November 2011 - Pharo-dev - lists.pharo.org

[Pharo-project] FS and binary files

by Johannes Rasche

Hi folks, I just started to using FS instead of FileDirectory. On readStream sent to FSReference a FSAnsiiStream turns out, which is not suitable for reading images ( for instance with PNGReadWriter). What did I wrong? Johannes

Nov. 9, 2011

Re: [Pharo-project] GlorpDBX PostgreSQL Patch

by Mariano Martinez Peck

Thanks Sven. We let the repo as global read because we usually receive unintentional commits. Do you think it is worth it to have a kind of DBXTalkInbox ? Alan, do you think this is useful also for Glorp trunk ? Thanks! On Tue, Nov 8, 2011 at 5:59 AM, Sven Van Caekenberghe <sven(a)beta9.be> wrote: > Hi, > > I have a patch for Glorp(DBX) PostgreSQL to support the newer bytea hex > format (see > file:///usr/local/pgsql/share/doc/html/datatype-binary.html#AEN5037). I > can't write to the repository, so here it is: > > PostgreSQLPlatform>>#convertSQLStringToByteArray: aString for: aType > aString isNil ifTrue: [ ^ nil ]. > ^ (aString beginsWith: '\x') > ifTrue: [ self convertHexSQLStringToByteArray: aString ] > ifFalse: [ self convertEscapeSQLStringToByteArray: aString ] > > PostgreSQLPlatform>>#convertEscapeSQLStringToByteArray: aString > | aStream str | > aStream := (ByteArray new: aString size // 4) writeStream. > str := aString readStream. > [str atEnd] whileFalse: [ |nextChar| > nextChar := str next. > aStream nextPut: (nextChar = $\ > ifTrue: > [str peek = $\ > > ifTrue: [str next asInteger] > > ifFalse: [Number readFrom: (str next: 3) base: 8]] > ifFalse: > [nextChar asInteger]) > ]. > aStream close. > ^aStream contents > > PostgreSQLPlatform>>#convertHexSQLStringToByteArray: aString > ^ ByteArray > new: (aString size // 2) - 1 > streamContents: [ :out | | in | > (in := aString readStream) skip: 2. > [ in atEnd ] whileFalse: [ > out nextPut: (Number readFrom: (in next: 2) > base: 16) ] ] > > #convertEscapeSQLStringToByteArray: is the old code, unmodified; > #convertHexSQLStringToByteArray: is new. > This might be of use upstream as well, I don't really know. > > If it would help, I could save this as an MC package somewhere. > > Regards, > > Sven > > > -- Mariano http://marianopeck.wordpress.com

Nov. 9, 2011

Re: [Pharo-project] KeyboardEvent differences between platforms

by Sean P. DeNigris

Fernando olivero-2 wrote: > > 1. any modifier key WinVMs produce an DOWN and UP event, and MAC VM's > dont produce anything. > Where did you put your hook? In 1.3-rc with Jenkins Carbon JIT VM, InputEventSensor>>handleEvent: sees #(1 102468 341 741 0 8 0 1) "Cmd pressed" #(1 102704 341 741 0 0 0 1) "Cmd unpressed" #(1 103210 341 741 0 4 0 1) "Alt pressed" #(1 103456 341 741 0 0 0 1) "Alt unpressed" #(1 105053 341 741 0 2 0 1) "Ctrl pressed" #(1 105215 341 741 0 0 0 1) "Ctrl unpressed" but they're classified as mouse events (first element). See also http://forum.world.st/Task-List-td4014056.html Sean -- View this message in context: http://forum.world.st/KeyboardEvent-differences-between-platforms-tp3995799… Sent from the Pharo Smalltalk mailing list archive at Nabble.com.

Nov. 9, 2011

Re: [Pharo-project] KeyboardEvent differences between platforms

by Fernando Olivero

Hi Guille, this is what i've captured by hooking up a GauchoWorld to listen to vm events. Below you can find the sequences of keyDown, KeyChar, KeyUp, that results from different keyboard presses, in both Mac and Win VM's. In short, when pressing 1. any modifier key WinVMs produce an DOWN and UP event, and MAC VM's dont produce anything. 2. a modified keystroke ( modifer + key ), WinVMS produce a DOWN,DOWN,CHAR,UP,UP, whilst MacVm's produce a DOWN,CHAR,UP. 3. regarding the content of the Control key presses, both WinVm's and MacVms, produce an incorrect value of the keychar when they announce the CHAR event. Not when they announce DOWN and UP. 4. On command, alt, shift both VM's produce the correct values on the CHAR event. Fernando pd1: I've didn't collect results for the Linux VM. pd2: DATA collected VM "Sensor Events" format: #(type stamp ascii type buttons unicode 0 1 ) -------------------------------------------------------------------------------------------- MAC OS X VM ----------------------------------------------------------------------- SINGLE Keystrokes Control: nothing Command: nothing Shift: nothing ----------------------------------------------------------------------- Modified KEYSTROKES Commandl+$f - Down, keyStroke, Up #(2 523126 3 1 8 0 0 1) #(2 523126 102 0 8 102 0 1) #(2 523126 3 2 8 0 0 1) Control+$c - Down #(2 156689 0 2 2 0 0 1) ALT+$a - Down, keyStroke, Up #(2 165404 0 1 4 0 0 1) #(2 165404 140 0 4 229 0 1) #(2 165483 0 2 4 0 0 1) SHIFT + $s, Down, KeyStroke,Up #(2 734260 1 1 1 0 0 1) #(2 734260 83 0 1 83 0 1) #(2 734362 1 2 1 0 0 1) ----------------------------------------------------------------------- Unmodified KEYSTROKES Right Arrow-down, keystroke,up #(2 2329743 125 1 0 0 0 1) #(2 2329743 31 0 0 31 0 1) #(2 2329838 125 2 0 0 0 1) -------------------------------------------------------------------------- WINDOWS VM Shift- Down and Up #(2 123006507 16 1 1 16 0 1) #(2 123007958 16 2 0 16 0 1) Control- Down and Up #(2 123110575 17 1 2 17 0 1) #(2 123111589 17 2 0 17 0 1) ALT- Down and Up #(2 123205408 18 1 8 18 0 1) #(2 123208201 18 2 0 18 0 1) Control+$a - Down, Down, keyStroke, Up, Up #(2 123584803 17 1 2 17 0 1) #(2 123585037 65 1 2 65 0 1) #(2 123585037 1 0 2 1 0 1) #(2 123585099 17 2 0 17 0 1) #(2 123585115 65 2 0 65 0 1) Control+$b - Down, Down, keyStroke, Up, Up #(2 126935049 17 1 2 17 0 1) #(2 126935377 66 1 2 66 0 1) #(2 126935377 2 0 2 2 0 1) #(2 126935439 66 2 2 66 0 1) #(2 126935501 17 2 0 17 0 1) Control+$h - Down, Down, keyStroke, Up, Up #(2 126503769 17 1 2 17 0 1) #(2 126503878 72 1 2 72 0 1) #(2 126503878 8 0 2 8 0 1) #(2 126503925 17 2 0 17 0 1) #(2 126504003 72 2 0 72 0 1) ALT+$h - Down, Down, keyStroke, Up, Up #(2 126738160 18 1 8 18 0 1) #(2 126738316 72 1 8 72 0 1) #(2 126738316 104 0 8 104 0 1) #(2 126738379 72 2 8 72 0 1) #(2 126738410 18 2 0 18 0 1) SHIFT+$h - Down, Down, keyStroke, Up, Up #(2 126866175 16 1 1 16 0 1) #(2 126866721 72 1 1 72 0 1) #(2 126866721 72 0 1 72 0 1) #(2 126866830 16 2 0 16 0 1) #(2 126866830 72 2 0 72 0 1) On Sun, Nov 6, 2011 at 3:07 PM, Guillermo Polito <guillermopolito(a)gmail.com> wrote: > In TextEditor>>initializeShiftCmdKeyShortcuts the shortcuts are set in a way > that it creates three handlers for each shortcut to make it work xD. > > The main difference here is that in Linux, ubuntu 11.04 the Ctrl+a comes > with the right keyCharacter, but in Windows and MacOS it does not :(.Â We've > tested that on host OSs with friends :). > > I've captured the problem I'm experimenting so far and wrote down an ugly > hack in this method. > > KeyboardEvent>>modifiedCharacter > Â Â Â self flag: #hack. > Â Â Â "Hack me.Â When Ctrl is pressed, the key ascii value is not right and we > have to do this shit" > Â Â Â ^(self controlKeyPressed and: [ (#(MacOSX Windows) includes: Smalltalk > os current platformFamily) and: [ keyValue <= 26 ]]) > Â Â Â Â Â Â Â ifTrue: [ (self keyValue + $a asciiValue - 1) asCharacter ] > Â Â Â Â Â Â Â ifFalse: [ self keyCharacter ] > > > But I'd like to replace that by: > > KeyboardEvent>>modifiedCharacter > Â Â Â ^self keyCharacter > > :P > > > Are there any technical reasons why the Sensor behaves like that?Â How > difficult is it to make it work the same for all three vms? > > Thanks! > Guille >

Nov. 9, 2011

Re: [Pharo-project] NewTextMorph

by Hilaire Fernandes

Hi Fernando, Thanks for the feedback. I also saw these classes were not used, so I sticked to TextMorph. Hilaire Le 09/11/2011 10:17, Fernando Olivero a Ã©crit : > Hi Hillary, the classes you mention were created by me when Pharo > wasnt using the CUIS editors, they allowed me to plug them into Pharo, > coming up with a text morph form scratch. > > Nobody used them since them, except me in Gaucho. So i've discontinued > maintaing them in Pharo. I would suggest you to use the other text > morphs classes (TextMorph, PluggableTextMorph, etcâ¦), which are > maintained and used currently in the system. > > I will create an issue to remove them from the system, because now is > confusing, to have more than one option. > > Fernando > > On Sat, Nov 5, 2011 at 4:20 PM, Hilaire Fernandes > <hilaire.fernandes(a)edu.ge.ch> wrote: >> I want to plug a multiple line text morph a DrGeo canvas. The morph >> should display the text (yes) and be editable on place when the user ask >> for. >> >> Which class should I use ? >> EditableTextMorph, EntryFieldMorph or NewTextMorph >> >> >> The comments of the class EntryFieldMorph wrote about example1 and >> example2 methods bit I dont see it (I am using Pharo 1.3) >> >> Hilaire >> >> >> -- >> Education 0.2 -- http://blog.ofset.org/hilaire >> >> >> > > -- Education 0.2 -- http://blog.ofset.org/hilaire

Nov. 9, 2011

Re: [Pharo-project] NewTextMorph

by Fernando Olivero

Hi Hillary, the classes you mention were created by me when Pharo wasnt using the CUIS editors, they allowed me to plug them into Pharo, coming up with a text morph form scratch. Nobody used them since them, except me in Gaucho. So i've discontinued maintaing them in Pharo. I would suggest you to use the other text morphs classes (TextMorph, PluggableTextMorph, etcâ¦), which are maintained and used currently in the system. I will create an issue to remove them from the system, because now is confusing, to have more than one option. Fernando On Sat, Nov 5, 2011 at 4:20 PM, Hilaire Fernandes <hilaire.fernandes(a)edu.ge.ch> wrote: > I want to plug a multiple line text morph a DrGeo canvas. The morph > should display the text (yes) and be editable on place when the user ask > for. > > Which class should I use ? > EditableTextMorph, EntryFieldMorph or NewTextMorph > > > The comments of the class EntryFieldMorph wrote about example1 and > example2 methods bit I dont see it (I am using Pharo 1.3) > > Hilaire > > > -- > Education 0.2 -- http://blog.ofset.org/hilaire > > >

Nov. 9, 2011

Re: [Pharo-project] [Moose-dev] mooseentitystate

by jannik.laval

Hi Stef, It is not so simple :) Take this example from Moose: === LANNode methodDictionary collect:[:e | e timeStamp ifNotEmpty: [:s | s subStrings first ]] === LANNode has no author, so if you take the first substring you will have the date of creation. I have not a good solution, but for my scripts, I test the first character of the substring: if it is a letter, this is an author, else, there is no author. === ((myString subStrings) first at:1) isLetter === Cheers, On Nov 6, 2011, at 21:48 , StÃ©phane Ducasse wrote: > http://code.google.com/p/pharo/issues/detail?id=4971 > >> >>> In moose ? because in pharo this is like that and the changes I added where just to make sure that all the clients do not have to deal with timestamp stupid parsing. >>> >> > > > _______________________________________________ > Moose-dev mailing list > Moose-dev(a)iam.unibe.ch > https://www.iam.unibe.ch/mailman/listinfo/moose-dev --- Jannik Laval

Nov. 9, 2011

Re: [Pharo-project] Slides Pharo talk Smalltalks

by Schwab,Wilhelm K

Nice presentation! I look forward to Stef's annotations and the video. One thing jumped off the page: the intersection of OpenGL, Sig and NativeBoost. It says "done on Windows" which I *hope* means that Linux and Mac are planned??? Being able to embed 3D worlds in "normal" desktop software would be huge (virtual reality/visualization, simulation, etc.). Question: does this suggest that NativeBoost should advance in the FFI roadmap? AFAIK, I do not care whether FFI, Alien or NativeBoost becomes the first complete FFI solution for us. I mostly want to see callbacks and some level of multi-threading. It would be nice if existing FFI method definitions continued to work, but that's negotiable. On threads, the most important objective is to be able to call into a library without blocking the entire image. If callbacks are not possible from such a function, I could learn to live with that - it's more than we have now. If allowed, I would expect that callbacks would be limited to the calling Process in some way. One could/should use queues, semaphores or deferred actions to make things felt in the UI. Bill ________________________________________ From: pharo-project-bounces(a)lists.gforge.inria.fr [pharo-project-bounces(a)lists.gforge.inria.fr] on behalf of StÃ©phane Ducasse [stephane.ducasse(a)inria.fr] Sent: Tuesday, November 08, 2011 4:54 PM To: Pharo-project(a)lists.gforge.inria.fr Subject: Re: [Pharo-project] Slides Pharo talk Smalltalks I should release my annotated version because it explains the vision behind each itemâ¦ Should fish it. On Nov 8, 2011, at 8:35 PM, Sven Van Caekenberghe wrote: > > On 08 Nov 2011, at 18:25, Marcus Denker wrote: > >> Here are the Slides of the Pharo talk at Smalltalks2011: >> >> PDF: http://www.marcusdenker.de/talks/11Smalltalks/2011-Smalltalks-Denker-Pharo.… >> SlideShare: http://www.slideshare.net/MarcusDenker/pharo-one-year-later >> >> All talks where filmed, so there will be a video later at http://www.fast.org.ar/smalltalks2011 > > Beautiful and cool slides, as always, Marcus. > > Yes the Pharo community did a lot of work this passed year, and the future looks really exciting ! > > It feels good to be part of this. > > Thx for sharing. > > Sven > > > >

Nov. 9, 2011

Re: [Pharo-project] About implementors and senders

by Schwab,Wilhelm K

Guille, all, When in doubt, look at Dolphin. It has the best IDE I have ever used. That's all the more true of 5.x. In Pharo, the way one gets to implementors and senders could be better. I am both amazed by the system's ability to find a method name in selected text, and disappointed by that fact that I have to select text vs. just getting a list of methods from which to choose. Choose senders or implementors in Dolphin (context menu in the method lists), and it provides a list (the selected method in bold at the top), and the locally sent/referenced methods below that. Dolphin also has a command that uses a text prompter to enter the selector, and it allows wildcards. Pharo's MessageNames/Finder tools do at least as much as Dolphin's version of the command. My only gripe with Pharo's version of the search is that I sometimes get too many matches; regex will probably fix it if I would just work at it<g>. Overall, Pharo is on course to do this really well. Regardless of how one obtains a method list, Dolphin does a (much) better job of presenting the results, mostly by providing sortable multi-column lists. Looking for #add: over the entire system can be painful, but #add: defined or used in a particular package or class can be very manageable. Dolphin makes that easy to see by sorting on the appropriate package/class/selector column. Bill ________________________________ From: pharo-project-bounces(a)lists.gforge.inria.fr [pharo-project-bounces(a)lists.gforge.inria.fr] on behalf of Guillermo Polito [guillermopolito(a)gmail.com] Sent: Monday, November 07, 2011 10:11 PM To: Pharo Development Subject: [Pharo-project] About implementors and senders Hi! http://code.google.com/p/pharo/issues/detail?id=4547 That is a bug entry about the implementors and senders behavior when selecting this or that in different GUIs. So, I think the whole UI has to be rethought (in little steps), and arguing about this is healthy. So, the places where searching implementors and senders today should be: 1) a text editor 2) a message/method list 3) an inspector ? 4) ? 1) In a text editor the behavior is obvious for me :) 2) This is the discussion in the issue. The browser asks about senders/implementors of messages sent in the selected method. The senders/implementors browser just browses again the message. I think that searching the implementors of #something while I'm browsing the implementors of #something has no sense, hehe. But the other behavior (the one in the browser) does not convince me either... 3, 4 and etcetera) I'm not sure about anything, hehe. Cheers, Guille

Nov. 9, 2011

Re: [Pharo-project] Loading Projects

by Jeff Gray

I happily feel like an idiot in English and it's my first language :-) -- View this message in context: http://forum.world.st/Loading-Projects-tp4016416p4018128.html Sent from the Pharo Smalltalk mailing list archive at Nabble.com.

Nov. 9, 2011