Squeak-dev November 2008

squeak-dev@lists.squeakfoundation.org

119 participants
135 discussions

Re: [Pharo-project] Alien FFI
by John M McIntosh 25 Nov '08

25 Nov '08

Ok, after much mindless poking, cross checking and testing, retesting, writing sunits today I can say the Alien FFI logic is mostly ported into Pharo. In the http://www.squeaksource.com/Alien.html MC library I would suggest someone with an eye towards "Compiler, CompileMethod, ParseNode" carefully look at the changes in "Alien-Core" to ensure I've not trashed some important part of the Pharo Compiler or Traits. I can't say if the changes are destructive or passive? you will find: For VMMaker folks Alien-VMMaker-Support which provides the changes to the latest VMMaker from squeaksource to build a Alien aware VM. Alien-VMMaker-Plugins which is the plugin logic For users of Alien Alien-Core-First - Load this first otherwise you'll get a parsing error later Alien-Core - Load this next, Compiler changes etc.. Alien-LibC - LibC support for os-x/unix/windows Alien-CoreTest - SUnits Optional stuff then is Alien-MacOSX-Core - Objective-C interface Alien-MacOSX-AppKit- Objective-C application api Alien-MacOSX-Examples- Objective-C examples Alien-Win32 - Windows examples. Lastly load Alien-Last-Class-Initialize - To ensure all the initializers were run, just in case MC loses it's mind... Oh and you need a plugin and a VM. It's midnight again and I don't have one compiled up for distribution, Monday's chore... Anyway you should be able to load the source code, point/learn about how it's used. Someone will have to build a Unix one, and a Windows one. For the moment there is NO powerpc version since that requires a bit of icky assembler. Same for the iPhone btw (icky ARM assembler). On 23-Nov-08, at 4:20 AM, Stéphane Ducasse wrote: > the future looks shining :) > > Stef -- = = = ======================================================================== John M. McIntosh <johnmci(a)smalltalkconsulting.com> Corporate Smalltalk Consulting Ltd. http://www.smalltalkconsulting.com = = = ========================================================================

5 10

[ANN] OSProcess 4.3.7 and CommandShell 4.3 on SqueakMap
by David T. Lewis 25 Nov '08

25 Nov '08

These updates are now released on SqueakMap and Universes, and include the latest updates to OSProcess and CommandShell from SqueakSource. For OSProcess: - At system shutdown, break dependency between ThisOSProcess and the OSProcessAccessor. This is done explicitly at shutdown in order to prevent possible problems when an image is restarted on another platform type, in which case a new ThisOSProcess instance is created and the old instance could still have an unwanted dependency on an OSProcessAccessor (bug found by Andreas Raab). - Add WindowsProcess>>waitForCommand: (implementation by Keith Hodges). For CommandShell: - Fix improper handling of executable path resolution for executables not in current PATH list, and remove incorrect permission to execute a file in current (or Squeak default) working directory if not explicitly in the unix PATH. If an executable file is not found in the PATH, an ExpressionEvaluator is created instead of a PipeableOSProcess, such that the ExpressionEvaluator writes a suitable error message to stderr of the command shell and presents an error exit status. This fixes a problem that was exposed on OS X (thanks to Scott Wallace). - Fix initialization bug when OSProcess is not present in the image. - Eliminate unnecessary WeakArray extension. - Update unit tests for OS X. Dave

1 0

Re: [squeak-dev] Re: [Pharo-project] Alien FFI
by John M McIntosh 24 Nov '08

24 Nov '08

On 24-Nov-08, at 6:37 AM, David Pennell wrote: > John, > > Its probably worthwhile to find a place to post Eliot's > documentation files for Alien. > > -david Documentation files and a start on a wiki page is at: http://wiki.squeak.org/squeak/6100 -- = = = ======================================================================== John M. McIntosh <johnmci(a)smalltalkconsulting.com> Corporate Smalltalk Consulting Ltd. http://www.smalltalkconsulting.com = = = ========================================================================

1 0

Better VM <-> plugin API
by Igor Stasenko 24 Nov '08

24 Nov '08

---------- Forwarded message ---------- From: Igor Stasenko <siguctua(a)gmail.com> Date: 2008/11/21 Subject: Re: [squeak-dev] Better VM <-> plugin API To: "Greg A. Woods, Planix, Inc." <woods(a)planix.ca> 2008/11/21 Greg A. Woods; Planix, Inc. <woods(a)planix.ca>: > > On 21-Nov-2008, at 11:50 AM, Igor Stasenko wrote: > >> I was pretty unhappy with some places in VM, where it loads a function >> from plugin by calling ioLoadFunction... >> Also, even more painful to see when one plugin wants to call another >> plugin function. > > I know very little about the VM, but I agree what you describe is ugly and > inelegant, even at a very generic level. > >> Now, plugins first, should declare the atoms they would want to use or >> provide. This can be done once at plugin initialization stage, for >> instance: >> >> static sqInt bitBlitAtom = makeAtom('ioBitBlt'); >> >> Now, after acquiring atom, plugin can register a service fn under this >> atom: >> >> registerService(bitBlitAtom, (void*) bitBlt); >> and upon unloading >> unregisterService(bitBlitAtom); >> >> now, any other plugin can do: >> >> bitBltFn = getService(bitBlitAtom); > > Everything looked good until I got here. > > How is another plugin supposed to access this 'bitBlitAtom' handle that is > effectively always in the first plugin's private storage and literally (as > you've declared it above) also in the other plugin module's private symbol > namespace? (i.e. if it's compiled as a "static" variable in the first > plugin module then there's no way the second plugin can use the same symbol > to refer to the same storage) Its not in a private storage. When you telling makeAtom(), you receiving an atom id for given name which is shared among all plugins. To be more clear, i think there is no need in register/unregister functions, instead there could be only single function: void * setAtomValue(int atom, void* value); which assigns atom new value and returns old one. Atom values are global and seen throughout VM. There is no privacy of any kind - one plugin can assign value to atom, and other can read value from it, or vice versa. So, there is 3 functions: makeAtom setAtomValue getAtomValue if we take an analogy from smalltalk, then its similar to system dictionary with 3 operations: atom := 'something' asSymbol. "intern new symbol, or get existing one" value := Smalltalk at: atom. "get value" Smalltalk at: atom put: value. "set value" > > Perhaps you meant to add a getAtom() function to your API? Then the second > plugin would first do: > > static sqInt theAtom; > > theAtom = getAtom("ioBitBlt"); > > And then it could do the call above as: > > if (theAtom != 0) > bitBltFn = getService(theAtom); > >> if (bitBltFn) { bitBltFn( a, b,c blabla); } > > -- > Greg A. Woods; Planix, Inc. > <woods(a)planix.ca> > > -- Best regards, Igor Stasenko AKA sig. -- Best regards, Igor Stasenko AKA sig.

2 2

Etoys 4.0 and Monticello
by K. K. Subramaniam 24 Nov '08

24 Nov '08

Hi, I got a "Key not found" error while trying to install Monticello in my Etoys squeak-dev-4.0.image (Squeak 4.0-2192). The error also appears in squeak-dev-3.0 images. The debug log is attached. I got stuck in a loop while trying to register a project in Squeaksource using Etoys dev image? Etoys dev image does not have Monticello browser. SqueakMap Package loader, when opened, reported that its client version is too old and I agreed to an update. But the upgrade needed Monticello to be installed. During its installation, I got the above error. Any help is appreciated. TIA .. Subbu

5 9

FileStream purports ascii text to be html
by Mr. Timothy C. Johnson 24 Nov '08

24 Nov '08

Hi, I'll try to file this on Mantis but I don't know if it's a bug, design flaw, deliberate choice, my ignorance, or what. I searched the list archives via Nabble and could not find anything related. In a Browser, choose "printOut" from the class pane menu. This will create a file in your image's directory named className.html. However, it is not an HTML file. It is plaintext. Safari and Opera display the file with no linefeeds. The ultimate method being called by Browser>>printOutClass is FileStream>>writeSourceCodeFrom:baseName:isSt:useHtml: which responds to a "true" value for #useHtml: by simply adding .html to a filename, rather than actually writing out a proper HTML file. This is deceptive. Perhaps the file name should be kept as .txt so a user opening the file by double-clicking is not presented by a web browser full of linefeedless code. - timj --- Browser>>printOutClass "Print a description of the selected class onto a file whose name is the category name followed by .html." Cursor write showWhile: [classListIndex ~= 0 ifTrue: [self selectedClass fileOutAsHtml: true]] ---- FileStream>>writeSourceCodeFrom:baseName:isSt:useHtml: [...] fileName := useHtml ifTrue: [baseName, '.html'] ifFalse: [baseName, extension]. f := FileStream newFileNamed: fileName. f ifNil: [^ self error: 'Cannot open file']. (converter isMemberOf: UTF8TextConverter) ifTrue: [f binary. UTF8TextConverter writeBOMOn: f]. f text. f converter: converter. f nextPutAll: aStream contents. f close.

1 0

Re: [squeak-dev] Re: Bug in 3.10.2 - browser / moving classes
by Mr. Timothy C. Johnson 24 Nov '08

24 Nov '08

Jerome Peace wrote: >> >>I had two browser windows open (standard 3.10.2 default browsers). > >?? Full browsers of Hierarchical browsers?? They were not Hierarchal browsers, AFAIK. Just the standard browser you get from World menu -> open... -> Class browser. > >The reason I ask is that there is a known problem with the later > which will sometimes confuse the browser. >I usually see it when creating a new class. >You may have found a new way to bump into it. Perhaps? > > >>They were both open to the same class category. In one browser, I >>moved a number of methods out of that category into another one (via >>drag-and-drop). After moving about 5 or 6 methods, I got an error: >>subscript out of bounds, in World. I selected "proceed" and I looked >>over at the other browser (the one in which I hadn't made any changes), >>and its classes list pane had become an invalid rectangle (a red >>rectangle with a yellow X inside). Browsing to other categories in the >>troubled browser does not restore the class category pane to operation. > >Invalid drawing panes "latch up". >To fix them it is necessary to not only fix the underlying problem > but to release the latch via: >the red menu>debug>start drawing again. Ah! I didn't know that, thanks. >For browsers it almost always easier to just start fresh. Yes, that did work out well. >> A new browser is fine. >> >>Using 3.10.2 on latest Mac PPC VM. > > >If you are willing to put some effort into getting this tracked and fixed, >please add a report to mantis > describing a way to from a fresh start reproduce the problem. >Enough specifics to reliably reproduce the problem > is a starting point for others to explore and play with it. > >Here are the FAQs: [ ... ] Thanks Jerome. Mantis has always been mysterious to me. I will see if I can duplicate the problem and add a report. - timj

1 0

smalltalk and Web stuff
by Friedrich 24 Nov '08

24 Nov '08

well I know this is way to general. But I'd like nevertheless get some pointers. AFAIKG is Seaside the "first" Web framework for Smalltalk, I've just downloaded AIDA/Web and am playing around it a bit. But let me as this question nevertheless. What alternatives are there for a "pure" Smalltalk Web presence? Currently my systems are mixture of Apache for static stuff and some other web servers to which Apache proxies (please bear with me that it's Ruby based ;-). Now I'd like to have the following things: 1) normal pages 2) forum software 3) bug software 4) shop pages 5) a wiki and blog would be nice 6) support for RCS systems I know that the later may be quite weak for Smalltalks, but well the world does not always talk Smalltalk ;- Any hints and impressions and expressions would be very welcome With best regards Friedrich

9 24

Re: Bug in 3.10.2 - browser / moving classes
by Jerome Peace 24 Nov '08

24 Nov '08

[squeak-dev] Bug in 3.10.2 - browser / moving classes Hi Tim, *** >Mr. Timothy C. Johnson tjohnson at iwu.edu >Sun Nov 23 21:46:26 UTC 2008 > >Hi, > >I had two browser windows open (standard 3.10.2 default browsers). ?? Full browsers of Hierarchical browsers?? The reason I ask is that there is a known problem with the later which will sometimes confuse the browser. I usually see it when creating a new class. You may have found a new way to bump into it. >They were both open to the same class category. In one browser, I >moved a number of methods out of that category into another one (via >drag-and-drop). After moving about 5 or 6 methods, I got an error: >subscript out of bounds, in World. I selected "proceed" and I looked >over at the other browser (the one in which I hadn't made any changes), >and its classes list pane had become an invalid rectangle (a red >rectangle with a yellow X inside). Browsing to other categories in the >troubled browser does not restore the class category pane to operation. Invalid drawing panes "latch up". To fix them it is necessary to not only fix the underlying problem but to release the latch via: the red menu>debug>start drawing again. For browsers it almost always easier to just start fresh. > A new browser is fine. > >Using 3.10.2 on latest Mac PPC VM. If you are willing to put some effort into getting this tracked and fixed, please add a report to mantis describing a way to from a fresh start reproduce the problem. Enough specifics to reliably reproduce the problem is a starting point for others to explore and play with it. Here are the FAQs: FAQ: Is this a known issue ? Where is the place to report bugs (or check if some have already been raised) ? The best place for this info would be to start a Mantis report. (You can get a mantis acct freely and easily). A good place to start is: http://bugs.squeak.org/my_view_page.php Mantis provides a patient persistent way to focus on an issue. I use it to accumulate data on a problem until a solution can be found. It provides a place to alert the community to a problem; -accumulate facts and clues from the analysis; -publish preposed solutions and get feedback; -get solutions harvested and included into the main stream. This FAQ is available at the Squeak wiki site. http://wiki.squeak.org/squeak/Mantis%20FAQ%20and%20Tips aka: http://wiki.squeak.org/squeak/5912 Mantis FAQ and Tips *** - TimJ

1 0

Detecting DNU sent by VM, not by code
by Igor Stasenko 24 Nov '08

24 Nov '08

Hello list, is there any way to detect that DNU message was sent by VM , as result of failed method lookup, not intentionally by language side? I need to have a pseudo-class which intercepts _any_ message you sending to its instance, including #doesNotUnderstood: sent directly by language side. Of course, i can't do anything with messages which handled directly by VM , such as #class , but at least i want to make sure that rest of messages is intercepted with guarantee. -- Best regards, Igor Stasenko AKA sig.

2 2

← Newer
1
2
3
4
5
6
7
8
...
14
Older →

Jump to page: