Squeak-dev May 2018

squeak-dev@lists.squeakfoundation.org

44 participants
189 discussions

The Trunk: Morphic-tpr.1437.mcz
by commits＠source.squeak.org 23 May '18

23 May '18

tim Rowledge uploaded a new version of Morphic to project The Trunk: http://source.squeak.org/trunk/Morphic-tpr.1437.mcz ==================== Summary ==================== Name: Morphic-tpr.1437 Author: tpr Time: 22 May 2018, 5:33:32.199768 pm UUID: d675d118-fb16-4878-b0a7-1756fb5b8aae Ancestors: Morphic-kfr.1436 Use the new DirectoryChooser to find projects. It lacks the interesting decorations of the older version; this may upset people enough to want to revert this. =============== Diff against Morphic-kfr.1436 =============== Item was changed: ----- Method: MorphicProject>>findAFolderForProject:label: (in category 'utilities') ----- findAFolderForProject: aProject label: dialogLabel "Find a folder for saving or loading a project" + ^DirectoryChooserDialog findAFolderForProject: aProject label: dialogLabel - ^FileList2 modalFolderSelectorForProject: aProject !

1 0

The Trunk: Network-tpr.226.mcz
by commits＠source.squeak.org 23 May '18

23 May '18

tim Rowledge uploaded a new version of Network to project The Trunk: http://source.squeak.org/trunk/Network-tpr.226.mcz ==================== Summary ==================== Name: Network-tpr.226 Author: tpr Time: 22 May 2018, 5:17:49.262652 pm UUID: ca28ea3c-7b03-4b6e-b1ad-26c1c2efa2ba Ancestors: Network-ul.225 If one has a server directory included in the FileList but do not enter the password when it is requested, a never ending cycle can be entered. The #poenFTPClient method will ask if yu want to give up and if you do, will try open a notifier. Either proceeding or abandoning will just throw you back into the drug den of horror. A simple improvement - though not really a full solution - is to not open the notifier. At least that way you get back to the real world. Added comment to explain a little of this in FTPClient>login, which is where the real problem lies; the rather odd seeming decision that #requestPassword: should return an empty string on cancel means that we can never tell whether the intent was an actual empty password (permitted and used in some places) or a cancellation. As it is, the login is attempted (and takes up user time) with the empty string, and if that fails to login correctly, the user is asked if they meant to cancel. This is a very poor UI. Much ncer would be for a cancellation to be recognised correctly and for things to be, well, cancelled. There aren't really any good reasons for returning an in-band value when an out-of-band action was requested. It's also not the best of ideas for non-UI systems to be explicitly requesting UI actions rather than using a more indirect method such as exception raising. =============== Diff against Network-ul.225 =============== Item was changed: ----- Method: FTPClient>>login (in category 'private') ----- login self initiateSession. self user ifNil: [^self]. ["repeat both USER and PASS since some servers require it" self sendCommand: 'USER ', self user. "331 Password required" self lookForCode: 331. + + self sendCommand: 'PASS ', self password."will ask user, if needed. An extra delight is that either accepting an empty string or cancelling the dialog will return an empty string. So how do we know whether to try to log in with an empty password or give up?" - "will ask user, if needed" - self sendCommand: 'PASS ', self password. "230 User logged in" ([self lookForCode: 230.] on: TelnetProtocolError do: [false]) == false ] whileTrue: [ (LoginFailedException protocolInstance: self) signal: self lastResponse] ! Item was changed: ----- Method: ServerDirectory>>openFTPClient (in category 'dis/connect') ----- openFTPClient | loginSuccessful | client ifNotNil: [client isConnected ifTrue: [^client] ifFalse: [client := nil]]. client := FTPClient openOnHostNamed: server. loginSuccessful := false. [loginSuccessful] whileFalse: [ [loginSuccessful := true. client loginUser: self user password: self password] on: LoginFailedException do: [:ex | | what | passwordHolder := nil. what := UIManager default chooseFrom: #('enter password' 'give up') title: 'Would you like to try another password?'. + what = 1 ifFalse: [ ^nil]. - what = 1 ifFalse: [self error: 'Login failed.'. ^nil]. loginSuccessful := false]]. client changeDirectoryTo: directory. ^client!

