- Pharo-dev - lists.pharo.org

Re: [Pharo-dev] ARM Cogit

by Aliaksei Syrel

Hi Any news on Pharo VM for android? ;) Cheers, Alex On Thu, Apr 23, 2015 at 7:54 AM, stepharo <stepharo(a)free.fr> wrote: > This is really excellent! > Keep pushing this. > I'm eager to play with a raspberry. > We should continue to revisit the UI elements to consume less in any case. > Because I'm convinced that we are doing too many things at the wrong place. > > Stef > > Le 22/4/15 21:42, Eliot Miranda a Ã©crit : > > Hi All, >> >> thanks to hard work by Tim Rowledge and Lars Wasserman we now have a >> functional JIT for the ARM in the simulator. We're still probably some >> weeks away from a functional ARM JIT VM; issues like instruction cache >> flushing on real hardware can be tricky to solve. But we're at least ready >> to start trying to compile the JIT VM for ARM. Doug, see that I added a >> build directory for build.linux32ARM/squeak.cog.spur. Hopefully a fast VM >> for Raspberry Pi and Android is only a few weeks away. >> -- >> best, >> Eliot >> > > >

Oct. 6, 2015

Re: [Pharo-dev] UserManager removed - alternative?

by Aliaksei Syrel

Hi Stef, Thanks for explanation! I agree and looking forward to see a new solution :) Cheers, Alex On Tue, Oct 6, 2015 at 8:26 AM, stepharo <stepharo(a)free.fr> wrote: > Hi Aliak > > The design of the UserManager was ****plain**** bad. It was reintroducing > this ugly facade singleton pattern (monolitic and global) > like the old Preferences class that we fight a lot with alain and > finally replace by setting. > It was a mistake that I did not have the time to review the code. (BTW we > were pushing too much code in the system without > a good review). Now this period is over. > > The solution of the new approach should be the same than settings. > > - first any part that should be user managed should do it locally and > introduce its own hooks to do so > but be self-contained: exactly like Setting. A class uses class > variables and check its class variables > when doing something. > > - It means that as a client of the part I can change some accesses. > > - then we decorate such user-managed hooks and provide tools. > > Stef > > > > > > Bloc is broken because UserManager is no longer in the image. (it was > removed on 14th of August) > For some reason there is a check: > > UserManager default canDropOSFile > > > this check is used in #dropFiles: > > What is alternative or does it make sense to remove the check? > > Cheers, > Alex > > >

Oct. 6, 2015

Re: [Pharo-dev] Slack?

by Esteban A. Maringolo

2015-10-06 15:12 GMT-03:00 stepharo <stepharo(a)free.fr>: > Not sure that yet another tool will help me to be more productive. > Don't count me in it. I don't think it aims at making somebody more productive, but instead to gather people with similar interests and avoid querying the mailing list for trivial things. Let's see if it works, if not, closing it is instantaneous. Esteban A. Maringolo

Oct. 6, 2015

Re: [Pharo-dev] Slack?

by Esteban A. Maringolo

2015-10-06 13:03 GMT-03:00 Dimitris Chloupis <kilon.alios(a)gmail.com>: > Quite frankly, the reason why IRC is so extremely popular, is because it > does its job very well and has some very nice clients. I am active at #Pharo > because I use several IRC channels, #Blender , #C++ , #Opengl , #MACOS , > #unrealengine , #emacs , #blendercoders and #blenderpython. So the > incredible power of IRC is not its features, but its community. You don't have to give up your familiar tools. I set up the IRC gateway, so you can access the slack channel without leaving your IRC client. I tried it today and it worked right away. Regards. Esteban A. Maringolo

Oct. 6, 2015

Re: [Pharo-dev] About Magritte Seaside

by Dale Henrichs

On 10/6/15 11:43 AM, stepharo wrote: > >> Stef, >> >> I hope you and Christophe are considering cross platform >> considerations when implementing "new dependencies declarations at >> the package level" ... MonticelloConfigurations had "package level >> dependencies" and the implementation was extremely rigid to the point >> where it wasn't possible to substitute an alternate >> (platform-speicific) package in the package graph without creating >> new package versions of all packages in the graph ... Of course I am >> not aware of the approach that Christophe, but be aware that "there >> be dragons there":) > > Hi dale. > > I will let christophe answers because I do not know the details. I > know that he is working hard on it. > Ideally I would like to have a ConfigurationOf for each package > because it would reduce the bloat of the actual configurationOf and > also be closer to the package. But if I say that people will start > shouting that I'm mad :) (I'm not fool and I'm right :). > so this is why we are doing it inside the package but we will see. > First the core bootstrap based on dependencies ;) > > Stef > Stef, Basing things on a "configuration" external to the package structure itself will avoid the MetacelloConfiguration limitations. From your description, it sounds like Christophe might be "deconstructing" the project configuration in some way --- I would be interested in more details ... Thanks, Dale

Oct. 6, 2015

Re: [Pharo-dev] About Magritte Seaside

by stepharo

> Stef, > > I hope you and Christophe are considering cross platform > considerations when implementing "new dependencies declarations at the > package level" ... MonticelloConfigurations had "package level > dependencies" and the implementation was extremely rigid to the point > where it wasn't possible to substitute an alternate > (platform-speicific) package in the package graph without creating new > package versions of all packages in the graph ... Of course I am not > aware of the approach that Christophe, but be aware that "there be > dragons there":) Hi dale. I will let christophe answers because I do not know the details. I know that he is working hard on it. Ideally I would like to have a ConfigurationOf for each package because it would reduce the bloat of the actual configurationOf and also be closer to the package. But if I say that people will start shouting that I'm mad :) (I'm not fool and I'm right :). so this is why we are doing it inside the package but we will see. First the core bootstrap based on dependencies ;) Stef

Oct. 6, 2015

Re: [Pharo-dev] Mongo - find object by id

by stepharo

This is great I will reread soon the mongo chapter. I should start to code a real app (if only I could have the time). stef Le 7/9/15 09:00, Sabine Manaa a Ã©crit : > Hi Stephane, > > my changes have been integrated into the document by Johan and I > checked as ok, so it is done. > > Regards > Sabine > > 2015-08-18 12:49 GMT+02:00 Sabine Manaa <manaa.sabine(a)gmail.com > <mailto:manaa.sabine@gmail.com>>: > > Hi Stephane, > > ok, I have put all my suggestions in a pull request. Please feel > free to change, delete .... it is only a suggestion. I am > interested in feedback! > > Regards > Sabine > > 2015-08-17 14:53 GMT+02:00 Sabine Manaa <manaa.sabine(a)gmail.com > <mailto:manaa.sabine@gmail.com>>: > > Hi Steph, > > thanks. I am editing my suggestions and tell you when finished. > > Regards > Sabine > > 2015-08-16 19:29 GMT+02:00 stepharo <stepharo(a)free.fr > <mailto:stepharo@free.fr>>: > > >> Hi Stephane, >> >> I have collected my points to add to the book. >> >> Can you give me an entry point how to do a pull request >> on the paper please? >> Pillar? Editor? Script? Sounds complicated for the first >> time. >> > If you can use pillar this is better, else we will do a > pass and convert it to pillar > Since the chapter is for the book that should be out I'm > not that editing it directly is a good. > Now what we can do is to work on on a separate file that > Johan or Esteban can merge to the chapter once > it is finished. > > What I was thinking about is to have a series of extras > section > with for example some of the problems and solutions you > encountered. > > > How to handle loop within your domain? > > How to find an object by id? > > ... > > The original file is here: > > https://github.com/SquareBracketAssociates/EnterprisePharo/tree/master/Voya… > > > I created a VoyageAdditionalText.pier that you can edit. > > And this is easy > - you can go to the page > - fork the folder > - edit in the web the file and send a pull request. > > if this is not working, we will add you to the project and > you can directly edit this file. > >> Regards >> Sabine >> >> >> >> 2015-08-13 16:17 GMT+02:00 stepharo <stepharo(a)free.fr >> <mailto:stepharo@free.fr>>: >> >> >>> Hi Stephane, >>> >>> sure, I am happy to give something back to the >>> community. I dont have so much time because of my >>> baby but I will try my best. >>> >>> Where shall I put it in? >> >> You can do pull requests in the paper. This way they >> will be close to the original version. >> >> >>> Sabina >>> >>> you know what I would really love from you .... >>> this is that you reread the Voyage chapter and >>> make a list of the problems you faced and that >>> are not described in the chapter. >>> You can write it in any bad ways and we can do a >>> edit phase after, >>> >>> Why do I ask that? Because I'm stupid in Mongo >>> and I would love to have a good doc. >>> And only the people facing the problems can spot >>> them nicely. >>> I often write doc on things that I do not know >>> but this is not working well because I'm slow in >>> that case and .... I could be much better >>> doing something else where I'm good at. >>> >>> Stef >>> >>> Le 12/8/15 09:40, Sabine Manaa a Ã©crit : >>> >>> Hi, >>> >>> on mongo console, this brings me a persons >>> information from mongo db >>> db.Persons.find({"_id" : >>> ObjectId("75b900000000000000000000")}) >>> >>> I try to query the same person from pharo >>> and I do not succeed. >>> >>> This makes my image hanging: >>> RKAPerson selectMany: {('_id' -> (OID >>> value:'75b900000000000000000000'))} >>> >>> also other variations e.g. >>> RKAPerson selectMany: {('_id' -> >>> ('75b900000000000000000000')). } >>> asDictionary. >>> >>> do not work >>> >>> Other queries like >>> RKAPerson selectMany:{('lastName' -> >>> ('Maier')).} asDictionary. >>> or >>> RKAPerson selectMany:{('nextTripNumber' -> >>> (57)).} asDictionary. >>> >>> work fine. >>> >>> How can I query for an object with the >>> object id? >>> >>> Regards >>> Sabine >>> >>> >>> >>> -- >>> View this message in context: >>> http://forum.world.st/Mongo-find-object-by-id-tp4842231.html >>> Sent from the Pharo Smalltalk Developers >>> mailing list archive at Nabble.com. >>> >>> >>> >>> >> >> > > > >

