- Pharo-dev - lists.pharo.org

Re: [Pharo-project] Feedback about SmallLint

by Mariano Martinez Peck

On Sat, Apr 21, 2012 at 5:01 PM, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr > wrote: > thanks for the report. > Mariano so far we did not change any of the rules provided by RB. For now > we will run small lint and collect information. > > Ok, I understand. But since I am SmallLintying my packages I guess that my feedback does not hurt for future work (when you do start changing rules). Some more feedback: 6) RBAbstractClassRule is confusing. *"Checks for references to classes that have subclassResponsibility methods. Such references might be creating instances of the abstract class or more commonly being used as the argument to an isKindOf: message which is considered bad style."* In my case, it found FLBenchmarkRunner because it has INSTANCE side messages with subclassResponsability. However, all references to the class FLBenchmarkRunner send class side messages to it. So.... maybe It can exclude class side messages that do not end up sending #new or #basicNew? 7) how does the Manifiesto class envolves while the real packages are changed (class and methods are renamed or removed) ? If a method class is removed then I guess it is easy to remove the metadada associated to them. But what about if a method change? you cannot be sure if the rule exclusion still make sense or not. So, I guess you should remove the metadata as well ? 8) RBMethodHasNoTimeStampRule seems to have a problem with Traits because it shows me lots of methods without timestamp and they all result to come from a trait. 9) Sometimes, clicking on a method raises an error. Find attached PhaorDebug.log 10) RBSubclassResponsibilityNotDefinedRule could show me WHICH are the leaf subclasses that do NOT implement the method ... otherwise it is complicated to find it. 11) I don't like storing the Manifiest class in the default/main package. I changed ManifiestFuel to be in Fuel-SmallLint rather than Fuel but then, as soon as I run the rules again, it moves back the class to Fuel. Can't it check if the class already exist in Smalltalk globals and if true just use it and don't move it ? 12) It would be nice to group rules by criteria. I saw you have #group in each rule so I guess it is a matter of time to have the UI aware of that, right? Cheers > Stef > > > > > Hi guys. After our meeting at Lille, I did a pass of SmallLint over Fuel > project. I have to say that I enjoyed the tool and that. I love the > manifiesto idea. I love to store metada and to be able to reject rules and > scenarios to avoid having false positives again in the future. In addition, > it looks like I had less false positive than with other tools. However, > there were a couple of rules that I don't understand or may suggest > something: > > > > 1) First, apart form showing the package between parenthesis, I would > also show the protocol (category). Because some rules play with that, so it > is handy if you can directly see them there, otherwise you need to browse > the code. > > > > 2) RBContainsRule is not clear for me. " 'Checks for the common code > fragment: "(aCollection detect: [:each | ''some condition''] ifNone: [nil]) > ~= nil". contains: can simplify this code to "aCollection contains: [:each > | ''some condition'']". Not only is the contains: variant shorter, it > better signifies what the code is doing.'" > > That is not very true. #contains answers true or false, while > detect:ifNone:[] answers whatever, and in your example it is nil. So, the > answer will be different and hence it is not the same. Ok, in your example, > you compare after with == nil... but... I don't really see a use case. For > example, in my case it detected: > > > > classNamed: className > > > > ^ (migrations > > detect: [:m | m sourceClassName = className ] > > ifNone: [ ^ self globalClassNamed: className ]) > > targetClass. > > > > > > 3) RBLawOfDemeterRule: maybe you can exclude string concatenation? (#,) > because it detected a lot of messages like this: > > signalWith: aGlobalName and: aSelector > > ^ self signal: 'Method ', aGlobalName, '>>#', aSelector, ' not > found.' > > > > Also, what about excluding messages like #asXXX. For example, here it > detected "self signature asByteArray". Is that completly breaking the law > of demeter? mmm I don't think so. > > > > verifySignatureFrom: aDecoder > > > > | streamSignature | > > streamSignature := ByteArray new: self signature size. > > aDecoder nextEncodedBytesInto: streamSignature. > > (self signature asByteArray = streamSignature) ifFalse: [ > > FLBadSignature > > signalCurrentSignature: self signature > > streamSignature: streamSignature ]. > > > > > > 4) RBSentNotImplementedRule: I didn't understand why it has selected: > > > > largeIdentityHash > > > > <primitive: 75> > > self primitiveFailed > > > > both messages are implemented :( > > > > > > 5) RBClassInstVarNotInitializedRule could be smarter and search not only > for #initialize but for #initializeXXX. Lots of times, I initialize my > objects with data, using a particular nitialize method, such as: > > initializeWithBehaviors: someBehaviors extensionMethods: > someExtensionMethods > > so the rule says there is no initialize...but I have that one ;) > > > > > > That's all the feedback so far. It helped me to find some things to > improve so thanks! > > > > > > -- > > Mariano > > http://marianopeck.wordpress.com > > > > > -- Mariano http://marianopeck.wordpress.com

