-before-insert.

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 10, 2020 07:57 AM

COGMTVM:
Nuke the old V3 GC lock hack.
Restore disownCount cuz aio.c uses it (to reduce noise when using the repl image).
Use DisownVMForThreading as the disown flag in primitiveFileReadPinningAndDisowning rather than DisownVMForFFICall because this flag is used for the fail-on-ffi-exceptin machinery.
Get the Vm to compile, making sure funcitons used in asserts are not always inlined (& hence inlined away).

by Eliot Miranda (eliot.miranda@gmail.com) at November 10, 2020 04:03 AM

COGMTVM: The Slang change to super sends won't work. Use a clumsier approach.

by Eliot Miranda (eliot.miranda@gmail.com) at November 10, 2020 12:54 AM

COGMTVM:
Change the lock to Cogit>>tryLockVMOwnerTo: from tryLockVMOwner: & tryLockVMToIndex:.
Clean up various asserts; in partcular match simulator stack depth checks in disownVM: and ownVM:.

Have print[Stack]CallStackOf: handle the active process (suspendedContext isNil).

Slang: give flexibility to the simulator by only performing a translation-time super expension if the super selector is the same as the current method's.

by Eliot Miranda (eliot.miranda@gmail.com) at November 10, 2020 12:38 AM

Hack in a short cut to the CogBytecodeMethodInspector in machine code frame inspectors if the shift key is pressed (we're missing a CogMethodInspector for now).

by Eliot Miranda (eliot.miranda@gmail.com) at November 10, 2020 12:25 AM

Unlink on Supervisor deregistration.

by Tony Garnock-Jones (tonyg@leastfixedpoint.com) at November 09, 2020 08:33 PM

Improve Supervisor: it now can serve as a registry/directory/nameservice.

by Tony Garnock-Jones (tonyg@leastfixedpoint.com) at November 09, 2020 11:10 AM

Adds documentation for active globals. Thanks to Dave (dtl) for taking care!

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 09, 2020 07:52 AM

Use the lastDeferredUIMessage mechanism for avoiding updating the debugger UI too frequently. Alas this only solves the run to here case; it doesn't fix the rapid clicking issue. Hence this is in the inbox. Christoph, this may be useful to you.

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 09, 2020 07:31 AM

Fix an inadvertet truncation of flag field printing, and add the description of the file-times-in-utc flag bit in the image header flags.

by Eliot Miranda (eliot.miranda@gmail.com) at November 09, 2020 07:13 AM

Ugh, the 2 bit shift in the image flags will be the end of me. Get the file times in utc flag set code correct. Bah...

by Eliot Miranda (eliot.miranda@gmail.com) at November 09, 2020 06:55 AM

Add an image header flag that selects UTC times form the FilePlugin and FileAttributesPlugin, rather than local tme. e.g. to set (until we have a proper accessor) use

Smalltalk vmParameterAt: 48 put: ((Smalltalk vmParameterAt: 48) bitOr: 128)

by Eliot Miranda (eliot.miranda@gmail.com) at November 09, 2020 06:25 AM

Make sure initialiseModule invokes faInitialiseModule so faInitialiseModule can do its thing.

by Eliot Miranda (eliot.miranda@gmail.com) at November 09, 2020 06:03 AM

Name: System-mt.1191
Author: mt
Time: 5 November 2020, 4:34:33.039798 pm
UUID: 0d08f4e4-d4fb-db49-ad96-1f37768ba637
Ancestors: System-mt.1190

Adds KR as "Korakurider" to the list of contributors.

by Dave Lewis (lewis@mail.msen.com) at November 07, 2020 03:42 PM

Name: System-mt.1190
Author: mt
Time: 5 November 2020, 4:23:31.287798 pm
UUID: 832d3d0e-dc06-a14f-ac49-d7dc709680ea
Ancestors: System-mt.1189

Speeds up String >> #translated and locale switching as follows:
- Only call #localeChanged(Gently) on classes that implement it, just like #cleanUp:
- For the InternalTranslator, make translation-in-all-domains a no-op; 122µs -> 1.3µs
- For the GetTextTranslator, make translation-in-all-domains fetch all available .mo files once, then directly enumerate that cache until a result is found; roughly 50µs -> 8µs
- Avoid loading all .mo files on #localeChanged(Gently); only do that for translation-in-all-domains

by Dave Lewis (lewis@mail.msen.com) at November 07, 2020 03:41 PM

Name: Kernel-eem.1361
Author: eem
Time: 2 November 2020, 10:40:38.260046 am
UUID: 7c6efb93-d626-4a62-86f4-c258f8cf199d
Ancestors: Kernel-eem.1360

I couldn't resist...

by Dave Lewis (lewis@mail.msen.com) at November 07, 2020 03:14 PM

Name: Kernel-eem.1360
Author: eem
Time: 2 November 2020, 10:10:50.874417 am
UUID: 2041d4e1-746a-49f9-9098-8fbd6ba2f787
Ancestors: Kernel-eem.1359, Kernel-ct.1359

Merge Kernel-ct.1359 from inbox
Author: ct
Time: 1 November 2020, 7:50:24.183428 pm
UUID: 3b12b316-79f3-de41-8765-8296d964b821
Ancestors: Kernel-mt.1353

Revise and extend Context #runSimulated: implementation. Remove restriction to blocks that do not have a method return. Add support for exception signaling during the execution, which caused unterminated simulation of the calling process in the past. Support argless contextAtEachStep blocks.

Benchmarks:
code:
[Context runSimulated: [100@100 corner: 200@200]] bench
before:
16.7 ms/run
after:
19.8 ms/run
I think this should be okay, given the fact that the primary purpose of simulation is providing explorability but not efficiency ...

eem: Fix a bug in the use of runSimulated:contextAtEachStep:. Presumably because of the pre-closure block model existing senders were of the form
thisContext sender
runSimulated: aBlock
contextAtEachStep: [....]
but this caused the jump at the end of runSimulated:contextAtEachStep: to short-cut any processing done after runSimulated:contextAtEachStep:. So rewrite all users of runSimulated:contextAtEachStep: in this form:
thisContext
runSimulated: aBlock
contextAtEachStep: [....]
and consequently have tallyInstruction: rallyMethods: et al answer their intended results.

by Dave Lewis (lewis@mail.msen.com) at November 07, 2020 03:13 PM

Name: Kernel-eem.1359
Author: eem
Time: 30 October 2020, 12:10:07.699375 pm
UUID: 1ec9a445-37fa-4256-84d8-9bdf8b52eb7d
Ancestors: Kernel-eem.1358

Merge Kernel-ct.1358
Author: ct
Time: 28 October 2020, 11:37:24.817544 pm
UUID: c9af6acb-61cf-af42-b5fa-3afa8451bb24
Ancestors: Kernel-mt.1353

Fixes a simulation bug that occurs when executing ProtoObject >> #doesNotUnderstand:. See KernelTests-ct.388.

by Dave Lewis (lewis@mail.msen.com) at November 07, 2020 03:09 PM

Name: Kernel-eem.1358
Author: eem
Time: 30 October 2020, 12:04:50.705166 pm
UUID: 9e6daea9-3c8f-4ffa-99ab-e76c50af41d0
Ancestors: Kernel-eem.1356, Kernel-ct.1357

Merge Kernel-ct.1357
Author: ct
Time: 28 October 2020, 9:30:35.577652 pm
UUID: efd94ae0-26ab-0c4f-b854-52f414ff2d3e
Ancestors: Kernel-eem.1354

Implement missing simulation of objects as methods.

In the past, it was not possible to debug/simulate code that used objects as methods properly. (Thanks to Marcel for the hint!) This very simple commit adds support of the OaM protocol [1] to the simulation machinery. Now you can debug all tests in TestObjectsAsMethods as you would expect, instead of crashing your image!

Update: Uploaded a third time, this time with Kent Beck block formatting preserved.

[1] "The [Objects as Methods] contract is that, when the VM encounters an ordinary object (rather than a compiled method) in the method dictionary during lookup, it sends it the special selector #run:with:in: providing the original selector, arguments, and receiver.". DOI: 10.1145/2991041.2991062.

by Dave Lewis (lewis@mail.msen.com) at November 07, 2020 03:07 PM

Name: Kernel-eem.1356
Author: eem
Time: 30 October 2020, 7:55:52.516141 am
UUID: 589e93d6-d597-46f2-a09c-9d646902f1e1
Ancestors: Kernel-eem.1355

selectorJustSentOrSelf (used in process termination) must be able to handle being at the start of a method.

by Dave Lewis (lewis@mail.msen.com) at November 07, 2020 03:01 PM

Name: Kernel-eem.1355
Author: eem
Time: 28 October 2020, 6:16:21.972202 pm
UUID: 14e40036-578b-4645-88e4-77be34a6d96b
Ancestors: Kernel-eem.1354

Speed-up cull:[cull:*] by supplying the primitive to handle teh case where the number of block arguments matches the number of arguments to cull:[cull:].
Nuke the obsolete pre-closure simulation code for block closure activation.

by Dave Lewis (lewis@mail.msen.com) at November 07, 2020 02:59 PM

Name: Kernel-eem.1354
Author: eem
Time: 27 October 2020, 7:12:42.644138 pm
UUID: 0e8cc5f2-05b0-483a-9d7a-b06d376c1d00
Ancestors: Kernel-mt.1353

Have Cotext>>swapReceiver: do what it says it will (c.f. swapSender:). Change the one sender not to reply on it; Context>>restartWithNewReceiver: can simply assign to receiver directly with the new Context design.

by Dave Lewis (lewis@mail.msen.com) at November 07, 2020 01:58 PM

Name: Kernel-mt.1353
Author: mt
Time: 15 October 2020, 2:29:01.197368 pm
UUID: e4947d59-b541-7441-9662-67bb4052fbdb
Ancestors: Kernel-mt.1352

Complements System-mt.1180. See http://forum.world.st/Proposal-Resetting-pragma-preferences-via-nil-or-reset-message-td5123490.html

by Dave Lewis (lewis@mail.msen.com) at November 07, 2020 01:57 PM

Name: Compiler-eem.451
Author: eem
Time: 30 October 2020, 12:01:09.305583 pm
UUID: 30f155c7-a9c4-4cb8-bac6-f5089bbf7b36
Ancestors: Compiler-ct.450

just recategorizations.

by Dave Lewis (lewis@mail.msen.com) at November 07, 2020 01:54 PM

Name: Compiler-ct.450
Author: ct
Time: 24 October 2020, 12:16:36.359256 am
UUID: ea5bdd56-7ceb-ae4f-a2d7-1b9d11cda893
Ancestors: Compiler-ct.449

Implements #environment on SyntaxErrorNotification. Sorry I hit the Accept button to soon...

by Dave Lewis (lewis@mail.msen.com) at November 07, 2020 01:53 PM

Name: Compiler-ct.449
Author: ct
Time: 24 October 2020, 12:12:47.879256 am
UUID: f4dcd413-ed53-bf47-8762-a0e1c23f7a7f
Ancestors: Compiler-tobe.448

Redesigns SyntaxErrorNotification to hold a CompilationCue instead of 'inClass' and 'code', in orderto allow for recompilation of syntax errors where other compilation parameters have been specified, e.g. context or environment.

For example, the following syntax error could not be resolved before this patch:

Compiler new
evaluate: 'aCue yourself:'
in: thisContext sender
to: thisContext sender receiver.

by Dave Lewis (lewis@mail.msen.com) at November 07, 2020 01:52 PM

Name: Compiler-tobe.448
Author: tobe
Time: 17 October 2020, 9:11:04.555834 am
UUID: 0827d6fb-39db-4734-8440-9c0af90b39e5
Ancestors: Compiler-eem.447

Report pc for closureCreationNodes of full blocks

See BlockNode>>#emitCodeForValue:encoder: for the analogous call without full blocks

by Dave Lewis (lewis@mail.msen.com) at November 07, 2020 01:51 PM

Name: System-mt.1189
Author: mt
Time: 4 November 2020, 4:29:05.65862 pm
UUID: d73f04df-f789-f043-8cf0-370cd8332876
Ancestors: System-mt.1188

Updates list of contributors.

by Dave Lewis (lewis@mail.msen.com) at November 07, 2020 01:48 PM

Name: System-mt.1188
Author: mt
Time: 4 November 2020, 11:50:14.56762 am
UUID: 57645a78-976f-ed45-b29d-0a3263b10d3d
Ancestors: System-mt.1187

Minor clean up and speed up.

by Dave Lewis (lewis@mail.msen.com) at November 07, 2020 01:46 PM

Name: System-mt.1187
Author: mt
Time: 2 November 2020, 2:42:51.691816 pm
UUID: fb19b021-de9d-284a-bdea-0d3fd23508c4
Ancestors: System-tpr.1186

Minor code clean-up. Thanks to Christoph (ct)!

by Dave Lewis (lewis@mail.msen.com) at November 07, 2020 01:45 PM

Name: System-tpr.1186
Author: tpr
Time: 25 October 2020, 6:04:37.06521 pm
UUID: 63e37c9d-9d2c-4164-bf71-a0452572ed62
Ancestors: System-tpr.1181, System-mt.1185

Remove the no-longer needed (mis)calculation of the window label. Accompanies Tools-tpr.1008

Hopefully this has merged correctly and won't result in chaos...

by Dave Lewis (lewis@mail.msen.com) at November 07, 2020 01:43 PM

Avoid clearing the image's update URL when resetting all pragma preferences -- which happens during release bundling.

Note that setting the update URL cannot be part of setting the default preferences in the ReleaseBuilder because (update) repository switching is implemented at a different stage. So just ignoring this request works fine. ... we may want to move this preference to ReleaseBuilder anyway ... maybe #updateMapName, too.

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 07, 2020 01:04 PM

Refactor and comment Context >> #runSimulated:contextAtEachStep: per request.

Branch commit, depends on Kernel-ct.1362 indeed.

November 07, 2020 12:48 PM

Adds KR as "Korakurider" to the list of contributors.

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 05, 2020 03:34 PM

Complements System-mt.1190

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 05, 2020 03:24 PM

Speeds up String >> #translated and locale switching as follows:
- Only call #localeChanged(Gently) on classes that implement it, just like #cleanUp:
- For the InternalTranslator, make translation-in-all-domains a no-op; 122µs -> 1.3µs
- For the GetTextTranslator, make translation-in-all-domains fetch all available .mo files once, then directly enumerate that cache until a result is found; roughly 50µs -> 8µs
- Avoid loading all .mo files on #localeChanged(Gently); only do that for translation-in-all-domains

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 05, 2020 03:23 PM

Accommodate new (?) restrictions on storability of block argument variables.

by Tony Garnock-Jones (tonyg@leastfixedpoint.com) at November 05, 2020 01:50 PM

Complements Collections-ct.920.

November 04, 2020 07:02 PM

Proposal: Honor scaleFactor for default spacings in text constants. With this change, tabs in every text editor are finally displayed hi-dpi-sensitive.

Note that I found this method rather by incident than by expertise! I strongly advise against merging this into the Trunk without prior review by someone who has a bit more knowledge of this domain than me. ;-) Is this the right place for such a change, or should we patch the senders of every constant instead?

Requires System-ct.1188.

November 04, 2020 07:02 PM

Show number of truncated items in inspectors

November 04, 2020 06:39 PM

Revive StoryboardBookMorph

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 04, 2020 03:51 PM

More cleanup to StoyboardBookMorph

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 04, 2020 03:51 PM

Make FatBitsPaint menu a separate stay up menu

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 04, 2020 03:50 PM

Width 3000 was a little excessive for some images

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 04, 2020 03:50 PM

Merges some quality-of-life fixes for FatBitsPaint and storyboard books. Thanks to Karl (kfr)!

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 04, 2020 03:49 PM

Fixes dropping for non-world paste-up morphs.

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 04, 2020 03:45 PM

Updates list of contributors.

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 04, 2020 03:29 PM

Don't hide the ProjectViewMorph and it's owner when loading

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 04, 2020 03:27 PM

In progress indication for project loading, remove target-owner-visibility magic as suggested by Karl (kfr). When this indicator is used as part of InternalThreadNavigationMorph or ProjectViewMorph, hiding them (and keeping them hidden for quick workBlocks) makes no sense. Those are more than just layout containers.

(Note that this is unrelated to system-progress indication.)

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 04, 2020 03:26 PM

Add inst. var. lastParenLocation to TextEditor #stateArray (used to transfer state to a newly installed editor). This ensures that #clearParens works as expected in all circumstances.

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 04, 2020 03:12 PM

Fix TextEditor>>#hasMultipleLinesSelected to consider actual lines, instead of wrapped lines.

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 04, 2020 03:11 PM

Provide smoother scrolling in response to mouse wheel events

Instead of delivering the events when wheelDelta reaches 120, make this threshold a Preference (minimumWheelDelta)

Reminder: 120 represents a single notch for traditional mouse wheel with notches, but trackpads can deliver much smaller deltas

Rather than accumulating the wheelDeltas into MouseWheelState, do it when we #generateMouseWheelEvent:
Indeed, small deltas will come in packets of successive events, and it's more efficient to regroup then, exactly like we do with mouse trails...

Also MouseWheelState did ignore time outs (long delays between deltas) and other state changes (buttons/modifiers), which was not ideal.

Directly get those states from the raw eventBuffer, like any other mouse event. This requires integration of tose 2 PR:
https://github.com/OpenSmalltalk/opensmalltalk-vm/pull/461 for Windows
https://github.com/OpenSmalltalk/opensmalltalk-vm/pull/462 for OSX and linux

Honour larger wheelDeltas too in ScrollPane>>#mouseWheel: event handling
Honour horizontal mouse wheels too in ScrollPane
Note that the scroll pane still won't deliver sub-scrollDeltaWidth/scrollDeltaHeight with those changes, the number of scroll deltas is still rounded to 1.

With patched VM, and following settings, I get a reasonnable scrolling experience on OSX:

HandMorph minimumWheelDelta: 20.
Smalltalk sendMouseWheelEvents: true.

NOTE: it is tricky to add/remove inst. var. to such HandMorph because obsolete CompiledMethod pointing to old inst. var. offset are still active on the stack and can mess things up (unless you do not load the package from UI but rather from command line...).

Since accessing 'lastMouseEvent' in such obsolete CompiledMethod is causing grief, preserve its offset by swapping inst. var. order with 'targetOffset' which is less likely used.

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 04, 2020 03:11 PM

When "Mouse over for keyboard focus" is used, release the prior keyboard focus upon restarting the image, so that the Global Command Keys will be honored until the first focus shift.

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 04, 2020 03:10 PM

Add register, inst var, and temp var name lookup in ARMv8 disassembly decoration.

by Eliot Miranda (eliot.miranda@gmail.com) at November 03, 2020 06:23 PM

Name: Kernel-mt.1352
Author: mt
Time: 14 October 2020, 1:58:43.251569 pm
UUID: 72187ee9-c851-894c-b496-d910ac5c5865
Ancestors: Kernel-eem.1351

Rename #doWithIndex: to #withIndexDo:. See http://forum.world.st/The-Inbox-60Deprecated-ct-80-mcz-td5120706.html

by Dave Lewis (lewis@mail.msen.com) at November 03, 2020 01:47 AM

Name: Kernel-eem.1351
Author: eem
Time: 12 October 2020, 10:01:07.200133 pm
UUID: 16ddb529-e8ff-4c75-87cc-9b05ea9a23f2
Ancestors: Kernel-mt.1350

Rename EventSensorConstants WindowEventStinks to WindowEventScreenChange. (Why is EventSensorConstants still in Kernel?).

by Dave Lewis (lewis@mail.msen.com) at November 03, 2020 01:47 AM

Name: Kernel-mt.1350
Author: mt
Time: 11 October 2020, 1:36:58.093652 pm
UUID: bf00c27b-c143-6747-b06a-0a16460949be
Ancestors: Kernel-eem.1349

Refactoring 'Active' variables -- Step 2 of 2. Removes all uses of Active(World|Hand|Event) by replacing those with "self current(World|Hand|Event)" or "Project current world" when required to not add/cement Morphic dependency.

See http://forum.world.st/Changeset-Eliminating-global-state-from-Morphic-td5121690.html

by Dave Lewis (lewis@mail.msen.com) at November 03, 2020 01:46 AM

Name: Kernel-eem.1349
Author: eem
Time: 9 October 2020, 7:49:50.4607 pm
UUID: 6db35c96-122e-414f-bf75-c74d07149f49
Ancestors: Kernel-eem.1348

Neaten CompiledMethod>>#setSourcePointer:, avoiding unnecessary copies and becomes, bu=y refactoring CompiledMethodTrailer>>#createMethod:class:header: into copyToMethod:. Nuke the unsent method dropSourcePointer which is a duplicate of the much older (but also un sent) method zapSourcePointer.

by Dave Lewis (lewis@mail.msen.com) at November 03, 2020 01:45 AM

Name: Kernel-eem.1348
Author: eem
Time: 9 October 2020, 7:18:24.987162 pm
UUID: 006721f7-6f3b-44fa-86f3-8c661d950fa4
Ancestors: Kernel-eem.1347

Fix the wrong outer code bug (see CompilerTest>>testValidFullBlockMethod) and recompile all offending methods.

by Dave Lewis (lewis@mail.msen.com) at November 03, 2020 12:44 AM

Name: System-mt.1179
Author: mt
Time: 14 October 2020, 2:07:40.555569 pm
UUID: 8a7e511d-87d7-6042-8478-87417dd33e12
Ancestors: System-mt.1178

Rename #doWithIndex: to #withIndexDo:. See http://forum.world.st/The-Inbox-60Deprecated-ct-80-mcz-td5120706.html

by Dave Lewis (lewis@mail.msen.com) at November 03, 2020 12:34 AM

Name: System-mt.1178
Author: mt
Time: 11 October 2020, 1:36:25.615652 pm
UUID: fd8d0758-0935-c34c-8cb9-64d908b75f70
Ancestors: System-eem.1177

Refactoring 'Active' variables -- Step 2 of 2. Removes all uses of Active(World|Hand|Event) by replacing those with "self current(World|Hand|Event)" or "Project current world" when required to not add/cement Morphic dependency.

See http://forum.world.st/Changeset-Eliminating-global-state-from-Morphic-td5121690.html

by Dave Lewis (lewis@mail.msen.com) at November 03, 2020 12:33 AM

Name: Collections-mt.917
Author: mt
Time: 14 October 2020, 1:55:10.898569 pm
UUID: 46e13bee-89c0-4d4e-8e9c-8c54ec54ff2b
Ancestors: Collections-mt.916

Rename #doWithIndex: to #withIndexDo:. See http://forum.world.st/The-Inbox-60Deprecated-ct-80-mcz-td5120706.html

by Dave Lewis (lewis@mail.msen.com) at November 03, 2020 12:31 AM

Name: Collections-mt.916
Author: mt
Time: 14 October 2020, 1:01:18.401265 pm
UUID: 49aef966-6542-d34b-881e-a19161e040d4
Ancestors: Collections-ul.915

Prepare refactoring "(do|collect)WithIndex -> withIndex(Do|Collect)". See http://forum.world.st/The-Inbox-60Deprecated-ct-80-mcz-td5120706.html

by Dave Lewis (lewis@mail.msen.com) at November 03, 2020 12:30 AM

Name: Kernel-eem.1347
Author: eem
Time: 9 October 2020, 12:12:35.47215 pm
UUID: a8ab437c-2eac-4257-a403-12d21ea16588
Ancestors: Kernel-eem.1346

Reduce compileCue:environment:classified:withStamp:notifying:logSource: to compileCue:classified:withStamp:logSource: by deriving parameteras from the cue, as intended.

by Dave Lewis (lewis@mail.msen.com) at November 03, 2020 12:25 AM

Name: Kernel-eem.1346
Author: eem
Time: 9 October 2020, 11:55:45.364947 am
UUID: 99f2d20b-c9bf-4bec-a8f3-c681d619ec12
Ancestors: Kernel-eem.1345

Have compileCue:environment:classified:withStamp:notifying:logSource: pay attention to methodTrailer if stored in a CompilationCue.

by Dave Lewis (lewis@mail.msen.com) at November 03, 2020 12:24 AM

Name: Compiler-eem.447
Author: eem
Time: 9 October 2020, 12:11:07.767302 pm
UUID: aee4bc8a-7732-41ee-a2e8-d793572b08a4
Ancestors: Compiler-eem.446

Initialize source correctly in CompilationCue and answer it if available in stringToLog.

by Dave Lewis (lewis@mail.msen.com) at November 03, 2020 12:22 AM

Name: Compiler-eem.446
Author: eem
Time: 9 October 2020, 11:54:51.071572 am
UUID: 6f9f349c-5461-45b5-a707-6150decce9ef
Ancestors: Compiler-eem.445

