Squeak-dev July 2001

squeak-dev@lists.squeakfoundation.org

168 participants
308 discussions

Re: Jerry Archibald, 1940-2001
by Sam Adams 01 Jul '01

01 Jul '01

I first met Jerry at OOPSLA 86 on the exhibit floor, where he was watching me demo an implementation of John Sowa's conceptual graphs in Tektronix Smalltalk. His encouragement then and over the next year helped cement my dedication to a career using Smalltalk. Thanks, Jerry. Sam S. Adams, IBM Distinguished Engineer, IBM Research tie line 444-0736, outside 919-254-0736, email: ssadams(a)us.ibm.com <<Hebrews 11:6, Proverbs 3:5-6, Romans 1:16-17, I Corinthians 1:10>>

1 0

Re: GraphicsFile guru needed!
by Joern Eyrich 01 Jul '01

01 Jul '01

Have you tried Chris Reuter's stuff? it files into my image with the latest updates without problems and seems to work ok. You mentioned in an earlier post that you cannot access chris's web site. The link from the sqfixes archive is no longer active, seemingly, but I can access http://www.csclub.uwaterloo.ca/~cgreuter/downloads/, try that. If it doesn't work, contact me off-list and I'll send it to you. John Hinsley wrote: > Anyone who knows anything about GraphicsFiles got any ideas about this > one? I'm trying to get Squeak 3.1alpha to accept Duane Maxwell's > XPMReadWriter which was written for 2.7. [...]

1 0

Re: BankAccount Tutorial
by Joern Eyrich 01 Jul '01

01 Jul '01

Doug Way wrote: > Cool. Still, the problem remains that most people will go look at the out-of-date one on www.squeak.org. Maybe that one could point to the Swiki instead. (on the other hand, sometimes it's nice to have a more stable tutorial maintained on a non-Swiki page... but then again, if it's not maintained... :) ) Exactly! I volunteer to correct things like these that are brought up on the list, if someone gives me an account. > Speaking of the Swiki, is the new look related to it moving to a Linux box, or is it still running on a Mac? Also, I tried to edit the tutorial page that Joern copied, and it wouldn't let me edit the text, but there was no indication that it was read-only. Are certain colored pages read-only? when I try to telnet to minnow, it says it's an i686 running Read Hat 7.1. I also had the problem that Netscape (4.7 Win32) wouldn't let me edit it. It seems it is limited to 30000 characters in the text edit box, and the tutorial is slightly larger. IE (5.5) doesn't have that problem (and lets you search for text in the edit box...)

6 7

Re: BankAccount Tutorial
by Joern Eyrich 01 Jul '01

01 Jul '01

John Hinsley wrote: > Well done Joern, well done John! Thanks, but putting it this way really overrates what I did. All of the hard parts were done by others. John M. did the original tutorial and someone else on the list (Bob Arning?) figured out what the new code should look like. I only put it together.

1 0

Re: What is happening with SqueakFoundation
by Karl Ramberg 01 Jul '01

01 Jul '01

Cees de Groot wrote: > > Stephane Ducasse <ducasse(a)iam.unibe.ch> said: > >Does anybody have any news about SqueakFoundation effort? > >I hope something will come out of that. > > > She's softly sleeping, waiting for SqC to find a home and thus the time > to kiss her awake again... > > The most important issues that are on the agenda of SqF are: > - Web consolidation (bring website, Swiki, mailinglists, ... under > a single umbrella, both administratively and URL-wise); > - A move from the "canonical" version of Squeak from SqC to SqF so SqF > can do the boring management (release mgt, bugfixing, ...). We really need someone collect enhancements, bugfixes etc from the list to incorporate into the release stream. It's a looooooong time since it happened last. Karl > > IMHO, both need considerable attention from SqC and probably a lot of time to > hash out details of the cooperation and responsibilities. It's summer time, > and SqC (as far as I know) is still "on the move". So it's probably better to > wait until both are over so people can concentrate again. > > In the meantime, we are about to move this mailing list to a new home...

