From alanrogersmith at gmail.com Thu Aug 2 06:15:17 2007 From: alanrogersmith at gmail.com (Alan Smith) Date: Thu Aug 2 06:15:22 2007 Subject: Getting Real for Desktop Software Message-ID: Hello everyone, Yesterday I finished reading 'Getting Real,' the book by 37Signals. For those of you who haven't read it, I highly recommend it, to buy a copy or read the free online version go to http://getreal.37signals.com/ In 'Getting Real' there is major emphasis on keeping a web app simple and clean. One example was 37Signals did not add text coloring to their multi-person chat app (Campfire) because it wasn't necessary. That got me thinking, that's fine for a web app where there is a monthly fee instead of a one-time registration fee and so you don't have to add features to get people to buy it but, what about desktop apps? If you don't add little things like that how are you going to keep people interested in it? How do all of you write apps? I want to keep my apps sleek and thin but I don't want them to become stagnant. Is there a book like 'Getting Real' for desktop apps? If there is that would be fantastic, I'd love to read it. Cheers. Alan -- My blog: cocoalatte.info // Things I've said ------------------------- "Maturity resides in the mind." "Silence is the Universe's greatest gift." "When the World realizes that personal beliefs are not something to argue or fight over, it shall evolve." From andreas at harmless.de Thu Aug 2 07:30:18 2007 From: andreas at harmless.de (Andreas Mayer) Date: Thu Aug 2 07:55:24 2007 Subject: [ANN] AMSerialPort - new version Message-ID: <7CCA1C4F-DF92-4B7A-BCAD-43E6F91EF373@harmless.de> AMSerialPort is a collection of classes to access serial ports. This version features an API that better matches Cocoa design principles, adds a wider range of read and write methods and is able to detect the addition and removal of ports. For more information see: http://www.harmless.de/cocoa.php Regards, Andreas From smokinmonkey at yahoo.com Fri Aug 3 19:39:20 2007 From: smokinmonkey at yahoo.com (Michael Dupuis) Date: Fri Aug 3 19:39:23 2007 Subject: NSPopUpButtonCell within NSTextFieldCell: disappearing text Message-ID: <273229.4787.qm@web31910.mail.mud.yahoo.com> Greetings, I'm trying to implement a "complex" NSTextFieldCell, based originally on ImageAndTextCell from the DraggableOutlineView demo app, and I've got most of what I need working, but I'm having one problem I can't seem to resolve: When I have an NSPopupButtonCell within my NSTextFieldCell, popping up the menu for the button causes the text field text to vanish until the menu tracking is done. I could see the text of the NSTextFieldCell maybe not getting redrawn potentially, but why it gets blanked out completely while the menu is displayed, and then comes back perfectly when the menu is dismissed I can't understand. A little code (some details omitted for brevity: in drawWithFrame: inView: I draw my image, just like the sample code, then call super: [super drawWithFrame:cellFrame inView:controlView]; then draw my popup button: [popUpButtonCellOne drawWithFrame: NSMakeRect(POPUP_X_OFFSET, cellFrame.origin.y + POPUP_Y_OFFSET, 150.0, 24.0) inView:controlView]; popUpButtonCellOne is retained by a static right now and is created like this: if (popUpButtonCellOne == nil) { popUpButtonCellOne = [[NSPopUpButtonCell alloc] init]; [popUpButtonCellOne addItemWithTitle:@"One"]; [popUpButtonCellOne addItemWithTitle:@"Two"]; [popUpButtonCellOne addItemWithTitle:@"Three"]; [popUpButtonCellOne setBordered:NO]; } Ok, so with just this, the popup button never gets a mouse down when you click on the cell, so to remedy that, I tried a bunch of things and settled on implementing trackMouse:::: (I know, I can use something other than NSBezierPath for this) NSPoint point = [controlView convertPoint:[theEvent locationInWindow] fromView:nil]; NSBezierPath *path = [NSBezierPath bezierPathWithRect: NSMakeRect(POPUP_X_OFFSET, cellFrame.origin.y + POPUP_Y_OFFSET, 150.0, 24.0)]; if ([path containsPoint:point]) { [popUpButtonCellOne trackMouse:theEvent inRect: NSMakeRect(POPUP_X_OFFSET, cellFrame.origin.y + POPUP_Y_OFFSET, 150.0, 24.0) ofView:controlView untilMouseUp:flag]; } returning YES when I'm done. This all works, except for the problem I mentioned. Is there something I'm overlooking, or misunderstanding? It is possible that I should be passing the mouse down to the popup in another method other than trackMouse? TIA, Michael ____________________________________________________________________________________ Be a better Globetrotter. Get better travel answers from someone who knows. Yahoo! Answers - Check it out. http://answers.yahoo.com/dir/?link=list&sid=396545469 From alanrogersmith at gmail.com Fri Aug 3 20:11:26 2007 From: alanrogersmith at gmail.com (Alan Smith) Date: Fri Aug 3 20:11:29 2007 Subject: NSPopUpButtonCell within NSTextFieldCell: disappearing text In-Reply-To: <273229.4787.qm@web31910.mail.mud.yahoo.com> References: <273229.4787.qm@web31910.mail.mud.yahoo.com> Message-ID: Michael, If you don't need the subclass to be a cell then I suggest making things easier for yourself and subclassing NSTextField. That way you can add the pop-up button as a sub-view. If you would explain why you are attaching a pop-up button to a text field it would be greatly appreciated. It sounds like NSComboBox might be the kind of thing you are looking for. If it helps any to know how NSComboBox works, the blue button with the triangle on it is in fact a NSButton, or maybe a NSButtonCell, I forget which. When the button is pressed a transparent window with a table in it is popped up near the button. The editing part is a NSTextField, again, it might be a cell, NSTextFieldCell. Perhaps you'd find it easier to take this route and start from scratch. >From here it looks like if you are hooking up a NSPopUpButton to a NSTextField and NSComboBox isn't the sort of thing you are trying to accomplish you need to redesign your UI. By the way, it has been a long time since I've seen NSComboBox used in a modern application. If I have, they were disguised so I didn't notice. Let me know why you need to do this, I'm curious. Cheers, Alan -- My blog: cocoalatte.info // Things I've said ------------------------- "Maturity resides in the mind." "Silence is the Universe's greatest gift." "When the World realizes that personal beliefs are not something to argue or fight over, it shall evolve." From smokinmonkey at yahoo.com Fri Aug 3 20:27:08 2007 From: smokinmonkey at yahoo.com (Michael Dupuis) Date: Fri Aug 3 20:27:11 2007 Subject: NSPopUpButtonCell within NSTextFieldCell: disappearing text In-Reply-To: Message-ID: <421885.70295.qm@web31911.mail.mud.yahoo.com> --- Alan Smith wrote: > If you would explain why you are attaching a pop-up > button to a text > field it would be greatly appreciated. Ah yes, that might help, I didn't mention what the cell was for. It's an NSTableView cell that has several labels and at least one popup menu in it. It's definitely not an NSComboBox, it's a table cell that displays a summary of information, and allows some simple editing. Think Safari's Downloads window, or something similar, where the cell is made up of an image, two labels, and a couple of buttons. ____________________________________________________________________________________ Shape Yahoo! in your own image. Join our Network Research Panel today! http://surveylink.yahoo.com/gmrs/yahoo_panel_invite.asp?a=7 From dunham at mac.com Fri Aug 3 23:17:17 2007 From: dunham at mac.com (David Dunham) Date: Fri Aug 3 23:17:28 2007 Subject: Updating Align Left/Center/etc. for NSTextView Message-ID: <956FDF62-8A30-4A26-813D-D9E250E3260D@mac.com> I'm doing something much like TextEdit, where I want to have a menu control text alignment. I want to place a checkmark (or a hyphen) next to the appropriate menu item(s) depending on whether the selected text is left aligned, centered, etc. TextEdit doesn't do this. So I don't see an obvious way to do this. To make things worse, I also want to update a toolbar icon (much like the one in NSRulerView) to also show alignment. Any suggestions? The menu part is easy (I think), but I'm not sure how to efficiently track the selection changing. (Or am I worrying too much about observing every keystroke?) David Dunham www.pensee.com/dunham/ Imagination is more important than knowledge. -- Albert Einstein From alanrogersmith at gmail.com Sat Aug 4 05:33:58 2007 From: alanrogersmith at gmail.com (Alan Smith) Date: Sat Aug 4 05:34:02 2007 Subject: NSPopUpButtonCell within NSTextFieldCell: disappearing text In-Reply-To: <421885.70295.qm@web31911.mail.mud.yahoo.com> References: <421885.70295.qm@web31911.mail.mud.yahoo.com> Message-ID: Michael, I suggest not subclassing NSTextFieldCell but NSCell. That way you don't have to worry about overriding some of NSTextFieldCell's behavior. If you override NSCell just create a NSTextFieldCell and NSPopUpButtonCell as instance variables and draw them in drawWithFrame:: Good luck! Peace, Alan -- My blog: cocoalatte.info // Things I've said ------------------------- "Maturity resides in the mind." "Silence is the Universe's greatest gift." "When the World realizes that personal beliefs are not something to argue or fight over, it shall evolve." From sstevenson at mac.com Sat Aug 4 15:16:49 2007 From: sstevenson at mac.com (Scott Stevenson) Date: Sat Aug 4 15:17:12 2007 Subject: Updating Align Left/Center/etc. for NSTextView In-Reply-To: <956FDF62-8A30-4A26-813D-D9E250E3260D@mac.com> References: <956FDF62-8A30-4A26-813D-D9E250E3260D@mac.com> Message-ID: <28787C3A-CAD6-4204-9803-935081DD8124@mac.com> On Aug 3, 2007, at 11:17 PM, David Dunham wrote: > Any suggestions? The menu part is easy (I think), but I'm not sure > how to efficiently track the selection changing. (Or am I worrying > too much about observing every keystroke?) Does NSTextView's -textViewDidChangeSelection: do what you want? You'll probably also what to look at - textViewDidChangeTypingAttributes:. - Scott From dunham at mac.com Sat Aug 4 21:53:21 2007 From: dunham at mac.com (David Dunham) Date: Sat Aug 4 21:53:30 2007 Subject: Updating Align Left/Center/etc. for NSTextView In-Reply-To: <956FDF62-8A30-4A26-813D-D9E250E3260D@mac.com> References: <956FDF62-8A30-4A26-813D-D9E250E3260D@mac.com> Message-ID: On 3 Aug 2007, at 23:17, David Dunham wrote: > The menu part is easy (I think) Hmm, harder than I was hoping. I not only have to determine if I should use typingAttributes, but most of the time the selection's paragraph is of NSNaturalTextAlignment. So how do I determine the "default alignment for script?" I see this question was asked in 2004, with no answer. Yet somehow an NSTextView's ruler manages to show it correctly. David Dunham A Sharp, LLC Voice/Fax: 206 783 7404 http://a-sharp.com "People seem to misinterpret complexity as sophistication" -- Niklaus Wirth From dunham at mac.com Sun Aug 5 19:53:45 2007 From: dunham at mac.com (David Dunham) Date: Sun Aug 5 19:53:51 2007 Subject: Updating Align Left/Center/etc. for NSTextView In-Reply-To: <956FDF62-8A30-4A26-813D-D9E250E3260D@mac.com> References: <956FDF62-8A30-4A26-813D-D9E250E3260D@mac.com> Message-ID: > most of the time the selection's paragraph is of > NSNaturalTextAlignment. So how do I determine the "default > alignment for script?" This part at least I figured out -- I'd imported some of the text, and used [NSParagraphStyle defaultParagraphStyle]. This uses NSNaturalTextAlignment. In my case it's probably better to be explicit. David Dunham A Sharp, LLC Voice/Fax: 206 783 7404 http://a-sharp.com "People seem to misinterpret complexity as sophistication" -- Niklaus Wirth From smokinmonkey at yahoo.com Mon Aug 6 04:19:54 2007 From: smokinmonkey at yahoo.com (Michael Dupuis) Date: Mon Aug 6 04:19:57 2007 Subject: NSPopUpButtonCell within NSTextFieldCell: disappearing text In-Reply-To: Message-ID: <39302.92336.qm@web31911.mail.mud.yahoo.com> --- Alan Smith wrote: > I suggest not subclassing NSTextFieldCell but > NSCell. That way you > don't have to worry about overriding some of > NSTextFieldCell's > behavior. If you override NSCell just create a > NSTextFieldCell and > NSPopUpButtonCell as instance variables and draw > them in > drawWithFrame:: Thanks Alan, that's pretty much what I wound up doing. I was hoping to use NSTextFieldCell to get a bunch of stuff "for free" but ultimately, it didn't work. I'd still be interested in the "why?" of it not working though. Michael ____________________________________________________________________________________ Got a little couch potato? Check out fun summer activities for kids. http://search.yahoo.com/search?fr=oni_on_mail&p=summer+activities+for+kids&cs=bz From cmhofman at gmail.com Mon Aug 6 05:38:57 2007 From: cmhofman at gmail.com (Christiaan Hofman) Date: Mon Aug 6 05:39:03 2007 Subject: NSPopUpButtonCell within NSTextFieldCell: disappearing text In-Reply-To: <39302.92336.qm@web31911.mail.mud.yahoo.com> References: <39302.92336.qm@web31911.mail.mud.yahoo.com> Message-ID: My guess is actually that you forgot to override selectWithFrame:... and editWithFrame:... This does everything you asked for: @implementation MyPopUpTextFieldCell - (id)initTextCell:(NSString *)aString { if (self = [super initTextCell:aString]) { popUpCell = [[NSPopUpButtonCell alloc] initTextCell:@"" pullsDown:NO]; [popUpCell setBordered:NO]; [popUpCell addItemsWithTitles:[NSArray arrayWithObjects:@"One", @"Two", @"Three", nil]]; [popUpCell selectItemAtIndex:0]; } return self; } - (id)initWithCoder:(NSCoder *)decoder { if (self = [super initWithCoder:decoder]) { if (popUpCell == nil) { popUpCell = [[NSPopUpButtonCell alloc] initTextCell:@"" pullsDown:NO]; [popUpCell setBordered:NO]; [popUpCell addItemsWithTitles:[NSArray arrayWithObjects:@"One", @"Two", @"Three", nil]]; [popUpCell selectItemAtIndex:0]; } } return self; } - (id)copyWithZone:(NSZone *)aZone { MyPopUpTextFieldCell *copy = [super copyWithZone:aZone]; copy->popUpCell = [popUpCell copyWithZone:aZone]; return copy; } - (void)dealloc { [popUpCell release]; [super dealloc]; } - (void)drawWithFrame:(NSRect)cellFrame inView:(NSView *)controlView { NSRect popUpRect, textRect; NSDivideRect(cellFrame, &popUpRect, &textRect, POPUP_WIDTH, NSMinXEdge); [super drawWithFrame:textRect inView:controlView]; [popUpCell drawWithFrame:popUpRect inView:controlView]; } - (void)selectWithFrame:(NSRect)aRect inView:(NSView *)controlView editor:(NSText *)textObj delegate:(id)anObject start:(int)selStart length:(int)selLength { NSRect popUpRect, textRect; NSDivideRect(aRect, &popUpRect, &textRect, POPUP_WIDTH, NSMinXEdge); [super selectWithFrame:textRect inView:controlView editor:textObj delegate:anObject start:selStart length:selLength]; } - (void)editWithFrame:(NSRect)aRect inView:(NSView *)controlView editor:(NSText *)textObj delegate:(id)anObject event:(NSEvent *) theEvent { NSRect popUpRect, textRect; NSDivideRect(aRect, &popUpRect, &textRect, POPUP_WIDTH, NSMinXEdge); [super editWithFrame:textRect inView:controlView editor:textObj delegate:anObject event:theEvent]; } - (BOOL)trackMouse:(NSEvent *)theEvent inRect:(NSRect)cellFrame ofView:(NSView *)controlView untilMouseUp:(BOOL)untilMouseUp { NSPoint point = [controlView convertPoint:[theEvent locationInWindow] fromView:nil]; NSRect popUpRect, textRect; NSDivideRect(cellFrame, &popUpRect, &textRect, POPUP_WIDTH, NSMinXEdge); if (NSPointInRect(point, popUpRect)) return [popUpCell trackMouse:theEvent inRect:popUpRect ofView:controlView untilMouseUp:untilMouseUp]; return NO; } @end Christiaan On 6 Aug 2007, at 1:19 PM, Michael Dupuis wrote: > > --- Alan Smith wrote: > >> I suggest not subclassing NSTextFieldCell but >> NSCell. That way you >> don't have to worry about overriding some of >> NSTextFieldCell's >> behavior. If you override NSCell just create a >> NSTextFieldCell and >> NSPopUpButtonCell as instance variables and draw >> them in >> drawWithFrame:: > > Thanks Alan, that's pretty much what I wound up doing. > I was hoping to use NSTextFieldCell to get a bunch of > stuff "for free" but ultimately, it didn't work. I'd > still be interested in the "why?" of it not working > though. > > Michael > > > > ______________________________________________________________________ > ______________ > Got a little couch potato? > Check out fun summer activities for kids. > http://search.yahoo.com/search?fr=oni_on_mail&p=summer+activities > +for+kids&cs=bz > _______________________________________________ > MacOSX-dev mailing list > MacOSX-dev@omnigroup.com > http://www.omnigroup.com/mailman/listinfo/macosx-dev From cmhofman at gmail.com Tue Aug 7 03:33:52 2007 From: cmhofman at gmail.com (Christiaan Hofman) Date: Tue Aug 7 03:33:54 2007 Subject: PDFKit changes shortcuts, bug workaround? Message-ID: <3CB0F2B0-787F-471E-B0A5-4376D3EF310F@gmail.com> PDFKit on Tiger has a very weird bug where it changes the keyboard shortcut Opt-Cmd-P to Cmd-, when it opens a password protected PDF (really). More in detail: I open a password protected PDF (with a password that is not the empty string) in a PDFView, type in the password, and after that, hitting Opt-Cmd-P activates the menu item with shortcut Cmd-, (Show Preferences). Obviously, this is really annoying, in particular as both shortcuts are used in my PDFKit based app. So my question here is if anyone on this list would have an idea what it is that PDFKit changes behind my back. And most of all, how I could fix this. Any help would greatly be appreciated. Christiaan From dunham at mac.com Mon Aug 13 23:20:44 2007 From: dunham at mac.com (David Dunham) Date: Mon Aug 13 23:20:49 2007 Subject: Text alignment Message-ID: <51605BA6-E60F-469B-891C-9F3BC252A372@mac.com> I've been trying to update an alignment menu for an NSTextView, to show the current alignment (left, center, justified, right). I thought this was working, but an NSTextView with multiple paragraphs always shows as NSJustifiedTextAlignment. In my validateMenuItem: I have } else if (cmd == @selector(alignJustified:)) { sel = [self selectedRange]; if (sel.length == 0) p = [[self typingAttributes] objectForKey: NSParagraphStyleAttributeName]; else p = [[self textStorage] attribute: NSParagraphStyleAttributeName atIndex: sel.location effectiveRange: nil]; [anItem setState: [p alignment] == NSJustifiedTextAlignment]; } and this is indeed set in the dictionary (p). Elsewhere, I'm using -[NSText alignment] to get the alignment. It too is returning NSJustifiedTextAlignment when an NSTextView has multiple alignments (in different paragraphs). There doesn't have to be any justified text at all for this to happen. I'm guessing this is probably a bug. Has anyone else run into it? Is there a workaround for Tiger? (I haven't tried under Leopard yet.) David Dunham www.pensee.com/dunham/ Imagination is more important than knowledge. -- Albert Einstein From mailist at ericgorr.net Tue Aug 14 10:36:33 2007 From: mailist at ericgorr.net (Eric Gorr) Date: Tue Aug 14 10:36:44 2007 Subject: sysctl hw.model Message-ID: I was just wondering if anyone knew where a complete list of strings which can be returned by 'sysctl hw.model' is located? If you're unfamiliar with sysctl, it does have a man page. Thank you. From appleslack at gmail.com Tue Aug 14 12:50:29 2007 From: appleslack at gmail.com (Jason Slack) Date: Tue Aug 14 12:50:33 2007 Subject: [OT] Any Developers have a second... Message-ID: Hello All, If any seasoned software developer has a few minutes, I would like to get a few opinions. I am by far not new to software development, it is the only real skill I have and the only career I have ever had. I have a new task in front of me that I have never had to deal with. Recently I have used Qt, by Trolltech, CPLAT and wxWidgets as cross platform GUI toolkits. A venture capitalist friend of mine approached me last week about developing a new cross platform GUI toolkit exercising some HIG and other principals that he feels are important. I have never developed a cross-platform GUI toolkit and am sort of stumped a bit on the implementation details. I was hoping someone could provide thoughts on accomplishing this task. Thanks Jason From mah at jump-ing.de Tue Aug 14 13:34:50 2007 From: mah at jump-ing.de (Markus Hitter) Date: Tue Aug 14 13:35:05 2007 Subject: [OT] Any Developers have a second... In-Reply-To: References: Message-ID: <3F6C0EA5-5ADC-4918-8850-EA47E695A7DF@jump-ing.de> Am 14.08.2007 um 21:50 schrieb Jason Slack: > I have never developed a cross-platform GUI toolkit and am sort of > stumped a > bit on the implementation details. If it's just about the new HIG ideas, I'd try to implement these on top of or as a fork of another cross platform toolkit. If this doesn't fit, you can try to find another common nominator, like X11 or Cairo. If this still doesn't fit, you have to implement all the API on each of the platforms. Watching the GNUstep community, I'd expect several years of work. Markus - - - - - - - - - - - - - - - - - - - Dipl. Ing. Markus Hitter http://www.jump-ing.de/ From llothar at web.de Tue Aug 14 14:17:00 2007 From: llothar at web.de (Lothar Scholz) Date: Tue Aug 14 14:17:50 2007 Subject: [OT] Any Developers have a second... In-Reply-To: References: Message-ID: <33105825.20070815041700@web.de> Hello Jason, Wednesday, August 15, 2007, 2:50:29 AM, you wrote: JS> A venture capitalist friend of mine approached me last week about developing JS> a new cross platform GUI toolkit exercising some HIG and other principals JS> that he feels are important. Something that i wanted to do for a long time :-) JS> I have never developed a cross-platform GUI toolkit and am sort of stumped a JS> bit on the implementation details. Well there is one point where all the existing toolkits suffer at the moment: No toolkit was designed with all three platforms in mind. So they don't work too well with Apple Systems. So look carefully at Windows and the X11 toolkits. Look carefully at TCL/TK. The event handling is still unreached, the Text and Canvas widgets are a good start but need to be hugely improved. They were state of the art for such a long time (i only miss the model<->view separation). Also look for layout managers. IMHO Cocoa sucks for it explicit positioning - well maybe you need this for a design computer but it makes everything much much harder (I18n oder customized font sizes). If you implement the widgets, take the way in the middle because both other ways have shown that they are very far from perfect. Use simple widgets from the native widget set, but write more complex widgets in a cross plattform way. This will handle the most calls. Make sure you can integrate 3rd Party native controls like the webbrowser (for example IExplorer or GTK/KDE widgets) easily. This is what IMHO sucks most at the moment. And try to write a Custom Control Widget Library that is cross toolkit useable so that some 3rd Party widget vendors can use it to write sellable YourToolkit/MFC/QT/Delphi/Cocoa/RealBasic Widgets. With this you will a very strategy point. And come back with an offer if you are looking for an oversea employee :-) Love to burn venture capital. -- Best regards, Lothar mailto:llothar@web.de From dunham at mac.com Tue Aug 14 14:27:17 2007 From: dunham at mac.com (David Dunham) Date: Tue Aug 14 14:27:27 2007 Subject: [OT] Any Developers have a second... In-Reply-To: References: Message-ID: <840F6E9A-CD48-4372-B4B4-0AE8BCBA0898@mac.com> On 14 Aug 2007, at 12:50, Jason Slack wrote: > A venture capitalist friend of mine approached me last week about > developing > a new cross platform GUI toolkit exercising some HIG and other > principals > that he feels are important. This is a little vague. > I have never developed a cross-platform GUI toolkit and am sort of > stumped a > bit on the implementation details. Well, I've worked on a cross-platform game engine, though I don't know if that's quite the same thing. It doesn't use platform widgets at all. On the other hand, a development system like REALbasic (or I believe Qt) does end up using platform widgets. You need to spill the beans as to which approach you're talking. GameHouse Studios +1 206 926 5722 www.gamehouse.com "They said it couldn't be done but sometimes it doesn't work out that way." -- Casey Stengel From macosxdevlist at personal.fishh2o.com Tue Aug 14 13:49:51 2007 From: macosxdevlist at personal.fishh2o.com (SD) Date: Tue Aug 14 14:52:01 2007 Subject: [OT] Any Developers have a second... In-Reply-To: References: Message-ID: Given that Markus is probably right about this being many years of work, make sure your business reasons are VERY sound. Also, you should think about what would happen if interfaces changed and we started using speech or touch. Just some random thoughts. SD ______________________________________________________________________ Previous message from Jason Slack on 8/14/07 at 12:50 PM -0700 ********************************************************************** >Hello All, > >If any seasoned software developer has a few minutes, I would like to get a >few opinions. I am by far not new to software development, it is the only >real