Squeak-dev December 1999

squeak-dev@lists.squeakfoundation.org

146 participants
451 discussions

RE: InstantMessaging alpha release...
by Raab, Andreas 27 Dec '99

27 Dec '99

Scott and Dean, Excuse my ignorance (I'm living here between two connections - one where I can read email and one where I can do everything else and right now I'm on email so I haven't actually check the instant messaging stuff) but I think that ICQ is using UDP rather than TCP (since there is no point in having zillions of connections open at the same time for a few messages). And AFAIK UDP is not supported on the Mac. Andreas > -----Original Message----- > From: Scott Wallace [mailto:Scott.Wallace@disney.com] > Sent: Thursday, December 23, 1999 5:20 PM > To: Dean Pu Mao > Cc: recipient list not shown > Subject: Re: InstantMessaging alpha release... > > > Dean, > > What a terrific thing to be doing! This will make a qualitative > difference to the experience of using Squeak, particularly for those > of us who spend a high percentage of our waking hours in a Squeak > image. > > I got your code, got a registered ICQ number from Mirabalis, and > already had an AIM identity, so I fired up an image, filed in your > code, and executed the precipitating do-it. > > However, it was not successful. I got the "Messaging" window, but > when I hit the "Connect" button, I quickly fell into a debugger, > "primitive failed", in Socket.primSocket:connectTo:port:. > > Do I perhaps need some custom plug-in, or some custom vm, in order to > use this stuff successfully? Or could it be that the debug-notifier > is the "normal" way of responding to a request to connect to a user > who is presently offline? > > I'm eager to start using this, but for the moment am stuck... > > -- Scott > > > At 5:58 PM -0500 12/23/99, Dean Pu Mao wrote: > >Some of you here wanted my instant messaging stuff, so here > it is! Let > me > >know how you like it. Better yet, send me an instant message how you > like > >it! ICQ and AIM messaging are both supported. ICQ contact list is > >supported, but the AIM buddy list is not yet supported. > You'll need a > >registered ICQ number from Mirabilis and a registered Screen > Name from > >AOL to correctly use my client. > > > >Get the changesets here: http://instantmessaging.swiki.net > > > >Dino

5 4

Re: [BUG?] Text misaligned on 16 bit color depth
by Bob Arning 26 Dec '99

26 Dec '99

On Sun, 26 Dec 1999 14:43:11 +0100 Stefan Matthias Aust <sma(a)3plus4.de> wrote: >Okay, it seems that I don't get tranparency with 8 bit screen depth, but I >can live with that limitation. But WHY is the text indented by a few >pixels to the right on 16 bit screen depth. On 32 bits, it's correct again. Stefan, That's a bug in BitBlt. Here is my report from earlier this year: Subject: [BUG] drawing translucent text Date: Jun 24 1999 From: arning(a)charm.net Changing a TextMorph from opaque to translucent changes the position where the text is drawn. This effect can be seen by creating a new TextMorph. Open an inspector on this morph and in the inspector alternately evaluate: self color: (Color black alpha: 0.5) and self color: Color black Wonder of wonders! The text moves when the color changes! After much poking around, I have discovered that CharacterScanner>>scanCharactersFrom:to:in:rightX:stopConditions:displaying: fails the primitive (103) whenever translucent text (combination rule 31) is drawn. What causes the shift in position is that destX is incremented by the width of the first character even though nothing has been drawn. The Smalltalk code after the primitive then begins drawing the entire string from the wrong position. My guess is that the error lies in BitBltSimulation>>copyBits in the fragment below: (combinationRule = 30) | (combinationRule = 31) ifTrue: ["Check and fetch source alpha parameter for alpha blend" interpreterProxy argCount = 1 ifTrue: [sourceAlpha _ interpreterProxy stackIntegerValue: 0. (interpreterProxy failed not and: [(sourceAlpha >= 0) & (sourceAlpha <= 255)]) ifTrue: [interpreterProxy pop: 1] ifFalse: [^ interpreterProxy primitiveFail]] ifFalse: [^ interpreterProxy primitiveFail]]. which is sent from BitBltSimulation>>scanCharacters in the fragment below: scanDisplayFlag ifTrue: [self copyBits]. destX _ nextDestX. interpreterProxy storeInteger: BBDestXIndex ofObject: bitBltOop withValue: destX. No test is made here for the #copyBits failing, so destX gets updated even though there was nothing drawn. Since the above code is in the VM, it's a bit hard to debug. While a patch could be applied at the Smalltalk level (wrapping the primitive 103 in another method which would preserve destX in the event of failure), it would be better if the VM could be fixed. I also do not know if this problem appears under other circumstances (like non-text drawing). Cheers, Bob

2 1

[ENH;BUG;FIX]Compilation of Recent Changes to Slang Code Generators
by Andrew C. Greenberg 26 Dec '99

26 Dec '99

There has been a blizzard of changesets published for the Slang compiler and plugin code generation systems, resulting in large part from a flurry of activity (and hence great bug reports) on various fronts. Thanks to everyone for their excellent help and suggestions, particularly Stephan Rudlof. For the convenience of those who have not been keeping current, I have compiled recently posted changes into a single changeset, which incorporates changes made to all of the following: TestInterpreterPlugin3.1.cs TestInterpreterPlugin4.1.cs TestInterpreterPlugin5.1.cs CCodeGenerator.1.cs CCodeGenerator2.4.cs TestCodeGenerator.1.cs Happy holidays everyone!