1 0

Re: BankAccount Tutorial
by John Hinsley 01 Jul '01

01 Jul '01

Rosemary Michelle Simpson wrote: > > On Sat, 30 Jun 2001, Joshua D. Boyd wrote: > > > On Sat, 30 Jun 2001, John Hinsley wrote: > > > > > This seems a good compromise, but I wonder if people go to Mark's book > > > because of Squeak or to Squeak because of Mark's book? > > > > I bought Mark's book because of squeak. However, friends who don't squeak > > have shown interest in buying/borrowing the book, so their introduction > > would be through Mark. However, we are discussing a tutorial seperate > > from that book, so if people can get to the tutorial, then they should be > > able to also upgrade their squeak. > > However, it makes sense to stay with the Squeak 2.8 that is on the CD > while you are working with the book and its exercises/example code. When > learning a new language, especially one as different from most as is > Squeak, it seems wise to not change the language version in mid-stream. > Having said that I must confess that I jumped all the way to the 3.1 alpha > because I wanted the "class comments with it" facility. Still, I kept the > 2.8 image and use it in conjunction with the book. Again, that seems a good compromise: if I was sensible I might go that way. Not being a compromising kinda person I think I'll try the code with 3.1! I may well fall flat on my face (not unusual!) but I can always chicken out/ask for help. In some ways, as soon as we start to use something with a lead time as long as a book, or a course, we're stuck in this kind of bind. If I try and do it my awkward way it's because I'm awkward and have really bad memories of doing Java with the old AWT (which, I later discovered, no-one "really" uses -- and no wonder!) and learning C on DOS and discovering that none of that non-ANSI compliant code they taught you, none of those calls to DOS kludge libraries, could be made to work with any degree of ease on any operating system where C still held any degree of sway. Ouch! It still hurts. Cheers John -- ****************************************************************************** Marx: "Why do Anarchists only drink herbal tea?" Proudhon: "Because all proper tea is theft." ******************************************************************************

3 2

Passing argument to BitBlt
by Karl Ramberg 01 Jul '01

01 Jul '01

I made a experimental hue saturation brightness rule to BitBlt (attached) Now I wonder how to pass arguments to it. My first thought was using the halftoneForm, so I made a integer by printing this hue _ 10. saturation _ 0.0. brightness _ 0.05. hueMask _ 16rFFF. svMask_ 16rFF. hShift _ 16. sShift _ 8. halfToneForm _ ((hue bitAnd: hueMask) bitShift: hShift) + (((saturation * svMask) rounded bitAnd: svMask) bitShift: sShift) + ((brightness * svMask) rounded bitAnd: svMask). -> 655373 And then the doIt: s _ Form fromFileNamed:'path/to/image'. half _ Form extent: 1@1 depth: 32 bits: 655373 . (BitBlt destForm: Display sourceForm: s halftoneForm: half combinationRule: 39 destOrigin: 0 @ 0 sourceOrigin: 0 @ 0 extent: s extent clipRect: Display computeBoundingBox)copyBitsSimulated . But for the rule to access the bits I have to use bits _ halftoneForm bits. inside the BitBlt rule. Now I wonder if this will work if I want to compile the new BitBlt rule in a C compiler ? I also wonder if BitBlt can be compiled as a external plugin. I tried once a few month back and Squeak hang at start up. Is there another way to pass the argument ? Karl 'From Squeak3.1alpha of 7 March 2001 [latest update: #4164] on 1 July 2001 at 8:16:38 am'! !BitBltSimulation methodsFor: 'combination rules' stamp: 'kfr 7/1/2001 08:13'! hsv: sourceWord with: destWord | r g b brightness max min saturation span hue i f p q t red green blue componentMax componentMask redShift greenShift rgb adjustHue adjustSaturation adjustBrightness bits | bits _ halftoneBase object. adjustHue _ (bits bitShift: -16) bitAnd: 16rFFF. adjustSaturation _ (bits bitShift: -8) bitAnd: 16rFF. adjustBrightness _ bits bitAnd: 16rFF. adjustSaturation _ (adjustSaturation/ 255) asFloat. adjustBrightness _ (adjustBrightness/ 255) asFloat. r _ (sourceWord bitShift: -16) bitAnd: 16rFF. g _ (sourceWord bitShift: -8) bitAnd: 16rFF. b _ sourceWord bitAnd: 16rFF. r _ (r/ 255) asFloat. g _ (g/ 255) asFloat. b _ (b/ 255) asFloat. max _ ((r max: g) max: b). min _ ((r min: g) min: b). "(max = 1) ifTrue:[self halt]." brightness _ max . saturation _ ((max - min) max: 0.0) min: 1.0. saturation _ ((saturation + adjustSaturation) asFloat max: 0.0) min: 1.0. brightness _ ((brightness + adjustBrightness) asFloat max: 0.0) min: 1.0. span _ (max - min) asFloat. span = 0.0 ifTrue: [ hue _ 0.0 ] ifFalse:[ r = max ifTrue: [ hue _ ((g - b) asFloat / span) * 60.0] ifFalse: [g = max ifTrue: [ hue _ 120.0 + (((b - r) asFloat / span) * 60.0). ] ifFalse: [ hue _ 240.0 + (((r - g) asFloat / span) * 60.0) ]]. hue < 0.0 ifTrue: [ hue _ 360.0 + hue ]]. hue _ hue + adjustHue. (hue < 0.0 or: [hue >= 360.0]) ifTrue: [hue _ hue - ((hue quo: 360.0) asFloat * 360.0)]. hue _ hue / 60.0. i _ hue asInteger. "integer part of hue" f _ hue fractionPart. "fractional part of hue" p _ (1.0 - saturation) * brightness. q _ (1.0 - (saturation * f)) * brightness. t _ (1.0 - (saturation * (1.0 - f))) * brightness. 0 = i ifTrue: [ red _ brightness. green_t .blue_p ]. 1 = i ifTrue: [ red_ q. green_brightness. blue_p ]. 2 = i ifTrue: [ red _ p. green_ brightness. blue_t ]. 3 = i ifTrue: [ red _ p. green_ q. blue _ brightness ]. 4 = i ifTrue: [ red_ t. green_ p. blue_brightness ]. 5 = i ifTrue: [ red_ brightness. green_p. blue_q ]. componentMax _ 255.0. componentMask _ 255. redShift _ 16. greenShift _ 8. rgb _ (((red * componentMax) rounded bitAnd: componentMask) bitShift: redShift) + (((green * componentMax) rounded bitAnd: componentMask) bitShift: greenShift) + ((blue * componentMax) rounded bitAnd: componentMask). ^ rgb.! ! !BitBltSimulation methodsFor: 'initialize-release' stamp: 'kfr 6/30/2001 09:19'! initBBOpTable self cCode: 'opTable[0+1] = (int)clearWordwith'. self cCode: 'opTable[1+1] = (int)bitAndwith'. self cCode: 'opTable[2+1] = (int)bitAndInvertwith'. self cCode: 'opTable[3+1] = (int)sourceWordwith'. self cCode: 