May 2022 - Pharo-dev - lists.pharo.org

Re: Please check older Issue tracker entries!

by serge.stinckwich＠gmail.com

Thank you Marcus for your effort to engage more people from the community on managing the issue tracker entries more efficiently ð Regards, Sent from my iPhone > On 18 May 2022, at 14:50, Marcus Denker <marcus.denker(a)inria.fr> wrote: > > ï»¿Hi, > > One problem with a public issue tracker: it accumulates lots of issuesâ¦ > > This sadly means that issues will not be looked at one by oneâ¦ there are > just too many. > > One thing that is interesting to do is to check the issue that *you* submitted. > > - Is the issue still relevant? > - is the title good and in sync with the issue? > - Has it been fixed already ? (it happens!) > > Then, for every issue there is a ânext stepâ. It can be a good thing to spell out > very explicitly what the next step is âWe need to create a Pull Request for the > suggested fixâ. > > Or maybe what is needed is a summary of the huge discussion in the > issue tracker entry. > > Another idea: try to find others to help. E.g. > > - send a mail to the mailinglist. > - ask on discord. > > Maybe you can find a small group that works on this issue together. > > > > Marcus > > >

May 18, 2022

Please check older Issue tracker entries!

by Marcus Denker

Hi, One problem with a public issue tracker: it accumulates lots of issuesâ¦ This sadly means that issues will not be looked at one by oneâ¦ there are just too many. One thing that is interesting to do is to check the issue that *you* submitted. - Is the issue still relevant? - is the title good and in sync with the issue? - Has it been fixed already ? (it happens!) Then, for every issue there is a ânext stepâ. It can be a good thing to spell out very explicitly what the next step is âWe need to create a Pull Request for the suggested fixâ. Or maybe what is needed is a summary of the huge discussion in the issue tracker entry. Another idea: try to find others to help. E.g. - send a mail to the mailinglist. - ask on discord. Maybe you can find a small group that works on this issue together. Marcus

May 18, 2022

ANN: Introducing PhEPs (Pharo Enhancement Proposals)

by Esteban Lorenzano

Hello, Some of you were present at PharoDays 2022 were we made a presentation explaining this new way of handling the evolution of Pharo. In that talk we explained that this is an attempt of having the community involved in more deep way, with a clear (and visible) decision process, taking the model the Python community has been using with success (so why not us? :P). In a nutshell, the idea is that everybody wanting to work on any kind of enhancement for Pharo can now send their proposal and start a discussion process (a discussion arbitrated by the Pharo Board), and if this discussion leads to an accepted enhancement, the work will be integrated as part of the Pharo Environment. How it works? - There is a repository: https://github.com/pharo-project/pheps - and a detailed description of the process: https://github.com/pharo-project/pheps/blob/main/phep-0001.md You will have all the information you need there, but here is a summary of the process (you can also find it in the guidelines PhEP): 0. (Optional, but recommended) the proposer discusses the idea in pharo-dev list or discord, with members of the team, and gather ideas. 1. The proposer will clone the PhEP repository (pharo-project/pheps) and adds a proposal draft following the required stucture: - the file will not have an ID, hence it needs to be called `phep-proposal.md` - the proposal will have the header and structure explained in the section **Submitting a PhEP** (header, abstract, motivation and description). - the proposal will also have a "champion" (which by default is the author itself), responsible for the implementation of the proposal in case is been accepted. 2. The proposal will be submitted as PR to the PhEP repository. 3. A mail will be send to `pharo-dev` mailing list, including the link and the abstract of the proposal. 4. There will be a discussion process and in case of acceptance of the proposal, an ID will be assigned by the reviewer. 5. As final stage, the author renames the `phep-proposal.md` file as `phep-ID.md` and includes the ID in the phep file header (this step can be performed by the reviewer if he decides so). 6. The reviewer will merge the PR In the next days I will be sending a couple fo PhEPs to add to those already there, to get the process started. Do not hesitate to participate! Cheers, Esteban ps: Yes, I am aware "pheps" sounds fun on english :)

May 17, 2022

[review needed] Better structure for CleanBlock Hierachy

by Marcus Denker

Hi, For now, CleanBlockClosure was a subclass of FullBlockClosure, which is not nice (as CleanBlockClosure is not a FullBlockClosure) The PR https://github.com/pharo-project/pharo/pull/11225 improves that a bit: â¢ make CleanBlockClosure a subclass of BlockClosure â¢ BlockClosure is now tagged abstract â¢ override #receiver in CleanBlockClosure â¢ push down #sender â¢ Improve comments ( fixes Check BlockClosure comments for mention of startpc #11217 ) Both CleanBlockClosure and FullBlockClosure add the ivar âreceiverâ, as the vm needs it (it is nil for CleanBlockClosure). This we should look at that later. e.g. have it defined in BlockClosure once, but that has to be done as itâs own PR as it is not a change doable from the image. The change should be save to merge, as CleanBlockClosure is, for now, just used in the tests that turn it on explicitly. PR for review: https://github.com/pharo-project/pharo/pull/11225 Marcus

May 17, 2022

Consider watching the GitHub repo

by Marcus Denker

Hi, GitHub creates a kind of âspontaneous mini mailinglistsâ for every issue and every PR. For one, this is nice as all the discussion is at one place and can be found easily, and it makes sure those who are involved in the issue get the messages (e.g. the one who created an issue tracker entry). But: you will only see those by default. If you want to get an impression of things happening, it might be interesting to âWatchâ the repository. You can do that by clicking on the âWatchâ button on https://github.com/pharo-project/pharo Marcus

May 16, 2022

[Review Needed] Resolves #11184: printStringLimitedTo:Using: will fail to actually respect limit

by Marcus Denker

Review Needed: Resolves #11184: printStringLimitedTo:Using: will fail to actually respect limit https://github.com/pharo-project/pharo/pull/11185 "Changes Object >> printStringLimitedTo:Using so that it actually properly respects the limit, and that it doesn't add '...etc...' when exactly at the limit and thus not needed."

May 13, 2022

This week (19/2022) on the Pharo Issue Tracker

by Marcus Denker

We merged 14 Pull Requests and fixed 17 Issues. There are always more issues ready to be reviewed https://github.com/pharo-project/pharo/pulls Feature ======= - 6720-Dictionary--newFrom-logic-can-be-improved- #11176 https://github.com/pharo-project/pharo/pull/11176 - Fixes #11165 begins with and ends with should return true on empty string #11173 https://github.com/pharo-project/pharo/pull/11173 Clean Blocks ============ - Fix-messagetally-tests-Cleanblocks #11209 https://github.com/pharo-project/pharo/pull/11209 Bugs ==== - 6323-RubSmalltalkEditor-meta-j-shortcut-not-working #11205 https://github.com/pharo-project/pharo/pull/11205 - FIX issue #10933: Scaled decimals comparison result depends on the number creation method #10951 https://github.com/pharo-project/pharo/pull/10951 - Implement EpInverseVisitor>>#visitTraitModification: #11208 https://github.com/pharo-project/pharo/pull/11208 - 11198-Pharo11-Reflectivity-do-not-add-temp-for-value-if-not-generating-the-primtive-Wrapper #11199 https://github.com/pharo-project/pharo/pull/11199 Cleanup ======= - 11216-categorize-ZnStaticFileDecoratorDelegate-classdecorateservingFilesFrom #11224 https://github.com/pharo-project/pharo/pull/11224 - 11162-Check-ReToDoRule #11164 https://github.com/pharo-project/pharo/pull/11164 - Fix for the issue #11115: Feature Envy clipRect on AthensCairoCanvas #11117 https://github.com/pharo-project/pharo/pull/11117 - 11144-LocaleAnnouncementTest-seems-to-be-unused #11189 https://github.com/pharo-project/pharo/pull/11189 Pharo10 ======= - Remembering scroll position when collapsing and expanding tree table items #11202 https://github.com/pharo-project/pharo/pull/11202 - Making completion support $_ in class and method names #11197 https://github.com/pharo-project/pharo/pull/11197 - 11198-Pharo10-Reflectivity-do-not-add-temp-for-value-if-not-generating-the-primtive-Wrapper #11200 https://github.com/pharo-project/pharo/pull/11200

May 13, 2022

Re: Question about MethodDictionary in TraitedClass (methodDict/localMethodDict)

by Marcus Denker

> On 12 May 2022, at 11:27, Andrei Chis <chisvasileandrei(a)gmail.com> wrote: > > Hi, > > We were profiling the loading of code that uses many Traits. We saw something like this (this is a tally for `TraitedMetaclass>>#addAndClassifySelector:withMethod:inProtocol:`): > <Screenshot 2022-05-11 at 17.17.37.png> > > From what we saw in `TraitedMetaclass>>#addAndClassifySelector:withMethod:inProtocol:` a method is first added to `localMethodDict` and then it calls the super method that adds the selector to the `methodDict` of the actual class in `Behaviour>>addSelectorSilently:withMethod: `. Because of this we basically get an overhead as a method is added to two method dictionaries and adding a method to a MethodDictionary is slower, as it does operations like `flushCache` and `cachePragmas`. > > Should the `localMethodDict` also be an instance of MethodDictionary or could it be a simple dictionary? > I think it can be normal Dictionary: MethodDictionary we just meed where there VM executes code (for flushCache) and for lookup. The only downside is that it then uses a little bit more memory (MethodDictionary does not use associations), but that should be no problem. Marcus

May 12, 2022

Question about MethodDictionary in TraitedClass (methodDict/localMethodDict)

by Andrei Chis

Hi, We were profiling the loading of code that uses many Traits. We saw something like this (this is a tally for `TraitedMetaclass>>#addAndClassifySelector:withMethod:inProtocol:`): [image: Screenshot 2022-05-11 at 17.17.37.png] >From what we saw in `TraitedMetaclass>>#addAndClassifySelector:withMethod:inProtocol:` a method is first added to `localMethodDict` and then it calls the super method that adds the selector to the `methodDict` of the actual class in `Behaviour>>addSelectorSilently:withMethod: `. Because of this we basically get an overhead as a method is added to two method dictionaries and adding a method to a MethodDictionary is slower, as it does operations like `flushCache` and `cachePragmas`. Should the `localMethodDict` also be an instance of MethodDictionary or could it be a simple dictionary? Cheers, Andrei

May 12, 2022

About Pull Request Reviews: Multiple Reviews are best!

by Marcus Denker

Hi, If you see that a PR that you wanted to review already has a positive review, do not think âah, nothing to be done for meâ. Because imagine if the PR would be bad: you would add a comment how to improve in addition to the one already positive review. You should do the same even if you just add another âYes, looks goodâ. The reason is that *you* know that you looked at the PR and did not find a problem, but nobody else has a way to see that. Second (or even third, 4thâ¦) positive reviews are very valuable, they validate the first review *and* they are vote for getting that PR merged fast. You can find the PRs in need of a review here: https://github.com/pharo-project/pharo/pulls Marcus

May 12, 2022