2 1

[BUG?] Text misaligned on 16 bit color depth
by Stefan Matthias Aust 26 Dec '99

26 Dec '99

--=====================_946212191==_ Content-Type: text/plain; charset="us-ascii" Merry Christmas, Perhaps there's some other person also working with Squeak these days who can answer this question. I had a hard time to understand how to render paragraphs of text. I finally managed to create my RgMessageMorph (see attachment) as I wanted but there's one problem left. Depending on screen depth, my text is aligned differently. The message morph shows text which fades away on a half-transparent background. (I wanted something that's cooler than a simple transcript ;-) Okay, it seems that I don't get tranparency with 8 bit screen depth, but I can live with that limitation. But WHY is the text indented by a few pixels to the right on 16 bit screen depth. On 32 bits, it's correct again. The class comment has some example code to try out. I observed this on 2.6. bye --=====================_946212191==_ Content-Type: text/plain; name="RgMessageMorph.st" ; x-mac-type="65417070" ; x-mac-creator="43534F6D" Content-Disposition: attachment; filename="RgMessageMorph.st" Content-Transfer-Encoding: imap_stub 0,2660,2,5542,0, --=====================_946212191==_ Content-Type: text/plain; charset="us-ascii" -- Stefan Matthias Aust // Bevor wir fallen, fallen wir lieber auf. --=====================_946212191==_--

1 0

[ENH;BUG;FIX]Compilation of Recent Changes to Slang Code Generators
by Andrew C. Greenberg 26 Dec '99

26 Dec '99

--============_-1265935502==_============ Content-Type: text/plain; charset="us-ascii" ; format="flowed" There has been a blizzard of changesets published for the Slang compiler and plugin code generation systems, resulting in large part from a flurry of activity (and hence great bug reports) on various fronts. Thanks to everyone for their excellent help and suggestions, particularly Stephan Rudlof. For the convenience of those who have not been keeping current, I have compiled recently posted changes into a single changeset, which incorporates changes made to all of the following: TestInterpreterPlugin3.1.cs TestInterpreterPlugin4.1.cs TestInterpreterPlugin5.1.cs CCodeGenerator.1.cs CCodeGenerator2.4.cs TestCodeGenerator.1.cs Happy holidays everyone! --============_-1265935502==_============ Content-Type: text/plain; name="RecentSlangChanges.1.cs" ; x-mac-type="65417070" ; x-mac-creator="43534F6D" Content-Disposition: attachment; filename="RecentSlangChanges.1.cs" Content-Transfer-Encoding: imap_stub 0,2658,2.2,53754,0, --============_-1265935502==_============--

1 0

Newbie Alert: How do I save a class I have made to a disk file
by Mayuresh A Kathe 26 Dec '99

26 Dec '99

Hi, I have created a small clock, would like to know how to save the class to disk file, this file can then be used when I want at any other place. Also, once I have saved the file, how do I put it back into Squeak? Awaiting your reply. Bye. -- Mayuresh A Kathe Think Different. Think Better. mayuresh(a)vsnl.com

1 0

Re: Another Question
by Blacktooth 26 Dec '99

26 Dec '99

> >On the Class side of this attached object in the Initialize-Release > >Category is a method called 'buildSkillReferences'. > > To be honest, I don't understand that method. It's much too big. It had changed it many times, sometimes getting very close to the result I needed, but I never got there exactly. I got frustrated, and went to brute force mode. Things got really ugly from there. To make a long story short, using 'removeFirst' and 'addLast' and doing away entirely with the 'do:' solved my problem. The offending piece of code now looks like this: refQue := OrderedCollection new. storage := OrderedCollection new. tempRefs keysDo: [ :refKey | tempDeps := OrderedCollection newFrom: (tempRefs at: refKey). tempDeps do: [ :i | refQue add: i. ]. [ refQue isEmpty ] whileFalse: [ temp := (refQue first). storage addLast: temp. (tempRefs includesKey: temp) ifTrue: [ (tempRefs at: temp) do: [ :i | refQue addLast: i ].]. refQue removeFirst. ]. self skillReferences at: refKey put: storage asArray. storage := OrderedCollection new. refQue := OrderedCollection new. ]. A simple queue was all I needed. BT

2 2

RE: CCodeGenerator: suggestion
by Raab, Andreas 25 Dec '99

25 Dec '99

> My suggestions came from the vision to be able to > automatically compile arbitrary ST classes to speed > up them if they are finished... Do you > know the work of ST/X from Claus Gittinger? We've invited Claus to OOPSLA this year and he has given a series of *very* impressive demonstrations. It's a lot of fun to see people like Claus, Dan, Elliot, and Ian in one and the same room (but be careful, their discussion topics tend to give an uninitiated VM builder some decent headaches ;-) > > It also gets confusing when "the same" construct works sometimes but > > not others. > > This is absolutely true. > > Therefore I think most important - before further enhancements - are > error messages from the compiler if it something cannot compile! You're absolutely right - it's extremely annoying if the code generator can't translate certain expressions correctly and one has to find out in the C code. Any volunteers?! > A short summary: > > There are two main developing lines: > > - To be able to compile as much existing - _not_ C'ish like - > ST code as possible with as less changes as possible, and > - to give the programmer maximal flexibility in writing C'ish ST with > supporting as much C constructs as possible to gain maximal speed. I > know that it is already possible with > self cCode: 'some C code' inSmalltalk: [some ST code]. > , but this style isn't good for debugging purposes. It's also bad for a number of other reasons - just extremely convenient in some cases... I'd love to see most of this going away if there is a clear way of doing it. > P.S.: Here it is 20:41:42 now and I want to go out, so this is my last > mail for a while... Have fun! Andreas

1 0

[Q] Is it difficult to call plugins with receiver as argument?
by Stephan Rudlof 25 Dec '99

25 Dec '99

Dear Andrew, dear Squeakers! I'm just before merging my plugin routines (LargeInteger arithmetics) with the current system. For testing purposes it would be very nice to have a method in TestInterpreterPlugin like perform: selector receiver: receiver withArguments: argArray in addition to the existing one perform: selector withArguments: argArray .. Do you think this is an easy or a difficult task? Motivation: I want to replace e.g. Integer #digitAdd: by my own one in LargePositiveInteger with the following properties: - a plugin primitive should be called from there, - if it fails it should call 'super digitAdd:'. It would be nice if the receiver of the primitive could be set to LargeInteger, but then it is fussy to test the plugin in Smalltalk simulation mode, because it cannot called from elsewhere (wrong receiver!). Of course it is possible to workaround by calling the primitive with 'self' as additional argument what is my current approach. But I'd prefer the other way... There is a ContextPart method doPrimitive: primitiveIndex receiver: receiver args: arguments but this seems to be quiet far away from TestInterpreterPlugin. Any comments? Greetings, Stephan P.S.: Andrew: _really_ the last one for today... -- Stephan Rudlof (sr(a)evolgo.de) "Genius doesn't work on an assembly line basis. You can't simply say, 'Today I will be brilliant.'" -- Kirk, "The Ultimate Computer", stardate 4731.3

1 0

RE: Organizing VM sources, CCodeGenerator: suggestions and questions
by Raab, Andreas 25 Dec '99

25 Dec '99

Stephan, > isBytes: oop > ^ (interpreterProxy isIntegerObject: oop) not and: > [interpreterProxy > isBytes: oop] > > has led to > > int isBytes(int oop) { > return (!((oop & 1))) && (interpreterProxy->isBytes(oop)); > } > > . Compiling works only without inlining, otherwise it runs into an > infinite recursion. That's a bug in the code generator, but trivial to fix. Just change your method into isBytes: oop self inline: false. ^(interpreterProxy isIntegerObject: oop) not and: [interpreterProxy isBytes: oop] The code generator doesn't take the receiver into account which is somewhat ugly and should be fixed but I don't know a simple way of fixing this. > The plugin runs in simulation, but in calling this function > in C crashes the VM. > > The problem here is that this function interferes with the same > signatured function from 'sqVirtualMachine.c'. sqVirtualMachine.c must *never* be compiled with the plugin itself. It merely defines the interpreter interface and is compiled with the VM not with the plugin! All you need in the plugin is the definition of the interface (e.g., from sqVirtualMachine.h) since you don't know beforehand with what version of the interpreter you're running. > This leads me to the following suggestions: > > - A compiler warning would be nice, if someone overrides a predefined > function. See above - this shouldn't be necessary if in a plugin. > - Wouldn't it be good to make all user plugin functions which are not > exported 'static' as default? Yes and no ... sometimes it's very convenient if one has outside access to these functions and so far I haven't found a case where this causes any problems. > - Wouldn't it be better to make 'int isBytes(int oop);' and similar > functions in 'sqVirtualMachine.c' static? For plugins they are called > indirectly by 'interpreterProxy->isBytes(oop)' anyway. sqVirtualMachine.c does *not* define these functions. Again, sqVirtualMachine.c is just for the VM - not for the plugin. > Couldn't there arise any compatibility problems between different > plugins in the current situation? Not that I know of. > Greetings, Merry Christmas, Und ebenfalls schoene Weihnachten und ein frohes neues J2K (oder wie heisst das in Deutsch?!) Andreas

2 1

← Newer
1
...
4
5
6
7
8
9
10
...
46
Older →

Jump to page: