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

Re: [Pharo-project] [Vm-dev] New Cog builds available: included Freetype plugin on macs by default

by Mariano Martinez Peck

On Tue, Jul 5, 2011 at 5:04 PM, Igor Stasenko <siguctua(a)gmail.com> wrote: > > Hello, > > i updated the ConfigurationOfCog and added verison 2.3 which now also > loads freetype plugin into vmmaker image. > > I also made sure that plugin automatically builds on macs (no other > platforms yet) and included it by default into CMakeVMMaker configs. > > There was some hiccups, i had to deal with: > The autoconf toolchain (or most probably the configuration scripts for > freetype library) don't like the spaces in path. > So i have to rename all jobs on Hudson to avoid using spaces (because > hudson creating a subdirectory for jobs with same name). > > The renamed jobs are: > https://pharo-ic.lille.inria.fr/hudson/job/Cog-Mac-Carbon/ > https://pharo-ic.lille.inria.fr/hudson/job/Cog-Mac-Cocoa/ > https://pharo-ic.lille.inria.fr/hudson/job/CogMT-Mac-Cocoa/ > https://pharo-ic.lille.inria.fr/hudson/job/StackVM-Mac-Carbon/ > https://pharo-ic.lille.inria.fr/hudson/job/StackVM-Mac-Cocoa/ > > Hi Igor. You have been working so much that I am now lost ;) are all those jobs already building from blessed? > There is another idiosyncrasy: outdated toolchain for autoconf. > I had to install mac ports on slave server. > See > http://code.google.com/p/cog/wiki/FT2Plugin > for details. > > ( i tested Carbon builds, now they are including FT2Plugin). > > Would that mean that http://code.google.com/p/pharo/issues/detail?id=4374 is fixed? because last comment of Alain says it is not. > Esteban, could you please verify that it also works for Cocoa configs? > I changed configs to include FT2Plugin by default in Cocoa configs as > well, > but i haven't tested builds. > > So, currently following builds should contain FT2Plugin by default: > > https://pharo-ic.lille.inria.fr/hudson/job/Cog-Mac-Carbon/ > https://pharo-ic.lille.inria.fr/hudson/job/StackVM-Mac-Carbon/ > > Freetype library are built from scratch by downloading library sources > directly from git repository (maybe its a bit too much, but it works). > See /unixbuild/FT2Plugin/makeFreetype.sh > > this script is then used by cmake config to build libfreetype.a which > is statically linked with freetype plugin. > > -- > Best regards, > Igor Stasenko AKA sig. > -- Mariano http://marianopeck.wordpress.com

July 6, 2011

[Pharo-project] [update 1.4] #14027

by Marcus Denker

14027 ----- Issue 4502: Fix for TimeProfileBrowser http://code.google.com/p/pharo/issues/detail?id=4502 -- Marcus Denker -- http://www.marcusdenker.de INRIA Lille -- Nord Europe. Team RMoD.

July 6, 2011

[Pharo-project] [update 1.3] #13282

by Marcus Denker

13282 ----- Issue 4502: Fix for TimeProfileBrowser http://code.google.com/p/pharo/issues/detail?id=4502 -- Marcus Denker -- http://www.marcusdenker.de INRIA Lille -- Nord Europe. Team RMoD.

July 6, 2011

Re: [Pharo-project] About locating home

by StÃ©phane Ducasse

thanks! I'm not sure that we should read the code but we can definitively use the list of known locations. Stef On Jul 5, 2011, at 7:07 PM, Eliot Miranda wrote: > > > On Tue, Jul 5, 2011 at 1:58 AM, Marcus Denker <marcus.denker(a)inria.fr> wrote: > > On Jul 5, 2011, at 10:53 AM, Igor Stasenko wrote: > > > > > i'm not sure if there are any prims which allowing you to access > > environment vars. > > I remember i made an example for FFI how to read them.. but that means > > it will work only using FFI. > > I personally think we should (mid-term) get one *really* good and fast FFI and > than move as many things as possible from the VM to the image... > > I agree with this. However, in this case... > > In the Qwaq VM (now open source but under GPL :( :( ) there is a FileDialogPlugin and it contains primitiveGetFileLocation > "Primitive. Query for a common file location. > Arguments: > location: String describing the common file location. > Return value: The path to the designated location. > Known locations: > 'home' - the user's home directory > 'desktop' - the user's desktop directory > > 'temp' - the temp directory to use > 'preferences' - the place to store (per user) app preferences > 'applications' - the directory for installing applications > 'fonts' - the directory to install fonts in the system > > 'documents' - the users documents folder > 'music' - the users default location for music > 'pictures' - the users default location for pictures > 'videos' - the users default location for videos > " > > and there are implementations of fileDialogGetLocation for Mac Windows and linux. This could serve as a spec. I know you can't use the code directly. But I'm sure it would be fine to read the code and use it as a spec. At least the functionality has been thought through. I'll also ask Andreas whether there's a chance that the code could be released under MIT. > > Security (what people in the past used as an excuse to not do this) we should > deal with on another level. > > Marcus > > -- > Marcus Denker -- http://www.marcusdenker.de > INRIA Lille -- Nord Europe. Team RMoD. > > > > > > -- > best, > Eliot >

July 6, 2011

Re: [Pharo-project] [Tiling Window Manager ] DNU when topWindow is Nil

by laurent laffont

And there's some tests ! Great !!! Thank you !!!!! I've set ConfigurationOfTilingWindowManager version 1.1 as stable for Pharo 1.3. I've created version 1.2 to start working on Pharo 1.4. Laurent. On Tue, Jul 5, 2011 at 6:19 PM, Bernardo Ezequiel Contreras < vonbecmann(a)gmail.com> wrote: > hi Laurent, > i added a new version because every move gives you a DNU when topWindow is > Nil. > could you check it? it may need same changes. > Steps to reproduce: > # install TWM > # close all windows > # press any move button > Actual Result: a DNU bounds > Expected Result: nothing because it's nothing to move. > > Thanks >

July 5, 2011

Re: [Pharo-project] how can I get bit representation of numbers

by Nicolas Cellier

If you want to know the number of bits used to represent a SmallInteger, just evaluate: SmallInteger maxVal highBit + 1. SmallInteger maxVal highBit tells the highest bit which can be used to represent a positive SmallInteger, and + 1 accounts for the sign bit of the SmallInteger (0 for positive, 1 for negative). Then you can just rewrite: bitRepresentation ^(self bitRepresentationLength to: 1 by: -1) collect: [:i | Character value: $0 charCode + (self bitAt: i)] as: String Integer>>bitRepresentationLength ^self digitLength * 8 "make sure positive integer bitRepresentation always begins with 0" + (self positive ifTrue: [1] ifFalse: [0]) SmallInteger>>bitRepresentationLength "always use as many bits as the native format to represent a SmallInteger" ^self class maxVal highBit + 1 Nicolas 2011/7/5 StÃ©phane Ducasse <stephane.ducasse(a)inria.fr>: > Salut nicolas :) > >> For positive integers, we have an infinite serie of 0 bits, but we >> don't care, we just don't print them. >> The problem with two complement is that you have an infinite serie of >> leading 1 bits... > > Yes :) > >> Otherwise, you can access the bit at any rank with bitAt: >> For example, you could use >> bitRepresentation >> Â Â Â ^(self digitLength * 8 + 1 to: 1 by: -1) collect: [:i | Character >> value: $0 charCode + (self bitAt: i)] as: String > > Included in my question is the fact that small integer are encoded on 31 bits (may be I'm wrong) > so printing 32 bits represent LargeINteger and so far I want small integetr so was my assumption correct? > and what would be a good method to return bitString of a Small Integer.s > > 2 raisedTo: 29 > Â Â Â Â returns 536870912 > > 536870912 class > Â Â Â Â returns SmallInteger > > 2 raisedTo: 30 > Â Â Â Â returns 1073741824 > > 1073741824 class > Â Â Â Â returns LargePositiveInteger > > -1073741824 class > Â Â Â Â returns SmallInteger > > 2 class maxVal > Â Â Â Â returns 1073741823 > > -1 * (2 raisedTo: (31-1)) > Â Â Â Â returns -1073741824 > > (2 raisedTo: 30) - 1 > Â Â Â Â returns 1073741823 > > >> >> The first bit will aways be the sign with the + 1 trick. >> >> Nicolas >> >> 2011/7/4 Sven Van Caekenberghe <sven(a)beta9.be>: >>> >>> On 04 Jul 2011, at 20:16, StÃ©phane Ducasse wrote: >>> >>>> I would like to see the two complement representation of numbers. >>> >>> This is what I do, for reading/writing unsigned or two complement signed integer from/to byte streams. >>> >>> Note that two complement is only defined for a specific number size, 8, 16, 32 bits. >>> >>> unsignedToSigned: integer size: size >>> Â Â Â Â ^ integer < (2 raisedTo: size - 1) >>> Â Â Â Â Â Â Â Â ifTrue: [ integer ] >>> Â Â Â Â Â Â Â Â ifFalse: [ (self twoComplement: integer size: size) negated ] >>> >>> signedToUnsigned: integer size: size >>> Â Â Â Â ^ integer negative >>> Â Â Â Â Â Â Â Â ifTrue: [ self twoComplement: integer size: size ] >>> Â Â Â Â Â Â Â Â ifFalse: [ integer ] >>> >>> twoComplement: integer size: size >>> Â Â Â Â | mask | >>> Â Â Â Â mask := (2 raisedTo: size) - 1. >>> Â Â Â Â ^ mask bitAnd: ((integer abs bitXor: mask) + 1) >>> >>> These are also very handy in this context (I believe I once submitted that as an issue): >>> >>> integerFromByteArray: bytes >>> Â Â Â Â | integer | >>> Â Â Â Â integer := 0. >>> Â Â Â Â bytes withIndexDo: [ :each :index | >>> Â Â Â Â Â Â Â Â integer := integer + (each bitShift: (bytes size - index) * 8) ]. >>> Â Â Â Â ^ integer >>> >>> and Integer>>#asByteArrayOfSize: >>> >>> Once you have a byte representation, you can render it as bits as well. >>> >>> Sven >>> >>> >>> >>> >> > > >

July 5, 2011

Re: [Pharo-project] About locating home

by Eliot Miranda

On Tue, Jul 5, 2011 at 1:58 AM, Marcus Denker <marcus.denker(a)inria.fr>wrote: > > On Jul 5, 2011, at 10:53 AM, Igor Stasenko wrote: > > > > > i'm not sure if there are any prims which allowing you to access > > environment vars. > > I remember i made an example for FFI how to read them.. but that means > > it will work only using FFI. > > I personally think we should (mid-term) get one *really* good and fast FFI > and > than move as many things as possible from the VM to the image... > I agree with this. However, in this case... In the Qwaq VM (now open source but under GPL :( :( ) there is a FileDialogPlugin and it contains primitiveGetFileLocation "Primitive. Query for a common file location. Arguments: location: String describing the common file location. Return value: The path to the designated location. Known locations: 'home' - the user's home directory 'desktop' - the user's desktop directory 'temp' - the temp directory to use 'preferences' - the place to store (per user) app preferences 'applications' - the directory for installing applications 'fonts' - the directory to install fonts in the system 'documents' - the users documents folder 'music' - the users default location for music 'pictures' - the users default location for pictures 'videos' - the users default location for videos " and there are implementations of fileDialogGetLocation for Mac Windows and linux. This could serve as a spec. I know you can't use the code directly. But I'm sure it would be fine to read the code and use it as a spec. At least the functionality has been thought through. I'll also ask Andreas whether there's a chance that the code could be released under MIT. > > Security (what people in the past used as an excuse to not do this) we > should > deal with on another level. > > Marcus > > -- > Marcus Denker -- http://www.marcusdenker.de > INRIA Lille -- Nord Europe. Team RMoD. > > > -- best, Eliot

July 5, 2011

Re: [Pharo-project] About locating home

by Igor Stasenko

On 5 July 2011 15:40, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr> wrote: > Thanks igor. > So this is more complex I thought :) > > Stef > > On Jul 5, 2011, at 10:53 AM, Igor Stasenko wrote: > >> Here what they returning (on macs) >> >> TTT class>>getSecuredir >> Â Â Â <primitive: 'primitiveGetSecureUserDirectory' module: 'SecurityPlugin'> >> Â Â Â self primitiveFailed. >> >> TTT getSecuredir >> '/foobar/tooBar/forSqueak/bogus/' >> >> >> TTT class>>primitiveGetUntrustedUserDirectory >> Â Â Â <primitive: 'primitiveGetUntrustedUserDirectory' module: 'SecurityPlugin'> >> Â Â Â self primitiveFailed. >> >> TTT primitiveGetUntrustedUserDirectory >> '/Users/sig/Library/Preferences/Croquet/Internet/Untrusted' >> >> as you can see none of them could be called 'home', which in my case should be >> /Users/sig > > what colin did was to do parent parent parent parent :) > yeah, but this is err.. lame. :) Because there is no any guarantee that it won't change. So, direct request, served by VM to answer a user's home dir would be preferable. > Stef > > > >> >> On unix system(s), this is easy - one could just use HOME environment >> variable. On windows.. its different. >> >> i'm not sure if there are any prims which allowing you to access >> environment vars. >> I remember i made an example for FFI how to read them.. but that means >> it will work only using FFI. >> >> >> On 5 July 2011 10:47, Igor Stasenko <siguctua(a)gmail.com> wrote: >>> There are two primitives defined in SecurityPlugin: >>> >>> primitiveGetSecureUserDirectory >>> primitiveGetUntrustedUserDirectory >>> >>> i'm not sure what they return, but this is all it has. >>> >>> On 4 July 2011 18:01, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr> wrote: >>>> do yo have an idea of the primitive? >>>> >>>> Stef >>>> >>>>>> >>>>>> FS uses SecurityManager as a fix point to identify home. >>>>>> Now we do not have Security manager so I'm wondering if we should not remove home/document/desktop from FS, >>>>>> else any ideas how we can replace the entry point a securityManager? >>>>>> >>>>> I think you can place a primitives which providing these paths somewhere in FS. >>>>> >>>>>> Stef >>>>>> >>>>> >>>>> >>>>> >>>>> -- >>>>> Best regards, >>>>> Igor Stasenko AKA sig. >>>>> >>>> >>>> >>>> >>> >>> >>> >>> -- >>> Best regards, >>> Igor Stasenko AKA sig. >>> >> >> >> >> -- >> Best regards, >> Igor Stasenko AKA sig. >> > > > -- Best regards, Igor Stasenko AKA sig.

July 5, 2011

Re: [Pharo-project] About locating home

by StÃ©phane Ducasse

Yes so far I just want to make the tests of FS green. Stef On Jul 5, 2011, at 10:58 AM, Marcus Denker wrote: > > On Jul 5, 2011, at 10:53 AM, Igor Stasenko wrote: > >> >> i'm not sure if there are any prims which allowing you to access >> environment vars. >> I remember i made an example for FFI how to read them.. but that means >> it will work only using FFI. > > I personally think we should (mid-term) get one *really* good and fast FFI and > than move as many things as possible from the VM to the image... > > Security (what people in the past used as an excuse to not do this) we should > deal with on another level. > > Marcus > > -- > Marcus Denker -- http://www.marcusdenker.de > INRIA Lille -- Nord Europe. Team RMoD. > >

July 5, 2011

Re: [Pharo-project] About locating home

by StÃ©phane Ducasse

Thanks igor. So this is more complex I thought :) Stef On Jul 5, 2011, at 10:53 AM, Igor Stasenko wrote: > Here what they returning (on macs) > > TTT class>>getSecuredir > <primitive: 'primitiveGetSecureUserDirectory' module: 'SecurityPlugin'> > self primitiveFailed. > > TTT getSecuredir > '/foobar/tooBar/forSqueak/bogus/' > > > TTT class>>primitiveGetUntrustedUserDirectory > <primitive: 'primitiveGetUntrustedUserDirectory' module: 'SecurityPlugin'> > self primitiveFailed. > > TTT primitiveGetUntrustedUserDirectory > '/Users/sig/Library/Preferences/Croquet/Internet/Untrusted' > > as you can see none of them could be called 'home', which in my case should be > /Users/sig what colin did was to do parent parent parent parent :) Stef > > On unix system(s), this is easy - one could just use HOME environment > variable. On windows.. its different. > > i'm not sure if there are any prims which allowing you to access > environment vars. > I remember i made an example for FFI how to read them.. but that means > it will work only using FFI. > > > On 5 July 2011 10:47, Igor Stasenko <siguctua(a)gmail.com> wrote: >> There are two primitives defined in SecurityPlugin: >> >> primitiveGetSecureUserDirectory >> primitiveGetUntrustedUserDirectory >> >> i'm not sure what they return, but this is all it has. >> >> On 4 July 2011 18:01, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr> wrote: >>> do yo have an idea of the primitive? >>> >>> Stef >>> >>>>> >>>>> FS uses SecurityManager as a fix point to identify home. >>>>> Now we do not have Security manager so I'm wondering if we should not remove home/document/desktop from FS, >>>>> else any ideas how we can replace the entry point a securityManager? >>>>> >>>> I think you can place a primitives which providing these paths somewhere in FS. >>>> >>>>> Stef >>>>> >>>> >>>> >>>> >>>> -- >>>> Best regards, >>>> Igor Stasenko AKA sig. >>>> >>> >>> >>> >> >> >> >> -- >> Best regards, >> Igor Stasenko AKA sig. >> > > > > -- > Best regards, > Igor Stasenko AKA sig. >

July 5, 2011