skill I have and the only career I have ever had. > >I have a new task in front of me that I have never had to deal with. >Recently I have used Qt, by Trolltech, CPLAT and wxWidgets as cross platform >GUI toolkits. > >A venture capitalist friend of mine approached me last week about developing >a new cross platform GUI toolkit exercising some HIG and other principals >that he feels are important. > >I have never developed a cross-platform GUI toolkit and am sort of stumped a >bit on the implementation details. > >I was hoping someone could provide thoughts on accomplishing this task. > >Thanks > >Jason >_______________________________________________ >MacOSX-dev mailing list >MacOSX-dev@omnigroup.com >http://www.omnigroup.com/mailman/listinfo/macosx-dev -- ========================================== SD WARNING: Programming may be habit forming. From jfieber at slis.indiana.edu Tue Aug 14 18:58:37 2007 From: jfieber at slis.indiana.edu (John Fieber) Date: Tue Aug 14 19:31:52 2007 Subject: [OT] Any Developers have a second... In-Reply-To: References: Message-ID: <4687F227-8B41-45F9-B956-CD762ABF4E8D@slis.indiana.edu> On Aug 14, 2007, at 12:50 PM, Jason Slack wrote: > I have a new task in front of me that I have never had to deal with. > Recently I have used Qt, by Trolltech, CPLAT and wxWidgets as cross > platform > GUI toolkits. > > A venture capitalist friend of mine approached me last week about > developing > a new cross platform GUI toolkit exercising some HIG and other > principals > that he feels are important. You can shoot for an xplat API that yields platform native look-and- feel. The gotcha is around the edges where one platform has a really different way of doing something than the other and you have to choose between substantially different behavior and encroaching platform specific code in the application around that, or you introduce something to one platform that may not really fit well. Or worse, you choose one platform as the model and blindly replicate on the other(s). Or you can (mostly) ignore the platform and develop your own UI architecture that works almost exactly the same on either side. Contemporary xplat Adobe applications, particularly the A/V applications (After Effects, Premiere Pro, Soundbooth, etc.) and Lightroom, are pretty good examples of this. They drop into platform native UI for the more mundane dialog boxes, but the core UI you interact with 98% of the time is an almost identical experience regardless of the platform. And the gap between the HIG and what Apple does with their own applications leaves a lot of latitude for interpretation, and no matter how good the toolkit is, some folks will still produce applications with a horrible user experience, and the HIG is no substitute for understanding how actual users interact with your program. -john From chris.garaffa at gmail.com Tue Aug 14 19:51:43 2007 From: chris.garaffa at gmail.com (Chris Garaffa) Date: Tue Aug 14 19:51:46 2007 Subject: sysctl hw.model In-Reply-To: References: Message-ID: <6cf7ded20708141951w243bfa87tff726145948cffbf@mail.gmail.com> On 8/14/07, Eric Gorr wrote: > I was just wondering if anyone knew where a complete list of strings > which can be returned by 'sysctl hw.model' is located? I've used the list on this page in an in-house inventory app I wrote and have never had a machine come by as unidentified: http://www.cocoadev.com/index.pl?MacintoshModels -- Chris From atomicbird at gmail.com Tue Aug 14 20:08:45 2007 From: atomicbird at gmail.com (Tom Harrington) Date: Tue Aug 14 20:08:49 2007 Subject: sysctl hw.model In-Reply-To: <6cf7ded20708141951w243bfa87tff726145948cffbf@mail.gmail.com> References: <6cf7ded20708141951w243bfa87tff726145948cffbf@mail.gmail.com> Message-ID: <68162e550708142008y18b1678sbfc5b71736931626@mail.gmail.com> On 8/14/07, Chris Garaffa wrote: > On 8/14/07, Eric Gorr wrote: > > I was just wondering if anyone knew where a complete list of strings > > which can be returned by 'sysctl hw.model' is located? > > I've used the list on this page in an in-house inventory app I wrote > and have never had a machine come by as unidentified: > > http://www.cocoadev.com/index.pl?MacintoshModels The