September 2012 - Pharo-dev - lists.pharo.org

Re: [Pharo-project] Changing permanently the mouse cursor

by Hilaire Fernandes

In the Squeaksource DrGeoII repo. I add a package DrGeoII-Hacks with modified class to let me turn the mouse cursor as a hudge hand. When the user clic, the hand is changed a bit to show the tap. I plane to use it for videotutorial of drgeo on tablet. Hilaire

Sept. 17, 2012

Re: [Pharo-project] Topics for a term project in CG

by HernÃ¡n Morales Durand

On 17/09/2012 7:16, Dennis Schetinin wrote: > Hello! > > My student is looking for a topic for her term project/paper in area of > Computer Graphics and UI maybe. I didn't pay enough attention to this > area for some time, so I ask if there are some tasks we could help to > solve here? Please note, she has no strong background in CG. One topic > could be something about Morphic, but what is it exactly? Or any other > idea is welcome also. Do you know many other environments where Morphic was applied successfully outside the Squeak niche? GTK 2 is a very popular GUI library (by far much more users than Morphic) and there is a ported implementation for GNU Smalltalk: http://bioskop.fr/gtk_tutorial/ And there are some starting points for Squeak here: http://squeakgtk.pbworks.com/w/page/10714189/FrontPage http://blog.summer.squeak.org/2008/05/introduction-to-squeak-gtk-project.ht… Or even better, you may try Qt, some initiatives and comments exists for other Smalltalks: https://github.com/thephred/gst-qt http://www.cincomsmalltalk.com/userblogs/runarj/blogView?showComments=true&… Best regards, HernÃ¡n > > And one more thing: our university imposes some additional requirements > thus not every practically useful topic can be accepted, I'll have to > select so more ideas is better. > > TIA for any help. > > > Best regards, > Dennis Schetinin >

Sept. 17, 2012

Re: [Pharo-project] Topics for a term project in CG

by Igor Stasenko

On 17 September 2012 12:16, Dennis Schetinin <chaetal(a)gmail.com> wrote: > Hello! > > My student is looking for a topic for her term project/paper in area of > Computer Graphics and UI maybe. I didn't pay enough attention to this area > for some time, so I ask if there are some tasks we could help to solve here? > Please note, she has no strong background in CG. One topic could be > something about Morphic, but what is it exactly? Or any other idea is > welcome also. > > And one more thing: our university imposes some additional requirements thus > not every practically useful topic can be accepted, I'll have to select so > more ideas is better. > Of course! We have a lot of topics related to GC: You should know , what Athens is, it is a vector graphics framework we're currently working on. so, pick one (or two): + UI Widgets Framework for Athens + integrating Athens with Morphic + zoomable interface using Athens + SVG importer for Athens + Animation framework + OpenGL backend for Athens i can guide and mentor, of course :) > TIA for any help. > > > Best regards, > Dennis Schetinin -- Best regards, Igor Stasenko.

Sept. 17, 2012

Re: [Pharo-project] An idea about visual feedback in Athens

by Igor Stasenko

On 17 September 2012 16:13, Fernando Olivero <fernando.olivero(a)usi.ch> wrote: > Igor, the way i've handled the "mouse hit" issue, is by asking the > parent shape (morph) to take care of a mouse hit, and let him handle > or forward the hit. > > The topmost shape are sent the mouseDown: aLocalPoint, then each shape > can choose to react, if its a composite shape or by a specific > behavior. > > The topmost shape is always "a scene shape", which knows the bounds in > global coordinates, then internally it applies local transforms to > eventually get to the inner shape which was hit. The separation > between scene shapes and shapes, allows this mechanism, similar to the > one you propose, to cache the global state. > > I prefer this approach, instead of the "framework" level as you > propose, because is more of a top-down collaboration, always dealing > with local coordinates. It greatly simplifies the UI framework, > avoiding the messy "god classes" and complex methods you mentioned and > many more in Morphic. > yes, but you see that you must go over all hierarchy and transform coordinates accordingly, as you nesting into sub-shapes. but the point is, that at the moment when you drawing things , you know the coordinates so, if you "remember" them at that point, you can do a direct check without need to do any transformations. > > Fernando -- Best regards, Igor Stasenko.

Sept. 17, 2012

Re: [Pharo-project] Changing permanently the mouse cursor

by Igor Stasenko

On 17 September 2012 15:46, Fernando Olivero <fernando.olivero(a)usi.ch> wrote: > HandMorph is the software cursor, which now only gets drawn when the > hand is drawn, which only happens when its carrying a morph. > > Cursor is the hardware cursor, the one handled by the VM and drawn > most of the time. > > IMO, we need to change the VM, to not handle the hardware cursor > anymore, so we can rely only on the software cursor. So we can always > be sure that HandMorph or (the equivalent using other UI framework) is > the cursor being drawn. > why throwing hardware cursors? we just must make sure that a cursor appearance controlled by a single point, not multiple ones, like hand and Cursor class. > Fernando > > On Mon, Sep 17, 2012 at 12:00 PM, Igor Stasenko <siguctua(a)gmail.com> wrote: >> On 17 September 2012 11:40, Hilaire Fernandes >> <hilaire.fernandes(a)edu.ge.ch> wrote: >>> Hello, >>> >>> One can change temporary the mouse cursor with something like: >>> >>> World currentHand showTemporaryCursor: ThemeIcons eyedropperIcon >>> >>> But depending on the widgets the mouse is over, its cursor may change again. >>> >>> Beside hacking the HandMorph class is there a way to change permanently >>> the mouse cursor? >>> >> Good question. >> >> AFAIK, last time i seen it in code, the point where cursor is >> controlled is in Cursor class. >> Hand resets cursor by sending #normal message to it. >> >> It would be really nice to have changeable default cursor. >> The Cursor global used everywhere in system. We should find the way >> how to avoid that. >> >>> Thanks >>> >>> Hilaire >> >> -- >> Best regards, >> Igor Stasenko. >> > -- Best regards, Igor Stasenko.

Sept. 17, 2012

[Pharo-project] pharo-2.0-tests Â» win - Build # 354 - Still Failing!

by jenkins-pharo.ci.inria.fr＠jenkins-server.lille.inria.fr

Sept. 17, 2012

Re: [Pharo-project] An idea about visual feedback in Athens

by Fernando Olivero

Igor, the way i've handled the "mouse hit" issue, is by asking the parent shape (morph) to take care of a mouse hit, and let him handle or forward the hit. The topmost shape are sent the mouseDown: aLocalPoint, then each shape can choose to react, if its a composite shape or by a specific behavior. The topmost shape is always "a scene shape", which knows the bounds in global coordinates, then internally it applies local transforms to eventually get to the inner shape which was hit. The separation between scene shapes and shapes, allows this mechanism, similar to the one you propose, to cache the global state. I prefer this approach, instead of the "framework" level as you propose, because is more of a top-down collaboration, always dealing with local coordinates. It greatly simplifies the UI framework, avoiding the messy "god classes" and complex methods you mentioned and many more in Morphic. Fernando On Sun, Sep 2, 2012 at 10:12 AM, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr> wrote: > Yes we should really refactor this hierarchy. > We should finish first the event model cleaning at VM/Image border then we work on this one. > > On Aug 31, 2012, at 9:21 PM, Igor Stasenko wrote: > >> and to give an idea, how current Morphic doing that, a starting point is >> MorphicEventDispatcher >> >> personally, i find this code too complex to understand, given heavy >> use of case statements and branches. >> >> But i figured out, that this method: >> >> >> dispatchEvent: anEvent with: aMorph >> "Dispatch the given event for a morph that has chosen the receiver to >> dispatch its events. The method implements a shortcut for repeated >> dispatches of events using the same dispatcher." >> anEvent type == lastType ifTrue:[^self perform: lastDispatch with: >> anEvent with: aMorph]. >> "Otherwise classify" >> lastType := anEvent type. >> anEvent isMouse ifTrue:[ >> anEvent isMouseDown ifTrue:[ >> lastDispatch := #dispatchMouseDown:with:. >> ^self dispatchMouseDown: anEvent with: aMorph]]. >> anEvent type == #dropEvent ifTrue:[ >> lastDispatch := #dispatchDropEvent:with:. >> ^self dispatchDropEvent: anEvent with: aMorph]. >> anEvent isWindowEvent ifTrue:[ >> lastDispatch := #dispatchWindowEvent:with:. >> ^self dispatchWindowEvent: anEvent with: aMorph]. >> lastDispatch := #dispatchDefault:with:. >> ^self dispatchDefault: anEvent with: aMorph >> >> --- >> >> actually can be replaced by one-liner: >> >> dispatchEvent: anEvent with: aMorph >> morph := aMorph. >> ^ anEvent sentTo: self >> >> .. yeah.. my favorite thing: double-dispatch. >> >> -- >> Best regards, >> Igor Stasenko. >> > >

Sept. 17, 2012

Re: [Pharo-project] Changing permanently the mouse cursor

by Fernando Olivero

HandMorph is the software cursor, which now only gets drawn when the hand is drawn, which only happens when its carrying a morph. Cursor is the hardware cursor, the one handled by the VM and drawn most of the time. IMO, we need to change the VM, to not handle the hardware cursor anymore, so we can rely only on the software cursor. So we can always be sure that HandMorph or (the equivalent using other UI framework) is the cursor being drawn. Fernando On Mon, Sep 17, 2012 at 12:00 PM, Igor Stasenko <siguctua(a)gmail.com> wrote: > On 17 September 2012 11:40, Hilaire Fernandes > <hilaire.fernandes(a)edu.ge.ch> wrote: >> Hello, >> >> One can change temporary the mouse cursor with something like: >> >> World currentHand showTemporaryCursor: ThemeIcons eyedropperIcon >> >> But depending on the widgets the mouse is over, its cursor may change again. >> >> Beside hacking the HandMorph class is there a way to change permanently >> the mouse cursor? >> > Good question. > > AFAIK, last time i seen it in code, the point where cursor is > controlled is in Cursor class. > Hand resets cursor by sending #normal message to it. > > It would be really nice to have changeable default cursor. > The Cursor global used everywhere in system. We should find the way > how to avoid that. > >> Thanks >> >> Hilaire > > -- > Best regards, > Igor Stasenko. >

Sept. 17, 2012

Re: [Pharo-project] Sending email with Pharo

by GastÃ³n Dall' Oglio

Hello Hilaire. Zodiac? I used zodiac to send email from contact forms in web sites made with Seaside, and works fine. Although sometimes they go to spam :/ See http://stackoverflow.com/questions/12293670/zodiac-mail-sending-in-pharo-sm… But, instead to use ZdcSecureSMTPClient class>>sendUsingGMailAccount:password:to:message: , I created a slightly modified version of that method, due to the smtp server I use the "from" field in mail message is different to "user" in smtp client. E.g. for "from" user1(a)mailserver.com the smtp "user" look like user1.mailserver. This is the method: ************************* sendServer: server port: port user: user password: password from: from to: to subject: subject body: body | smtpClient mailMessage | mailMessage := MailMessage empty. mailMessage setField: 'from' toString: from. mailMessage setField: 'to' toString: to. mailMessage setField: 'subject' toString: subject. mailMessage body: (MIMEDocument contentType: 'text/plain' content: body). (smtpClient := ZdcSecureSMTPClient new) user: user; password: password. smtpClient openOnHost: (NetNameResolver addressForName: server) port: port. ^ smtpClient mailFrom: from to: { to } text: mailMessage text; quit; close; yourself ******************************** Regards 2012/9/15 Hilaire Fernandes <hilaire.fernandes(a)edu.ge.ch> > Dear all, > > A DrGeo user asked me if it will be possible to share DrGeo sketch by > Email. I guess he means, sending an email with an attached DrGeo file. > > My question is: > - can we send email with Pharo? > - if yes, can we attach file to such email? > > Any pointer, if any, appreciated > > Thanks > > Hilaire > > PS: if any of you have any idea how to share a file between different > DrGeo instances running in different host, I will greatly appreciate > your ideas/advices. > > >

Sept. 17, 2012

[Pharo-project] [update 2.0] #20292

by Marcus Denker

20292 ----- Issue 6630: [Enh] Pointers only via STRONG references http://code.google.com/p/pharo/issues/detail?id=6630 Issue 6674: Make HTTPProgress a proper object http://code.google.com/p/pharo/issues/detail?id=6674 Issue 6676: PseudoClassOrganizer DNU #categoriesSorted http://code.google.com/p/pharo/issues/detail?id=6676 -- Marcus Denker -- http://marcusdenker.de

Sept. 17, 2012