Squeak-dev July 2018

squeak-dev@lists.squeakfoundation.org

28 participants
119 discussions

The Trunk: Collections-cmm.803.mcz
by commits＠source.squeak.org 26 Jul '18

26 Jul '18

Chris Muller uploaded a new version of Collections to project The Trunk: http://source.squeak.org/trunk/Collections-cmm.803.mcz ==================== Summary ==================== Name: Collections-cmm.803 Author: cmm Time: 26 July 2018, 4:01:33.780963 pm UUID: d6d9dcd4-56c1-4242-9bf8-18d2caa66516 Ancestors: Collections-dtl.802, Collections-cmm.801 Merge. =============== Diff against Collections-dtl.802 =============== Item was changed: ----- Method: Text class>>string:runs: (in category 'private') ----- string: aString runs: anArray + ^self basicNew setString: aString setRuns: anArray! - ^self basicNew setString: aString setRunsChecking: anArray! Item was changed: ----- Method: Text>>setString:setRunsChecking: (in category 'private') ----- setString: aString setRunsChecking: aRunArray + "Check runs and do the best you can to make them fit..." + - | stringSize runsSize | string := aString. + "check the runs" aRunArray ifNil: [^ aString asText]. + (aRunArray isKindOf: RunArray) ifFalse: [^ aString asText]. + aRunArray runs size = aRunArray values size ifFalse: [^ aString asText]. + aRunArray size = aString size ifFalse: [^ aString asText]. + runs := aRunArray.! - "Check runs and do the best you can to make them fit..." - aRunArray runs size = aRunArray values size ifFalse: [^ aString asText]. "raise error here?" - runsSize := aRunArray size. - stringSize := string size. - runs := stringSize = runsSize - ifTrue: [aRunArray] - ifFalse: [ stringSize > runsSize - ifTrue: [aRunArray addLast: {} times: stringSize - runsSize] - ifFalse: [aRunArray copyFrom: 1 to: stringSize]].!

2 1

[TEST PLANS] We need your support for making Squeak better
by Marcel Taeumel 26 Jul '18

26 Jul '18

Dear Squeakers, we, the Squeak community, are aiming to do a new release soon. To ensure as high a quality release as possible we'd like to test it adequately. This implies testing across all supported platforms, and possibly on some unsupported ones too. :-) While we have a fairly strong test suite, it does not cover all areas of the system, and not all tests are included such as FFI tests are in its separate package. And of course currently we have about 100 expected failures, 40 unexpected failures, and 10 errors in the base test suite. So we, the board, would like - volunteers to offer to TEST PLATFORMS, different versions of Windows, Raspberry Pi, the many flavors of Linux, etc. Once we have a release candidate we will want to do a clean install on the platform and then run tests. - the community to develop a set of TEST PLANS for testing those parts of the system not tested by the test suite - a REVIEW of the test failures and errors. Are any obsolete? Are any platform-specific? Can we triage them into those that are tractable with little effort, those that are tractable with a lot of effort, those that are insoluble, etc...? Of the expected failures, which are due to fixable bugs, fundamental limitations, etc? So, the data we need to collect includes volunteers, a platform matrix, test plans, and reviews of existing failures and errors. *** To complement Squeak's automated tests, we want to create a MANUAL TEST PLAN. Such a manual plan can be used by any volunteering Squeaker to put any upcoming release "to the acid test." ;-) As a result, our release process gets more robust, and the final release artifact can become of higher quality. As a first step, we want YOU to write down some frequent interactions that you do on a regular basis. You know, things like "I do always do this when adding a class" or "I really need this to do my taxes". It may be a keyboard shortcut, a familiar entry in the world menu, or that lovely button in the upper right corner of tool so-and-so. We want to know! How does it work? Well, we want to guide you a little bit so that things do not get out of hand ... content-wise. ;-) Here are the rules: 1) Give that interaction a name (or title) to express the goal you want to achieve. 2) Briefly describe the starting context. Is the world empty? 3) Divide that interaction into no more than 5 steps so that everybody can click or push his or her way through the Squeak image as quickly as possible. 4) Briefly add some details about the things you should now see or experience. Here is an example: GOAL: "I want to calculate 3+4 in Squeak." CONTEXT: "The world is empty." STEPS - Left click into the world, a menu should appear. - Choose "workspace" from the upper items in the menu. - A workspace window appears, and I click into it's text area. - I type "3+7" and hit [CMD]+[P]. RESULT: "I see a 7." Our idea is to generalize and summarize the most common interactions into a manual test plan for everybody to play around with and report back to the list. When you create a new test, please send it to the list by replying to this email thread. :-) Best, Marcel (for the Squeak Oversight Board)

4 20

Re: [squeak-dev] [Vm-dev] VM crashes while installing Seaside in Squeak 5.2a (trunk)
by H. Hirzel 26 Jul '18

26 Jul '18

Hello The note that the installation of Seaside crashes is of early June. There was actually a crash reported by me in the thread [Test][Squeak5.2a][Metacello] Installation causes VM crash Tobias provided a fixed VM which made the Metacello installation no longer crash. As Seaside uses Metacello this is surely a contributing factor to the crash of the Seaside installation. Maybe the only one, maybe not. I also do not know if the fix done by Tobias has been included into the All-In-One builds for Squeak. So my question is: What are the news about installing Seaside in Squeak 5.2a? Regards Hannes

1 2

The Trunk: Tools-eem.827.mcz
by commits＠source.squeak.org 26 Jul '18

26 Jul '18

Eliot Miranda uploaded a new version of Tools to project The Trunk: http://source.squeak.org/trunk/Tools-eem.827.mcz ==================== Summary ==================== Name: Tools-eem.827 Author: eem Time: 25 July 2018, 3:12:22.772831 pm UUID: 005a00a2-4a38-4aa5-b361-569231851539 Ancestors: Tools-cmm.826 Implement tempsAndValues in ContextVariablesInspector, allowing 'all temp vars' to display integers in the default integer base. [Leave Context>>tempsAndValues for now; someone can clean up the duplication later if desired.] =============== Diff against Tools-cmm.826 =============== Item was changed: ----- Method: ContextVariablesInspector>>selection (in category 'selecting') ----- selection "Refer to the comment in Inspector|selection." + ^selectionIndex + caseOf: { + [0] -> ['']. + [1] -> [object]. + [2] -> [object actualStackSize > 0 ifTrue: [object top] ifFalse: ['']]. + [3] -> [self tempsAndValues] } + otherwise: + [object debuggerMap namedTempAt: selectionIndex - 3 in: object]! - selectionIndex = 0 ifTrue:[^'']. - selectionIndex = 1 ifTrue: [^object]. - selectionIndex = 2 ifTrue: [^object actualStackSize > 0 ifTrue: [object top] ifFalse: ['']]. - selectionIndex = 3 ifTrue: [^object tempsAndValues]. - ^object debuggerMap namedTempAt: selectionIndex - 3 in: object! Item was added: + ----- Method: ContextVariablesInspector>>tempsAndValues (in category 'selecting') ----- + tempsAndValues + "Answer a string of the temporary variables and their current values" + | debuggerMap integerClasses aStream | + aStream := WriteStream on: (String new: 100). + debuggerMap := object debuggerMap. + integerClasses := Integer allSubclasses. + (debuggerMap tempNamesForContext: object) doWithIndex: + [:title :index | | temp | + aStream nextPutAll: title; nextPut: $:; space; tab. + temp := debuggerMap namedTempAt: index in: object. + (integerClasses identityIndexOf: (object objectClass: temp)) ~= 0 + ifTrue: [temp storeOn: aStream base: self defaultIntegerBase] + ifFalse: [object print: temp on: aStream]. + aStream cr]. + ^aStream contents!

1 0

[ANN] A new way of view Squeak Wiki
by Edgar De Cleene 25 Jul '18

25 Jul '18

http://190.193.255.76:9090/1 <http://190.193.255.76:9090/1> Ip could change , use squeaks.org, wait 3 seconds and when the welcome shows, type /nnn to right of port for see wiki page. Enjoy Edgar @morplenauta

1 0

Dynabook concept, critic view
by Hilaire 25 Jul '18

25 Jul '18

Hi, Some of you may be interested on the point of view of a teacher-developer on the Dynabook concept. Here is a little writing about it http://blog.drgeo.eu/post/2018/The-Dynabook-Concept Hilaire -- Dr. Geo http://drgeo.eu

5 7

[ANN] Changing the Trunk version from 6.0alpha to 5.2alpha
by Marcel Taeumel 24 Jul '18

24 Jul '18

Hi, there. We are going to change the current version number of the Trunk to "5.2alpha" to denote our plans for the upcoming Squeak release. After several discussions within the board and with our release manager Edgar, we came to conclusion that it makes sense to postpone the "6.0" a little bit to finish some bigger concerns such as the use of ephemerons and Etoys clean-up. Best, Marcel

20 55

Dynabook learning models
by Hilaire 24 Jul '18

24 Jul '18

As a follow up of my previous posts on the subject, a critic view on the Dynabook under the perspective of its learning models http://blog.drgeo.eu/post/2018/The-Dynabook-and-its-learning-models Hilaire -- Dr. Geo http://drgeo.eu

1 0

Cleaning unneeded UpdateStreamDownloader
by Edgar J. De Cleene 24 Jul '18

24 Jul '18

I wish we have .cs as in old days , but some people like using clumsy tools. So if we do not use .cs anymore why have UpdateStreamDownloader and related to lo download .cs from server ? Edgar @morplenauta

6 15

Improving SqueakToolsHelp class>>#fontSizeSummary reveals problems in Text & styles
by tim Rowledge 23 Jul '18

23 Jul '18

I wanted to make the help page for the font summary show the actual fonts; it seems an appropriate thing to do. Some ... interesting ... issues came up whilst looking into it. Changing TextStyle class>>#fontSizeSummaryContents wasn't too hard. It's easy enough to stick in TextFontReferences that point to a relevant font and the resulting text can be displayed easily enough. It seems a bit strange the TextStyle>>knownTextStyles returns a list of symbols rather than the actual styles, but whatever. That's where it stopped being simple. Upon opening a help browser the fonts were all flattened out to the default font of the default style. It turns out that in the building of the pluggable text morph the text font gets overwritten, although initially in the code it might look otherwise. See MorphicToolBuilder>>#buildPluggableText: and note where it says 'aSpec font ifNotNil:[:f| widget font: f]'. Since the spec #font method (PluggableTextSpec>>#font) returns the default font unless some other has been set, we will normally get the preferences set default font - Deja Vu san 9 in my image. This font is then blasted over the text with #beAllFont: which actually does a great deal of manipulating of the text's runs and produces a rather obese collection of runs all with the same values, rather than consolidating into a single run (modulo having other types of text attribute involved of course). Looking at how the TextMorph>>#beAllFont: and Text>>#addAttribute: works is quite interesting too, since we get into a rather complex and ripe for improved implementation subsystem where I see mis-use of #isKindOf: etc. in places where some double-dispatching would be much nicer. Changing any of this seems like more change than we really want just before a release so it will have to wait and the font summary help page can't really be improved much until then. To make things a bit more irritating it is quite possible to generate the multi-font text and paste it into a text morph and see the desired effect! And yet again the supposedly randomly chosen sigline summarises the entire thing rather well. "Random" my ass.... tim -- tim Rowledge; tim(a)rowledge.org; http://www.rowledge.org/tim This is all a lot simpler and a lot more complicated than you could possibly imagine

2 1

← Newer
1
2
3
4
5
6
7
8
...
12
Older →