Have Parser pay attention to encoderClass if stored in a cue.

by Dave Lewis (lewis@mail.msen.com) at November 03, 2020 12:21 AM

Name: Compiler-eem.445
Author: eem
Time: 9 October 2020, 11:49:44.532447 am
UUID: f890a5ec-6748-4c88-bbaa-fe311e572c51
Ancestors: Compiler-eem.444

Allow CompilationCue to carry encoderClass and methodTrailer to more fully control code generation.

by Dave Lewis (lewis@mail.msen.com) at November 03, 2020 12:07 AM

Name: Kernel-eem.1345
Author: eem
Time: 9 October 2020, 11:24:29.39536 am
UUID: a2fbce4f-7c3e-4945-ae44-2f7cb7d40cc7
Ancestors: Kernel-tonyg.1344

Refactor ClassDescription>>#compile:environment:classified:withStamp:notifying:logSource: into compileCue:environment:classified:withStamp:notifying:logSource: to allow clients to more easily install methods with dfferent btecode sets (for tests).

Provide accessors for CompiledCode's primary and secondary bytecodeSetEncoderClasses.

Correct a slip; integers should be compared using #=, not #==.

by Dave Lewis (lewis@mail.msen.com) at November 03, 2020 12:04 AM

Name: System-eem.1177
Author: eem
Time: 8 October 2020, 11:16:38.655114 am
UUID: b65f0705-bbc0-45b7-8f9c-6cc95281cbd5
Ancestors: System-mt.1176

More accurate description of vm parameter 48 (the image header flags).

by Dave Lewis (lewis@mail.msen.com) at November 02, 2020 11:44 PM

Name: Kernel-tonyg.1344
Author: tonyg
Time: 8 October 2020, 11:35:47.705963 pm
UUID: d68b8946-be04-4e8c-9a2d-76e31736aa4b
Ancestors: Kernel-dtl.1343

Avoid explicit mention of Promise in a handful of places, preferring instead some means of specifying "the currently-relevant kind of Promise". Useful for subclasses of Promise.

by Dave Lewis (lewis@mail.msen.com) at November 02, 2020 11:42 PM

Name: Compiler-eem.444
Author: eem
Time: 9 October 2020, 10:59:39.150557 am
UUID: 52facf73-6ea8-4abc-8053-242eacfbb0ec
Ancestors: Compiler-eem.443

Stage one of refactoring of CompilationCue so that compile:environment:classified:withStamp:notifying:logSource: can be refactored into compileCue:environment:classified:withStamp:notifying:logSource: to allow easy installation of methods with different encoder classes (for tests of full vs embedded blocks).

by Dave Lewis (lewis@mail.msen.com) at November 02, 2020 11:40 PM

Name: System-mt.1176
Author: mt
Time: 8 October 2020, 4:56:28.487753 pm
UUID: 31fb9dbd-8015-fb4c-a900-dc9167c278fe
Ancestors: System-topa.1174

Push Morphic-specific code into Morphic project.

by Dave Lewis (lewis@mail.msen.com) at November 02, 2020 11:20 PM

Name: System-topa.1174
Author: topa
Time: 24 September 2020, 11:14:16.007835 pm
UUID: 8f692d9f-4973-4aac-b181-9a2643089f8a
Ancestors: System-eem.1173

Make it so that Theme fonts are looked up on usage, not on theme creation.

by Dave Lewis (lewis@mail.msen.com) at November 02, 2020 11:19 PM

Complements SUnit-ct.131.

November 02, 2020 07:00 PM

Don't save #lastStoredRun results into the source package. See also ReleaseBuilder class >>#discardUserObjects.

Thanks to tobe & mt for the advice! :-)

November 02, 2020 07:00 PM

I couldn't resist...

by Eliot Miranda (eliot.miranda@gmail.com) at November 02, 2020 06:40 PM

Merge Kernel-ct.1359 from inbox
Author: ct
Time: 1 November 2020, 7:50:24.183428 pm
UUID: 3b12b316-79f3-de41-8765-8296d964b821
Ancestors: Kernel-mt.1353

Revise and extend Context #runSimulated: implementation. Remove restriction to blocks that do not have a method return. Add support for exception signaling during the execution, which caused unterminated simulation of the calling process in the past. Support argless contextAtEachStep blocks.

Benchmarks:
code:
[Context runSimulated: [100@100 corner: 200@200]] bench
before:
16.7 ms/run
after:
19.8 ms/run
I think this should be okay, given the fact that the primary purpose of simulation is providing explorability but not efficiency ...

eem: Fix a bug in the use of runSimulated:contextAtEachStep:. Presumably because of the pre-closure block model existing senders were of the form
thisContext sender
runSimulated: aBlock
contextAtEachStep: [....]
but this caused the jump at the end of runSimulated:contextAtEachStep: to short-cut any processing done after runSimulated:contextAtEachStep:. So rewrite all users of runSimulated:contextAtEachStep: in this form:
thisContext
runSimulated: aBlock
contextAtEachStep: [....]
and consequently have tallyInstruction: rallyMethods: et al answer their intended results.

by Eliot Miranda (eliot.miranda@gmail.com) at November 02, 2020 06:10 PM

Extend tests for Context #runSimulated:. Depends on Kernel-ct.1359.

by Eliot Miranda (eliot.miranda@gmail.com) at November 02, 2020 04:53 PM

Indicate instance messages when instance-side reference is browsed.

Otherwise the browser may reach an invalid state, where the class button is still selected, even though an instance message is browsed:

b := Browser open.
b browseReference: (Color >> #asColor) asCodeReference.
b browseReference: (Color class >> #green) asCodeReference.
b browseReference: (Color >> #asColor) asCodeReference.

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 02, 2020 04:10 PM

Merged.

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 02, 2020 04:09 PM

self world returns nil when the morph is not visible/ opened yet

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 02, 2020 04:06 PM

Improve first impression of Etoys features, which are always accessible through a morph's halo. Thanks to Karl (kfr) for taking care!

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 02, 2020 04:04 PM

Do the most basic installs so Etoys appears with right icons. A few more icons for BookMorph. Thanks to Karl (kfr)!

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 02, 2020 04:02 PM

Complements Morphic-mt.1710.

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 02, 2020 03:38 PM

Let file-drag/drop events from the host system arrive as regular (in-image) drag/drop events for a uniform look-and-feel.

Thanks to Christoph (ct)!

For backwards compatibility, DropFilesEvent stays functional until the package 60Deprecated is unloaded. Users should move away from Morph >> #(handle|wants|.)dropFiles to use #acceptDroppingMorph:event: (etc.) and check #dragTransferType for #filesAndDirectories.

Note that, in the long term, this unification (to always use DropEvent) makes sense because it will help extract event handling from Morphic to a more generic, object-oriented, GUI-framework-agnostic mechanism to be used in -- for example -- Morphic, MVC, and SqueakShell.

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 02, 2020 03:37 PM

Minor code clean-up. Thanks to Christoph (ct)!

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 02, 2020 01:42 PM

Complements Morphic-mt.1709. Thanks to Christoph (ct)!

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 02, 2020 01:28 PM

Complements Morphic-mt.1709. Thanks to Christoph (ct)!

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 02, 2020 01:27 PM

Adds an is-check for the drag-and-drop helper TransferMorph, which is the check usually needed to discriminate between "has meta data" or "drop morph as is" in all kinds of tools and widgets.

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at November 02, 2020 01:26 PM

Extend tests for Context #runSimulated:. Depends on Kernel-ct.1359.

November 01, 2020 06:54 PM

Revise and extend Context #runSimulated: implementation. Remove restriction to blocks that do not have a method return. Add support for exception signaling during the execution, which caused unterminated simulation of the calling process in the past. Support argless contextAtEachStep blocks.

Benchmarks:
code:
[Context runSimulated: [100@100 corner: 200@200]] bench
before:
16.7 ms/run
after:
19.8 ms/run
I think this should be okay, given the fact that the primary purpose of simulation is providing explorability but not efficiency ...

November 01, 2020 06:51 PM

Fine-tune message adding in a MessageTrace. When the user is asked for one of many selectors and cancels the menu, don't add any message. Reuse the selector search known from browsing implementors for browsing senders as well. Improve multilingual support.

November 01, 2020 05:29 PM

Cogit: implement the VM lock/unlock funcitons on x86_64. Recategorize x86 processor-specific instruction generators in the correct category. Change the order of fields in 64-bit CogAbstractInstructions to get better packing (affects CogARMv8Instruction).

Compute numTarmpolines correctly for COGMTVM.
Improve the map documentation in initializeAnnotationConstants.

by Eliot Miranda (eliot.miranda@gmail.com) at October 25, 2020 03:04 AM

Update MultiProcessor to current disassembly API.

by Eliot Miranda (eliot.miranda@gmail.com) at October 25, 2020 02:51 AM

Refactoring to start reviving CoInterpreterMT.

by Eliot Miranda (eliot.miranda@gmail.com) at October 24, 2020 03:24 AM

Simplifications in set/void register state.
State to revive MultiProcessor.

by Eliot Miranda (eliot.miranda@gmail.com) at October 24, 2020 03:21 AM

Adds regression test for Kernel-ct.1355 (source logging after SyntaxError). Should I invest further complexity into this test for the purpose of keeping the production source logs clean?

October 23, 2020 10:40 PM

Fixes source logging when fixing a SyntaxError. In the past, the old source was stored in the logs instead of the new one.

October 23, 2020 10:39 PM

empty log message

October 23, 2020 10:18 PM

Removes unintended mock category from package organization.

October 23, 2020 10:17 PM

Implements #environment on SyntaxErrorNotification. Sorry I hit the Accept button to soon...

October 23, 2020 10:16 PM

Complements Compiler-ct.449 (CompilationCue in SyntaxErrorNotification). Fixes syntax highlighting in SyntaxError window.

October 23, 2020 10:15 PM

Redesigns SyntaxErrorNotification to hold a CompilationCue instead of 'inClass' and 'code', in orderto allow for recompilation of syntax errors where other compilation parameters have been specified, e.g. context or environment.

For example, the following syntax error could not be resolved before this patch:

Compiler new
evaluate: 'aCue yourself:'
in: thisContext sender
to: thisContext sender receiver.

October 23, 2020 10:12 PM

CoInterpreterMT: eliminate use of reenterInterpreter, thsi is StackInterpreter only now.

FilePlugin: use methodReturnInteger: rather than methodReturnValue: ... intergerObjectOf::

Simulators: start to simulate the interrupt-driven inputSemaphore input event mechanism (with an egregious hack in ioRelinquishProcessorForMicroseconds: that we can probably get rid of when we understand the code better). The mystery is that the changes here do allow events to get in quickly (e.g. Run tests in a TestRunner) but the screen menu resolutely refuses to appear (?!?!). Move ioGetNextEvent:, queueForwardedEvent: & ioProcessEvents up into StackInterpreter (as doNotGenerate). Use DoubleWordArray for the input event buffer on 64-bits.

by Eliot Miranda (eliot.miranda@gmail.com) at October 23, 2020 09:40 PM

Provide a much simpler much more ratuional mechanism for squashing duplicate mouse move events in the simulator. Simply peek last and update the last event's position if it's a duplicate of the next event.

by Eliot Miranda (eliot.miranda@gmail.com) at October 23, 2020 09:10 PM

Promote annotation requests to be a pragma preference. Needs PreferenceBrowser-mt.108.

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at October 23, 2020 01:49 PM

Make views on text prefs more flexible to support, for example, arrays with literals.

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at October 23, 2020 01:48 PM

Make format of "unknown author" match "no timestamp".

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at October 23, 2020 10:10 AM

Fixes inverted author index to be case-sensitive to improve lookup quality for overlaps such as "JM" and "jm".

