Squeak-dev September 2005

squeak-dev@lists.squeakfoundation.org

157 participants
266 discussions

RE: Is Wrapping Safe? [Re: Is Set growth thread-safe?]
by Peter Crowther 23 Sep '05

23 Sep '05

> From: [...] jwalsh(a)bigpond.net.au > Thanks Ross, you have given me the partial answer to a > problematic which has bugged me for a while: > It started off as "Is Smalltalk safe from Windows?", > Then progressed to "Is Smalltalk safe from 'C' ?", > Now its "Is Smalltalk Safe from Smalltalk". > This was originally a problem which could not ever happen to > Smalltalk. > Encapsulation and Message passing meant the Method was safe, > from intrudere. ... it's not. Object has plenty of debugging methods for poking at slots on arbitrary objects. It's relatively simple to overwrite a CompiledMethod's bytecodes and even simpler just to replace the method with one you compile up. There is nothing mystical about objects; they're just data. In order to run a debugger, *something* has to break encapsulation. > Can someone please convince me that Smalltalk is safe from > 'C' hackers. On the contrary, I can convince you the other way. Object memory in Smalltalk is merely bytes at locations in memory; the VM uses a C-language stack for some of its work, which can be attacked; and the virtual machine is a chunk of compiled C. If the VM can do it, attacking code can do it; and the VM is the only way of creating any effect in Smalltalk. Therefore, all the effects that you see in Smalltalk can be duplicated by C code, because they are *created* by C code. There is nothing mystical about objects, image, or anything else. - Peter

2 1

Is Wrapping Safe? [Re: Is Set growth thread-safe?]
by jwalsh＠bigpond.net.au 22 Sep '05

22 Sep '05

Thanks Ross, you have given me the partial answer to a problematic which has bugged me for a while: It started off as "Is Smalltalk safe from Windows?", Then progressed to "Is Smalltalk safe from 'C' ?", Now its "Is Smalltalk Safe from Smalltalk". This was originally a problem which could not ever happen to Smalltalk. Encapsulation and Message passing meant the Method was safe, from intrudere. Can someone please convince me that Smalltalk is safe from 'C' hackers. JW ---- Ross Boylan <RossBoylan(a)stanfordalumni.org> wrote: > Thanks everyone for setting me straight. It seems reasonable that > classes are not generally thread-safe, but in the original issue that > got me into this, WeakIdentityKeyDictionary, it seems problematic. > Research: http://minnow.cc.gatech.edu/squeak/1891 > http://www.laputan.org/brant/brant.html :- MethodWrappers for Squeak Robert Hirschfeld (hirschfeld(a)acm.org) wrote WARNING! Incorrectly using the MethodWrappers port may result in an inconsistent image that can't be trusted anymore. Therefore, before experimenting with the MethodWrappers port, please save and backup your image. SNIP Wrappers to the Rescue John Brant, Brian Foote, Ralph E. Johnson, and Donald Roberts Department of Computer Science University of Illinois at Urbana-Champaign Urbana, IL 61801 {brant, foote, johnson, droberts}(a)cs.uiuc.edu Abstract. Wrappers are mechanisms for introducing new behavior that is executed before and/or after, and perhaps even in lieu of, an existing method. This paper examines several ways to implement wrappers in Smalltalk, and compares their performance. Smalltalk programmers often use Smalltalk’s lookup failure mechanism to customize method lookup. Our focus is different. Rather than changing the method lookup process, we modify the method objects that the lookup process returns. We call these objects method wrappers. We have used method wrappers to construct several program analysis tools: a coverage tool, a class collaboration tool, and an interaction diagramming tool. We also show how we used method wrappers to construct several extensions to Smalltalk: synchronized methods, assertions, and multimethods. Wrappers are relatively easy to build in Smalltalk because it was designed with reflective facilities that allow programmers to intervene in the lookup process. Other languages differ in the degree to which they can accommodate change. Our experience testifies to the value, power, and utility of openness. ETC

1 0

Re: [ANN]A directory with Companies using Smalltalk
by Chris Muller 22 Sep '05

22 Sep '05

Glad to see this list, thanks Noury! A similar list can be found at http://www.goodstart.com/whoswho.shtml but it's older and looks like spammers have polluted the top-third.

1 0

Re: Is Set growth thread-safe?
by Chris Muller 22 Sep '05

22 Sep '05

> It's not safe. You need to protect concurrent access to all objects > that are not specifically designed for thread-safety (like > SharedQueues): Actually SharedQueues protect themselves, or at least the code appears intent on doing so (it has its own 'accessProtect' Semaphore). But yes, the standard "non-shared" classes like Set do not.

1 0

Re: OmniBase + Linux
by Ramiro Diaz Trepat 22 Sep '05

22 Sep '05

I get an ODBLockingConflict exception when I run OmniBaseTester test1. The version of OSProcess that is installed is 3.2.5 It's weird that this happens only to me. It must be something silly that is wrong in my configuration. I am attaching the stack trace. Thanks again. r. On 9/22/05, Avi Bryant <avi.bryant(a)gmail.com> wrote: > > On Sep 21, 2005, at 8:09 PM, Ramiro Diaz Trepat wrote: > > > Hello Avi, > > Yes, I did install OSProcess. Indeed, I did it on a fresh image > > (3.9). > > I'd love to have OmniBase working on Linux, right now I am only > > booting Windows because of this problem. > > Hm... what's the exception, and which test do you get it for? > > Locking on linux is definitely problematic compared to on windows: > we've had some deadlocks in production with high loads. But the > basic tests do normally pass. > > Is Win32 locking emulation turned on for OSProcess? It should be by > default... > > Avi > > > > > > On 9/21/05, Avi Bryant <avi.bryant(a)gmail.com> wrote: > > > >> > >> On Sep 21, 2005, at 11:28 AM, Ramiro Diaz Trepat wrote: > >> > >> > >>> Hello everyone, > >>> I've tryied OmniBase on Linux without success. When I run the > >>> tests in OmniBaseTester I always get an exception relative to > >>> locking: > >>> ODBLockxxxx. This does not happen in Windows. > >>> Is there a particular version of the VM or image to run OmniBase > >>> on Linux ? > >>> Thanks in advance. > >>> > >> > >> Do you have OSProcess loaded? It's a requirement for OminBase on > >> Unix. > >> > >> Avi > >> > >> > > > > > > -- > > http://neosmt.com > > > > -- http://neosmt.com

1 0

I love it that way? [Re: How much is my Squeak Image like Sourdough bread?]
by jwalsh＠bigpond.net.au 22 Sep '05

22 Sep '05

My (off-topic?, sorry) answer to this, > It's somewhat ironic that Alan Kay, who urged his > group to "Burn The Disk Packs" > (http://www.smalltalk.org/smalltalk/TheEarlyHistoryOfSmalltalk_V.html) > is using an image which has basically survived from the early eighties > :-) But I love it that way. is this, as I see it of course: It is because Smalltalk is so aMAZEing.<http://en.wikipedia.org/wiki/Maze> that most, including myself "love it that way". It is the original (apriori) goodie.<http://en.wikipedia.org/wiki/Apriori_algorithm >. The Object, toward which Smalltalkers Orient themselves, is fixed, like the fence surrounding a Kindergarten. That fence plays the limiting role which Reason itself is supposed to supply, but which, in children, is still too underdeveloped. Alan Kay (correct me if I am wrong), being a professional child behaviorist and a pragmatic knew that, all to well. He sought to nurture the Positive Imagination by limiting the Negative Reason. So he gathered together some other adult "children" from a University to help him build a kind of Turing Machine <http://en.wikipedia.org/wiki/Turing_Machine> to frame his ideas, and encapsulate the imagination. A machine which could squease the immagination like a lemon, "until the pips squeak". He also knew that the child within us could never be extinguished: That at each stage of our development, through to old age, we will behave just like a child. Our Imagination growing ever richer with the years, but our capacity to use our inner Reason, as nature determined, held back. That fence, formerly for our protection, then becomes our undoing. Reflect upon the egg. The shell is so designed by nature, strong enough to protect the growing chicken within. Too strong, it will not allow the chicken to escape, and it will die, before it can take breath of air. JW ---- Hans-Martin Mosner <hmm(a)heeg.de> wrote: > Austin King wrote: > > > Greetings Squeakers, > > I am interested in the history of the Squeak 3.8 image. > > > > Metaphorically how much is my Squeak image like sourdough bread? > > That's an interesting comparison (especially as I rather like sourdough > bread...) > .. > :-) But I love it that way. > > Cheers, > Hans-Martin >

2 1

[GOODIE] X10 CM11a interface
by William E Harford 22 Sep '05

22 Sep '05

Below is a copy of an email sent to this lists a few years back announcing a module to the CM11a X10 interface. I would like to rewrite my home automation server in Smalltalk using Squeak the current perl code is a bit of a bear to work with and the code for the interface to CM11a would give a huge boost to that process. If you have any information on the whereabouts of the code or the author please enlighten. Thanks Will. ______________________________________________________________ A MESSAGE SENT TO Squeak-Dev on Thu Jan 24 16:52:45 CET 2002 Download it from: http://spair.swiki.net/X10 "Change Set: X10-CM11a Date: 24 January 2002 Author: Stephen Pair This change set provides a model of the X10 CM11A module. It allows you to use Squeak to send and receive X10 commands over your house electrical wiring (you need the X10 CM11a module attached to a serial port of course). It also provides a model of the CM11a EEPROM memory to allow you to download timers and macros to the module for disconnected operation. I have not documented this stuff in the slightest. I wrote it to program a CM11a module that I have because the X10.com ActiveHome software would not install or run properly on my Win2k machine. I use this to program my CM11a (so I know it works...at least on some level). There is no GUI interface...just the model. I am releasing it in this immature state mainly because I don't plan to spend any more time on it anytime soon. If you would like help getting this stuff to work, send me an email: spair at acm.org. The following is an example of the EEPROM memory setup: Turns A2 lights on at 5pm and off at midnight, also A4 on is a macro that dims A3 to 50% ----- |memory macro1 timer1 macroInit1 | memory := X10Memory new. command1 := X10BrightnessCommand new houseCode: $A; addDevice: 3; function: #dim; brightness: 0.5; yourself. command2 := X10BrightnessCommand new houseCode: $A; addDevice: 1; function: #dim; brightenFirst: true; brightness: 0; yourself. command3 := X10BasicCommand new houseCode: $A; addDevice: 2; function: #on; yourself. command4 := X10BasicCommand new houseCode: $A; addDevice: 2; function: #off; yourself. macro1 := X10Macro new delay: 0; addCommand: command1; yourself. macro2 := X10Macro new delay: 15; addCommand: command2; yourself. macro3 := X10Macro new delay: 0; addCommand: command3; yourself. macro4 := X10Macro new delay: 0; addCommand: command4; yourself. timer1 := X10TimerInitiator new dayRange: (1 to: 366); days: #(Sunday Monday Tuesday Wednesday Thursday Friday Saturday); startTime: (Time fromString: '17:00:00'); stopTime: (Time fromString: '23:59:00'); startMacro: macro3; stopMacro: macro4; yourself. macroInit1 := X10MacroInitiator new houseCode: $A; deviceCode: 4; macro: macro1; yourself. memory addMacro: macro1. memory addMacro: macro2. memory addMacro: macro3. memory addMacro: macro4. memory addTimerInitiator: timer1. memory addMacroInitiator: macroInit1. memory. ------- And the following will download the memory to a CM11AModule (on serial port 1): (CM11AModule new: 1) writeEeprom: memory You can also issue direct X10 commands: (CM11AModule new: 1) address: 'A2'; function: #on; address: 'A3'; function: #dim by: 1. And, you may open the transcript to monitor any signals issued by other X10 devices. Note: You should not create a new CM11AModule every time you want to do something with it (as I have shown in these examples), create one, then keep it around for issue commands and configuring it. "

1 0

re: Running InterpreterSimulator and introduction to Spoon
by Craig Latta 22 Sep '05

22 Sep '05

Hi Luc-- (I hope you don't mind my including the Spoon and Squeak lists in my response. I think the simulator is of interest in those places.) > I like to play with the interpreter simulator a little more before > learning more about Spoon. > > What version of VM and image do you use to get InterpreterSimulator up > and running? I tried the 3.7.7 VM and 3.7 and 3.8 stock release on > linux without success. I've been using Squeak 3.2 for Spoon (that was the current version when I started in August 2002, and my plan is to move everything from there to a Spoon 1 final system). > As experiment, I followed the guidelines described in > http://aspn.activestate.com/ASPN/Mail/Message/squeak/1196663 and > http://minnow.cc.gatech.edu/squeak/2465 > and got the following trace back > > Subscript is out of bounds: 0 > Array at: > functionPointerFor inClass: > addNewMethodToClass > internalFindNewMethod > commonSend > normalSend > secondExtendedSendBytecode > dispatchOn in: > test So it looks like we've accomplished our annual breaking of the simulator? :) If you decide to give it a go with 3.2, you might also want my simulator fixes[1]. thanks, -C [1] http://tinyurl.com/dmjo8 -- Craig Latta improvisational musical informaticist www.netjam.org Smalltalkers do: [:it | All with: Class, (And love: it)]

2 3

Re: How much is my Squeak Image like Sourdough bread?
by Austin King 22 Sep '05

22 Sep '05

Jecel thank you for the specific reference in the green book and the interesting response.

1 0

The Weekly Squeak No.4: 12 September - 18 September
by Giovanni Corriga 22 Sep '05

22 Sep '05

Dear squeakers, Hello, and welcome back to the The Weekly Squeak, a weekly report on what's going on in the world of Squeak. The feedback on the past issues has been great, and has moved us to try and improve the format of this newsletter. So, don't miss the next issue, which will sport the first in a series of interviews to the Squeak developers! Meanwhile, in this issue you may find: 1. New releases of AST and RefactoringEngine 2. Developer initials, identity, and SqueakMap 3. Coordinators' summer report 4. Videos, videos everywhere... 5. A security model for SmallWiki 2 6. VerySmallTalk PhD Position Ecole des Mines de Douai 7. Test VMMaker branch for 38b4 on SVN server 8. On the Squeak License 9. A new name for SmallWiki 2? 10. Some clarifications on the Squeak Foundation 11. Two articles on Squeak 12. Your articles on Smalltalk now worth more 13. New Mac VM 3.8.9b1 14. A preview of the next issue 15. Upcoming Squeak Chats You'll find the complete report at http://minnow.cc.gatech.edu/squeak/5740 . The past issues archive can be found at http://minnow.cc.gatech.edu/squeak/5751 .

2 1

← Newer
1
...
4
5
6
7
8
9
10
...
27
Older →

Jump to page: