July 2010 - Pharo-dev - lists.pharo.org

[Pharo-project] [update 1.2] #12045
by stephane ducasse July 10, 2010

July 10, 2010

Some fixes from squeak that accumulated in our bug tracker. 12045 ----- - Issue 2300: ServerDirectory cleanup - Issue 2226: shuffled and shuffle - Issue 2238: tests for FileStream behavior. - Issue 2129: Cursor with Alpha from OLPC CursorWithAlpha biggerNormal showWhile:[Sensor waitButton]. Stef

1 0

[Pharo-project] [update 1.2] #12044
by StÃ©phane Ducasse July 10, 2010

July 10, 2010

12044 ----- - Issue 2652: better flush of class name cache when removing key - Issue 2568: Remove ancient j3 support - Issue 2650: Easier doit of commented expression Stef

1 0

[Pharo-project] New Engineer position for Moose
by StÃ©phane Ducasse July 10, 2010

July 10, 2010

We got a one year engineer position to support the development of a demonstratior for Moose. The ***constraints*** linked with the position are that the person should have got his master in 2009 or 2010. The starting date is 1 of December 2010. The salary should be around 3500 Euros bruto ~> 2300 Euros before taxes (but after health insurance). The job will probably consist in - getting Moose running into a web browser - having a better reporting system. - building a dashboard - porting part of codecity What this of course means is that it's kind of a cool job: work in a research lab, but no paper writing, no teaching, no administration. In addition: -> live in France (one hour from Paris, 1h20 from London, 35 min from Brussels) -> try a lot of good beers (our office is just 10km from the border to Belgium) -> enjoy french food -> code all the day in Smalltalk and a bit of C By september we should decide if we take this year in 2010 or 2011 (for decembre 2011 the person should have got a master in 2010 or 2011). Please send your cv. More about: - RMoD: http://rmod.lille.inria.fr - INRIA Lille: http://www.inria.fr/lille/ - INRIA in General: http://www.inria.fr - Engineers at INRIA: http://bit.ly/cLe6q4 - Pharo: http://pharo-project.org - Lille: - http://en.wikipedia.org/wiki/Lille - http://www.mairie-lille.fr/en - http://wikitravel.org/en/Lille Stef

1 0

[Pharo-project] Are random refactorings welcome?
by Javier PimÃ¡s July 9, 2010

July 9, 2010

Hi all, Some time ago I found out that half of the time spent in opening a browser was spent in ScrollBars code (you can see it by doing TimeProfileBrowser onBlock: Object browse). So I looked at the code and as it was messy I made a big refactory and cut browser opening time a bit. The thing is I never submitted that code because it wasn't fully working, and as I didn't know if this kind of random refactorings were welcome I didn't take the time to finish it. I found it some days ago rotting in my filesystem. Then the question is, would you be willing to include the code in next iterations of pharo? I can rebase the code to make it work Pharo1.1 and send it for review. Regards, Javier. -- Javier PimÃ¡s Ciudad de Buenos Aires

3 2

[Pharo-project] About String>>translated
by Guillermo Polito July 9, 2010

July 9, 2010

Hi! I was looking at String >> translated and i've found the following: String>>translated "answer the receiver translated to the default language" ^ NaturalLanguageTranslator translate: self NaturalLanguageTranslator class>>translate: aString ^ aString Sooooo, we have NaturalLanguageTranslator hardcoded, and it does nothing :). What do you think about enhance it to allow multi-language in pharo? How would you do it? The most simple way I'm thinking is a set of dictionarys... And a browser to browse them, he. Cheers, Guille

3 2

[Pharo-project] Spy?
by Johan Brichau July 9, 2010

July 9, 2010

Hi, I tried to download SPY in a Pharo1.0 image using its Metacello conf, but it told me that I needed MMProbe and ActivityMonitor first. Is SPY working in Pharo 1.0 ? Or not? thx! ---------------------------- Johan Brichau johan.brichau(a)uclouvain.be

3 3

[Pharo-project] Local time zone is UTC per default on macs
by Max Leske July 9, 2010

July 9, 2010

I filed a bug concerning unsatisfactory timezone offset on macs: http://code.google.com/p/pharo/issues/detail?id=2660. I guess John McIntosh is responsible for this, but I couldn't find his e-mail address on this list. Max

2 1

[Pharo-project] [BetaTesting] [ANN] Pharo-1.1-11406-rc3dev10.07.1
by Mariano Martinez Peck July 9, 2010

July 9, 2010

Hi folks. I am glad to announce the Pharo 1.1 RC3. It has not too much changes in compare to RC2. The main change is the menu structure. I also updated to newest version of external packages. Here is the image: https://gforge.inria.fr/frs/download.php/27241/Pharo-1.1-11406-rc3dev10.07.… We would really appreaciate feedback and help to fix the failing tests. Cheers Mariano

6 13

[Pharo-project] Polymorph question - vertical tabs
by Torsten Bergmann July 9, 2010

July 9, 2010

Does polymorph also support vertical tab groups similar to: http://www.orbifold.net/default/wp-content/uploads/roundlefttab.jpg Thx T. -- GRATIS fÃ¼r alle GMX-Mitglieder: Die maxdome Movie-FLAT! Jetzt freischalten unter http://portal.gmx.net/de/go/maxdome01

2 1

[Pharo-project] Polymorph question - white background on tab group
by Torsten Bergmann July 9, 2010

July 9, 2010

|window tabs contentMorph | window := StandardWindow new model: self. window title: 'Test'. tabs := TabGroupMorph new. tabs addPage: EllipseMorph new label: 'Page 1'. tabs addPage: RectangleMorph new label: 'Page 2'. tabs selectedPageIndex: 1. window addMorph: tabs fullFrame: (LayoutFrame fractions: (0@0 corner: 1@1) offsets: (0@0 corner: 0 @ 0)). window openInWorld. ^window Why is the tab background still white and not aligned with the parent panes color? -- GRATIS fÃ¼r alle GMX-Mitglieder: Die maxdome Movie-FLAT! Jetzt freischalten unter http://portal.gmx.net/de/go/maxdome01

2 1

[Pharo-project] [OT] Mouseless
by Torsten Bergmann July 9, 2010

July 9, 2010

While Squeak keeps the mouse logo others think about getting mouseless: Interesting times with interesting solutions ... http://www.pranavmistry.com/projects/mouseless/ Bye T. -- GMX DSL: Internet-, Telefon- und Handy-Flat ab 19,99 EUR/mtl. Bis zu 150 EUR Startguthaben inklusive! http://portal.gmx.net/de/go/dsl

1 0

[Pharo-project] [Last Call] [JOB] Pharo Engineer at INRIA
by Marcus Denker July 9, 2010

July 9, 2010

[REPOST] Please note: DEADLINE 15 July* INRIA RMoD is looking for a full-time engineer for helping with Pharo development. As researchers, we need a really great system to explore new ideas with. The job of the engineer is to help us by improving the infrastructure we use for our work. What this of course means is that it's kind of a cool job: work in a research lab, but no paper writing, no teaching, no administration. In addition: -> live in France (one hour from Paris, 1h20 from London, 35 min from Brussels) -> try a lot of good beers (our office is just 10km from the border to Belgium) -> enjoy french food -> code all the day in Smalltalk and a bit of C >From what the DHR services told us, the salary will depend on years of expertise and diplomas. We can think that 2700 Euros (before final taxes) should be a good working number for a candidate of the experience we are looking for. In France the employer pays already social security and pension. So the salary already includes lots (and lots) of deductions compared to e.g. a german brutto salary. Holidays: 45 days per year (plus standard bank-holidays). Lille is an active city but not expensive (lots of students...), centrally located, with good connections by train to everywhere (including Paris CDG and Brussels airports). The job will be to continue making Pharo better: - better tools - better network support - better compiler infrastructure - improve the VM - help with day-to-day bug fixing and constant improvement Therefore we are looking for a candidate with some years of Smalltalk (possible Pharo/Squeak) background. Duration: 30 months Starting date: October 2010 Location: Lille (no remote job possible) Official Job Description and Application via the INRIA website http://bit.ly/9z42MY Deadline: 15 July 2010 Web-Post of this call: http://rmod.lille.inria.fr/web/pier/blog/2010-06-07 More about: - RMoD: http://rmod.lille.inria.fr - INRIA Lille: http://www.inria.fr/lille/ - INRIA in General: http://www.inria.fr - Engineers at INRIA: http://bit.ly/cLe6q4 - Pharo: http://pharo-project.org - Lille: - http://en.wikipedia.org/wiki/Lille - http://www.mairie-lille.fr/en - http://wikitravel.org/en/Lille Do not hesitate to contact us for more information. -- Marcus Denker -- http://www.marcusdenker.de INRIA Lille -- Nord Europe. Team RMoD.

1 0

[Pharo-project] FFI and global structs (I think)
by Schwab,Wilhelm K July 9, 2010

July 9, 2010

Andreas, I am using FFI to make calls into the Gnu Scientific Library and am having problems that I do not understand. They provide "types" for various algorithms/functions/etc. which appear to be global structs. Suppose I want to use Haar wavelets; then I have interest in what I believe is a struct exported as gls_wavelet_haar. My simple-minded understanding of such things is that "the name is address" and that dlsym()/GetProcAddress() should provide a void pointer to the correct place in memory. My attempts at doing that have ended badly. If I instead export a function of my own (from a library that links to GSL), something like gls_wavelet_type* WaveletHaar() { return gsl_wavelet_haar; } and call it something like waveletTypeHaar < cdecl: GslWaveletType* WaveletHaar()> then all is well. If I use GslWaveletType fromHandle: with an address from dlsym(), ugliness (segment fault) follows. Am I missing something really fundamental, or should #fromHandle: be able to do the job? Bill

1 0

[Pharo-project] Issues with ObjectiveCPlugin.bundle 1.1.3
by John M McIntosh July 9, 2010

July 9, 2010

Well *cough*, I might have written a couple of dozen tests to check the Objective-C bridge, but obviously if I don't run them, why something broken creeps in. It seems that I managed to make a code change that broke how the unsigned short data is fetched from a objective-c message send. As I was working on the Cog Stack VM for the iPhone today I tripped over the fact that I had somehow managed not to run the test suite. Obviously if I had it would have told me something was busted. I've fixed this and shortly I'll look at building ObjectCPlugin.bundle 1.1.4 and Mac Carbon VM 5.7b2 People who have built their own plugin should fire up VMMaker and rebuild a copy from the latest SqueakSource update -- =========================================================================== John M. McIntosh <johnmci(a)smalltalkconsulting.com> Twitter: squeaker68882 Corporate Smalltalk Consulting Ltd. http://www.smalltalkconsulting.com ===========================================================================

1 0

[Pharo-project] Pharocasts: live testing
by laurent laffont July 8, 2010

July 8, 2010

Hi, Unit tests represents the security belt of software development. But having a security belt doesn't mean it's always fastened. You take a look at a piece code, have fun hacking it, submit the new code without checking all tests are green... it's easy to forgot there's tests here. Autotest automatically runs tests for you. Test Driven Development ensures test exhaustivity, reflexion on design and so on. You stop and run tests very often to check the health of your code. In your car you don't stop driving, hands off the wheel in order to check the speed. Autotest provides a real time dashboard so you don't stop coding. Testing is always here. Now you can drive. http://pharocasts.blogspot.com/2010/07/live-testing-with-autotest.html Cheers, Laurent Laffont http://pharocasts.blogspot.com/ http://magaloma.blogspot.com/

1 0

[Pharo-project] should we move pharo to use webclient?
by StÃ©phane Ducasse July 8, 2010

July 8, 2010

to replace the old code? Stef

8 17

[Pharo-project] [update 1.2] #12043
by StÃ©phane Ducasse July 8, 2010

July 8, 2010

12043 ----- - Issue 2595: Changing color/fillStyle by morph halo menu

1 0

[Pharo-project] Updating VM links for Pharo 1.1
by Mariano Martinez Peck July 8, 2010

July 8, 2010

Hi. Henry made me remember that we need to update the VMs links for Pharo 1.1. We are in RC4 and very very near of the stable release. Thus, I suggest to fix this problem. You can see in http://code.google.com/p/pharo/issues/detail?id=2646 Mac: 4.2.5 Windows: 4.0.2 Linux: 4.0.3.2202 What I am not sure if we should put these links in a particular section of Pharo 1.1 or just update the current VMs links. The problem of doing the last one is that these VMs will be downloaded for both images: 1.0 and 1.1. Are there problems with it ? The only thing I can image is if the vm has FreeType compiled. Cheers Mariano

6 10

[Pharo-project] simple point function seems strange
by StÃ©phane Ducasse July 8, 2010

July 8, 2010

Hi (10@20) rotateBy: #left centerAt: 0 @0 returns 20@ -10 and my math tells me that this is a rotation to the right. Am I correct? Then why I'm wrong? screen coordinates vs. math coordinate? Stef

3 4

[Pharo-project] Teleplace Cog VMs are now available
by Eliot Miranda July 8, 2010

July 8, 2010

Hi All, it gives me great pleasure to announce that the Teleplace Cog VMs are now available. Huge thanks to all at Teleplace who have given me the opportunity to build Cog and release it as open source, been willing guinea pigs braving its bugs, and providing indispensable participation in getting Cog to its current state. Huge thanks are also due to the original Back To The Future team whose VMMaker Cog extends to write the VM, and to Peter Deutsch from whom I've taken many ideas. This release contains two VMs. The Stack VM, is a cross-platform interpreter that uses context-to-stack mapping to achieve modest performance gains. The Cog VM is a just-in-time compiler that currently supports only x86 that builds upon the Stack VM to achieve substantial performance improvements. The release is in the form of a Monticello package containing the VMMaker source and a tarball containing the platform sources, the generated sources and a Squeak 4.1 image containing the VMMaker sources. Download both at http://ftp.squeak.org/Cog/VMMaker-oscog.11.mcz http://ftp.squeak.org/Cog/OpenSourceCog.tar.gz Cog VMs: The Cog VMs are Squeak/Croquet VMs that run closure Squeak/Croquet/Pharo/Cuis images. The VMs support existing plugin source but will require plugins to be recompiled as the VM_PROXY_MAJOR plugin api has been extended. This release contains two distinct VMs, the StackInterpreter and the Cogit. The StackInterpreter is a fully-portable plug-in replacement for the current closure Squeak VMs and images. The Stack VM uses context-to-stack mapping and a somewhat improved garbage collector to achieve modest but useful performance gains in the 10% to 15% range. The StackInterpreter is intended to supersede the Squeak VM on platforms where the Cogit cannot be used. The Cogit extends the StackInterpreter with a just-in-time compiler that uses aggressive inline caching techniques to deliver substantial performance gains in the 3x to 15x range, depending on benchmark. The Cogit currently supports only x86 and the floating-point primitives and parts of the platform support code depend on SSE2. I hope members of the community will attempt to port it, e.g. to ARM, PowerPC and x86-64. The Cogit (excuse the pun) is so named because it is both an interpreter and a JIT, choosing not to generate machine code for large methods, interpreting them instead, the default policy being not to JIT methods with more than 60 literals. The Cog VM requires a few minor image changes all in image/NecessaryImageChangesForCogToWork.1.cs. The JIT's machine-code SmallInteger primitives insist on a SmallInteger receiver so the primitives in LargePositiveInteger = ~= bitAnd: bitOr: butShift: and bitXor: cannot be used and these methods must be deleted. The Cogit inlines the address of the Character instance table, Smalltalk specialObjectsArray at: 25, into the machine-code at: primitive for faster ByteString>>at: and so the table cannot be rebuilt in SmalltalkImage>>recreateSpecialObjectsArray. The new version preserves the existing table. Both VMs maintain floats in platform order to ease implementation of machine code floating-point primitives, and hence internally are in little-endian order instead of big-endian in current Squeak images. While the VMs convert float order automatically on load they do require special accessing primitives Float>>basicAt: & Float>>basicAt:put: that undo the reversal and answer Float contents in big-endian order so that e.g. Float>>hash is unchanged. The methods assume these primitives can fail, allowing the code to be used on current Squeak VMs. The image/VMMaker-Squeak4.1.image is a Squeak 4.1 image, runnable with the current Squeak VMs, that contains these changes, and can hence also be run with a Cog VM. But beware, once an image has been saved on Cog it cannot be run by an existing Squeak VM, because existing VMs cannot undo the Float order change. Platform Subsystem: Most of the platform subsystem is unchanged but there are some important changes that need description. The biggest change is the heartbeat and the clock in platforms/unix/vm/sqUnixHeartbeat.c and platforms/win32/vm/sqWin32Heartbeat.c. The Cog VMs avoid the slow and variable interruptCheckCounter, folding the event check into the stack overflow check on frame build. The heartbeat, typically 500Hz or 1KHz, changes the stackLimit to a value that will always fail. On the next frame building send the VM will enter stack overflow handling that, as a side effect, will also check for events. This is more efficient than the update of interruptCheckCounter and much more regular. If one is running code that executes long-running primitives (e.g. large integer arithmetic) the counter approach will result in too low an interrupt check frequency, and conversely if one is running normal code the interrupt check frequency can be very high. The heartbeat also maintains a 64-bit microsecond clock, UTC microseconds from 1901, from which the backward-compatible millisecond and second clocks are derived. Primitives exist to answer UTC microseconds and local microseconds. Updating the clock in the heartbeat results in a 1 or 2 millisecond resolution but avoids the cost of accessing the OS time on every prim tie which we've found important for performance at Teleplace. The 64-bit microsecond clocks provide a unified time basis and eliminate wrapping (for the next 54,000 years at least). I hope community images will move to these clocks. It's worked well in VisualWorks. Another significant change is in the external semaphore table support code. This is now lock-free at the cost of having to specify a maximum number of external semaphores at start-up (default 256). The support code for the lock-free data structures are processor-specific and is currently implemented only for x86 and gcc-compatible compilers; see platforms/Cross/vm/{sqAtomicOps.h,sqMemoryFence.h}. There is also improved crash reporting code that prints a primitive log and a C backtrace in addition to the Smalltalk backtrace. See platforms/Mac OS/vm/sqMacMain.c, platforms/unix/vm/sqUnixMain.c, platforms/win32/vm/sqWin32Intel.c & platforms/win32/vm/sqWin32Backtrace.c. Finally there is support for the QVMProfiler, a pc-sampling profiler for profiling at the VM level. See platforms/unix/vm/sqUnixVMProfile.c and platforms/win32/vm/sqWin32VMProfile.c. The profiler itself is in the VMMaker image described below in Qwaq-VMProfiling. There are also changes to do with Teleplace-specific extensions to the HostWindowPlugin but these are not essential to Cog. VMMaker and Slang: The image/VMMaker-Squeak4.1.image Squeak 4.1 image contains the complete Cog VMMaker with necessary support code for simulation. This image was used to generate the sources in the src and stacksrc directories. Cog's VMMaker is substantially revised and extended from the current VMMaker. It supports multiple classes, not just Interpreter and superclasses, because both context-to-stack mapping and the Cogit are too complex to write monolithically. Classes can specify ancilliaryClasses and ancilliaryStructClasses, such as CoInterpreterStackPage, CogMethod and CogAbstractInstruction. The Monticello package version is included in the header of all generated files and constitutes the version stamp for generated code. Code is generated in sorted order so that minor changes in the Smalltalk source produce correspondingly minor changes in the generated code. The gnuification step is built-in to VMMaker. No effort has been made to maintain 64-bit compatibility. Apologies, this was unaffordable. The VMMaker generates a single source tree used by all platforms. Instead of deciding at generation time whether to use the Interpreter struct the generated code depends on the SQ_USE_GLOBAL_STRUCT define which can be overridden in platform makefiles. All plugins live in src/plugins and platform makefiles along with plugins.int and plugins.ext files in the build subdirectories decide which plugins are built as external or internal. The VM Generation Workspace from Workspace.text workspace contains dots to generate the sources. We no longer use the VMMakerTool since there should be nothing platform-specific in the generated sources (if we add ports to other ISAs all their source can be included and selected as required by the platform makefiles). Since the Cogit generates x86 machine code simulation is much more complex. There is a support plugin, platforms/Cross/plugins/BochsIA32Plugin that depends on a large simulation of the x86 family implemented in C++ (see processors/IA32/bochs) and on Alien. I use the simulator frequently (but note that I haven't had time to build a working version for Squeak 4.1). I have tested Cog simulation in this image, running on the image/VMMaker-Squeak4.1.image itself. The VM Simulation Workspace in the VMMaker image contains an example doit that starts the simulator. Be patient, even on a fast machine unhibernating the Squeak display background image takes nearly a minute. Native fonts do not (yet) simulate correctly, but the system runs. But note that I have only attempted to build and run the simulator on Mac OS X. I expect Bochs can be built on linux and win32 but I have not tried. By the way, I've not described how to run the Bochs simulator on the current Squeak VM. That's because the plugin depends on the heartbeat to break out of simulation occasionally via a new interpreterProxy entry point setInterruptCheckChain. As this isn't supported by the current Squeak VMs the plugin would require modification. So to simulate first build either of the Cog VMs and then run the simulation with it. There are a number of unpublished changes to the base other than those in NecessaryImageChangesForCogToWork.1.cs. This is partly laziness on my part, partly avoiding publishing things in advance of Cog. These changes are better motivated once Cog is in use. There are changes to the "translated primitives" (see implementors of translatedPrimitives) which replace messages with method tags for generation directives. The Cog VMMaker uses Object>>perform:with:with:with:with: & Object>>perform:with:with:with:with:with: during simulation, and Collection>>#fold: & SquenceableCollection>>#copyUpThrough: during generation. Object>>inline: and Object var:declareC:, which are mispackaged in Kernel in Squeak 4.1 are obsolete (method tags being used instead) and have been removed. I have changed Integer>>hex and Integer>>hex8 back to their original semantics as of 3.8. Backward compatibility is important and one can easily add new selectors if one wants different functionality. VMMaker was here first ;) Tarball: The top-level directories in the tarball are src the tree for the Cog generated sources including all plugins stacksrc/vm the directory containing the Stack VM source (plugins can be taken from above) platforms the usual svn platform tree but including Cog specific changes such as the heartbeat processors the tree containing simulation support code, i.e. the bochs C++ x86 simulation library, along with a potential ARM, PowerPC & MIPS simulator, Skeye. image the Cog-prepared Squeak 4.1 VMMaker image scripts some svn scripts to revert unchanged plugins that haven't really changed cygwinbuild the win32 build directory winbuild the old win32 build directory for minnow gcc 2.95. Not entirely obsolete as the cygwin build as yet fails to generate a functional FFIPlugin macbuild the CoreVM.xcodeproj and support build projects for Mac OS X 10.5 or better unixbuild the build directory for linux Building Cog: Each build directory above contains a HowToBuild file that describes building in more detail. The build directories only contain Cogit makefiles. f you want to build a Stack VM you're on your own but this is very close to the existing Squeak VM build. Status: The Cogit VM has been our sole VM at Teleplace for nearly a year. We do occasionally find bugs and there are almost certainly areas of functionality that we have not touched (for example I know that co-routining does not yet work). If you find a bug please try and create a reproducible test case and let me know. I can't promise to take a look or fix it but I am motivated to do so and will try my best as time allows. Better still if you find and fix bugs be sure to let me know. License (MIT): All contributions from Teleplace in this release are Copyright (c) 2010 Teleplace, Inc. Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. Eliot Miranda June 2010

8 8

[Pharo-project] [update 1.2] #12041
by StÃ©phane Ducasse July 7, 2010

July 7, 2010

12041 ----- - Issue 2642: Improve the PopupChoiceDialogWindow/PopupChoiceOrRequestDialogWindow

3 3

[Pharo-project] [update 1.2] #12042
by StÃ©phane Ducasse July 7, 2010

July 7, 2010

Trying some enh of scriptloader on a simple enh.... 12042 ----- - Issue 2657: some Point enhancements

1 0

[Pharo-project] Monticello & proxy with security.
by Diogenes Moreira July 7, 2010

July 7, 2010

hi folks. right now, i'm behind a squid proxy with security (user & pass).. in the preferences, I find the proxy configuration, but only have ip and port. Exists a way to set user and password in that connection??? Thanks in advance.

3 3

[Pharo-project] [update 1.2] #12041
by StÃ©phane Ducasse July 7, 2010

July 7, 2010

12041 ----- - Issue 2635: Part 11

1 3

[Pharo-project] Fix the other dependencies of asContextWithSender:
by stephane ducasse July 7, 2010

July 7, 2010

Eliot should this fix be applied to pharo? http://code.google.com/p/pharo/issues/detail?id=2655 Is there a dependency? Stef

3 6

[Pharo-project] Suggestions for hotel/apartment for ESUG2010 in Barcelona?
by Stephan Eggermont July 7, 2010

July 7, 2010

Does anyone have suggestions for a good (and cheap) place to stay during ESUG2010 in Barcelona? The conference is not in the city centre, and the only hotel which is close by (NH CornellÃ ) has no places available. We thought of renting a large appartment somewhere along the metro line 5 and sharing. Anyone interested? Stephan & Diego

1 0

[Pharo-project] [update 1.2] #12040
by StÃ©phane Ducasse July 7, 2010

July 7, 2010

12039 ----- Issue 2635: Preparation. 12040 ----- Issue 2635: Uppercase extensions - first mail.

1 0

[Pharo-project] feedup of selecting doit or typing " above and below expressions in comments?
by Torsten Bergmann July 7, 2010

July 7, 2010

>This was removed on request of Stef and I completely agree on this: >for me it is time consuming to be forced to search for the selector in >a sorted list when the selector is written/selected right in front of >me. Maybe I dont understand what you wrote - but if I understand you correctly OB had the same behavior as in Core and this was removed from OB on request of Stef? Again: when I have a method foo self sendMessage: myMessage with: text. self sendSecondMessage: myMessage. A) In Core: I'm able to browse the senders/implementors of all three selectors using the sorted list provided by the "Senders of..."/"Implementors of..." menu: #(foo, sendMessage:with:, sendSecondMessage:) where the methods selector (#foo) of the selected method is explicitly at the top. B) In Pharo (with OB): I can only check the senders/implementors of #foo which is limiting and makes it hard to search for s/i of the other messages sent from within the method body. => So I'm limited with Pharo(OB) compared to Standard browser in Core. Bye T. -- GMX DSL: Internet-, Telefon- und Handy-Flat ab 19,99 EUR/mtl. Bis zu 150 EUR Startguthaben inklusive! http://portal.gmx.net/de/go/dsl

3 2

[Pharo-project] Pharo on OLPC XO kid notebook
by Hilaire Fernandes July 7, 2010

July 7, 2010

OLPC moto is based from day one on the use of dynamic language to design the content coming with their notebook. It is to write content but also to allow kids to examine source code, edit it and test it ; as we are doing continuously in Smalltalk. They even think about a source-code key in the keyboard to instantaneously open a source code browser exposing the code of a running application. Now that Pharo 1.1 is in good sigh and as we are not ashame any more to show how Smalltalk looks like we could also produce a release of Pharo for OLPC XO Sugar environment. It is definitely a good opportunity to bring an interesting tool to neutral mind (kids). In the other hand, it will definitely help to bring awareness about Pharo. For example, without any action of my part, DrGeoII was downloaded about 180'000 times in one year for XO. Any takers? Hilaire PS: I have a couple of XO machines for testing purpose.

4 9

[Pharo-project] I updated the how to contribute web page
by stephane ducasse July 7, 2010

July 7, 2010

please comment :) http://code.google.com/p/pharo/wiki/HowToContribute Stef

4 6

[Pharo-project] Markup Builder in Smalltalk
by Sudhakar Krishnamachari July 7, 2010

July 7, 2010

Accepted and committed as alpha: 1.13 in http://squeaksource.com/PharoGoodies/ modified api is: newBldr := XMLBuilder new. newBldr xmlPreprocessor; attributes: #(âa=â6â³â âb=â7â³â); test: âwelcome to buildersâ with: [ newBldr attributes: #('a="1"' 'b="2"'); trial: ''; attributes: #('a="3"' 'b="4"'); trial: 'testing..' ; trial ]. to get: '<?xml version="1.0" ?><test a="6" b="7" >welcome to builders<trial a="1" b="2" ></trial><trial a="3" b="4" >testing..</trial><trial></trial></test>' hope this aligns.. lots more stil to be cleaned for plugins, will do so by the next commit.. -Skrish ***** There is an interest! Please submit/publish it... Now What is the exact API I found cNode: a bit a bad name. What does it mesn childrenNode: newBldr := XMLBuilder new xmlPreprocessor. newBldr mbAttr: #(âa=â6â³â âb=â7â³â); test: âwelcome to buildersâ why new and xmlPreprocessor mbAttr: sucks also! Could you shine a bit the API to be reader friendly? Stef

2 1

[Pharo-project] Mysql Port of Josh Mcdonald Driver
by Sudhakar Krishnamachari July 7, 2010

July 7, 2010

Hi, Can anyone forward the latest version of Alan Reider's fixes of the Mysql driver. I cant access the squeak map I get link of.. -Skrish

2 1

[Pharo-project] feedup of selecting doit or typing " above and below expressions in comments?
by Torsten Bergmann July 6, 2010

July 6, 2010

Henry wrote: > Personally, being used to VisualWorks, I always get bitten by having to >select the entire selector for implementors/senders... :) +1 Another thing that is related to the above and to the different browsers: In Core: When I select a method I can choose "Senders of..." and "Implementors of..." to either search for senders/implementors of the selected method AND the messages sent into the selected method. In Pharo: with OB I can just do "senders" and "implementors" which looks for the senders/implementors of the selected method One is NOT able to search for senders/implementors of the messages sent within the selected method. Here I have to select the entire selector which is time consuming as described above. So Pharo with OB is (compared to core) really a step backwards here. -- GRATIS fÃ¼r alle GMX-Mitglieder: Die maxdome Movie-FLAT! Jetzt freischalten unter http://portal.gmx.net/de/go/maxdome01

3 3

[Pharo-project] Markup Builder in Smalltalk
by Sudhakar Krishnamachari July 6, 2010

July 6, 2010

I may perhaps be unaware of the XML/SGML builder generation from Seaside, love to be pointed out to.., The Html generation of Seaside uses explicit construct of classes ( WA**Tags) to do so.. Builders in dynamic manner use doesNotUnderstand: construct to generate the stream.. The strength in this approach is a lighter construction.. just one object creation for the entire sgml stream creation.. Extensibility too therefore is available by default.. any tag/ attribute can be streamed without any extension to creating a new class/ method.. The offside though is, it is streaming so no handling the DOM / tree for manipulations.. Infact I feel it can be a good alternative mode to current seaside markup streaming for the Html streaming ( for static content) too..but that is specialization of its use.. and not as yet worked out for handling the callbacks and all special abilities of Seaside html streaming. ***** Seaside provides that kind of functionality. Lukas 2010/7/6 Sudhakar Krishnamachari <skrishnamachari at gmail.com <http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project>>: >* Is there any general interest in Markup Builder...a la Groovy..? *>* *>* http://skrishnamachari.wordpress.com/2010/07/06/markupbuilder-in-smalltalk/ *>* *>* *>* *>* _______________________________________________ *>* Pharo-project mailing list *>* Pharo-project at lists.gforge.inria.fr <http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project> *>* http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project *>* * -- Lukas Renggli www.lukas-renggli.ch

3 2

[Pharo-project] Morph rendering with antialiasing
by Denis Kudriashov July 6, 2010

July 6, 2010

Hello. Is somebody implement morphs rendering with antialiasing? Maybe it is already possible? In practice I need circle button. But without antialiasing it seems really ugly. If there is no existed solution. How do you think It can be implemented? I know that fonts rendered with antialiasing. How its approach can be applyed for morphs? I notice browser buttons (senders, implementors..) has rounded corners. And its nice rendered. What secrete? Can it be done for any morph corner rounding?

2 1

[Pharo-project] Cleaning WriteBarrier (as used in Magma)
by Torsten Bergmann July 6, 2010

July 6, 2010

Hi Avi, how is life at Twitter ;) Would it be possible that you add me as a developer on http://www.squeaksource.com/WriteBarrier so I can clean up underscore assignment in the package. WriteBarrier is used in Magma (see ConfigurationOfMagma) which is installable right from a workspace in the upcoming Pharo 1.1 release. Currently this fails since by default underscore assignment is not allowed in Pharo. Thanks Torsten -- GMX DSL: Internet-, Telefon- und Handy-Flat ab 19,99 EUR/mtl. Bis zu 150 EUR Startguthaben inklusive! http://portal.gmx.net/de/go/dsl

5 4

[Pharo-project] Markup Builder in Smalltalk
by Torsten Bergmann July 6, 2010

July 6, 2010

> More or less exactly the same..except the syntax.. I think the syntax in my XMLWriter package is more oriented towards clean Smalltalk, while yours reminds me on some cryptic languages I used in the past ;) What does #mbAttr: mean, "markup builder" or "my bytes" or ...? I will never understand why people invent languages to save bytes in writing/typing...especially after autocompletion was invented. >see means of merging the frameworks for best use.. I doubt I will accept this syntax style. So it is better when you provide an own package. An additional idea: =================== I can also imagine another kind of XMLBuilder where one can use the DNU trick to write any tag (with all the pro and cons of the DNU hack): builder := XMLBuilder xml. builder person: [ builder name: 'Luke'; planet: 'Moon' number: 12] builder xml then results in <xml> <person> <name>Luke</name> <planet number="12">Moon</planet> </person> </xml> Bye T. -- GRATIS fÃ¼r alle GMX-Mitglieder: Die maxdome Movie-FLAT! Jetzt freischalten unter http://portal.gmx.net/de/go/maxdome01

1 0

[Pharo-project] Markup Builder in Smalltalk
by Sudhakar Krishnamachari July 6, 2010

July 6, 2010

More or less exactly the same..except the syntax.. but includes the CDATA, Comments, plainText, preproc etc.. instead of tag: with: .. I went off to.. {theTag}: theText cNode: [] mbAttr: taking a collection of assocs or strings.. for the attrib. tabs are integrated with valueStringTrimmed dummping it without tabs/crlf.. will see if I can check in this too as an alternative to see means of merging the frameworks for best use.. ***** Just check "XMLWriter-tbn.5" which can be found in "http://squeaksource.com/PharoGoodies/" <http://squeaksource.com/PharoGoodies/%22>

2 1

[Pharo-project] Markup Builder in Smalltalk
by Sudhakar Krishnamachari July 6, 2010

July 6, 2010

Is there any general interest in Markup Builder...a la Groovy..? http://skrishnamachari.wordpress.com/2010/07/06/markupbuilder-in-smalltalk/

3 2

Re: [Pharo-project] Release Core?
by Adrian Lienhard July 6, 2010

July 6, 2010

(moving this thread to pharo-dev) Context: green tests for Pharo 1.1 (currently at RC4). Thanks Mariano, I went through your list of tests. In summary, what needs to be done: - Fixes for Nile - Fix one failing Shout test - Adaptation/fixes of external packages: method wrappers, sound, morphic examples In detail: Needs fixing ----------- SHParserST80Test>>#testNumbers NSMultiByteBinaryOrTextStreamTest>>#testFileOut NSSmartRefStreamTest>>#testGlobal NSMCPseudoFileStreamTest>>#testFileOut NSMCPseudoFileStreamTest>>#testFileOutFileIn NSRWBinaryOrTextStreamTest>>#testFileOut NSRWBinaryOrTextStreamTest>>#testFileOutFileIn ReleaseTest>>#testUndeclared Undeclared classes and variables (these external packages need to be adapted or should not be loaded) --------- WatchMorph references ColorPickerMorph AbstractSound references FillInTheBlank EnvelopeEditorMorph, MidiInputMorph, and ScorePlayerMorph reference PopUpChoiceMorph KeyboardMorphForInput, MidiInputMorph, and ScorePlayerMorph reference SimpleSwitchMorph ArchiveViewer references StandardFileMenu ObjectAsMethodWrapper has undeclared variable named selector and wrappedClass Fixed, see attached cs (we had that already in 1.0 but apparently it was lost for 1.1) ----------- WeakSetInspectorTest>>#testSymbolTableM6812 Workes for me (when running each test alone)(?) ----------- ObjectAsOneTimeMethodWrapperTest>>#testInstallOnClassCategory MCInitializationTest>>#testWorkingCopy WeakRegistryTest>>#testFinalization FileStreamTest>>#testPositionPastEndIsAtEnd AllocationTest>>#testOneGigAllocation Cheers, Adrian On Jul 5, 2010, at 22:06 , Mariano Martinez Peck wrote: > On Mon, Jul 5, 2010 at 8:00 PM, Adrian Lienhard <alienhard(a)netstyle.ch>wrote: > >> I updated the download page. >> >> > thanks > > >> Cheers, >> Adrian >> >> BTW, what is the state of the tests? Last time I run them, there were a lot >> of failures... >> >> > much better. I fixed a lot of Nile with your suggestion and further > problems/fixes. > > My results so far are these: > > 9764 run, 9697 passes, 53 expected failures, 8 failures, 6 errors, 0 > unexpected passes > Failures: > AllocationTest>>#testOneGigAllocation > NSMultiByteBinaryOrTextStreamTest>>#testFileOut > SHParserST80Test>>#testNumbers > ReleaseTest>>#testUndeclared > FileStreamTest>>#testPositionPastEndIsAtEnd > WeakRegistryTest>>#testFinalization > MCInitializationTest>>#testWorkingCopy > ObjectAsOneTimeMethodWrapperTest>>#testInstallOnClassCategory > > Errors: > WeakSetInspectorTest>>#testSymbolTableM6812 > NSSmartRefStreamTest>>#testGlobal > NSMCPseudoFileStreamTest>>#testFileOut > NSMCPseudoFileStreamTest>>#testFileOutFileIn > NSRWBinaryOrTextStreamTest>>#testFileOut > NSRWBinaryOrTextStreamTest>>#testFileOutFileIn > > > > > >> ___________________ >> http://www.adrian-lienhard.ch/ >> >> On Jul 5, 2010, at 18:26 , StÃ©phane Ducasse wrote: >> >>> Yes >>> >>>> I have just created the image RC4. It is in >> https://gforge.inria.fr/frs/download.php/27254/Pharo-1.1-11409-rc4dev10.07.… >>>> >>>> Adrian, do you think we should update the link of 1.1 from the website >> to this one? >>>> >>>> Cheers >>>> >>>> Mariano >>> >> >>

5 4

[Pharo-project] python documentation rocks
by laurent laffont July 6, 2010

July 6, 2010

Hi, Python 2.7 has just been released, documentation is up to date. http://docs.python.org/ That's so great. I hope we will have a process for Pharo's documentation. Is it possible to automatically generate a changelog for Pharo 1.1 ? Cheers, Laurent

2 1

[Pharo-project] [update 1.2] #12037
by Marcus Denker July 6, 2010

July 6, 2010

12037 ----- Issue 2590: Remove PreferenceBrowser package Issue 2645: Remove System Package -- Marcus Denker -- http://www.marcusdenker.de INRIA Lille -- Nord Europe. Team RMoD.

2 1

[Pharo-project] DateAndTime>>asUnixTime should ignore timezones?
by Nick Ager July 6, 2010

July 6, 2010

I've filed a new issue: #2644<http://code.google.com/p/pharo/issues/detail?id=2644>

2 3

[Pharo-project] [update 1.2] #12038
by Marcus Denker July 6, 2010

July 6, 2010

12038 ----- Issue 2648: LanguageEnvironment>>setupSqueaklandSpecifics Issue 2647: Consistency by providing uppercase menu items in Tools menus Issue 2649: World tasklist navigation Issue 2644: DateAndTime>>asUnixTime should ignore timezones? -- Marcus Denker -- http://www.marcusdenker.de INRIA Lille -- Nord Europe. Team RMoD.

1 0

[Pharo-project] PluginBasedJPEGReadWriter anomaly
by Gary Chambers July 6, 2010

July 6, 2010

Hi all. Iv'e noticed a slight anomaly with the PluginBasedJPEGReadWriter when reading jpeg files (at least on Windows platforms). The upper left pixel appears to be white and the whole left-hand column of pixels have anomalous values. Attached, for reference, are screenshots when using the plugin, the original Smalltalk implementation and Windows picture viewer. Regards, Gary

2 2

[Pharo-project] World keyboard navigation for tasklist
by Gary Chambers July 6, 2010

July 6, 2010

New issue http://code.google.com/p/pharo/issues/detail?id=2649 Regards, Gary

1 0

[Pharo-project] Issue 2647 fixed for 1.2 - consistent menu items with uppercase for all tools
by Torsten Bergmann July 6, 2010

July 6, 2010

It is a change that affects the way people are used to the system - but I think we should integrate it to align with already cleaned up menus and known host platforms. "do it" -> "Do it" ... and all tool menu items starting with uppercase. Also the buttons I was able to find. Ready to be harvested for 1.2: http://code.google.com/p/pharo/issues/detail?id=2647 Bye T. -- GMX DSL: Internet-, Telefon- und Handy-Flat ab 19,99 EUR/mtl. Bis zu 150 EUR Startguthaben inklusive! http://portal.gmx.net/de/go/dsl

2 1

[Pharo-project] Changes not logged
by Lukas Renggli July 6, 2010

July 6, 2010

I noticed that "World > Tools > Recover lost changes ..." does not log any longer method removals. Also class definitions, class removals, and do-its are missing in the .changes file. I think this is since Pharo1.1rc3, but maybe earlier. Did anybody else notice that too? Lukas -- Lukas Renggli www.lukas-renggli.ch

2 1

[Pharo-project] feedup of selecting doit or typing " above and below expressions in comments?
by stephane ducasse July 6, 2010

July 6, 2010

with this little change we can execute simply run in the following line just by having the mouse on the line and pressing doit even if there are comments around.... "self run" We should stop to turn around the system all the system but fix it to make us more productive.

3 2

July 6, 2010

Begin forwarded message: > From: Geert Claes <geert.wl.claes(a)gmail.com> > Date: July 6, 2010 10:11:32 AM GMT+02:00 > To: Smalltalk Research <smalltalk-research(a)googlegroups.com> > Subject: [smalltalk-research] FYI: Smalltalk/Seaside developer > Reply-To: smalltalk-research(a)googlegroups.com > > Boris Popov on the Seaside group on LinkedIn wrote: > >> We are looking for another Smalltalk/Seaside developer to join our team at DeepCove Labs http://goo.gl/LyXE

1 0