March 2016 - Pharo-users - lists.pharo.org

Re: [Pharo-users] BaselineOf's and Loose Dependencies

by Dale Henrichs

On 3/21/16 9:48 AM, Sean P. DeNigris wrote: > Dale Henrichs-3 wrote >> First, congratulations! > Thanks!! > > > Dale Henrichs-3 wrote >> A baseline is associated with a "git version", so there is no need to >> "pin the version down in a specific version" > Yes, but I'm not talking about a version of my main project, but a version > of a dependent project. Right ... but you were also talking about "pinning the version of a dependent project in a version of your main project" .... each git commit is a new version of your main project, the equivalent of a new numbered version in a configuration, so to "pin the version of a dependent project", you simply change the version of the dependent project ... which I did in the example I provided ... > > > Dale Henrichs-3 wrote >> in a "git project" you >> have the option of changing the specific version associated with a >> baseline on "every commit" .... > Again, yes, but what I want to do (and what I understood as best practice in > the past) is to defer the linking to a specific version until I release a > stable version of my main project. Okay ... in a git environment, the master branch of your project is the stable version and you do development on a separate "topic branch" like `dev` or `issue_88`. > That is, while I'm developing, I can be > much looser with my tracking of loose dependencies, e.g. commonly it is > enough to specify that I always want #stable. Of course, symbolic > dependencies are not a good idea for release versions, so I will want to > override the loose dependency with a specific numbered version. As I've said elsewhere I think that a symbolic version scheme based on versions (#release3, #release3.1, #release3.1.2, etc.) is superior to #stable for Monticello-based projects (those using ConfigurationOf) and it your dependent project used this scheme, there would be no need to "override the loose dependency with a specific numbered version". If the dependent project doesn't use this scheme then you are stuck with using "specific numbered versions" ... That means that on branch `dev` you'll change the version of the project to #stable and do your work, when you are preparing to merge your work on `dev` into the master branch, you will choose a "specific numbered version", replace #stable with the the numbered version, test, merge to `master` branch. The next time you start work on a new topic branch, you can choose to use #stable, or simply select the latest numbered version explicitly and develop away ... before merging to master in this scheme, just double check if there is new version availble and if not, you're good to go ... > Unfortunately, while I'm no longer getting the mentioned error, my override > seems to be ignored - `ConfigurationOfBabyPhexample project stableVersion > record loadDirective.` reports that the #stable version of BabyMock (1.3) > will be loaded instead of the version 1.2 that I specified in my > configuration. Did you use the example I provided or did you continue to try to use the original spec? The spec you provided didn't make sense to me (I figured it was copy paste errors)... > > I created an issue in my GitHub project where it might be easier to > reference the code (https://github.com/seandenigris/Baby-Phexample/issues/1) > or we can continue to discuss here. Yeah, it will be much easier to get into specific details on github ... Dale >

March 22, 2016

Re: [Pharo-users] [Pillar] Logo for Pillar

by Ben Coman

On Fri, Mar 18, 2016 at 10:30 PM, Thibault ARLOING <thibault.arloing(a)hotmail.fr> wrote: > Hi, > > > With a mate, we've worked on a logo for Pillar. > > We just want feedbacks on it. I like the style. But as a logo I think it is too thin. When I think of pillar I think of http://tinyurl.com/pillar-images And for a logo, something like... http://www.presentation-process.com/wp-content/uploads/powerpoint-pillar-tu… Attached is a concept I knocked up. Its too detailed for a logo, and the font was plagiarized from http://tinyurl.com/hhc5a3d but may provide inspiration. cheers -ben

March 22, 2016

Re: [Pharo-users] Pharo launcher in homebrew cask

by Damien Pollet

FYI I was also thinking of creating a homebrew tap for pharo (VM etc). Still at the draft stage, unfortunately. On 22 March 2016 at 13:00, Damien Cassou <damien.cassou(a)inria.fr> wrote: > Filip Krikava <krikava(a)gmail.com> writes: > > > I find Pharo launcher incredibly useful and I was surprised not to find > it > > in homebrew cask [1]. Pharo itself is there. I have therefore created a > PR: > > https://github.com/caskroom/homebrew-cask/pull/19956 > > I don't know homebrew-cask but I imagine that it is better to reference > stable files on the web rather than continuously changing files. At > least, that's true for the package manager I package(d) for. If that's > also the case here, I would recommend to use a file from: > > http://files.pharo.org/platform/launcher/blessed/ > > Files there don't change. If you are interested, I can move OS X releases > there the same way I do it for Linux releases. > > -- > Damien Cassou > http://damiencassou.seasidehosting.st > > "Success is the ability to go from one failure to another without > losing enthusiasm." --Winston Churchill > >

March 22, 2016

Re: [Pharo-users] Pharo launcher in homebrew cask

by Damien Cassou

Filip Krikava <krikava(a)gmail.com> writes: > I find Pharo launcher incredibly useful and I was surprised not to find it > in homebrew cask [1]. Pharo itself is there. I have therefore created a PR: > https://github.com/caskroom/homebrew-cask/pull/19956 I don't know homebrew-cask but I imagine that it is better to reference stable files on the web rather than continuously changing files. At least, that's true for the package manager I package(d) for. If that's also the case here, I would recommend to use a file from: http://files.pharo.org/platform/launcher/blessed/ Files there don't change. If you are interested, I can move OS X releases there the same way I do it for Linux releases. -- Damien Cassou http://damiencassou.seasidehosting.st "Success is the ability to go from one failure to another without losing enthusiasm." --Winston Churchill

March 22, 2016

Re: [Pharo-users] What frameworks/libraries are available for creating programming languages with Pharo

by Dimitris Chloupis

yes exactly, something that takes advantage of the existing pharo AST so I dont have to reinvent IDE tools etc. Because in that case I may as well do this without Pharo because then it defeats the purpose of taking advantage of all this cool Pharo functionality and would be a ton extra effort. The way I see it , because Pharo syntax is so minimal should be relative easy to implement a language on top that its syntax is nothing more than Pharo objects. Looks different but underneath its plain old Smalltalk. This is why I mentioned Lisp. On Tue, Mar 22, 2016 at 11:11 AM ClÃ©ment Bera <bera.clement(a)gmail.com> wrote: > I think in addition of the parser/compilation chain, a language is about > having a good IDE. > > The Pharo IDE depends more and more on the AST and not on the source code, > so if your new language uses an AST polymorphic with the Smalltalk AST I > guess you can have IDE tools for free (at least partly). > > At least you can consider frameworks to build code browser such as Glamour > as an available library to create a programming language. > > In addition, Omnibrower, which is the default code browser in Pharo 1.4, > can browse the code of any language. You provide Omnibrowser a language > model and it browses your code based on your model. > > > > 2016-03-22 9:48 GMT+01:00 Dimitris Chloupis <kilon.alios(a)gmail.com>: > >> If lisp is ideal for creating new languages so is Smalltalk so I was >> wondering what kind of libraries there are out there for helping with the >> creation of new languages in Pharo. I know about SmaCC and PettitParser , >> what else is out there ? >> > >

March 22, 2016

Re: [Pharo-users] FLMethodChanged: Materialization error. Method DmQuartalsplan>>#initialize: changed its bytecodes.

by Johannes Brauer

Hi Mariano, i tried it with fuelReplacement but I get the exception: FLObjectNotFound: Serialization error. Unexpected reference to Quartalsplan: 1-1 in the graph. This usually happens when the graph changes during serialization Joh. Am 21.03.2016 um 20:10 schrieb Mariano Martinez Peck <marianopeck(a)gmail.com<mailto:marianopeck@gmail.com>>: On Mon, Mar 21, 2016 at 3:51 PM, Johannes Brauer <brauer(a)nordakademie.de<mailto:brauer@nordakademie.de>> wrote: Am 21.03.2016 um 19:19 schrieb Mariano Martinez Peck <marianopeck(a)gmail.com<mailto:marianopeck@gmail.com>>: HI Johannes, Could I see the source of the DmQuartalsplan>>#initialize: ? yes, of course: 'From Pharo4.0 of 18 March 2013 [Latest update: #40626] on 21 March 2016 at 6:19:38.568095 pm'! !DmQuartalsplan methodsFor: 'initialize-release' stamp: ' 15/1/14 15:24'! initialize: einQuartal self lehrveranstaltungen: (SortedCollection sortBlock: [:x :y | x modul modulNummer <= y modul modulNummer]). self quartal: einQuartal! ! I am not sure if the following will work. Please backup your image file and all your fuel files. But...what if you try to implement: DmQuartalsplan >> fuelReplacement | copy | copy := self copy. copy lehrveranstaltungen: copy lehrveranstaltungen asOrderedCollection. ^ copy And finally, at the image where you are materializing, you could implement: DmQuartalsplan >> fuelAfterMaterialization self lehrveranstaltungen: (self lehrveranstaltungen asSortedCollection: [:x :y | x modul modulNummer <= y modul modulNummer]) Once the migration is done, you can discard #fuelReplacement and #fuelAfterMaterialization if you want. I never tried this approach, so I am not sure ;) Joh. I think I workaround may be to do a Fuel replacement to replace the sortedCollection with an OrderedCollection and then a post initialize method. But again, probably seeing if the closure of the sorted collection is clean or not may have different behaviors. Cheers, On Mon, Mar 21, 2016 at 3:09 PM, Johannes Brauer <brauer(a)nordakademie.de<mailto:brauer@nordakademie.de>> wrote: Hi, I am trying to port a pharo2 application to pharo4. Therefore I have to migrate some fuel files from version 1.9 to 1.9.4. I've done this using the script shown in the EnterprisePharo Book in my pharo2 image, no problem. But trying to load the migrated fuel file into my pharo4 image I get the error message shown above. The source code of the initialize methods is the same in both images, of course. But the byte codes differ: Pharo2: 37 <70> self 38 <42> pushLit: SortedCollection 39 <8F 02 00 08> closureNumCopied: 0 numArgs: 2 bytes 43 to 50 â¦ Pharo4: 37 <70> self 38 <40> pushLit: SortedCollection 39 <8F 02 00 08> closureNumCopied: 0 â¦ What can I do? Johannes -- Mariano http://marianopeck.wordpress.com<http://marianopeck.wordpress.com/> -- Mariano http://marianopeck.wordpress.com<http://marianopeck.wordpress.com/>