1 0

The Trunk: Morphic-kfr.1436.mcz
by commits＠source.squeak.org 22 May '18

22 May '18

Karl Ramberg uploaded a new version of Morphic to project The Trunk: http://source.squeak.org/trunk/Morphic-kfr.1436.mcz ==================== Summary ==================== Name: Morphic-kfr.1436 Author: kfr Time: 22 May 2018, 10:58:37.053577 pm UUID: 31da9b7c-7013-354a-838c-91d19533e3ef Ancestors: Morphic-kfr.1435 Convert old Slider/ ScollBar sends from #scrollBarValue: to #vScrollBarValue: I don't think there will be a problem with horizontal scrolling... =============== Diff against Morphic-kfr.1435 =============== Item was added: + ----- Method: Slider>>convertToCurrentVersion:refStream: (in category 'accessing') ----- + convertToCurrentVersion: varDict refStream: smartRefStrm + + (setValueSelector = #scrollBarValue:) ifTrue:[self setValueSelector: #vScrollBarValue:]. + ^ super convertToCurrentVersion: varDict refStream: smartRefStrm!

1 0

[Please Test] Refactoring for BorderedMorph to use BorderStyle
by Marcel Taeumel 22 May '18

22 May '18

Hi, there. Please find attached a refactoring of BorderedMorph to make use of BorderStyle like regular morphs do. Maybe, in the future, we can get rid of BorderedMorph. It includes: - No instVar access to borderColor and borderWidth but message sends. - No #inset or #raised anymore when asking a morph for its #borderColor. Just colors. - Copying the prototypical border styles from the UI theme when used in morphs. - A post-load script that updates all your morphs in the image. Not that important but good for keeping your current tools opened. After a few days of no complaints, I will merge it into Trunk. Best, Marcel

8 33

Improvements to wiki.squeak.org
by K K Subbu 20 May '18

20 May '18

Hi, wiki.squeak.org/squeak is an excellent resource for Squeakers. However, it's interface is a bit dated and could do with some improvements. I hope I am not burdening maintainers with my site suggestions. * search could be a text field instead of a button. This will ease the interface for beginners * The search results prints lots of text (including the IP address of the machine from which the author last edited the text!). Printing the page title and modification timestamp (ISO) should be sufficient. VM Command Line Options, 2006-11-12 9:12 pm * Is it possible to tag pages (say with #obsolete or #beginner #developer). Pages with #obsolete tags could be listed at end of the search results making them more relevant. A ton of thanks to the team that keeps the wiki running. Regards .. Subbu

17 52

Re: [squeak-dev] Squeak board meeting notes 2018-05-16
by Bruce O'Neel 19 May '18

19 May '18

Hmm, maybe we could have another three letter acronym than SOB for the Squeak Oversight Board. Just SB would be good... > SOB meeting notes are much appreciated. Thx. > > Ken G. Brown > > > On May 16, 2018, at 12:09, Craig Latta wrote: > > > > > > Hi all-- > > > > The notes for the 2018-05-16 Squeak board meeting are at: > > > > https://tinyurl.com/y7vflrcz (docs.google.com) > > > > Thanks for any feedback! > > > > > > Greetings on behalf of your Squeak board of directors, > > Craig > > > > -- > > Craig Latta > > Black Page Digital > > Amsterdam :: San Francisco > > craig(a)blackpagedigital.com > > +31 6 2757 7177 (SMS ok) > > + 1 415 287 3547 (no SMS) > > > >