April 23, 2012

Re: [Pharo-project] ECompletion/Shout bug

by Esteban Lorenzano

yes, I can reproduce this one... taking care of it now :) On Apr 23, 2012, at 10:11 AM, StÃ©phane Ducasse wrote: > Note that the bug shows up if you edit the code in the result of the implementors. > Esteban can you reproduce the bug? > > Stef > > On Apr 23, 2012, at 9:18 AM, StÃ©phane Ducasse wrote: > >> Hi >> >> I get a bug with OCompletion/Ecompletion >> >> TextEditor>>changeEmphasis >> >> try to type >> >> menuList nextPut: >> >> and you get a bug. >> >> Now in addition I do not understand what I see in the debugger: >> >> As you see model returns a debugger and the debugger title show FlatMessageListBrowser >> >> Stef >> >> >> >> <Screen Shot 2012-04-23 at 9.08.45 AM.pdf> > >

April 23, 2012

Re: [Pharo-project] completion lets some traces

by Esteban Lorenzano

with ECompletion? No, I'm not having a fade out (I do have it with OCompletion, but that's the intended behavior) Esteban On Apr 22, 2012, at 10:44 PM, StÃ©phane Ducasse wrote: > Esteban > > don't you have the same problem than me with ecompletion? > I get a slow vanishing of propositions. > > Stef >

April 23, 2012

Re: [Pharo-project] ECompletion/Shout bug

by StÃ©phane Ducasse

Note that the bug shows up if you edit the code in the result of the implementors. Esteban can you reproduce the bug? Stef On Apr 23, 2012, at 9:18 AM, StÃ©phane Ducasse wrote: > Hi > > I get a bug with OCompletion/Ecompletion > > TextEditor>>changeEmphasis > > try to type > > menuList nextPut: > > and you get a bug. > > Now in addition I do not understand what I see in the debugger: > > As you see model returns a debugger and the debugger title show FlatMessageListBrowser > > Stef > > > > <Screen Shot 2012-04-23 at 9.08.45 AM.pdf>