March 22, 2016

Re: [Pharo-users] What frameworks/libraries are available for creating programming languages with Pharo

by ClÃ©ment Bera

I think in addition of the parser/compilation chain, a language is about having a good IDE. The Pharo IDE depends more and more on the AST and not on the source code, so if your new language uses an AST polymorphic with the Smalltalk AST I guess you can have IDE tools for free (at least partly). At least you can consider frameworks to build code browser such as Glamour as an available library to create a programming language. In addition, Omnibrower, which is the default code browser in Pharo 1.4, can browse the code of any language. You provide Omnibrowser a language model and it browses your code based on your model. 2016-03-22 9:48 GMT+01:00 Dimitris Chloupis <kilon.alios(a)gmail.com>: > If lisp is ideal for creating new languages so is Smalltalk so I was > wondering what kind of libraries there are out there for helping with the > creation of new languages in Pharo. I know about SmaCC and PettitParser , > what else is out there ? >

March 22, 2016

What frameworks/libraries are available for creating programming languages with Pharo

by Dimitris Chloupis

If lisp is ideal for creating new languages so is Smalltalk so I was wondering what kind of libraries there are out there for helping with the creation of new languages in Pharo. I know about SmaCC and PettitParser , what else is out there ?

March 22, 2016

Re: [Pharo-users] Interleaving strings

by HernÃ¡n Morales Durand

2016-03-21 17:43 GMT-03:00 Nicolai Hess <nicolaihess(a)gmail.com>: > What is your implementation with with:collection: ? > > I know this function from python, they call it "zip" > > we could implement this in SequenceableCollection > > SequenceableCollection>>zip: aCollection > | index | > index := 0. > self size > aCollection size ifTrue:[ ^ aCollection zip: self]. > ^ self class new: (aCollection size + self size) streamContents: > [:stream | > aCollection > do: [:each | index:=index + 1. stream nextPut: each ] > separatedBy: [(index > self size) ifFalse:[stream nextPut: > (self at:index) ]]] > > Another example, using a generator (I didn't used Generator before, but I > like the > idea to create an endless stream from a string, with some fill value > (empty string in this case)). > > |gen| > gen := Generator on:[:g | '123' do:[:k | g yield: k asString ] . [true] > whileTrue:[g yield:'']]. > String streamContents:[:stream | > 'abcdefg' do:[:c | stream nextPut: c] > separatedBy:[ stream nextPutAll:gen next]] > > > I use those all the time (filling matrices etc). It would be nice to have something like these: '2' repeatJoin: 4 ==> '2222' 2 repeatJoin: 4 ==> 2222 '2' repeat: 4 ==> #('2' '2' '2' '2') 2 repeat: 4 ==> #(2 2 2 2) What do you think? HernÃ¡n

March 22, 2016

Re: [Pharo-users] Interleaving strings

by Nicolai Hess

What is your implementation with with:collection: ? I know this function from python, they call it "zip" we could implement this in SequenceableCollection SequenceableCollection>>zip: aCollection | index | index := 0. self size > aCollection size ifTrue:[ ^ aCollection zip: self]. ^ self class new: (aCollection size + self size) streamContents: [:stream | aCollection do: [:each | index:=index + 1. stream nextPut: each ] separatedBy: [(index > self size) ifFalse:[stream nextPut: (self at:index) ]]] Another example, using a generator (I didn't used Generator before, but I like the idea to create an endless stream from a string, with some fill value (empty string in this case)). |gen| gen := Generator on:[:g | '123' do:[:k | g yield: k asString ] . [true] whileTrue:[g yield:'']]. String streamContents:[:stream | 'abcdefg' do:[:c | stream nextPut: c] separatedBy:[ stream nextPutAll:gen next]] 2016-03-15 2:09 GMT+01:00 Andy Burnett <andy.burnett(a)knowinnovation.com>: > I have two strings, which are typically of different lengths, and I want > to interleave them, i.e.: > > 'abcdefg' interleaveWith: '123' > > Would produce > > 'a1b2c3defg' > > I know how to write this procedurally, but I have been very impressed by > what Sven shows in his Medium articles, so I wondered if there is any > existing mechanism that would allow me to solve this elegantly? > with:collection: almost does it, but the collections have to be the same > size. > > Cheers > Andy >

March 21, 2016