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at October 23, 2020 10:09 AM

Tools-mt.1006

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at October 23, 2020 09:58 AM

Complements Tools-mt.1006

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at October 23, 2020 09:58 AM

empty log message

by zhang0xin (313366084@qq.com) at October 23, 2020 02:08 AM

Report pc for closureCreationNodes of full blocks

See BlockNode>>#emitCodeForValue:encoder: for the analogous call without full blocks

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at October 21, 2020 04:13 PM

Restore default name for update map to be 'update'. Not sure whether this one should also be ensured in ReleaseBuilder.

by Marcel Taeumel (marcel.taeumel@hpi.uni-potsdam.de) at October 21, 2020 04:02 PM

Have the Workspace thang deal with both ,text and .txt

by Eliot Miranda (eliot.miranda@gmail.com) at October 20, 2020 04:36 AM

...and remember to track the target directory if it is changed.

by Eliot Miranda (eliot.miranda@gmail.com) at October 20, 2020 12:56 AM

Integrate Javier Diaz-Reinoso's kindly contributed improvement to Workspace file editing.

by Eliot Miranda (eliot.miranda@gmail.com) at October 20, 2020 12:18 AM

Integrate Javier Diaz-Reinoso;s kindly contributed improvement to Workspace file editing.

by Eliot Miranda (eliot.miranda@gmail.com) at October 20, 2020 12:17 AM

Have the interpreters include sqImageFileAccess.h directly so that sq.h doesn't have to, and hence the entire VM isn't recompiled whenever sqImageFileAccess.h does.

Fix some C compiler warnings around invalidCompactClassError:.

by Eliot Miranda (eliot.miranda@gmail.com) at October 19, 2020 05:28 PM

And have the flags not recognised by a specific VM still get recorded in imageHeaderFlags, so set parameters will get set in the image header.

by Eliot Miranda (eliot.miranda@gmail.com) at October 19, 2020 12:32 AM

Rework the window label handing of MessageSet/Trace to be a bit more accurate, and to work within the normal style of labels.

Instead of assuming the size of the list passed to the MessageSet/Trace is correct, derive it from the post-processing list. We process the list to remove duplicates, which e.g. the #allUnimplementedCalls method produces.

This required adding an instvar to M-Set, removing one from M-Trace, setting the new one properly, dumping some dodgy code that used to half-assedly derive the old one, use the proper #changed: #windowTitle to, y'know, change the title, and some faffing with the messageList.

As a result we are better placed to make further improvements if/when we develop a way to correctly hightlight multi-part keywords (or indeed, multiple messages) within a single method, which would greatly improve many browsers.

This benefits from being accompanied by the single-method change in the System-tpr.1181 package

October 19, 2020 12:24 AM

Remove the no-longer needed (mis)calculation of the window label. Accompanies Tools-tpr.1004

October 19, 2020 12:23 AM

Fix an awful regression introduced in VMMaker.oscog-eem.2837. Have the setFullScreen fag also set the flag that is stored in the image header, hecne having the image start up as it was saved.

by Eliot Miranda (eliot.miranda@gmail.com) at October 18, 2020 11:58 PM

Add BIT_IDENTICAL_FLOATING_POINT to the CogVMFeatureFlags printing. Print flag fields in binary to the appropriate width.

by Eliot Miranda (eliot.miranda@gmail.com) at October 18, 2020 09:34 PM

OSProcess 4.6.22
Methods for accessing posix system configuration queries. Requires OSProcessPlugin 4.6.7 (VMConstruction-Plugins-OSProcessPlugin-dtl.53) or equivalent for primitive support. See convenience method OSProcess class>>gnuLinuxConf for example usage.

Supported queries are sysconf(3), confstr(3), pathconf(3), and fpathconf(3)

by Dave Lewis (lewis@mail.msen.com) at October 18, 2020 08:11 PM

OSProcess 4.6.22
Methods for accessing posix system configuration queries. Requires OSProcessPlugin 4.6.7 (VMConstruction-Plugins-OSProcessPlugin-dtl.53) or equivalent for primitive support. See convenience method OSProcess class>>gnuLinuxConf for example usage.

Supported queries are sysconf(3), confstr(3), pathconf(3), and fpathconf(3)

by Dave Lewis (lewis@mail.msen.com) at October 18, 2020 08:10 PM

OSProcess 4.6.22
Methods for accessing posix system configuration queries. Requires OSProcessPlugin 4.6.7 (VMConstruction-Plugins-OSProcessPlugin-dtl.53) or equivalent for primitive support. See convenience method OSProcess class>>gnuLinuxConf for example usage.

Supported queries are sysconf(3), confstr(3), pathconf(3), and fpathconf(3)

by Dave Lewis (lewis@mail.msen.com) at October 18, 2020 08:10 PM

OSProcessPlugin 4.6.7
Add new primitives for posix system configuration access via sysconf(3), confstr(3), pathconf(3), and fpathconf(3). For example usage, see OSProcess class>>gnuLinuxConf.

by Dave Lewis (lewis@mail.msen.com) at October 18, 2020 07:42 PM

Fix #graphQlTestsEngine.

by Chris Muller (ma.chris.m@gmail.com) at October 14, 2020 12:12 AM

Fix GraphQL tests package name.

by Chris Muller (ma.chris.m@gmail.com) at October 14, 2020 12:05 AM

Don't use #ma repository yet.

by Chris Muller (ma.chris.m@gmail.com) at October 14, 2020 12:01 AM

Excavating KryptOn!

by Chris Muller (ma.chris.m@gmail.com) at October 13, 2020 11:57 PM

Fix a dangerous part of the Chronology API.

- DateAndTime utcMicroseconds: Time utcMicrosecondClock offset: 0 looks reasonable, but it's very not. Notice how it's off, silently, by 68 years!!!
- Deprecate Time class>>#utcMicrosecondClock. Only soft-deprecate it for now due to its use in Kernel and Network packages. Will update them as needed.
- Replace it with Time class>>#primMicrosecondClockValue, where "prim" at least indicates implementation-specific properties may be present.
- Complement legacy Smalltalk-80 protocol, and expose the new microsecond-level precision in the VM for users with:
DateAndTime>>#asMicroseconds and
DateAndTime class>>#fromMicroseconds:
- Rename Time class>>#localMicrosecondClock to #localMicrosecondClockValue to 1) be consistent with microsecondClockValue and 2) avoid overloading the "clock" nomenclature. It's a value, not an object!
- Remove unused DateAndTime class>>#millisecondClockValue.

by Chris Muller (ma.chris.m@gmail.com) at October 13, 2020 11:46 PM

- Optimize for system compactness by ensuring the default internal array size of any HashedCollection is not initialized larger than it may ever need to be.
- Let #new: be used to define larger sizes than the minimum, and perform comparably with #new even if the minimum size is specified.

by Chris Muller (ma.chris.m@gmail.com) at October 13, 2020 09:22 PM

Ability to run and compare multiple benchmarks at once via:

{ baselineExpression.
altExpression1.
altExpression2.
"...etc..." } benchCompare

The standard bench string is reported for each, along with a percentage rate relative to the baseline, the first one.

by Chris Muller (ma.chris.m@gmail.com) at October 13, 2020 09:21 PM

Integrate fix from Stéphane Rollandin that keeps ProportionalSplitter bars in their proportionally-assigned place after resizing windows.

by Chris Muller (ma.chris.m@gmail.com) at October 13, 2020 09:00 PM

- Use environment bindingOf: rather than associationAt:.
- Fix thoroughSenders preference.
- Make ColorTheme a concrete, attributable object.
- Set the default theme to a new instance of ColorTheme.
- This will help us disengage from SmalllandTheme dependency.

by Chris Muller (ma.chris.m@gmail.com) at October 13, 2020 09:00 PM

Fall back to MorphicUIManager for Text instances.

October 12, 2020 10:30 AM

SnarlUIManager

October 12, 2020 10:28 AM

Updates for modern Squeak

October 12, 2020 10:17 AM

added link to squeakbooks.com

by tty (gettimothy@zoho.com) at October 11, 2020 10:03 AM

modified class comment

by tty (gettimothy@zoho.com) at October 11, 2020 09:57 AM

collision with TBSBootstrap apparently.

by tty (gettimothy@zoho.com) at October 11, 2020 09:53 AM

empty log message

October 11, 2020 09:28 AM

BitSyntax Help documentation.

by Tony Garnock-Jones (tonyg@leastfixedpoint.com) at October 06, 2020 01:54 PM

Rename package and monticello file to BitSyntax-Examples (from BitSyntaxExamples)

by Tony Garnock-Jones (tonyg@leastfixedpoint.com) at October 06, 2020 01:54 PM

Rename Monticello file, too.

by Tony Garnock-Jones (tonyg@leastfixedpoint.com) at October 06, 2020 01:54 PM

updated required package version to 2.

by Robert Withers (rabbit@churchofthesacrifice.org) at October 04, 2020 06:54 PM

updated required package version to 2.

by Robert Withers (rabbit@churchofthesacrifice.org) at October 04, 2020 06:53 PM

added comments and removed the PromisesLocalServiceProvider.

by Robert Withers (rabbit@churchofthesacrifice.org) at October 04, 2020 06:53 PM

added comments and removed the PromisesLocalServiceProvider.

by Robert Withers (rabbit@churchofthesacrifice.org) at October 04, 2020 06:53 PM

added a couple of comments.

by Robert Withers (rabbit@churchofthesacrifice.org) at October 04, 2020 06:52 PM

added a couple of comments.

by Robert Withers (rabbit@churchofthesacrifice.org) at October 04, 2020 06:52 PM

checkpoint

by Robert Withers (rabbit@churchofthesacrifice.org) at October 04, 2020 04:57 PM

checkpoint

by Robert Withers (rabbit@churchofthesacrifice.org) at October 04, 2020 04:57 PM

checkpoint

by Robert Withers (rabbit@churchofthesacrifice.org) at October 04, 2020 04:55 PM

checkpoint

by Robert Withers (rabbit@churchofthesacrifice.org) at October 04, 2020 04:55 PM

split out ProtocolState

by Robert Withers (rabbit@churchofthesacrifice.org) at October 04, 2020 04:54 PM

split out ProtocolState

by Robert Withers (rabbit@churchofthesacrifice.org) at October 04, 2020 04:54 PM

explicitly take protocol from trunk Promise

by Robert Withers (rabbit@churchofthesacrifice.org) at October 04, 2020 04:47 PM

explicitly take protocol from trunk Promise

by Robert Withers (rabbit@churchofthesacrifice.org) at October 04, 2020 04:47 PM

complete protocol removal

by Robert Withers (rabbit@churchofthesacrifice.org) at October 04, 2020 04:36 PM

complete protocol removal

by Robert Withers (rabbit@churchofthesacrifice.org) at October 04, 2020 04:35 PM

finish cycling out unused protocol

by Robert Withers (rabbit@churchofthesacrifice.org) at October 04, 2020 04:20 PM

Have up/down/page up/page down keys also exit from the completion dialog.

by Eliot Miranda (eliot.miranda@gmail.com) at September 23, 2020 03:33 PM

Fix facet binding in callbacks; add FacetWorkQueue utility

September 23, 2020 08:35 AM

Make bindActor automatically bindFacet in FacetActors

September 22, 2020 08:34 PM

DataspaceProxy>>newPresenceField:value:; FacetField>>changed.

September 22, 2020 04:23 PM

And-patterns.

September 22, 2020 04:23 PM

Adds version 2.0 to make FFI loadable through this configuration in Squeak 6.0alpha.

by Marcel Taeumel (marcel.taeumel@hpi.de) at September 15, 2020 12:20 PM

Adds 1.0.3 explicitely for 6.0alpha. (Note that Metacello needs to be informed that latest Squeak 5 is 5.3, not 5.1)

September 15, 2020 08:42 AM

Adds 1.0.3 for Squeak 5.2 and above.

September 15, 2020 08:21 AM

Clean up ActorBehavior spawn interface

September 14, 2020 10:00 AM

Fix a typo

by Eliot Miranda (eliot.miranda@gmail.com) at September 14, 2020 01:40 AM

- fix #unescapeUnicode when the encoded character value is outside of BMP (reported by Christoph Thiede on squeak-dev)
- #unescapeUnicode won't accept invalid unicode surrogate values any more
- added a test case to cover the positive case of #unescapeUnicode
- fixed Json class >> #mimeType

by Levente Uzonyi (leves@elte.hu) at September 05, 2020 11:05 PM

Added examples for toying with CSV

September 04, 2020 09:52 AM

Better during:monitor: linkage.

September 03, 2020 08:53 PM

Move to patch-based assertion changes

September 03, 2020 01:53 PM

Name: Kernel-eem.1335
Author: eem
Time: 21 July 2020, 12:33:42.401688 pm
UUID: 829e7c17-0cb2-4269-a6ca-d85bef9ecc2a
Ancestors: Kernel-mt.1334

Support for FloatArray; add 64-bit duals of the 32-bit float converters, so along side Float>>asIEEE32BitWord & Float class>>fromIEEE32Bit: we have Float>>asIEEE64BitWord & Float class>>fromIEEE64Bit:.

by Dave Lewis (lewis@mail.msen.com) at August 23, 2020 05:46 PM