November 2017 - Pharo-users

Re: [Pharo-users] Fwd: GitHub

by Ben Coman

As a backup, select your package in System Browser and choose <File Out>, so at least your code is safe. If you VM crashes, you can recover code using Epicia. From the World Menu > Tools > Recover changes (from memory) I'm not clear what you mean by "VM plante" cheers -ben On Thu, Nov 2, 2017 at 10:28 PM, hamdi gabsi <gabsi.hamdi(a)gmail.com> wrote: > > ---------- Forwarded message ---------- > From: hamdi gabsi <gabsi.hamdi(a)gmail.com> > Date: 2017-11-02 14:05 GMT+01:00 > Subject: GitHub > To: pharo-users(a)lists.pharo.org > > > Hello, > > I'am Hamdi, I am doing an internship with Rmod Team. I have a problem > with pushing my code in Github, In fact, Iam using Pharo 6 when I wont to > push my code my VM plante and I will be obliged to kill it, so I loose my > code. > > Thank you for helping me. > > >

Nov. 2, 2017

Re: [Pharo-users] Peeking at a stream

by Ben Coman

On Thu, Nov 2, 2017 at 9:28 PM, Andrew Glynn <aglynn42(a)gmail.com> wrote: > Perhaps #keepStateWhile or something similar might be better than > #unchangedDuring ? > Something like "keepState" was also my first impression, but "During" seems to be the convention for restoring things after a code block completes. Are there methods breaking that convention? cheers -ben *From: *Prof. Andrew P. Black <black(a)cs.pdx.edu> > *Sent: *Thursday, November 2, 2017 5:00 AM > *To: *Any question about pharo is welcome <pharo-users(a)lists.pharo.org> > *Subject: *[Pharo-users] Peeking at a stream > > > > I sometimes find, when working with a stream, that I need to "peek aheadâ > a few items, without disturbing the state of the stream. The message stream > peek gives me the next item, but doesnât generalize to multiple items, So > I end up writing code like this: > > > > findSomething > > | savedPosition result | > > savedPosition := stream position. > stream position: start. > âcode involving various manipulations of stream, including > stream next" > result := stream position + 1. > stream position: savedPosition. > > ^ result > > > > This code involves at least two temps, is not very easy to read, and risks > not resetting the stream properly if I do a return. > > > > It seems to me that a better solution would be what Kent Beck calls an > âexecute aroundâ method: > > > > findSomething > > ^ stream unchangedDuring: [ :s | > > âcode involving various manipulations of > stream, including stream next" > > stream position + 1 ] > > > > Iâm not sure that I like the name #unchangedDuring:; perhaps you can think > of a better one? > > > > The implementation of this would be > > > > PositionableStream >> #unchangedDuring: aBlock > > "execute aBlock with self as argument. Ensure > that when this method returns, > > I still have my current state." > > > > | savedPosition | > > savedPosition := self position. > ^ [ aBlock value: self ] ensure: [ self > position: savedPosition ] > > > > or perhaps > > > > PositionableStream >> #unchangedDuring: aBlock > "execute aBlock with self as argument. Ensure > that when this method returns, > > I still have my current state." > > > ^ [ aBlock value: self copy ] > > > > which might be better in the case that the stream is a file stream and > aBlock causes the buffer to be refreshed from the disk. > > > > - Do other Smalltalks have a method like this? > - Does it already exist in Pharo, somewhere that Iâve failed to look? > - Would it be a worthwhile addition to Pharo? > - What would be a good name? > > > > > > > > >

Nov. 2, 2017

Fwd: GitHub

by hamdi gabsi

---------- Forwarded message ---------- From: hamdi gabsi <gabsi.hamdi(a)gmail.com> Date: 2017-11-02 14:30 GMT+01:00 Subject: Re: GitHub To: pharo-users(a)lists.pharo.org For resolving my problem. I knew that it was when pushing my code, that's mean that the commit was well done. So I pushed it from my Terminal. I went to my Pharo local repository and I tapped git status then git push. After that in iceberg ,setting I introduced my user name and password. so that he remembered me. And it work from my VM no need to go to the Terminal. We can find answers in https://github.com/pharo-vcs/iceberg Thank you. 2017-11-02 14:06 GMT+01:00 hamdi gabsi <gabsi.hamdi(a)gmail.com>: > I am working with Pharo 6.0 > > 2017-11-02 14:05 GMT+01:00 hamdi gabsi <gabsi.hamdi(a)gmail.com>: > >> Hello, >> >> I'am Hamdi, I am doing an internship with Rmod Team. I have a problem >> with pushing my code in Github, In fact, when I wont to push my code my VM >> bug and I will be obliged to kill it, so I loose my code. >> >> Thank you for helping me. >> >> >

Nov. 2, 2017

Fwd: GitHub

by hamdi gabsi

---------- Forwarded message ---------- From: hamdi gabsi <gabsi.hamdi(a)gmail.com> Date: 2017-11-02 14:05 GMT+01:00 Subject: GitHub To: pharo-users(a)lists.pharo.org Hello, I'am Hamdi, I am doing an internship with Rmod Team. I have a problem with pushing my code in Github, In fact, Iam using Pharo 6 when I wont to push my code my VM plante and I will be obliged to kill it, so I loose my code. Thank you for helping me.

Nov. 2, 2017

Re: [Pharo-users] Exchanging information between 2 pharo applications (2 images running on two different computers)

by Denis Kudriashov

To load such code switch to old compiler in settings browser. It supports old FFI syntax. Then you can manually adopt all ffi calls to UFFI and commit new version. 2017-10-29 15:09 GMT+01:00 Paulo R. Dellani <dellani(a)pobox.com>: > Dear all, > > I would like to give ZeroMQ a try in Smalltalk, but am unable to > load it on a new image successfully. In Pharo 5, I loaded the package > ConfigurationOfZeroMQ from the repository in smalltalkhub > <http://smalltalkhub.com/#%21/%7Epanuw/zeromq> and > evaluated the following to load the code: > > ConfigurationOfZeroMQ loadBleedingEdge > > It seems to load everything well, but when I try to look at the source > code of the methods that call the C routines from libzmq using the > system browser, the system exhibits the following error message: > > UndefinedObject(Object)>>doesNotUnderstand: #keywords > RBFFICallPragma(RBPragmaNode)>>selectorParts > RubSHTextStylerST80(SHRBTextStyler)>>visitPragmaNode: > RBFFICallPragma(RBPragmaNode)>>acceptVisitor: > RubSHTextStylerST80(SHRBTextStyler)>>visitNode: > [ :each | self visitNode: each ] in RubSHTextStylerST80>>visitMethodNode: > in Block: [ :each | self visitNode: each ] > OrderedCollection>>do: > > (its a long stack, cutting here) > > When loading ZeroMQ on Pharo 6, a window pops-up with the title > 'Syntax Error: Literal constant expected' showing the source code > of a method defining a call to a routine from a C library: > > apiDeleteDC: aHDC > <apicall: Literal constant expected -> bool 'DeleteDC' (Win32HDC) > module:'gdi32.dll'> > ^self externalCallFailed > > If I update the syntax, re-writing the code above as > > apiDeleteDC: aHDC > ^self ffiCall: #( bool DeleteDC (Win32HDC aHDC) ) module:'gdi32.dll' > > the compiler accepts it and the process can go on, but eventually > the system will become unstable... > > Any ideas on how to solve this problems? Is this syntax still supported > in Pharo5 & 6? > > Cheers, > > Paulo > > On 10/25/2017 03:47 PM, Sebastian Heidbrink via Pharo-users wrote: > > ForwardedMessage.eml > > Subject: > Re: [Pharo-users] Exchanging information between 2 pharo applications (2 > images running on two different computers) > > From: > Sebastian Heidbrink <sheidev(a)yahoo.de> <sheidev(a)yahoo.de> > > Date: > 10/25/2017 03:47 PM > > To: > pharo-users(a)lists.pharo.org > Hi Cederik, > > you should have a look at http://smalltalkhub.com/#!/~panuw/zeromq > ZeroMQ is a networking library. Nice thing about it is that all the > networking related workload is dealt with in a second process. That can > save you some resources within Smalltalk. > > >

Nov. 2, 2017

Re: [Pharo-users] Peeking at a stream

by Andrew Glynn

Perhaps #keepStateWhile or something similar might be better than #unchangedDuring ? Sent from Mail for Windows 10 From: Prof. Andrew P. Black Sent: Thursday, November 2, 2017 5:00 AM To: Any question about pharo is welcome Subject: [Pharo-users] Peeking at a stream I sometimes find, when working with a stream, that I need to "peek aheadâ a few items, without disturbing the state of the stream. The message stream peek gives me the next item, but doesnât generalize to multiple items, Â So I end up writing code like this: Â Â findSomething | savedPosition result | savedPosition := stream position. stream position: start. âcode involving various manipulations of stream, including stream next" result := stream position + 1. stream position: savedPosition. ^ result This code involves at least two temps, is not very easy to read, and risks not resetting the stream properly if I do a return. It seems to me that a better solution would be what Kent Beck calls an âexecute aroundâ method: Â Â findSomething ^Â stream unchangedDuring: [ :s | âcode involving various manipulations of stream, including stream next" stream position + 1 ] Iâm not sure that I like the name #unchangedDuring:; perhaps you can think of a better one?Â The Â implementation of this would be PositionableStream >> #unchangedDuring: aBlock "execute aBlock with self as argument. Ensure that when this method returns, I still have my current state." | savedPosition | savedPosition := self position. ^Â [ aBlock value: self ] ensure: [ self position: savedPosition ] or perhaps PositionableStream >> #unchangedDuring: aBlock "execute aBlock with self as argument. Ensure that when this method returns, I still have my current state." ^Â [Â aBlock value: self copy ] which might be better in the case that the stream is a file stream and aBlock causes the buffer to be refreshed from the disk. â¢ Do other Smalltalks have a method like this? Â â¢ Does it already exist in Pharo, somewhere that Iâve failed to look?Â â¢ Would it be a worthwhile addition to Pharo? Â â¢ What would be a good name?

Nov. 2, 2017

Re: [Pharo-users] Peeking at a stream

by Denis Kudriashov

Hi. In XStreams this method called #explore: . 2017-11-02 9:59 GMT+01:00 Prof. Andrew P. Black <black(a)cs.pdx.edu>: > I sometimes find, when working with a stream, that I need to "peek aheadâ > a few items, without disturbing the state of the stream. The message stream > peek gives me the next item, but doesnât generalize to multiple items, So > I end up writing code like this: > > findSomething > | savedPosition result | > savedPosition := stream position. > stream position: start. > âcode involving various manipulations of stream, including stream next" > result := stream position + 1. > stream position: savedPosition. > ^ result > > This code involves at least two temps, is not very easy to read, and risks > not resetting the stream properly if I do a return. > > It seems to me that a better solution would be what Kent Beck calls an > âexecute aroundâ method: > > findSomething > ^ stream unchangedDuring: [ :s | > âcode involving various manipulations of stream, including stream next" > stream position + 1 ] > > Iâm not sure that I like the name #unchangedDuring:; perhaps you can think > of a better one? > > The implementation of this would be > > PositionableStream >> #unchangedDuring: aBlock > "execute aBlock with self as argument. Ensure that when this method > returns, > I still have my current state." > | savedPosition | > savedPosition := self position. > ^ [ aBlock value: self ] ensure: [ self position: savedPosition ] > > or perhaps > > PositionableStream >> #unchangedDuring: aBlock > "execute aBlock with self as argument. Ensure that when this method > returns, > I still have my current state." > > ^ [ aBlock value: self copy ] > > which might be better in the case that the stream is a file stream and > aBlock causes the buffer to be refreshed from the disk. > > > - Do other Smalltalks have a method like this? > - Does it already exist in Pharo, somewhere that Iâve failed to look? > - Would it be a worthwhile addition to Pharo? > - What would be a good name? > > > > >

Nov. 2, 2017

Re: [Pharo-users] Peeking at a stream

by henry

I use the approach you mention, restoring stream position, in my ParrotTalk code, both in Pharo and in Java. I am not peeking more than one character but I had to implement peek in Java and your code example is how I did this code. - HH On Thu, Nov 2, 2017 at 04:59, Prof. Andrew P. Black <[black(a)cs.pdx.edu]("mailto:black@cs.pdx.edu")> wrote: > I sometimes find, when working with a stream, that I need to "peek ahead" a few items, without disturbing the state of the stream. The message stream peek gives me the next item, but doesnât generalize to multiple items, So I end up writing code like this: > > findSomething > | savedPosition result | > savedPosition := stream position. > stream position: start. > "code involving various manipulations of stream, including stream next" > result := stream position + 1. > stream position: savedPosition. > ^ result > > This code involves at least two temps, is not very easy to read, and risks not resetting the stream properly if I do a return. > > It seems to me that a better solution would be what Kent Beck calls an "execute around" method: > > findSomething > ^ stream unchangedDuring: [ :s | > "code involving various manipulations of stream, including stream next"stream position + 1 ] > > Iâm not sure that I like the name #unchangedDuring:; perhaps you can think of a better one? > > The implementation of this would be > > PositionableStream >> #unchangedDuring: aBlock > "execute aBlock with self as argument. Ensure that when this method returns, > I still have my current state." > > | savedPosition |savedPosition := self position. > ^ [ aBlock value: self ] ensure: [ self position: savedPosition ] > > or perhaps > > PositionableStream >> #unchangedDuring: aBlock > "execute aBlock with self as argument. Ensure that when this method returns, > I still have my current state." > ^ [ aBlock value: self copy ] > > which might be better in the case that the stream is a file stream and aBlock causes the buffer to be refreshed from the disk. > > - Do other Smalltalks have a method like this? > - Does it already exist in Pharo, somewhere that Iâve failed to look? > - Would it be a worthwhile addition to Pharo? > - What would be a good name?

Nov. 2, 2017

Re: [Pharo-users] Peeking at a stream

by Guillermo Polito

Hi Andrew, My first hunch is that you don't want to change PositionableStream. Because it works on collections and files (because of inheritance). But not on other kind of streams that are not part of the hierarchy, such as sockets. And we are cleaning that part. Instead, some time ago Sven produced a really nice hierarchy of stream decorators. Check the package: Zinc-Character-Encoding-Core There you have ZnBufferedReadStream That implements buffering in an orthogonal way. Maybe it makes more sense to put such method there? Guille On Thu, Nov 2, 2017 at 9:59 AM, Prof. Andrew P. Black <black(a)cs.pdx.edu> wrote: > I sometimes find, when working with a stream, that I need to "peek aheadâ > a few items, without disturbing the state of the stream. The message stream > peek gives me the next item, but doesnât generalize to multiple items, So > I end up writing code like this: > > findSomething > | savedPosition result | > savedPosition := stream position. > stream position: start. > âcode involving various manipulations of stream, including stream next" > result := stream position + 1. > stream position: savedPosition. > ^ result > > This code involves at least two temps, is not very easy to read, and risks > not resetting the stream properly if I do a return. > > It seems to me that a better solution would be what Kent Beck calls an > âexecute aroundâ method: > > findSomething > ^ stream unchangedDuring: [ :s | > âcode involving various manipulations of stream, including stream next" > stream position + 1 ] > > Iâm not sure that I like the name #unchangedDuring:; perhaps you can think > of a better one? > > The implementation of this would be > > PositionableStream >> #unchangedDuring: aBlock > "execute aBlock with self as argument. Ensure that when this method > returns, > I still have my current state." > | savedPosition | > savedPosition := self position. > ^ [ aBlock value: self ] ensure: [ self position: savedPosition ] > > or perhaps > > PositionableStream >> #unchangedDuring: aBlock > "execute aBlock with self as argument. Ensure that when this method > returns, > I still have my current state." > > ^ [ aBlock value: self copy ] > > which might be better in the case that the stream is a file stream and > aBlock causes the buffer to be refreshed from the disk. > > > - Do other Smalltalks have a method like this? > - Does it already exist in Pharo, somewhere that Iâve failed to look? > - Would it be a worthwhile addition to Pharo? > - What would be a good name? > > > > > -- Guille Polito Research Engineer Centre de Recherche en Informatique, Signal et Automatique de Lille CRIStAL - UMR 9189 French National Center for Scientific Research - *http://www.cnrs.fr <http://www.cnrs.fr>* *Web:* *http://guillep.github.io* <http://guillep.github.io> *Phone: *+33 06 52 70 66 13

Nov. 2, 2017

Peeking at a stream

by Prof. Andrew P. Black

I sometimes find, when working with a stream, that I need to "peek aheadâ a few items, without disturbing the state of the stream. The message stream peek gives me the next item, but doesnât generalize to multiple items, So I end up writing code like this: findSomething | savedPosition result | savedPosition := stream position. stream position: start. âcode involving various manipulations of stream, including stream next" result := stream position + 1. stream position: savedPosition. ^ result This code involves at least two temps, is not very easy to read, and risks not resetting the stream properly if I do a return. It seems to me that a better solution would be what Kent Beck calls an âexecute aroundâ method: findSomething ^ stream unchangedDuring: [ :s | âcode involving various manipulations of stream, including stream next" stream position + 1 ] Iâm not sure that I like the name #unchangedDuring:; perhaps you can think of a better one? The implementation of this would be PositionableStream >> #unchangedDuring: aBlock "execute aBlock with self as argument. Ensure that when this method returns, I still have my current state." | savedPosition | savedPosition := self position. ^ [ aBlock value: self ] ensure: [ self position: savedPosition ] or perhaps PositionableStream >> #unchangedDuring: aBlock "execute aBlock with self as argument. Ensure that when this method returns, I still have my current state." ^ [ aBlock value: self copy ] which might be better in the case that the stream is a file stream and aBlock causes the buffer to be refreshed from the disk. Do other Smalltalks have a method like this? Does it already exist in Pharo, somewhere that Iâve failed to look? Would it be a worthwhile addition to Pharo? What would be a good name?

Nov. 2, 2017