April 23, 2012

Re: [Pharo-project] Feedback about SmallLint

by Simon Allier

Thanks for the feedback On Apr 21, 2012, at 2:00 PM, Mariano Martinez Peck wrote: > Hi guys. After our meeting at Lille, I did a pass of SmallLint over Fuel project. I have to say that I enjoyed the tool and that. I love the manifiesto idea. I love to store metada and to be able to reject rules and scenarios to avoid having false positives again in the future. In addition, it looks like I had less false positive than with other tools. However, there were a couple of rules that I don't understand or may suggest something: > > 1) First, apart form showing the package between parenthesis, I would also show the protocol (category). Because some rules play with that, so it is handy if you can directly see them there, otherwise you need to browse the code. > > 2) RBContainsRule is not clear for me. " 'Checks for the common code fragment: "(aCollection detect: [:each | ''some condition''] ifNone: [nil]) ~= nil". contains: can simplify this code to "aCollection contains: [:each | ''some condition'']". Not only is the contains: variant shorter, it better signifies what the code is doing.'" > That is not very true. #contains answers true or false, while detect:ifNone:[] answers whatever, and in your example it is nil. So, the answer will be different and hence it is not the same. Ok, in your example, you compare after with == nil... but... I don't really see a use case. For example, in my case it detected: > > classNamed: className > > ^ (migrations > detect: [:m | m sourceClassName = className ] > ifNone: [ ^ self globalClassNamed: className ]) > targetClass. > > > 3) RBLawOfDemeterRule: maybe you can exclude string concatenation? (#,) because it detected a lot of messages like this: > signalWith: aGlobalName and: aSelector > ^ self signal: 'Method ', aGlobalName, '>>#', aSelector, ' not found.' > > Also, what about excluding messages like #asXXX. For example, here it detected "self signature asByteArray". Is that completly breaking the law of demeter? mmm I don't think so. > > verifySignatureFrom: aDecoder > > | streamSignature | > streamSignature := ByteArray new: self signature size. > aDecoder nextEncodedBytesInto: streamSignature. > (self signature asByteArray = streamSignature) ifFalse: [ > FLBadSignature > signalCurrentSignature: self signature > streamSignature: streamSignature ]. > > > 4) RBSentNotImplementedRule: I didn't understand why it has selected: > > largeIdentityHash > > <primitive: 75> > self primitiveFailed > > both messages are implemented :( > > > 5) RBClassInstVarNotInitializedRule could be smarter and search not only for #initialize but for #initializeXXX. Lots of times, I initialize my objects with data, using a particular nitialize method, such as: > initializeWithBehaviors: someBehaviors extensionMethods: someExtensionMethods > so the rule says there is no initialize...but I have that one ;) > > > That's all the feedback so far. It helped me to find some things to improve so thanks! > > > -- > Mariano > http://marianopeck.wordpress.com >

April 23, 2012

Re: [Pharo-project] Redline Smalltalk ...

by StÃ©phane Ducasse

You mean in the redline mailing-list? Stef >> Redline Smalltalk has undergone some Big changes, maybe now is a good time >> for you to try it? http://redline.st > > Could you outline some of those changes? > > Cheers > Philippe > > >

April 23, 2012

Re: [Pharo-project] [ANN] Styled Text Editor for Cuis 4.0 Smalltalk

by StÃ©phane Ducasse

More precisely the model was changed (and need more effort) so that we can use the logic of systems like opengl, cairoâ¦ We are waiting that alain defends his HDR to help us adapting all the textMorph to the new model now if somebody want to have a look let us know. Stef On Apr 21, 2012, at 11:26 PM, StÃ©phane Ducasse wrote: > note that igor started to rewrite a part of paragraph editor or textmorph because Athens needs a good text model. > > Stef > On Apr 21, 2012, at 10:22 PM, phil(a)highoctane.be wrote: > >> Yeah VM works on iOS. >> >> My use case is private, nothing fancy. I just value having decent TextEditor and friends. >> >> For an OO thing like Smalltalk Pharo, it would be a shame to now have a top notch UI kit for editing text. >> >> I hope to be able to provide several decent samples so that newcomers don't have to go through the same pain as I go through now. >> >> Phil >> >> 2012/4/21 Bernhard Pieber <bernhard(a)pieber.com> >> Thanks for the encouragement! >> >> The necessary funding depends on the effort of course, and this in turn depends on the approach one uses for porting. We have an idea how to do the port to Pharo as an external package. Do you have a concrete use case in mind? >> >> Cheers, >> Bernhard >> >> P.S. Did you succeed building the Mac VM on Xcode 4.3 yet? >> >> Am 21.04.2012 um 21:17 schrieb "phil(a)highoctane.be" <phil(a)highoctane.be>: >> >>> Great news! >>> >>> I am currently busy looking after Text, TextAttributes, TextConstants, Paragraph... and god, it went through the grinder and it shows. >>> >>> I understand the reasons that led to such things but Collections-Text is in need of improvement for sure. >>> >>> How much funding would be necessary to get this port done? >>> >>> Phil >>> >>> 2012/4/21 Bernhard Pieber <bernhard(a)pieber.com> >>> Dear Smalltalkers, >>> >>> I am very happy to announce that the Styled Text Editor for the brand new Cuis 4.0 is now available on GitHub [1]. The Styled Text Editor was first presented by me at last year's ESUG in Edinburgh [2]. Thanks to ESUG the presentation was recorded [3]. >>> >>> The Styled Text Editor is a framework for rich text editing using styles as known from popular word processors like Apple Pages or Microsoft Word. It features paragraph and character styles, allowing easy text formatting using styles only. It is intended for applications where users need to work with good looking rich text in a simple and fast way. >>> >>> Rich text commonly refers to text with formatting information like different fonts, sizes, alignments, and emphasis. To make editing as simple as possible instead of individual formatting information end users apply styles to parts of the text. Editing is made fast by keyboard shortcuts for text navigation, selection, and styles selection. Where possible the feel of widely used rich text editors is supported. >>> >>> It includes many features like >>> - numbered and bulleted list paragraph styles >>> - support for images >>> - multiple level undo and redo >>> - text completion using various glossaries including a dictionary of English with about 166.000 words >>> - RTF clipboard (on OS X VMs with the ClipboardExtendedPlugin) >>> >>> The Styled Text Editor is developed in Cuis with the plan to eventually port it to Squeak and Pharo. The development of the Styled Text Editor inspired many changes to Cuis itself, and It is the first package to use the brand new DVCS based development process for external packages for Cuis 4.0. >>> >>> The idea and funding was provided by me, Bernhard Pieber and my company Software Generation. The implementation was done by Juan Vuletich, the mastermind behind Cuis. Thanks Juan for the close cooperation. It was and still is great fun to work with you. >>> >>> We are looking forward to feedback from you. Fork it on GitHub [1], create issues and send us pull requests. ;-) >>> >>> Cheers, >>> Bernhard Pieber >>> >>> [1] https://github.com/bpieber/Cuis-StyledTextEditor >>> [2] http://www.esug.org/wiki/pier/Conferences/2011/Schedule-And-Talks/StyledTex… >>> [3] http://www.youtube.com/watch?v=pUoVbvwspi8&list=PL813665D04A2E4C0A&index=7&… >>> >>> >>> >>> -- >>> Philippe Back >>> "Helping you hit the top 3 outcomes you really want to achieve" >>> >>> Mob: +32(0) 478 650 140 | Fax: +32 (0) 70 408 027 Mail: phil(a)highoctane.be | Web: http://philippeback.eu | Blog: >>> http://philippeback.be >>> >>> High Octane SPRL >>> rue cour Boisacq 101 >>> 1301 Bierges >>> >> >> >> >> -- >> Philippe Back >> "Helping you hit the top 3 outcomes you really want to achieve" >> >> Mob: +32(0) 478 650 140 | Fax: +32 (0) 70 408 027 Mail: phil(a)highoctane.be | Web: http://philippeback.eu | Blog: >> http://philippeback.be >> >> High Octane SPRL >> rue cour Boisacq 101 >> 1301 Bierges >> > >

April 23, 2012

[Pharo-project] ECompletion/Shout bug

by StÃ©phane Ducasse

Hi I get a bug with OCompletion/Ecompletion TextEditor>>changeEmphasis try to type menuList nextPut: and you get a bug. Now in addition I do not understand what I see in the debugger: As you see model returns a debugger and the debugger title show FlatMessageListBrowser Stef

April 23, 2012

Re: [Pharo-project] Redline Smalltalk ...

by Philippe Marschall

On 04/23/2012 08:05 AM, James Ladd wrote: > > Redline Smalltalk has undergone some Big changes, maybe now is a good time > for you to try it? http://redline.st Could you outline some of those changes? Cheers Philippe

April 23, 2012

[Pharo-project] Redline Smalltalk ...

by James Ladd

Redline Smalltalk has undergone some Big changes, maybe now is a good time for you to try it? http://redline.st

April 23, 2012