3 2

The Trunk: WebClient-Core-ul.115.mcz
by commits＠source.squeak.org 19 May '18

19 May '18

Levente Uzonyi uploaded a new version of WebClient-Core to project The Trunk: http://source.squeak.org/trunk/WebClient-Core-ul.115.mcz ==================== Summary ==================== Name: WebClient-Core-ul.115 Author: ul Time: 19 May 2018, 11:24:01.219213 am UUID: 64708317-dee5-446d-8a1f-f71335246095 Ancestors: WebClient-Core-topa.114 Decrease the priority of connection handler processes from 60 to 30, to avoid potential image lockups. =============== Diff against WebClient-Core-topa.114 =============== Item was changed: ----- Method: WebServer>>asyncHandleConnectionFrom: (in category 'handling') ----- asyncHandleConnectionFrom: aSocket "Handle an incoming connection asynchronously to avoid blocking the listener." | handlerProcess | handlerProcess := [ [ [ self handleConnectionFrom: aSocket ] on: Error do: [ :ex | aSocket destroy ] ] ensure: [ aSocket destroy. mutex critical: [ + "At this point handlerProcess is initialized, so it is safe to use that variable. Using Processor activeProcess is not possible, because this unwind block may be evaluated by another process. Doing this outside the #ensure: block is not safe." + connections remove: handlerProcess ifAbsent: nil ] ] ] newProcess. - "If we get here, then handlerProcess is initialized, so it's safe to use that variable. Using Processor activeProcess is not possible, because this block may be evaluated by another process. Doing this outside the #ensure: block is not safe." - connections remove: handlerProcess ifAbsent: [ ] ] ] ] newProcess. mutex critical:[connections add: handlerProcess]. + "Process priority has to be set to a lower level here, else it'll have lowIOPriority (60), and will block many important processes (UI, WeakArrayFinalization, etc.)" + handlerProcess + priority: Processor userBackgroundPriority; + resume! - - handlerProcess resume. - !

1 0

Saving the default image when one starts it for the first time
by tim Rowledge 19 May '18

19 May '18

I've long tried to argue that the release images ought to start up in state where they insist on being saved under a new, user chosen name as the first (or nearly) action. This way the release image (and changes) could be locked, guaranteeing that there is a clean image on-disk for the inevitable "Dang!" moment. This seems to me to apply doubly for the all-in-one systems since those release images are buried so deep in directories that you have to look up to see the Mohorovičić discontinuity far above you. What I claim ought to happen is that the current and very excellent preferences wizard should be extended to open a file saver dialog and insist on the image being saved. We would need to do some directory path parsing to make sure that we actually offer a realistic directory and not the /Users/tim/Documents/Squeak/Squeak-5.0-All-in-One.app/Contents/Resources/Squeak5.0-15113.image path I see on my iMac; I'd posit that find the all-in-one part and chopping just before it might be adequate. I can see one situation where this might cause a problem that needs some thought; when using a script to prepare an image eg the VMMaker build scripts. I see the 'NukePreferenceWizardMorph.st' in the vmmaker build scripts which removed the pref wizard ok, but it still leaves the image being saved over the original. Strictly speaking this is very unlikely to be a serious issue with the vmmaker regime because we almost always load the images into the mv/image working directory, but it's a poor example to have around. Changing to make the 'NukePreferenceWizardMorph.st' save the image under the new name and then dropping the copy in buildspurtrunkvmmakerimage.sh might be nicer? Next question is whether anyone can think of serious problems with making this change? And then, how to implement it cleanly? I *think* it might work to over-ride #delete in PreferenceWizardMorph which is sent when the chance to configure is declined right up-front, or when the user clicks on 'Done'. tim -- tim Rowledge; tim(a)rowledge.org; http://www.rowledge.org/tim Strange OpCodes: CAO: Compare Apples to Oranges

10 13

Squeak board meeting notes 2018-05-16
by Craig Latta 19 May '18

19 May '18

Hi all-- The notes for the 2018-05-16 Squeak board meeting are at: https://tinyurl.com/y7vflrcz (docs.google.com) Thanks for any feedback! Greetings on behalf of your Squeak board of directors, Craig -- Craig Latta Black Page Digital Amsterdam :: San Francisco craig(a)blackpagedigital.com +31 6 2757 7177 (SMS ok) + 1 415 287 3547 (no SMS)

8 7

UK Smalltalk User Group meeting - Monday May 21st
by Giovanni Corriga 18 May '18

18 May '18

The next meeting of the UK Smalltalk User Group Meeting will be on Monday, 21st May. We'll meet at our now usual venue The City Pride ( http://www.cityprideclerkenwell.co.uk/ ) at 7pm. For this month’s meeting, we’ll do a session of bug hunting: fix a few things in our favourite Smalltalk frameworks and tools. If you'd like to join us, you can just show up at the pub. You can also sign up in advance at the meeting's Meetup page: http://www.uksmalltalk.org/2018/05/bug-hunting-monday-may-25th.html . See you there!

1 0

← Newer
1
...
7
8
9
10
11
12
13
...
19
Older →

Jump to page: