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

Re: [Pharo-users] Documentation about Pharo Launcher?

by Damien Cassou

Offray Vladimir Luna CÃ¡rdenas <offray.luna(a)mutabit.com> writes: > I would like to made one of my images available through Pharo > Launcher. Anyone knows where I can find documentation about this? I > can't find anything on the StHub page for the project. I agree with both Serge and CÃ©drick. Their solution is easy and the best practice for any Pharo development (regardless of the launcher): - write a ConfigurationOf using Versionner - add a job to https://ci.inria.fr/pharo-contribution/ (just create an account at http://ci.inria.fr and request to join pharo-contribution). As soon as the job builds successfully, you will see it in the Launcher. -- Damien Cassou http://damiencassou.seasidehosting.st "Success is the ability to go from one failure to another without losing enthusiasm." --Winston Churchill

May 4, 2016

Re: [Pharo-users] Can you help me find this info?

by stepharo

Le 2/5/16 Ã 11:30, Victor RENE a Ã©crit : > Hello, > > I am new to Pharo and would like to understand a few things. > I can search by myself, but I do not know the code base yet, so: > > [Questions] > Pharo image: > Serialization, deserialization, where is the code? Are you talking about image or serialization of objects? > Pharo vm (Cog/Spur): > Rendering loop and event loops. done with SDL2, where is the code? Look at OSWindows in Pharo we are in transition. > Message passing, object hierarchy lookup, where is the code? in the VM > Pharo minimal: > How to get a working version? The builds I tried (4.0) crashed. Can you tell us more? We have different systems - miniimage which is built automatically by removing packges and this is working so may be you use the wrong vm. - bootstrapped kernels will be in production in Pharo 60 - specialized image we got down to 11k for 2+3 and 18k for 20 factorial this is experimental but hyper fun. [Purpose] > My goal is work out an image/vm with minimum overhead. For example, > let's say I want to build a modern multi-touch application. Yes we build that for Thales > Pharo's environment looks promising on the live coding/debugging side, > but what about the performance? It depends what you want to do. if this is real time, Pharo is not a real time os After Thales are migrating to Pharo for their "prototype" of new interactive UI. > Lack of responsiveness would not be > acceptable (think about design differences between Android/iOS). We should do a pass on the Android VM and its event. > > [Context] > I want to understand if the object oriented nature of Smalltalk is > the reason for the lack of apps of this nature (as far as I can tell). Lack of Marketing is the reason. Not the technology. Squeak run on PS2 in 300 k back in 2002 or befor. > For example, games have been done with Lisp: Jack and Daxter (on the PS2, > by Naughty Dogs). There are also game engines for the JVM > (https://www.lwjgl.org/) > Dedicated hardware has been built for either Lisp or Forth. I can't seem > to find any hardware built specifically to power Smalltalk software. > Yet as far as I can tell, it's the big time for software virtual machines, > because the new generation of hardware will be supporting them. > (https://news.ycombinator.com/item?id=10032295) I do not know. > > UPDATE: I am reading through "Evaluating message passing control > techniques > in Smalltalk" > (https://www.researchgate.net/publication/2440118_Evaluating_Message_Passing…) > I am getting a better picture of the whole, please direct to more such > papers. it is an old paper. When I arrived from CLOS I wanted to know if I could do the same. Now I know and we can certainly do better at the MOP level. > > Thanks, > > Victor RENE > Software engineer, > Game designer, Writer > > Blog: http://victor-rene.com > Projects: http://logiqub.com > > tel: +33 6 26 83 61 76 > email: victor-rene(a)outlook.com

May 4, 2016

Re: [Pharo-users] #ast vs. #parseTree

by Marcus Denker

> On 01 May 2016, at 14:45, Sean P. DeNigris <sean(a)clipperadams.com> wrote: > > Marcus Denker-4 wrote >> #ast returns a result from the cache, while #parseTree always gets a new >> oneâ¦ > > Ah, okay. What is the difference? That is, how/when is the cache updated? > -> added on first call of #ast -> cleaned on image shutdown. This is not yet really good, but works amazingly well for being so simple. I want to change it to be LRU and not grow too largeâ¦ clients that are interested in persisting the AST (e.g. due to annotations) do that now already to survive shutdown. E.g. if you put a breakpoint, the AST of *that* method will survive shutdownâ¦ so a LRU cache would not change that. > > Marcus Denker-4 wrote >> Yes, it could be better. Like always. > > :) I will add commentsâ¦ and the API should be improved, too, but not now. Marcus

May 4, 2016

Re: [Pharo-users] Securely downloading Pharo

by Peter UhnÃ¡k

I guess you could download it from the Jenkins job https://ci.inria.fr/pharo/job/Pharo-5.0-Update-Step-5-Publish/lastSuccessfu… although the job itself downloads over http from files.pharo.orgâ¦ but Jenkins should be fine enough if you are worried about MitM between you and Pharo. For SHA you still need a secure connection assuming the website wouldn't get compromised, or digitally signed releases assuming the signing won't be misconfigured, or the keys won't get compromised. But I agree that this should be solved, or at least improved; but then again whole smalltalkhub doesn't use https and you are sending your credentials in plaintext over the network (have fun committing from a hotel room wifi)â¦ security through obscurity can be incredibly efficient. Of course that means that we need someone with the time and knowledge that can invest effort into this. Peter On Wed, May 4, 2016 at 12:33 AM, Wilfred Hughes <me(a)wilfred.me.uk> wrote: > Is there any way of downloading Pharo securely? > > I'm trying to download Pharo itself over HTTPS, so I know I can trust the > data: > > $ wget https://files.pharo.org/platform/Pharo4.0-linux.zip > --2016-05-02 22:44:34-- > https://files.pharo.org/platform/Pharo4.0-linux.zip > Resolving files.pharo.org (files.pharo.org) 128.93.162.72 > Connecting to files.pharo.org (files.pharo.org)|128.93.162.72|:443... > connected. > OpenSSL: error:140770FC:SSL routines:SSL23_GET_SERVER_HELLO:unknown > protocol > Unable to establish SSL connection. > > The excellent pharo zeroconf script doesn't seem available over HTTPS > either: > > $ curl https://get.pharo.org/vm50 > curl: (35) error:140770FC:SSL routines:SSL23_GET_SERVER_HELLO:unknown > protocol > > Looking at the script itself, it's downloading files over HTTP from > files.pharo.org and executing them without verifying. I've explored > files.pharo.org, but I can't see any signatures or hashes (e.g. > sha256sum) of any of the files. > > The pharo homepage is largely available at https://pharo.org/ > (although some of the styling is missing due to being served over > HTTP). > > Have I missed something? Would it be possible to provide HTTPS and/or > sha256sums for downloads? > > > Wilfred > >

May 3, 2016

Securely downloading Pharo

by Wilfred Hughes

Is there any way of downloading Pharo securely? I'm trying to download Pharo itself over HTTPS, so I know I can trust the data: $ wget https://files.pharo.org/platform/Pharo4.0-linux.zip --2016-05-02 22:44:34-- https://files.pharo.org/platform/Pharo4.0-linux.zip Resolving files.pharo.org (files.pharo.org) 128.93.162.72 Connecting to files.pharo.org (files.pharo.org)|128.93.162.72|:443... connected. OpenSSL: error:140770FC:SSL routines:SSL23_GET_SERVER_HELLO:unknown protocol Unable to establish SSL connection. The excellent pharo zeroconf script doesn't seem available over HTTPS either: $ curl https://get.pharo.org/vm50 curl: (35) error:140770FC:SSL routines:SSL23_GET_SERVER_HELLO:unknown protocol Looking at the script itself, it's downloading files over HTTP from files.pharo.org and executing them without verifying. I've explored files.pharo.org, but I can't see any signatures or hashes (e.g. sha256sum) of any of the files. The pharo homepage is largely available at https://pharo.org/ (although some of the styling is missing due to being served over HTTP). Have I missed something? Would it be possible to provide HTTPS and/or sha256sums for downloads? Wilfred

May 3, 2016

Re: [Pharo-users] NoSWL Databases

by Stephan Eggermont

On 03/05/16 15:42, AWS wrote: > The current course is all about Database management and analysis. We > are currently looking at NoSQL databases and I would like to try out > one of these against some ideas I have for the project. Did you consider using Gemstone for this? It solves a rather large part of the common persistency problems, including a few difficult ones Stephan

May 3, 2016

Re: [Pharo-users] Can you help me find this info?

by Offray Vladimir Luna CÃ¡rdenas

On minimal images, you could take a look at Cuis: http://cuis-smalltalk.org/ Is not based on Pharo, but could give you some ideas/inspiration. Cheers, Offray On 02/05/16 04:30, Victor RENE wrote: > Hello, > > I am new to Pharo and would like to understand a few things. > I can search by myself, but I do not know the code base yet, so: > > [Questions] > Pharo image: > Serialization, deserialization, where is the code? > Pharo vm (Cog/Spur): > Rendering loop and event loops. done with SDL2, where is the code? > Message passing, object hierarchy lookup, where is the code? > Pharo minimal: > How to get a working version? The builds I tried (4.0) crashed. > > [Purpose] > My goal is work out an image/vm with minimum overhead. For example, > let's say I want to build a modern multi-touch application. > Pharo's environment looks promising on the live coding/debugging side, > but what about the performance? Lack of responsiveness would not be > acceptable (think about design differences between Android/iOS). > > [Context] > I want to understand if the object oriented nature of Smalltalk is > the reason for the lack of apps of this nature (as far as I can tell). > For example, games have been done with Lisp: Jack and Daxter (on the PS2, > by Naughty Dogs). There are also game engines for the JVM > (https://www.lwjgl.org/) > Dedicated hardware has been built for either Lisp or Forth. I can't seem > to find any hardware built specifically to power Smalltalk software. > Yet as far as I can tell, it's the big time for software virtual machines, > because the new generation of hardware will be supporting them. > (https://news.ycombinator.com/item?id=10032295) > > UPDATE: I am reading through "Evaluating message passing control > techniques > in Smalltalk" > (https://www.researchgate.net/publication/2440118_Evaluating_Message_Passing…) > I am getting a better picture of the whole, please direct to more such > papers. > > Thanks, > > Victor RENE > Software engineer, > Game designer, Writer > > Blog: http://victor-rene.com > Projects: http://logiqub.com > > tel: +33 6 26 83 61 76 > email: victor-rene(a)outlook.com

May 3, 2016

[ANN] Pharo Astares Distribution

by Torsten Bergmann

Hi, I summarized some of the goodies that I wrote like - DesktopManager - QuickAccess - HubCap - WebBrowser - MessageFlowBrowser - ... into an own kind of Pharo distribution that I call AstaresDistribution. So if you already use these goodies it is easier to load quickly into a fresh image. You can load directly from Catalog or by opening Spooter, enter "Astares" and hit ENTER to load. Alternatively you can download a prebuilt image using PharoLauncher from "Pharo Contributions Jenkins" -> "AstaresDistribution" or directly from [1]. Screenshot attached. Have fun T. [1] https://ci.inria.fr/pharo-contribution/job/AstaresDistribution

May 3, 2016

Re: [Pharo-users] Metaprogramming facilities

by Peter UhnÃ¡k

> > More importantly the autogenerated code is a bit stupid > > TBPost >> title: anObject > title := anObject > > this is stupid, so even for generated one I select the argument, press > ctrl+t (or right click + suggestions) and select 'rename' ... and rename it > into something meaningful like > > TBPost >> title: aString > title := aString > This (usually) doesn't apply to 6 (TDD), because the generated correctly names it based on the real value that is being passed to it.

May 3, 2016

Re: [Pharo-users] Metaprogramming facilities

by Peter UhnÃ¡k

1) write them by hand annoying typing required 2) use Class refactorings (right click on class -> refactoring -> class refactoring -> generate accessors; or ctrl+h+a) this will ask you to choose for all attributesâ¦ good initially, but really stupid if you are providing e.g. just getters (because then you would need to cancel genreation for all the attributes) 3) use attribute refactorings (right click on class -> refactoring -> inst var refactoring -> accessors -> select attribute) good for individual attributes 4) write the code compilation (ideally as part of some tool) MyClass compile: 'attr ^ attr' classified: 'accessing'. MyClass compile: 'attr: anObject attr := anObject' classified: 'accessing'. Much more typing, but good if you have it as part of some tool (so you can quickly script a gui that would generate it just by clicking on a checkbox or something) 5) use refactoring (this is what the System Browser uses internally) (RBCreateAccessorsForVariableRefactoring variable: #attr class: MyClass classVariable: false) execute 6) use TDD (test driven development) the Debugger will offer you to create the method and you can modify it as you work through (I don't know if TDD is already part of the broadcasted MOOC courses) Typing by hand is annoying, but it's literally two words for getter and four for setter. This mail is probably more words than all the accessors I've created this yearâ¦ so typing it by hand is no big deal. More importantly the autogenerated code is a bit stupid TBPost >> title: anObject title := anObject this is stupid, so even for generated one I select the argument, press ctrl+t (or right click + suggestions) and select 'rename' ... and rename it into something meaningful like TBPost >> title: aString title := aString But otherwise I regularly use all 1, 2, 4, 5, 6â¦ I never use 3) it's too annoying and distracting because I have to read the labelsâ¦ I can type in the getter/setter without breaking focus on what I am doingâ¦ but it just gave me an idea... :) Peter On Tue, May 3, 2016 at 9:05 PM, stepharo <stepharo(a)free.fr> wrote: > Code not tested but .... > > > #('title' 'text') > > do: [:each | > > | body | > > body := String new streamContents: > > [ :s | > > s nextPutAll: each ; cr; tab; > > nextPutAll: '^ '; > > nextPutAll: each ; cr]. > > TBPost compile: body classified: 'accessing' > > ] > > Le 2/5/16 Ã 17:35, Victor RENE a Ã©crit : > > In the TinyBlog mini-project of the MOOC, we start by creating the model. > However, since I basically finished the exercises, I wanted to go one step > further and > ask you guys, what is the correct way _not_ to write all gettters / > setters for the fields. > > > TBPost >> title > ^ title > TBPost >> title: anObject > title := anObject > TBPost >> text > ^ text > TBPost >> text: anObject > text := anObject > etc... > > > Do you do it with macros / templates / something else ? > I am having a bit of fun (although I would love that they unlock weeks 2 > to 7 for me). Thanks in advance. > > Victor RENE > Software engineer, > Game designer, Writer > > Blog: <http://victor-rene.com>http://victor-rene.com > Projects: <http://logiqub.com>http://logiqub.com > > tel: +33 6 26 83 61 76 > email: victor-rene(a)outlook.com > > > >

May 3, 2016