Oct. 6, 2015

Re: [Pharo-dev] [Reflectivity] more link options and enforcing link options per Class

by stepharo

+ 1 Symbols and strings and arrays are dead limited objects. Not even zomby objects. No behavior no fun. >>> >>> The latest image now has some more options for meta links. The >>> complete list is listed in the #defaultOptions method of >>> MetaLink class: >>> >>> defaultOptions >>> ^ #( >>> + optionInlineMetaObject "meta object is inlined by default." >>> + optionInlineCondtion "condition is inlined by >>> default.â (ups, typo, will fix) >>> - optionCompileOnLinkInstallation "generate compiledMethod on link >>> installation" >>> - optionOneShot "remove link after first activation" >>> - optionMetalevel "force level: 0 for the link" >>> - optionDisabledLink "links are active by default" >>> ) >>> >>> Some of the options make only sense because after this issue will be >>> integrated: >>> >>> https://pharo.fogbugz.com/f/cases/16695/allow-classes-to-force-meta-link-op… >>> >>> we will have a mechanism that classes can enforce options for all >>> links installed in them. This is done by adding >>> a method on the instance or class side like this: >>> >>> metaLinkOptions >>> ^{ >>> #exampleMethodWithMetaLinkOptionsViaClass -> #( + >>> optionCompileOnLinkInstallation) >>> } >>> >>> This allows e.g. a class to turn off links for one method or enforce >>> using meta-level awareness. >>> (for now the options are overridden when installing a link, I will >>> move this into the code generator >>> to have a real overlay semantics as a next step). >>> >>> Marcus >> >> Why the aversion to classes? >> Personally, I'd much prefer a MetaLinkOptions class with boolean >> instvars, and a class comment... >> >> Would work much better with the tools we have (finding users, etc.) >> and a class enforcing method such as: >> >> overrideMetaLinkOptions: options >> >> ^options >> oneShot: false; >> yourself >> > > Yes.. the current way starts to be too complexâ¦ could be a good next step. > > Marcus > > > > > > >

Oct. 6, 2015

Re: [Pharo-dev] About Magritte Seaside

by Dale Henrichs

On 10/6/15 5:10 AM, stepharo wrote: > > > Le 6/10/15 11:40, Stephan Eggermont a Ã©crit : >> On 06-10-15 11:11, stepharo wrote: >>> I will do a separate configuration of magritte seaside. >> >> You need to load the correct groups. Separate configurations are a >> bad practice and should be eliminated, not encouraged. > > I totally disagree. > And groups are not supported by the configurationBrowser >> The configuration has a lot of essential complexity, as many >> different combinations of components are needed in different >> situations. We need tools that recognize that essential complexity >> and support it, not tools that try to hide it and then have no >> practical usability. > > Well I disagree. And all the effort of christophe for the new > dependencies declarations at the package level just goes in that > direction. > I will do a MagritteSeaside Configuration that I will maintain and I > will also do a MagritteTutorial one. > Stef, I hope you and Christophe are considering cross platform considerations when implementing "new dependencies declarations at the package level" ... MonticelloConfigurations had "package level dependencies" and the implementation was extremely rigid to the point where it wasn't possible to substitute an alternate (platform-speicific) package in the package graph without creating new package versions of all packages in the graph ... Of course I am not aware of the approach that Christophe, but be aware that "there be dragons there":) Dale

Oct. 6, 2015

Re: [Pharo-dev] Slack?

by stepharo

Not sure that yet another tool will help me to be more productive. Don't count me in it. Le 6/10/15 12:23, Yuriy Tymchuk a Ã©crit : > Hi guys, > > Iâm thinking whether live chat can make sense for communication. And I > know that there is IRC, but for some reason itâs not working for me > (maybe because of clients). > > Iâve seen many teams using Slack [1]. Is anybody willing to try that > out. Or also maybe some have already created a pharo team on Slack? > > Cheers. > Uko > > [1]: https://slack.com/

Oct. 6, 2015