August 2021 - Pharo-users - lists.pharo.org

Re: Private Methods

by Jesus Mari Aguirre

Please keep Pharo simple, why do you need private methods...you can include then in a protocol named private...other language have it...yes...next addition will be namespaces...other...I don't know...at last we have Java El jue., 19 ago. 2021 9:00, Richard O'Keefe <raoknz(a)gmail.com> escribiÃ³: > Many years ago there was a proposal in the Squeak mailing list about > enforcing > a naming convention, "pvt", I implemented that in my Smalltalk system. > The > compiler enforces the rule that pvt.[A-Z].* message can only be sent to > (self|super) (basicNew|basicNew: n|new|new: n|pvtNew: n)? > in a class method or > (self|super) ((class (new|new: n)|pvtSpeciesNew: n|pvtClone)? > in an instance method. > There are currently > 9412 public selectors > 793 pvt* selectors and > 23 private* selectors, > where the last group is methods that I *want* to be private in some sense > but > cannot do with this machinery. (For example, calling a "private" method on > another object known to be of the same class.) > > I think the evidence shows that this works well enough to be useful, even > if it > isn't quite as expressive as I'd like. And what *that* means is that > this can be > done with a style check, using the machinery Pharo already has for style > checks. > > > > On Wed, 18 Aug 2021 at 08:14, Craig Johnson <craig(a)hivemind.net> wrote: > > > > Hi All, > > > > > > Just a newb off-the-wall question. > > > > Is there any good reason why we can't create a true private method in a > > Pharo class by putting that method inside an instance or class variable > > as a lambda (block). > > > > > > This would reduce one of my biggest bugbears with Pharo, namely the > > pollution of the global namespace with every single message name in the > > entire system. > > > > > > > > Craig >

Aug. 19, 2021

Re: Private Methods

by Richard O'Keefe

Many years ago there was a proposal in the Squeak mailing list about enforcing a naming convention, "pvt", I implemented that in my Smalltalk system. The compiler enforces the rule that pvt.[A-Z].* message can only be sent to (self|super) (basicNew|basicNew: n|new|new: n|pvtNew: n)? in a class method or (self|super) ((class (new|new: n)|pvtSpeciesNew: n|pvtClone)? in an instance method. There are currently 9412 public selectors 793 pvt* selectors and 23 private* selectors, where the last group is methods that I *want* to be private in some sense but cannot do with this machinery. (For example, calling a "private" method on another object known to be of the same class.) I think the evidence shows that this works well enough to be useful, even if it isn't quite as expressive as I'd like. And what *that* means is that this can be done with a style check, using the machinery Pharo already has for style checks. On Wed, 18 Aug 2021 at 08:14, Craig Johnson <craig(a)hivemind.net> wrote: > > Hi All, > > > Just a newb off-the-wall question. > > Is there any good reason why we can't create a true private method in a > Pharo class by putting that method inside an instance or class variable > as a lambda (block). > > > This would reduce one of my biggest bugbears with Pharo, namely the > pollution of the global namespace with every single message name in the > entire system. > > > > Craig

Aug. 19, 2021

Re: Private Methods

by David Pennington

I found this useful when creating commercial frameworks for VAST. For instance, in the printing framework, I used the idea of a turtle (from turtlegraphics) to create pages. All the turtle commands - turn(50) etc - were public but the code that actually created the printing code was all private. Hence purchases knew not to touch the private methods. David > On 18 Aug 2021, at 14:31, Esteban Maringolo <emaringolo(a)gmail.com> wrote: > > Hi Noury, > > In Dolphin Smalltalk and in IBM/VA Smalltalk, methods can be private, > it's a special attribute of the CompiledMethod, so the visibility > doesn't depend on a category name convention, and you can have both > public and private methods sharing a single category. I used that > attribute extensively in the previously described dialects, and it led > to good method categorization, in the case of Dolphin, it had the nice > thing of adding "Private - ..." to the method comment when you toggle > a method as private. > > I'm fine with the current state of everything being public, but if we > were to have that public/private distinction that would be the > simplest solution, having a <private> pragma could also work, but I > think that in terms of performance it could impact the lookup. I don't > know, however, if that is what the original request was after. > > Best regards, > > Esteban A. Maringolo > > On Wed, Aug 18, 2021 at 9:47 AM Noury Bouraqadi <bouraqadi(a)gmail.com> wrote: >> >> Hi Esteban, >> >> Which first "class attribute" are you referring to? Is it a pragma? Can you provide an example please ? >> >> Thanx >> Noury >> >> On Aug 17 2021, at 10:31 pm, Esteban Maringolo <emaringolo(a)gmail.com> wrote: >> >> I don't understand the request? >> >> Wouldn't tagging methods as private (as a first class attribute, and >> not only its category) be enough? >> >> What you ask for seems to be a splitted method dictionary of some sort. >> >> Having first class private methods would save you from autocomplete >> selector "pollution" of private methods. >> >> Regards! >> >> Esteban A. Maringolo >> >> On Tue, Aug 17, 2021 at 5:14 PM Craig Johnson <craig(a)hivemind.net> wrote: >>> >>> Hi All, >>> >>> >>> Just a newb off-the-wall question. >>> >>> Is there any good reason why we can't create a true private method in a >>> Pharo class by putting that method inside an instance or class variable >>> as a lambda (block). >>> >>> >>> This would reduce one of my biggest bugbears with Pharo, namely the >>> pollution of the global namespace with every single message name in the >>> entire system. >>> >>> >>> >>> Craig

Aug. 18, 2021

Re: Private Methods

by Gabriel Cotelli

As Esteban pointed out there are some dialects with proper support for that. But the semantics is the same as in Pharo, you can still send these messages. It is more a warning to the programmer, and the browser shows you these methods with a particular icon or has tabs so you can easily filter out the methods marked as private. A least in VA is done by attaching some metadata to the methods and not changing its source code. On Wed, Aug 18, 2021 at 10:32 AM Esteban Maringolo <emaringolo(a)gmail.com> wrote: > Hi Noury, > > In Dolphin Smalltalk and in IBM/VA Smalltalk, methods can be private, > it's a special attribute of the CompiledMethod, so the visibility > doesn't depend on a category name convention, and you can have both > public and private methods sharing a single category. I used that > attribute extensively in the previously described dialects, and it led > to good method categorization, in the case of Dolphin, it had the nice > thing of adding "Private - ..." to the method comment when you toggle > a method as private. > > I'm fine with the current state of everything being public, but if we > were to have that public/private distinction that would be the > simplest solution, having a <private> pragma could also work, but I > think that in terms of performance it could impact the lookup. I don't > know, however, if that is what the original request was after. > > Best regards, > > Esteban A. Maringolo > > On Wed, Aug 18, 2021 at 9:47 AM Noury Bouraqadi <bouraqadi(a)gmail.com> > wrote: > > > > Hi Esteban, > > > > Which first "class attribute" are you referring to? Is it a pragma? Can > you provide an example please ? > > > > Thanx > > Noury > > > > On Aug 17 2021, at 10:31 pm, Esteban Maringolo <emaringolo(a)gmail.com> > wrote: > > > > I don't understand the request? > > > > Wouldn't tagging methods as private (as a first class attribute, and > > not only its category) be enough? > > > > What you ask for seems to be a splitted method dictionary of some sort. > > > > Having first class private methods would save you from autocomplete > > selector "pollution" of private methods. > > > > Regards! > > > > Esteban A. Maringolo > > > > On Tue, Aug 17, 2021 at 5:14 PM Craig Johnson <craig(a)hivemind.net> > wrote: > > > > > > Hi All, > > > > > > > > > Just a newb off-the-wall question. > > > > > > Is there any good reason why we can't create a true private method in a > > > Pharo class by putting that method inside an instance or class variable > > > as a lambda (block). > > > > > > > > > This would reduce one of my biggest bugbears with Pharo, namely the > > > pollution of the global namespace with every single message name in the > > > entire system. > > > > > > > > > > > > Craig >

Aug. 18, 2021

Re: Private Methods

by Esteban Maringolo

Hi Noury, In Dolphin Smalltalk and in IBM/VA Smalltalk, methods can be private, it's a special attribute of the CompiledMethod, so the visibility doesn't depend on a category name convention, and you can have both public and private methods sharing a single category. I used that attribute extensively in the previously described dialects, and it led to good method categorization, in the case of Dolphin, it had the nice thing of adding "Private - ..." to the method comment when you toggle a method as private. I'm fine with the current state of everything being public, but if we were to have that public/private distinction that would be the simplest solution, having a <private> pragma could also work, but I think that in terms of performance it could impact the lookup. I don't know, however, if that is what the original request was after. Best regards, Esteban A. Maringolo On Wed, Aug 18, 2021 at 9:47 AM Noury Bouraqadi <bouraqadi(a)gmail.com> wrote: > > Hi Esteban, > > Which first "class attribute" are you referring to? Is it a pragma? Can you provide an example please ? > > Thanx > Noury > > On Aug 17 2021, at 10:31 pm, Esteban Maringolo <emaringolo(a)gmail.com> wrote: > > I don't understand the request? > > Wouldn't tagging methods as private (as a first class attribute, and > not only its category) be enough? > > What you ask for seems to be a splitted method dictionary of some sort. > > Having first class private methods would save you from autocomplete > selector "pollution" of private methods. > > Regards! > > Esteban A. Maringolo > > On Tue, Aug 17, 2021 at 5:14 PM Craig Johnson <craig(a)hivemind.net> wrote: > > > > Hi All, > > > > > > Just a newb off-the-wall question. > > > > Is there any good reason why we can't create a true private method in a > > Pharo class by putting that method inside an instance or class variable > > as a lambda (block). > > > > > > This would reduce one of my biggest bugbears with Pharo, namely the > > pollution of the global namespace with every single message name in the > > entire system. > > > > > > > > Craig

Aug. 18, 2021

Re: Private Methods

by Noury Bouraqadi

Hi Esteban, Which first "class attribute" are you referring to? Is it a pragma? Can you provide an example please ? Thanx Noury On Aug 17 2021, at 10:31 pm, Esteban Maringolo <emaringolo(a)gmail.com> wrote: > I don't understand the request? > > Wouldn't tagging methods as private (as a first class attribute, and > not only its category) be enough? > > What you ask for seems to be a splitted method dictionary of some sort. > Having first class private methods would save you from autocomplete > selector "pollution" of private methods. > > Regards! > Esteban A. Maringolo > On Tue, Aug 17, 2021 at 5:14 PM Craig Johnson <craig(a)hivemind.net> wrote: > > > > Hi All, > > > > > > Just a newb off-the-wall question. > > > > Is there any good reason why we can't create a true private method in a > > Pharo class by putting that method inside an instance or class variable > > as a lambda (block). > > > > > > This would reduce one of my biggest bugbears with Pharo, namely the > > pollution of the global namespace with every single message name in the > > entire system. > > > > > > > > Craig >

Aug. 18, 2021

Re: Private Methods

by Esteban Maringolo

I don't understand the request? Wouldn't tagging methods as private (as a first class attribute, and not only its category) be enough? What you ask for seems to be a splitted method dictionary of some sort. Having first class private methods would save you from autocomplete selector "pollution" of private methods. Regards! Esteban A. Maringolo On Tue, Aug 17, 2021 at 5:14 PM Craig Johnson <craig(a)hivemind.net> wrote: > > Hi All, > > > Just a newb off-the-wall question. > > Is there any good reason why we can't create a true private method in a > Pharo class by putting that method inside an instance or class variable > as a lambda (block). > > > This would reduce one of my biggest bugbears with Pharo, namely the > pollution of the global namespace with every single message name in the > entire system. > > > > Craig

Aug. 17, 2021

Private Methods

by Craig Johnson

Hi All, Just a newb off-the-wall question. Is there any good reason why we can't create a true private method in a Pharo class by putting that method inside an instance or class variable as a lambda (block). This would reduce one of my biggest bugbears with Pharo, namely the pollution of the global namespace with every single message name in the entire system. Craig

Aug. 17, 2021

UK Smalltalk User Group Meeting - Wednesday, August 25th

by Giovanni Corriga

The next meeting of the UK Smalltalk User Group will be on Wednesday, August 25th. Marcus Denker will talk about Variables in Pharo. We like to say that âEverything is an Objectâ in Smalltalk. This is true in many cases: Classes, methods, even the execution stack are reflectively available as objects. This talk shows how this idea can be extended to Variables and how Pharo implements first-class Variables for Globals, instance Variables, Class Variables, and even temporary variables. This presentation explores the Variable hierarchy, shows how variables simplify the compiler and how the reflective API provided by variables is used by the debugger. In a hands-on tutorial, we extend the language by defining new kinds of Variables. Marcus ( https://marcusdenker.de/ ) is a permanent researcher at INRIA Lille - Nord Europe ( http://www.inria.fr/lille/ ). Before, he was a postdoc at the PLEIAD lab/DCC University of Chile and the Software Composition Group, University of Bern. His research focuses on reflection and meta-programming for dynamic languages. Marcus Denker received a PhD in Computer Science from the University of Bern/Switzerland in 2008 and a Dipl.-Inform. (MSc) from the University of Karlsruhe/Germany in 2004. He co-founded Zweidenker GmbH ( https://zweidenker.de/ ) in 2009. Given the current COVID-19 restrictions, this will be an online meeting from home. If you'd like to join us, please sign up in advance on the meeting's Meetup page ( https://www.meetup.com/UKSTUG/events/cbklbrycclbhc/ ) to receive the meeting details. Donât forget to bring your laptop and drinks!

Aug. 16, 2021

Re: File-in Issue - Pharo 9

by David Pennington

Aug. 16, 2021

Re: Exercism v3 problems

by Roelof Wobben

I understand. Hopefully someone can and is willing to help me and we can give the user the experience he/she needs. I think that what you have said earlier, could be very in hand. That Syunit could be working standalone. > Op 15 augustus 2021 om 23:16 schreef "stephane.ducasse(a)free.fr" <stephane.ducasse(a)free.fr>: > > > Thanks for the info. > > I read a bit the âspecâ > I do not really get the test method cannot be passed? Anywayâ¦. > > > This is the body of the command that is being tested. For example, the following Ruby test: > > def test_duplicate_items_uniqs_list > cart = ShoppingCart.new > cart.add(:STARIC) > cart.add(:MEDNEW) > cart.add(:MEDNEW) > assert_equal 'Newspaper, Rice', cart.items_list > end > should return a test_code value of: > > "cart = ShoppingCart.new > cart.add(:STARIC) > cart.add(:MEDNEW) > cart.add(:MEDNEW) > assert_equal 'Newspaper, Rice', cart.items_list" > Looking at the other implementation can help. > > https://github.com/exercism?q=-test-runner > > Now my plate is more than full. So I cannot really help because I have too many things unfinished and that should be done. > > S > > > > On 15 Aug 2021, at 21:02, Roelof Wobben <r.wobben(a)home.nl> wrote: > > > > Yep, > > > > The spec and other information can be found here : https://github.com/exercism/docs/blob/main/building/tooling/test-runners/in… <https://github.com/exercism/docs/blob/main/building/tooling/test-runners/in…> > > > > I hoping we can get number 2 or 3 working. > > > > Roelof > > > > > > > > Op 15-8-2021 om 15:15 schreef stephane.ducasse(a)free.fr <mailto:stephane.ducasse@free.fr>: > >> > >>> Yes, > >>> > >>> There are already some test which are written with Sunit from v2 challenges. > >>> and for the new one I could write the tests. > >>> > >>> But what I mean is that exercismv 3 wants that the code written by a student is submitted and that then the tests are run and a json file with if a test is successfull or with the tests errors is made. > >> > >> We do that for our server (not Json). > >> Do you have a spec of what should be the output? > >> It could be occasion for us to extract a runner and strengthen the logic of the test execution. > >> S > >> > >>> > >>> On v2 the tests were on the same image as the code written by a student. > >>> > >>> Roelof > >>> > >>> > >>> > >>> Op 13-8-2021 om 22:14 schreef stephane.ducasse(a)free.fr <mailto:stephane.ducasse@free.fr>: > >>>> Hi > >>>> > >>>>> As far as I see it standalone means work without a image. > >>>>> all the other test-runners run in a vm as far as I know how exercism works. > >>>> > >>>> Nope. :) > >>>> You see in ruby or python or Javascript, you cannot run a software without a kernel (and a kernek is an image that cannot be saved and it often created > >>>> using some hardcoded C code - but this is the same). They just cannot save the kernel once loaded with libraries but this is the same. > >>>> > >>>> So if we can load tests and execute tests without clicking on a button interactively. > >>>> Yes we can. > >>>> > >>>> This is what we are doing since we have the jenkins test runner (since 2009 or before) and this is what you have when you > >>>> run tests on travis or github actions. > >>>> > >>>> So now the point is to know if their infrastructure needs a script to run the tests > >>>> - how to identify them > >>>> - how to execute them > >>>> > >>>> Because in Pharo we can write tests, identify them and execute them > >>>> Interactively but also from the command-line. > >>>> > >>>> Did you already write a test? > >>>> Because if not, it super super important and I will release some a booklet on Testing in Pharo. > >>>> and also a new chapter ofr Pharo by example 9 > >>>> > >>>> https://pharoweekly.wordpress.com/2021/08/04/ann-chapter-sunit-of-pharo-9-b… <https://pharoweekly.wordpress.com/2021/08/04/ann-chapter-sunit-of-pharo-9-b…> > >>>> > >>>> S > >>>> > >>>>> > >>>>> Roelof > >>>>> > >>>>> > >>>>> > >>>>> Op 13-8-2021 om 21:50 schreef stephane.ducasse(a)free.fr <mailto:stephane.ducasse@free.fr>: > >>>>>> Hello > >>>>>> > >>>>>> Can you be more specific because Pharo got since day 0 (and I would say even before) a test runner? > >>>>>> The SUnit framework is there already. > >>>>>> We were thinking to extract a runner object to make it first class from the core framework. > >>>>>> But beside that it is working, maintained, extended (now we have parametrised tests) and we have (I should publish it) supports for example (tests without setUp) > >>>>>> > >>>>>> Then what is standalone? > >>>>>> > >>>>>> S. > >>>>>>> > >>>>>>> Hello, > >>>>>>> > >>>>>>> > >>>>>>> For the new version of exercism which is due in a few weeks, the pharo track needs a test-runner, > >>>>>>> > >>>>>>> a example of that can be found here : https://github.com/exercism/ruby-test-runner/tree/main/lib <https://github.com/exercism/ruby-test-runner/tree/main/lib> <https://github.com/exercism/ruby-test-runner/tree/main/lib <https://github.com/exercism/ruby-test-runner/tree/main/lib>> > >>>>>>> > >>>>>>> > >>>>>>> first question : Can this be easily be done in pharo? > >>>>>>> > >>>>>>> second question : can it also be standalone? > >>>>>>> > >>>>>>> > >>>>>>> Regards, > >>>>>>> > >>>>>>> > >>>>>>> Roelof > >>>> > >>> > >> > >> > > >

Aug. 16, 2021

File-in Issue - Pharo 9

by Craig Johnson

Hi All, I just filed-out a small project that I playing with, and refreshed the image.Â The new image cannot file-in the package from the File Browser. "Instance of RubSmalltalkEditor did not understand #fileItIn". Is there another way to file-in the package? Thanks Craig

Aug. 15, 2021

Re: Exercism v3 problems

by Roelof Wobben

Aug. 15, 2021

Re: ABBench: Lightweight Microbenchmarking Library

by emdonahu＠gmail.com

Thanks. Yeah, that is to determine a stopping criterion: if one path is clearly faster it needs fewer samples to determine that with confidence. Evan

Aug. 15, 2021

Re: ABBench: Lightweight Microbenchmarking Library

by Kasper Osterbye

Cheers That looks really interesting - I saw that you use an underlying statistical test for the nul hypothesis. That is a cool way to approach it. Best, Kasper > On 12 Aug 2021, at 19.44, Evan Donahue <emdonahu(a)gmail.com> wrote: > > Hello, > > I've just released a utility I've been using for the past few years to help optimize Pharo code. It's a library that makes it easy to check whether an optimization has improved performance by wrapping the optimization and the old code in an if block and executing the optimized and unoptimized paths repeatedly under the same runtime conditions to determine which is faster. > > Basic usage is to wrap changes to the code in if blocks like so: > > ABBench a: [ ...existing code...] b: [...optimized code...] > > And then run the A/B test by printing: > > ABBench bench: [...some main method...] > > I have found it to make performance testing simple and quick enough that I actually do it. Perhaps someone else will find it useful. > > Install with: > > Metacello new > githubUser: 'emdonahue' project: 'ABBench' commitish: 'master' path: ''; > baseline: 'ABBench'; > load. > > More details can be found here: > > https://github.com/emdonahue/ABBench <https://github.com/emdonahue/ABBench> > > Cheers, > Evan

Aug. 14, 2021

Re: Pharo-SQLite3: closing connections immediately and deleting the database file

by vinref＠gmail.com

Hi I double-check changes to the db by running the SQLite3 client in a terminal. The memory db is only accessible by the same client/session so it wonât work. I also run multiple processes doing different things in the same image so each process needs its own connection/handle to SQLite3 so that wont work either. I am using DROP TABLE IF NOT EXISTS to drop all tables before the main schema and it works perfectly. But I did think of using a SQLite3 memory db to hold session info - i.e, cookies. Or I could just use a Dictionary. I havenât implemented this yet but does anyone have any thoughts? Vince

Aug. 13, 2021

Re: Pharo-SQLite3: closing connections immediately and deleting the database file

by JuliÃ¡n Maestri

As far as I know you need to tell it to open a file called :memory: and in Pharo-SQLite3 you should be able to use: SQLite3Connection memory I've never done this before in Pharo, but it's worth trying, it should be faster and easier to set up. https://www.sqlite.org/inmemorydb.html https://github.com/pharo-rdbms/Pharo-SQLite3/blob/7614fe8d0ac03fc94ed6be7e6… On Fri, 13 Aug 2021 at 00:26, Esteban Maringolo <emaringolo(a)gmail.com> wrote: > After I set up the tests I knew that was possible, but never knew how > to actually do it. > > Regards! > > Esteban A. Maringolo > > On Thu, Aug 12, 2021 at 8:55 PM JuliÃ¡n Maestri <serpi90(a)gmail.com> wrote: > > > > Not exactly what you are asking for, but did you consider using sqlite > in memory? > > > > On Sat, 7 Aug 2021 at 23:23, Esteban Maringolo <emaringolo(a)gmail.com> > wrote: > >> > >> I've been doing exactly that. It is, creating a new DB file for each > >> test, and I was having the same problem as you. > >> > >> I suggested a change to the driver to cover that situation: > >> https://github.com/pharo-rdbms/Pharo-SQLite3/pull/22/files > >> > >> Try applying that change and see if that works. > >> > >> I still prefer to recreate the database, it will give you a full fresh > >> start, and if you want you can halt before the deletion of the db, to > >> analyze its internal state. > >> > >> Best regards! > >> > >> > >> Esteban A. Maringolo > >> > >> On Sat, Aug 7, 2021 at 11:13 PM <vinref(a)gmail.com> wrote: > >> > > >> > Hi > >> > > >> > Actually I will just do a bunch of > >> > > >> > DROP TABLE IF EXISTS > >> > > >> > instead of dropping the db altogether. > >> > > >> > Thanks, Vince >

Aug. 13, 2021

Re: Pharo-SQLite3: closing connections immediately and deleting the database file

