- Pharo-users - lists.pharo.org

Re: [Pharo-users] Where do we go now ?

by Denis Kudriashov

2018-04-13 13:01 GMT+02:00 Marcus Denker <marcus.denker(a)inria.fr>: > > > > On 13 Apr 2018, at 12:40, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > > > > > > > >> On 13 Apr 2018, at 12:19, Joe Shirk <j.b.shirk(a)gmail.com> wrote: > >> > >> I've been a lurk-fan for a long time but this brings up something that > distressed me. Richard Eng, Smalltalk Renaissance hero loves to say > Smalltalk's grammar/syntax fits on a postcard. > >> > >> But the vocabulary doesn't. There is nothing English-like about the > always expanding bewildering library namespaces. > >> > > The package names that just use the âproject nameâ can be problematicâ¦ too > many words. e.g. âHiedraâ? No idea. (there are ideas of how to improve, I > will not list them here as this should > not turn into discussion about this issue). > > At some point we should force rule "All packages have comments". And indication with icon like we do for classes. With Calypso the package comment is always available. So it would be easy to find description. The way we present packages (and their granularity) is not ârightâ. > Namespaces are a problem in additionâ¦ > > So yes: we have a lot of thing to improve! > . > >> GT what? Oh a newbie might eventually figure out it means Glamorous > Toolkit. These are meaningless brands. In this drive to come up with > creative names for "just objects" that explain nothing at all, Smalltalk is > becoming like Java or PHP hell. > >> Just look at those examples through the eyes of a novice. The purity is > nowhere to be found. > >> :( > > > > You are right, but in 'the real world' it is no longer possible to > reserve the nice, simple names for just one variant. The prefixes are a > poor mans namespace mechanism. You have to read over them. > > > > Inspector, EyeInspector, GTInspector, ... > > > > I rather have cool alternatives and the development of new ideas than > 'one ring to rule them all' or no/slow progress. Remember that we develop > in a live system, changing things while testing them, this is often hard. > Alternative subsystems help a lot. > > It should be clear that what we have is what we managed to do, not what we > dreamed aboutâ¦ I, too, would like to have this clean, nice, small, amazing > systemâ¦ but it is not always easy. > > There is a lot we can (and will!) improve! > > Marcus > > >

April 13, 2018

Re: [Pharo-users] Where do we go now ?

by Esteban A. Maringolo

On 13/04/2018 09:05, Richard O'Keefe wrote: > There are a lot of subsystems in Pharo, and being a bear of > very little brain, I have a hard time relating Zinc, Calypso, > &c &c to, well, whatever they are.Â I presume there is > somewhere a list of topic/name/PFX triples for guidance. > Can some kind soul tell me where it is? I think it's not a matter of brain, and unless I also have a bear brain I find it hard to know what each thing means, in particular when there is a "flagship" framework (like Seaside) and the related extensions/libraries are called something different instead of, it is harder when the names are completely unrelated to the domain. I've been thinking of collecting the package names, what they do, an which class prefixes they use. But never had the willingness to actually do it, maybe now is a good moment. Regards, -- Esteban.

April 13, 2018

Re: [Pharo-users] Where do we go now ?

by Ben Coman

Unfortunately there probably isn't one list. Its hard to unlearn what is accumulated and easy to take for granted what we know is obvious to everyone. Maybe we need a "Glossary" at https://github.com/pharo-project/pharo/tree/master/wiki where newcomers can add items for others to fill in. cheers -ben On 13 April 2018 at 20:05, Richard O'Keefe <raoknz(a)gmail.com> wrote: > There are a lot of subsystems in Pharo, and being a bear of > very little brain, I have a hard time relating Zinc, Calypso, > &c &c to, well, whatever they are. I presume there is > somewhere a list of topic/name/PFX triples for guidance. > Can some kind soul tell me where it is? > > > On 13 April 2018 at 23:01, Marcus Denker <marcus.denker(a)inria.fr> wrote: > >> >> >> > On 13 Apr 2018, at 12:40, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: >> > >> > >> > >> >> On 13 Apr 2018, at 12:19, Joe Shirk <j.b.shirk(a)gmail.com> wrote: >> >> >> >> I've been a lurk-fan for a long time but this brings up something that >> distressed me. Richard Eng, Smalltalk Renaissance hero loves to say >> Smalltalk's grammar/syntax fits on a postcard. >> >> >> >> But the vocabulary doesn't. There is nothing English-like about the >> always expanding bewildering library namespaces. >> >> >> >> The package names that just use the âproject nameâ can be problematicâ¦ >> too many words. e.g. âHiedraâ? No idea. (there are ideas of how to improve, >> I will not list them here as this should >> not turn into discussion about this issue). >> >> The way we present packages (and their granularity) is not ârightâ. >> Namespaces are a problem in additionâ¦ >> >> So yes: we have a lot of thing to improve! >> . >> >> GT what? Oh a newbie might eventually figure out it means Glamorous >> Toolkit. These are meaningless brands. In this drive to come up with >> creative names for "just objects" that explain nothing at all, Smalltalk is >> becoming like Java or PHP hell. >> >> Just look at those examples through the eyes of a novice. The purity >> is nowhere to be found. >> >> :( >> > >> > You are right, but in 'the real world' it is no longer possible to >> reserve the nice, simple names for just one variant. The prefixes are a >> poor mans namespace mechanism. You have to read over them. >> > >> > Inspector, EyeInspector, GTInspector, ... >> > >> > I rather have cool alternatives and the development of new ideas than >> 'one ring to rule them all' or no/slow progress. Remember that we develop >> in a live system, changing things while testing them, this is often hard. >> Alternative subsystems help a lot. >> >> It should be clear that what we have is what we managed to do, not what >> we dreamed aboutâ¦ I, too, would like to have this clean, nice, small, >> amazing systemâ¦ but it is not always easy. >> >> There is a lot we can (and will!) improve! >> >> Marcus >> >> >> >

April 13, 2018

Re: [Pharo-users] Where do we go now ?

by Peter UhnÃ¡k

On Fri, Apr 13, 2018 at 2:05 PM, Richard O'Keefe <raoknz(a)gmail.com> wrote: > There are a lot of subsystems in Pharo, and being a bear of > very little brain, I have a hard time relating Zinc, Calypso, > &c &c to, well, whatever they are. I presume there is > somewhere a list of topic/name/PFX triples for guidance. > Can some kind soul tell me where it is? > Until we have mature package manager (similar to Cargo or npm), you have to use google. On the bright side, with the move to git(hub), people are much more likely to actually describe what their project does, and maybe even a bit of documentation. This was almost non-existent with SmalltalkHub. Peter

April 13, 2018

Re: [Pharo-users] Where do we go now ?

by Denis Kudriashov

With new Traits we have some issues which we are slowly fixing. Nice thing that fixes are quite simple and clean and covered by tests. I guess it was not like that before. So Pablo did very good job here. For this concrete case with implementors there is already issue 21652 <https://pharo.fogbugz.com/f/cases/21652/Smalltalk-allClassesAndTraits-dupli…>: the method allClassesAndTraits includes traits twice. 2018-04-13 11:30 GMT+02:00 Marcus Denker <marcus.denker(a)inria.fr>: > > > On 13 Apr 2018, at 11:26, Benoit St-Jean via Pharo-users < > pharo-users(a)lists.pharo.org> wrote: > > > *From: *Benoit St-Jean <bstjean(a)yahoo.com> > *Subject: **Re: [Pharo-users] Where do we go now ?* > *Date: *13 April 2018 at 11:26:09 CEST > *To: *Esteban Lorenzano <estebanlm(a)gmail.com> > *Cc: *Any question about pharo is welcome <pharo-users(a)lists.pharo.org> > *Reply-To: *Benoit St-Jean <bstjean(a)yahoo.com> > > > BTW, why put an .exe installer for Windows available when it crashes right > from the start? It just doesn't work at all. Period. For everyone. > > And I thought looking for senders of a method was something we mastered a > long time ago, like starting with Smalltalk-76. Am I supposed to assume > that everything, even basic functionalities, are all broken because it's > labeled "alphaâ ? > > > We have already a issue tracker entry with a description of a fix: > > https://pharo.fogbugz.com/f/cases/21518/New-Traits-brings- > wrong-behaviour-to-tagsForMethods > > TODO: turn that description into code and make a Pull Request. > > Marcus > >

April 13, 2018

Re: [Pharo-users] Where do we go now ?

by Richard O'Keefe

There are a lot of subsystems in Pharo, and being a bear of very little brain, I have a hard time relating Zinc, Calypso, &c &c to, well, whatever they are. I presume there is somewhere a list of topic/name/PFX triples for guidance. Can some kind soul tell me where it is? On 13 April 2018 at 23:01, Marcus Denker <marcus.denker(a)inria.fr> wrote: > > > > On 13 Apr 2018, at 12:40, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > > > > > > > >> On 13 Apr 2018, at 12:19, Joe Shirk <j.b.shirk(a)gmail.com> wrote: > >> > >> I've been a lurk-fan for a long time but this brings up something that > distressed me. Richard Eng, Smalltalk Renaissance hero loves to say > Smalltalk's grammar/syntax fits on a postcard. > >> > >> But the vocabulary doesn't. There is nothing English-like about the > always expanding bewildering library namespaces. > >> > > The package names that just use the âproject nameâ can be problematicâ¦ too > many words. e.g. âHiedraâ? No idea. (there are ideas of how to improve, I > will not list them here as this should > not turn into discussion about this issue). > > The way we present packages (and their granularity) is not ârightâ. > Namespaces are a problem in additionâ¦ > > So yes: we have a lot of thing to improve! > . > >> GT what? Oh a newbie might eventually figure out it means Glamorous > Toolkit. These are meaningless brands. In this drive to come up with > creative names for "just objects" that explain nothing at all, Smalltalk is > becoming like Java or PHP hell. > >> Just look at those examples through the eyes of a novice. The purity is > nowhere to be found. > >> :( > > > > You are right, but in 'the real world' it is no longer possible to > reserve the nice, simple names for just one variant. The prefixes are a > poor mans namespace mechanism. You have to read over them. > > > > Inspector, EyeInspector, GTInspector, ... > > > > I rather have cool alternatives and the development of new ideas than > 'one ring to rule them all' or no/slow progress. Remember that we develop > in a live system, changing things while testing them, this is often hard. > Alternative subsystems help a lot. > > It should be clear that what we have is what we managed to do, not what we > dreamed aboutâ¦ I, too, would like to have this clean, nice, small, amazing > systemâ¦ but it is not always easy. > > There is a lot we can (and will!) improve! > > Marcus > > >

April 13, 2018

Re: [Pharo-users] Progress bar for ZnClient>>#downloadTo:

by Cyril Ferlicot D.

On 13/04/2018 13:50, Sven Van Caekenberghe wrote: > > All good and well, I understand your point, but as a general rule, I don't want to add UI code to Zn, it is too fundamental for minimal/kernel//headless images. > > Hi, It could be useful to add this kind of methods via extensions in an UI package not in the default group of Zinc configuration maybe? And Pharo 7 could integrate the core of zinc in a low level layer during the bootstrap and the UI stuff in the tooling layer. -- Cyril Ferlicot https://ferlicot.fr

April 13, 2018

Re: [Pharo-users] Progress bar for ZnClient>>#downloadTo:

by Sven Van Caekenberghe

> On 12 Apr 2018, at 17:02, Ben Coman <btc(a)openinworld.com> wrote: > > > > On 12 April 2018 at 18:15, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > Ben, > > You disappoint me, as I would expect you to trace senders/users of #signalProgress: ;-) > > :P > Yeah I knew I was being slack. I couldn't really follow its operation and was a bit tired. > Actually maybe part of my confusion is that #signalProgress feels more like a command than a status. > I've a slight feeling it would be better as #signallingProgress. > > > ZnClient only signals certain Notifications which a UI around it should deal with. > > For example, > > [ :bar | > bar title: 'Downloading Sources...'. > [ > ZnClient new > url: 'http://files.pharo.org/sources/PharoV30.sources'; > signalProgress: true; > downloadTo: FileLocator temp ] > on: HTTPProgress > do: [ :progress | > progress isEmpty ifFalse: [ bar current: progress percentage ]. > progress resume ] ] asJob run. > > Thanks, that worked, but its a bit convoluted for my goldfish memory to remember each time. > I'd expect having that as a convenience method would be useful to many. > > ZnClient > showProgress: 'Downloading sources...' > during: [ :znClient | > znClient url: 'http://files.pharo.org/sources/PharoV30.sources'. > znClient downloadTo: FileLocator imageDirectory ]. > > ZnClient class >> showProgress: title during: clientBlock > |client| > client := ZnClient new. > client signallingProgress: true. > [ :bar | > bar title: title. > [clientBlock value: client] > on: HTTPProgress > do: [ :progress | > progress isEmpty ifFalse: [ bar current: progress percentage ]. > progress resume ] ] asJob run. > > Easily discoverable as the only class-side method of ZnClient, > and this would also make a useful howto reference. All good and well, I understand your point, but as a general rule, I don't want to add UI code to Zn, it is too fundamental for minimal/kernel//headless images. > But there are other UI approaches as well. > > Sven > > Thanks for you prompt response. > cheers -ben > > > > > On 12 Apr 2018, at 11:25, Ben Coman <btc(a)openinworld.com> wrote: > > > > I see ZnClient >> downloadTo: > > calls ZnClient >> downloadEntityTo: > > which uses #withProgressDo: > > which indicates a progress bar might appear. > > > > But... > > ZnClient new > > url: ' https://download.libsodium.org/libsodium/releases/libsodium-1.0.16-msvc.zip' ; > > downloadTo: FileLocator imageDirectory. > > doesn't show a progress bar. What is the recommended way to display one? > > > > cheers -ben

April 13, 2018

Re: [Pharo-users] Where do we go now ?

by Marcus Denker

> On 13 Apr 2018, at 12:40, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > > > >> On 13 Apr 2018, at 12:19, Joe Shirk <j.b.shirk(a)gmail.com> wrote: >> >> I've been a lurk-fan for a long time but this brings up something that distressed me. Richard Eng, Smalltalk Renaissance hero loves to say Smalltalk's grammar/syntax fits on a postcard. >> >> But the vocabulary doesn't. There is nothing English-like about the always expanding bewildering library namespaces. >> The package names that just use the âproject nameâ can be problematicâ¦ too many words. e.g. âHiedraâ? No idea. (there are ideas of how to improve, I will not list them here as this should not turn into discussion about this issue). The way we present packages (and their granularity) is not ârightâ. Namespaces are a problem in additionâ¦ So yes: we have a lot of thing to improve! . >> GT what? Oh a newbie might eventually figure out it means Glamorous Toolkit. These are meaningless brands. In this drive to come up with creative names for "just objects" that explain nothing at all, Smalltalk is becoming like Java or PHP hell. >> Just look at those examples through the eyes of a novice. The purity is nowhere to be found. >> :( > > You are right, but in 'the real world' it is no longer possible to reserve the nice, simple names for just one variant. The prefixes are a poor mans namespace mechanism. You have to read over them. > > Inspector, EyeInspector, GTInspector, ... > > I rather have cool alternatives and the development of new ideas than 'one ring to rule them all' or no/slow progress. Remember that we develop in a live system, changing things while testing them, this is often hard. Alternative subsystems help a lot. It should be clear that what we have is what we managed to do, not what we dreamed aboutâ¦ I, too, would like to have this clean, nice, small, amazing systemâ¦ but it is not always easy. There is a lot we can (and will!) improve! Marcus

April 13, 2018

Re: [Pharo-users] Where do we go now ?

by aglynn42＠gmail.com

Btw, in my fascination with messing around, the 32 bit version of Pharo 7 for Windows runs better on OS/2 v.5 (yes, it still exists, it was released last June). Probably because its Win32 subsystem is more compatible with Win32 apps than Windows 10. Andrew From: Benoit St-Jean <bstjean(a)yahoo.com> Sent: Friday, April 13, 2018 5:26 AM To: Esteban Lorenzano <estebanlm(a)gmail.com> Cc: Any question about pharo is welcome <pharo-users(a)lists.pharo.org> Subject: Re: [Pharo-users] Where do we go now ? BTW, why put an .exe installer for Windows available when it crashes right from the start? It just doesn't work at all. Period. For everyone. And I thought looking for senders of a method was something we mastered a long time ago, like starting with Smalltalk-76. Am I supposed to assume that everything, even basic functionalities, are all broken because it's labeled "alpha" ? ----------------- BenoÃ®t St-Jean Yahoo! Messenger: bstjean Twitter: @BenLeChialeux Pinterest: benoitstjean Instagram: Chef_Benito IRC: lamneth Blogue: endormitoire.wordpress.com "A standpoint is an intellectual horizon of radius zero". (A. Einstein) On Friday, April 13, 2018, 5:20:28 a.m. EDT, Esteban Lorenzano <estebanlm(a)gmail.com <mailto:estebanlm@gmail.com> > wrote: On 13 Apr 2018, at 11:07, Benoit St-Jean <bstjean(a)yahoo.com <mailto:bstjean@yahoo.com> > wrote: I'm on Windows 10, using Pharo 7.0 alpha 32 bit. and btwâ¦ which part of ALPHA you do not get? Esteban

April 13, 2018