'opTable[4+1] = (int)bitInvertAndwith'. self cCode: 'opTable[5+1] = (int)destinationWordwith'. self cCode: 'opTable[6+1] = (int)bitXorwith'. self cCode: 'opTable[7+1] = (int)bitOrwith'. self cCode: 'opTable[8+1] = (int)bitInvertAndInvertwith'. self cCode: 'opTable[9+1] = (int)bitInvertXorwith'. self cCode: 'opTable[10+1] = (int)bitInvertDestinationwith'. self cCode: 'opTable[11+1] = (int)bitOrInvertwith'. self cCode: 'opTable[12+1] = (int)bitInvertSourcewith'. self cCode: 'opTable[13+1] = (int)bitInvertOrwith'. self cCode: 'opTable[14+1] = (int)bitInvertOrInvertwith'. self cCode: 'opTable[15+1] = (int)destinationWordwith'. self cCode: 'opTable[16+1] = (int)destinationWordwith'. self cCode: 'opTable[17+1] = (int)destinationWordwith'. self cCode: 'opTable[18+1] = (int)addWordwith'. self cCode: 'opTable[19+1] = (int)subWordwith'. self cCode: 'opTable[20+1] = (int)rgbAddwith'. self cCode: 'opTable[21+1] = (int)rgbSubwith'. self cCode: 'opTable[22+1] = (int)OLDrgbDiffwith'. self cCode: 'opTable[23+1] = (int)OLDtallyIntoMapwith'. self cCode: 'opTable[24+1] = (int)alphaBlendwith'. self cCode: 'opTable[25+1] = (int)pixPaintwith'. self cCode: 'opTable[26+1] = (int)pixMaskwith'. self cCode: 'opTable[27+1] = (int)rgbMaxwith'. self cCode: 'opTable[28+1] = (int)rgbMinwith'. self cCode: 'opTable[29+1] = (int)rgbMinInvertwith'. self cCode: 'opTable[30+1] = (int)alphaBlendConstwith'. self cCode: 'opTable[31+1] = (int)alphaPaintConstwith'. self cCode: 'opTable[32+1] = (int)rgbDiffwith'. self cCode: 'opTable[33+1] = (int)tallyIntoMapwith'. self cCode: 'opTable[34+1] = (int)alphaBlendScaledwith'. self cCode: 'opTable[35+1] = (int)alphaBlendScaledwith'. self cCode: 'opTable[36+1] = (int)alphaBlendScaledwith'. self cCode: 'opTable[37+1] = (int)rgbMulwith'. self cCode: 'opTable[38+1] = (int)pixSwapwith'. self cCode: 'opTable[39+1] = (int)hsvwith'.! ! !BitBltSimulation class methodsFor: 'initialization' stamp: 'kfr 6/30/2001 09:40'! initializeRuleTable "BitBltSimulation initializeRuleTable" "**WARNING** You MUST change initBBOpTable if you change this" OpTable _ #( "0" clearWord:with: "1" bitAnd:with: "2" bitAndInvert:with: "3" sourceWord:with: "4" bitInvertAnd:with: "5" destinationWord:with: "6" bitXor:with: "7" bitOr:with: "8" bitInvertAndInvert:with: "9" bitInvertXor:with: "10" bitInvertDestination:with: "11" bitOrInvert:with: "12" bitInvertSource:with: "13" bitInvertOr:with: "14" bitInvertOrInvert:with: "15" destinationWord:with: "16" destinationWord:with: "unused - was old paint" "17" destinationWord:with: "unused - was old mask" "18" addWord:with: "19" subWord:with: "20" rgbAdd:with: "21" rgbSub:with: "22" OLDrgbDiff:with: "23" OLDtallyIntoMap:with: "24" alphaBlend:with: "25" pixPaint:with: "26" pixMask:with: "27" rgbMax:with: "28" rgbMin:with: "29" rgbMinInvert:with: "30" alphaBlendConst:with: "31" alphaPaintConst:with: "32" rgbDiff:with: "33" tallyIntoMap:with: "34" alphaBlendScaled:with: "35" alphaBlendScaled:with: "unused here - only used by FXBlt" "36" alphaBlendScaled:with: "unused here - only used by FXBlt" "37" rgbMul:with: "38" pixSwap:with: "39" hsv:with: ). OpTableSize _ OpTable size + 1. "0-origin indexing" ! !

1 0

Debugger desirable feature
by Ross Boylan 01 Jul '01

01 Jul '01

When in the debugger, pressing send will not get you out of a method. I think it would be good if it did. Is there any reason for the current implementation? While I'm at it, the debugger highlighting is also misleading around conditionals. if foo is false, code like this foo ifTrue: [ do something]. next command. has "do something" highlighted after evaluating the conditional, even though it will not execute that statement. At least, that's my recollection. When I try to evaluate the following method: silly |n| self halt. n_6. n _ Project allInstances size > 30 ifTrue: [n_5]. n _ n+10. ^n. I get a code simulation error at evaluating the +10. Other variants seem to work OK.

1 1

Jump to page: