- Pharo-users - lists.pharo.org

Re: Creating an SDL (OSWindow) window without showing the pharo interface

by Russ Whaley

Thanks! Canât wait to try it! On Sun, May 2, 2021 at 6:15 AM kmo <voxkmp(a)gmail.com> wrote: > Russ - > > The trick is to put the --no-quit after the eval not before it. I was > putting it in the wrong place. > > This should work: > > ./pharo Pharo.image eval --no-quit "OSWindowWorldMorph new open" > > This works with the latest headless VM and Pharo 9. > > Hope this helps. > > Ken > > > > > -- > Sent from: http://forum.world.st/Pharo-Smalltalk-Users-f1310670.html > -- Russ Whaley whaley.russ(a)gmail.com

May 2, 2021

Re: [Pharo-dev] [ANN] Freezing Pharo 9 development right now :)

by Noury Bouraqadi

Excellent news! On Apr 30 2021, at 11:04 pm, Esteban Lorenzano <estebanlm(a)netc.eu> wrote: > Hi, > > So, today I was checking how much it is missing for release and what is our status, and I realized if we do not freeze now, we will never release in a reasonable time. > So, even if there are a couple of cleanups and that to still do that should happen before, I will declare that Pharo 9.0 enters in freeze Pharo 9.0 right now (and I will sneak the cleanups anyway, but being more careful). > > What does this means? > No more features, just bugfixes > > No more fixes of things that can wait until Pharo 10 > > > > In two weeks I will open the Pharo 10 development branch, up to then, please help fixing what's missing (which is a lot) ! > Have a great weekend, > cheers! > Esteban

May 2, 2021

Re: Creating an SDL (OSWindow) window without showing the pharo interface

by kmo

Russ - The trick is to put the --no-quit after the eval not before it. I was putting it in the wrong place. This should work: ./pharo Pharo.image eval --no-quit "OSWindowWorldMorph new open" This works with the latest headless VM and Pharo 9. Hope this helps. Ken -- Sent from: http://forum.world.st/Pharo-Smalltalk-Users-f1310670.html

May 2, 2021

Re: Creating a Spec -GTK window without showing the pharo IDE

by kmo

Many thanks Estaban. All OK now. I forgot to add in the --no-quit but I would have put it in the wrong place anyway. I never thought it had to go after the eval. -- Sent from: http://forum.world.st/Pharo-Smalltalk-Users-f1310670.html

May 2, 2021

Re: Creating a Spec -GTK window without showing the pharo IDE

by StÃ©phane Ducasse

Thanks esteban I copied and pasted your email in the how to draft in Spec2 book. S > On 1 May 2021, at 20:47, Esteban Lorenzano <estebanlm(a)netc.eu> wrote: > > Hello, > > Remember: every time you execute "pharo-ui" you are invoking the UI (as the last part of the name says). If you want to NOT have the UI, you need to execute "pharo" : > > ./pharo Pharo.Image eval RunGtk execute > > BUT: this will not work because the image will evaluate "RunGtk execute" and then will exit (because it will evaluate it as a script). To avoid that you need to execute: > > ./pharo Pharo.Image eval --no-quit "RunGtk execute" > > That will work as you want. > > BUT, this is not how executing Spec applications is envisaged : > > I guess you defined an application (a children of SpApplication?) where you set your backend to make it a Gtk application ? > and you have override #start to do something like (MyPresenter newApplication: self) openWithSpec ? > > In that case, you just need to define in your application class (say is named MyApplication) : > > MyApplication class >> applicationName > ^ 'gtkapp' > > then, is enough to say: > > ./pharo Pharo.image run gtkapp > > which would be the "canonical" way to do it :) > > Esteban > > > On May 1 2021, at 1:12 pm, kmo <voxkmp(a)gmail.com> wrote: > After not lookng at it for I while, I tried the Gtk Spec bindings again on my > Xubuntu desktop and - this time - they worked. I was able to open a new > window using Spec-Gtk, the latest headless VM and the latest Pharo 9. > (Many thanks to all concerned). > > But I've immediately hit the same issue as I've raised before with SLD > support (OSWindow) - it seems impossible to open a Gtk window without the > whole Pharo IDE being shown as well. > > If I run the following command line - > > ./pharo-ui Pharo.Image eval RunGtk execute > > I get the Pharo IDE and a GTK window opened > > If I run > > ./pharo Pharo.Image eval RunGtk execute > > I just get the string RunGtk returned. No window opens. > > I presume this will work some time in the future. Otherwise what's the use > of Spec-Gtk (or OSWindow for that matter)? You might as well just use Spec > on Morphic. > > > > -- > Sent from: http://forum.world.st/Pharo-Smalltalk-Users-f1310670.html -------------------------------------------- StÃ©phane Ducasse http://stephane.ducasse.free.fr / http://www.pharo.org 03 59 35 87 52 Assistant: Aurore Dalle FAX 03 59 57 78 50 TEL 03 59 35 86 16 S. Ducasse - Inria 40, avenue Halley, Parc Scientifique de la Haute Borne, BÃ¢t.A, Park Plaza Villeneuve d'Ascq 59650 France

May 1, 2021

Re: Creating a Spec -GTK window without showing the pharo IDE

by Esteban Lorenzano

Hello, Remember: every time you execute "pharo-ui" you are invoking the UI (as the last part of the name says). If you want to NOT have the UI, you need to execute "pharo" : ./pharo Pharo.Image eval RunGtk execute BUT: this will not work because the image will evaluate "RunGtk execute" and then will exit (because it will evaluate it as a script). To avoid that you need to execute: ./pharo Pharo.Image eval --no-quit "RunGtk execute" That will work as you want. BUT, this is not how executing Spec applications is envisaged : I guess you defined an application (a children of SpApplication?) where you set your backend to make it a Gtk application ? and you have override #start to do something like (MyPresenter newApplication: self) openWithSpec ? In that case, you just need to define in your application class (say is named MyApplication) : MyApplication class >> applicationName ^ 'gtkapp' then, is enough to say: ./pharo Pharo.image run gtkapp which would be the "canonical" way to do it :) Esteban On May 1 2021, at 1:12 pm, kmo <voxkmp(a)gmail.com> wrote: > After not lookng at it for I while, I tried the Gtk Spec bindings again on my > Xubuntu desktop and - this time - they worked. I was able to open a new > window using Spec-Gtk, the latest headless VM and the latest Pharo 9. > (Many thanks to all concerned). > > But I've immediately hit the same issue as I've raised before with SLD > support (OSWindow) - it seems impossible to open a Gtk window without the > whole Pharo IDE being shown as well. > > If I run the following command line - > ./pharo-ui Pharo.Image eval RunGtk execute > I get the Pharo IDE and a GTK window opened > If I run > ./pharo Pharo.Image eval RunGtk execute > I just get the string RunGtk returned. No window opens. > I presume this will work some time in the future. Otherwise what's the use > of Spec-Gtk (or OSWindow for that matter)? You might as well just use Spec > on Morphic. > > > > -- > Sent from: http://forum.world.st/Pharo-Smalltalk-Users-f1310670.html >

May 1, 2021

Creating a Spec -GTK window without showing the pharo IDE

by kmo

After not lookng at it for I while, I tried the Gtk Spec bindings again on my Xubuntu desktop and - this time - they worked. I was able to open a new window using Spec-Gtk, the latest headless VM and the latest Pharo 9. (Many thanks to all concerned). But I've immediately hit the same issue as I've raised before with SLD support (OSWindow) - it seems impossible to open a Gtk window without the whole Pharo IDE being shown as well. If I run the following command line - ./pharo-ui Pharo.Image eval RunGtk execute I get the Pharo IDE and a GTK window opened If I run ./pharo Pharo.Image eval RunGtk execute I just get the string RunGtk returned. No window opens. I presume this will work some time in the future. Otherwise what's the use of Spec-Gtk (or OSWindow for that matter)? You might as well just use Spec on Morphic. -- Sent from: http://forum.world.st/Pharo-Smalltalk-Users-f1310670.html

May 1, 2021

[ANN] Freezing Pharo 9 development right now :)

by Esteban Lorenzano

Hi, So, today I was checking how much it is missing for release and what is our status, and I realized if we do not freeze now, we will never release in a reasonable time. So, even if there are a couple of cleanups and that to still do that should happen before, I will declare that Pharo 9.0 enters in freeze Pharo 9.0 right now (and I will sneak the cleanups anyway, but being more careful). What does this means? No more features, just bugfixes No more fixes of things that can wait until Pharo 10 In two weeks I will open the Pharo 10 development branch, up to then, please help fixing what's missing (which is a lot) ! Have a great weekend, cheers! Esteban

April 30, 2021

SQLite3 primitiveFailed after FFI call

by Esteban Maringolo

Hi all, I'm having a strange issue when executing a query with the SQLite3 driver (via GLORP). In: SQLite3Library>>#apiPrepare: dbHandle withSQLText: anSQLText with: textSize with: sHandle with: tail "int sqlite3_prepare_v2(sqlite3*, const char*, int, sqlite3_stmt**, const char **)" ^ self ffiCall: #(int sqlite3_prepare_v2 (sqlite3* dbHandle, String anSQLText, int textSize, sqlite3_stmt** sHandle, const char** tail)) After the FFI call, I get a primitive failed error when some part of the FFI machinery sends instVarAt: index to the undefined object. This is the call stack: PrimitiveFailed(Exception)>>signal PrimitiveFailed class(SelectorException class)>>signalFor: UndefinedObject(ProtoObject)>>primitiveFailed: UndefinedObject(ProtoObject)>>primitiveFailed UndefinedObject(Object)>>instVarAt: SQLite3Library>>apiPrepare:withSQLText:with:with:with: SQLite3Library>>prepare:on:with: SQLite3PreparedStatement>>prepare: SQLite3Connection(SQLite3BaseConnection)>>prepare: SQLite3Connection>>execute:with: SQLite3Driver>>basicExecuteSQLString:binding: SQLite3Driver>>basicExecuteSQLString: PharoDatabaseAccessor>>basicExecuteSQLString: PharoDatabaseAccessor>>executeCommandUnbound: QuerySelectCommand(DatabaseCommand)>>executeUnboundIn: If I run the same query via an external SQL tool (sqlite3) it works, returning a single row with a single value as expected. What's weird is that this doesn't fail always, since it's a very common query in the app. Any ideas? Esteban A. Maringolo

April 30, 2021

Re: We haven't had a design debate in a while - thoughts on CD.buy article...

by Erik Stel

Hi people, A bit late to the party, but I want to provide some info as well. I think an Article can perfectly respond to a #sell (or #sellQuantity:) message. Depending on the type of application it could know or interact with other objects responsible for Stock for example to provide an estimated delivery date and price. When a customer is not known through some context it could be #sellToCustomer: or #sellToAccount:. But mostly to have a delivery address and make sure payment is done. Not to have a customer do the buying (as an act), that is irrelevant for most (web)shops. There is often no need to have a person like object to fulfil the role of actor (dare I say 'manager') in the object interactions. Objects have behaviour and take on responsibilities of their own. Having a human like object moves responsibilities away from the objects and creates unneeded dependencies between objects. You'll quickly see conditional logic appear in those classes. Kids sometimes understand this way of looking at objects better than we adults. A kid does not find it weird to see a movie or play in which a teapot pours itself or where a CD plays itself. So I do understand that a CD understands #play as a message. Perfectly fine. We do not literally have to represent every physical object and its physical characteristics in our world as objects in our applications. It is the behaviour and responsibilities they posses that counts. (CRC cards are really quite helpful in OO) I highly recommend reading the book Object Thinking by Dave West. For sake of clarity: yes, we are partners in ObjectGuild. See the link to the book below. Don't be put back by the age of the book. http://davewest.us/product/object-thinking/ Cheers, Erik -- Sent from: http://forum.world.st/Pharo-Smalltalk-Users-f1310670.html

April 29, 2021