by Esteban Maringolo

After I set up the tests I knew that was possible, but never knew how to actually do it. Regards! Esteban A. Maringolo On Thu, Aug 12, 2021 at 8:55 PM JuliÃ¡n Maestri <serpi90(a)gmail.com> wrote: > > Not exactly what you are asking for, but did you consider using sqlite in memory? > > On Sat, 7 Aug 2021 at 23:23, Esteban Maringolo <emaringolo(a)gmail.com> wrote: >> >> I've been doing exactly that. It is, creating a new DB file for each >> test, and I was having the same problem as you. >> >> I suggested a change to the driver to cover that situation: >> https://github.com/pharo-rdbms/Pharo-SQLite3/pull/22/files >> >> Try applying that change and see if that works. >> >> I still prefer to recreate the database, it will give you a full fresh >> start, and if you want you can halt before the deletion of the db, to >> analyze its internal state. >> >> Best regards! >> >> >> Esteban A. Maringolo >> >> On Sat, Aug 7, 2021 at 11:13 PM <vinref(a)gmail.com> wrote: >> > >> > Hi >> > >> > Actually I will just do a bunch of >> > >> > DROP TABLE IF EXISTS >> > >> > instead of dropping the db altogether. >> > >> > Thanks, Vince

Aug. 13, 2021

Re: Pharo-SQLite3: closing connections immediately and deleting the database file

by JuliÃ¡n Maestri

Not exactly what you are asking for, but did you consider using sqlite in memory? On Sat, 7 Aug 2021 at 23:23, Esteban Maringolo <emaringolo(a)gmail.com> wrote: > I've been doing exactly that. It is, creating a new DB file for each > test, and I was having the same problem as you. > > I suggested a change to the driver to cover that situation: > https://github.com/pharo-rdbms/Pharo-SQLite3/pull/22/files > > Try applying that change and see if that works. > > I still prefer to recreate the database, it will give you a full fresh > start, and if you want you can halt before the deletion of the db, to > analyze its internal state. > > Best regards! > > > Esteban A. Maringolo > > On Sat, Aug 7, 2021 at 11:13 PM <vinref(a)gmail.com> wrote: > > > > Hi > > > > Actually I will just do a bunch of > > > > DROP TABLE IF EXISTS > > > > instead of dropping the db altogether. > > > > Thanks, Vince >

Aug. 12, 2021

ABBench: Lightweight Microbenchmarking Library

by Evan Donahue

Hello, I've just released a utility I've been using for the past few years to help optimize Pharo code. It's a library that makes it easy to check whether an optimization has improved performance by wrapping the optimization and the old code in an if block and executing the optimized and unoptimized paths repeatedly under the same runtime conditions to determine which is faster. Basic usage is to wrap changes to the code in if blocks like so: ABBench a: [ ...existing code...] b: [...optimized code...] And then run the A/B test by printing: ABBench bench: [...some main method...] I have found it to make performance testing simple and quick enough that I actually do it. Perhaps someone else will find it useful. Install with: Metacello new githubUser: 'emdonahue' project: 'ABBench' commitish: 'master' path: ''; baseline: 'ABBench'; load. More details can be found here: https://github.com/emdonahue/ABBench Cheers, Evan

Aug. 12, 2021

Exercism v3 problems

by r.wobben＠home.nl

Hello, For the new version of exercism which is due in a few weeks, the pharo track needs a test-runner, a example of that can be found here : https://github.com/exercism/ruby-test-runner/tree/main/lib first question : Can this be easily be done in pharo? second question : can it also be standalone? Regards, Roelof

Aug. 10, 2021

Re: Logging Frameworks

by Esteban Maringolo

Hi Gabriel, I found this Beacon-Server [1] that could work, it would be a good use case in a docker environment where you have several worker images talking to a single beacon-server-container. Regards! [1] https://github.com/grype/Beacon-Server Esteban A. Maringolo On Mon, Aug 9, 2021 at 9:10 AM Gabriel Cotelli <g.cotelli(a)gmail.com> wrote: > > Beacon is integrated as part of Pharo now, but the repository includes more logger options. Maybe some option there allows forwarding the logs to a centralized place. I'm also interested in other's experiences regarding logging options because we want to settle on some option as the default for our Pharo components. > > On Sun, Aug 8, 2021 at 11:29 AM Esteban Maringolo <emaringolo(a)gmail.com> wrote: >> >> What are the current options for logging? >> >> Beacon seems to be the most modern one, but the Github repository >> doesn't seem very active (maybe it forked somewhere else). >> >> Given that the logs might come from different worker images, is there >> something that can work via HTTP or similar? (it is, forwarding the >> logged events to a server). >> >> Regards! >> >> ps: I think I've asked this before, but couldn't find my previous >> question about it. >> >> >> Esteban A. Maringolo

Aug. 9, 2021

Re: Logging Frameworks

by Esteban Maringolo

That is my impression with Beacon, there was a lot of discussion about it and then never heard of it again. Is your code available somewhere? Regards! ps: There is also TinyLogger that works by means of a DynamicVariable, but I liked the idea of using objects instead of just strings for logging (which I'm manually doing now). Esteban A. Maringolo On Mon, Aug 9, 2021 at 9:16 AM Norbert Hartl <norbert(a)hartl.name> wrote: > > Beacon is a sad story. There was a lot of initiative and motivation for loggers which was killed. But it is still a good foundation to build something useful on top of it. I integrated a syslog option when I was working on beacon. In my projects Iâve done a sender to elasticsearch for logging. Beaconâs design allows all of that. After motivation died nobody put some love into it again. Thatâs all! > > Maybe time to pick it up again. But I personally have too much on my plate right now to take another attempt. > > Norbert > > > Am 09.08.2021 um 14:08 schrieb Gabriel Cotelli <g.cotelli(a)gmail.com>: > > Beacon is integrated as part of Pharo now, but the repository includes more logger options. Maybe some option there allows forwarding the logs to a centralized place. I'm also interested in other's experiences regarding logging options because we want to settle on some option as the default for our Pharo components. > > On Sun, Aug 8, 2021 at 11:29 AM Esteban Maringolo <emaringolo(a)gmail.com> wrote: >> >> What are the current options for logging? >> >> Beacon seems to be the most modern one, but the Github repository >> doesn't seem very active (maybe it forked somewhere else). >> >> Given that the logs might come from different worker images, is there >> something that can work via HTTP or similar? (it is, forwarding the >> logged events to a server). >> >> Regards! >> >> ps: I think I've asked this before, but couldn't find my previous >> question about it. >> >> >> Esteban A. Maringolo > >

Aug. 9, 2021

Re: Logging Frameworks

by Norbert Hartl

Beacon is a sad story. There was a lot of initiative and motivation for loggers which was killed. But it is still a good foundation to build something useful on top of it. I integrated a syslog option when I was working on beacon. In my projects Iâve done a sender to elasticsearch for logging. Beaconâs design allows all of that. After motivation died nobody put some love into it again. Thatâs all! Maybe time to pick it up again. But I personally have too much on my plate right now to take another attempt. Norbert > Am 09.08.2021 um 14:08 schrieb Gabriel Cotelli <g.cotelli(a)gmail.com>: > > Beacon is integrated as part of Pharo now, but the repository includes more logger options. Maybe some option there allows forwarding the logs to a centralized place. I'm also interested in other's experiences regarding logging options because we want to settle on some option as the default for our Pharo components. > > On Sun, Aug 8, 2021 at 11:29 AM Esteban Maringolo <emaringolo(a)gmail.com <mailto:emaringolo@gmail.com>> wrote: > What are the current options for logging? > > Beacon seems to be the most modern one, but the Github repository > doesn't seem very active (maybe it forked somewhere else). > > Given that the logs might come from different worker images, is there > something that can work via HTTP or similar? (it is, forwarding the > logged events to a server). > > Regards! > > ps: I think I've asked this before, but couldn't find my previous > question about it. > > > Esteban A. Maringolo

Aug. 9, 2021

Re: Logging Frameworks

by Gabriel Cotelli

Beacon is integrated as part of Pharo now, but the repository includes more logger options. Maybe some option there allows forwarding the logs to a centralized place. I'm also interested in other's experiences regarding logging options because we want to settle on some option as the default for our Pharo components. On Sun, Aug 8, 2021 at 11:29 AM Esteban Maringolo <emaringolo(a)gmail.com> wrote: > What are the current options for logging? > > Beacon seems to be the most modern one, but the Github repository > doesn't seem very active (maybe it forked somewhere else). > > Given that the logs might come from different worker images, is there > something that can work via HTTP or similar? (it is, forwarding the > logged events to a server). > > Regards! > > ps: I think I've asked this before, but couldn't find my previous > question about it. > > > Esteban A. Maringolo >

Aug. 9, 2021