SparklePlus project maintains a similar index as a property list at . I don't know of an official list from Apple. -- Tom Harrington atomicbird@gmail.com AIM: atomicbird1 From mailist at ericgorr.net Wed Aug 15 05:25:33 2007 From: mailist at ericgorr.net (Eric Gorr) Date: Wed Aug 15 05:25:39 2007 Subject: VRAM ( was Re: sysctl hw.model ) In-Reply-To: <68162e550708142008y18b1678sbfc5b71736931626@mail.gmail.com> References: <6cf7ded20708141951w243bfa87tff726145948cffbf@mail.gmail.com> <68162e550708142008y18b1678sbfc5b71736931626@mail.gmail.com> Message-ID: <1C413769-5740-43AC-9C37-8CDEC29D08A0@ericgorr.net> Thanks for the pointers to the two lists. Before I go off and use this information, I thought I would verify some information on this mailing list as well. The only thing I really need to know is whether or not VRAM is being shared with the 'real' ram like it is on a MacMini or MacBook. If there is a way to determine this, that would be the best solution for the problem I am attempting to solve. However, if there way a way to find actual (not shared) VRAM, that could work as well. Unfortunately, calls like aglDescribeRenderer or IORegistryEntrySearchCFProperty with kIOFBMemorySizeKey on the main display will not provide this information. There could be some other key I could search for with IORegistryEntrySearchCFProperty, but I am not certain what that would be. I was going to use the machine name only as a last resort. Thank you. From atomicbird at gmail.com Wed Aug 15 13:28:09 2007 From: atomicbird at gmail.com (Tom Harrington) Date: Wed Aug 15 13:28:18 2007 Subject: VRAM ( was Re: sysctl hw.model ) In-Reply-To: <1C413769-5740-43AC-9C37-8CDEC29D08A0@ericgorr.net> References: <6cf7ded20708141951w243bfa87tff726145948cffbf@mail.gmail.com> <68162e550708142008y18b1678sbfc5b71736931626@mail.gmail.com> <1C413769-5740-43AC-9C37-8CDEC29D08A0@ericgorr.net> Message-ID: <68162e550708151328l7c749a4cr24131eb8c0ea85e7@mail.gmail.com> On 8/15/07, Eric Gorr wrote: > The only thing I really need to know is whether or not VRAM is being > shared with the 'real' ram like it is on a MacMini or MacBook. If > there is a way to determine this, that would be the best solution for > the problem I am attempting to solve. I don't have a mini or MacBook to test on, but I'd guess that Apple's VideoHardwareDemo app at would demonstrate how to find that. Among other things it looks up the amount of video RAM-- maybe that'd be zero on those Macs? -- Tom Harrington atomicbird@gmail.com AIM: atomicbird1 From mailist at ericgorr.net Wed Aug 15 13:51:03 2007 From: mailist at ericgorr.net (Eric Gorr) Date: Wed Aug 15 13:51:21 2007 Subject: VRAM ( was Re: sysctl hw.model ) In-Reply-To: <68162e550708151328l7c749a4cr24131eb8c0ea85e7@mail.gmail.com> References: <6cf7ded20708141951w243bfa87tff726145948cffbf@mail.gmail.com> <68162e550708142008y18b1678sbfc5b71736931626@mail.gmail.com> <1C413769-5740-43AC-9C37-8CDEC29D08A0@ericgorr.net> <68162e550708151328l7c749a4cr24131eb8c0ea85e7@mail.gmail.com> Message-ID: <58C6107D-21A3-4248-A20C-C1FAC9D892C9@ericgorr.net> On Aug 15, 2007, at 4:28 PM, Tom Harrington wrote: > On 8/15/07, Eric Gorr wrote: >> The only thing I really need to know is whether or not VRAM is being >> shared with the 'real' ram like it is on a MacMini or MacBook. If >> there is a way to determine this, that would be the best solution for >> the problem I am attempting to solve. > > I don't have a mini or MacBook to test on, but I'd guess that Apple's > VideoHardwareDemo app at > would > demonstrate how to find that. Among other things it looks up the > amount of video RAM-- maybe that'd be zero on those Macs? Thanks for the reply. I have already checked that previously. Unfortunately, it is not zero. From mailist at ericgorr.net Wed Aug 15 14:18:47 2007 From: mailist at ericgorr.net (Eric Gorr) Date: Wed Aug 15 14:19:05 2007 Subject: VRAM ( was Re: sysctl hw.model ) In-Reply-To: <58C6107D-21A3-4248-A20C-C1FAC9D892C9@ericgorr.net> References: <6cf7ded20708141951w243bfa87tff726145948cffbf@mail.gmail.com> <68162e550708142008y18b1678sbfc5b71736931626@mail.gmail.com> <1C413769-5740-43AC-9C37-8CDEC29D08A0@ericgorr.net> <68162e550708151328l7c749a4cr24131eb8c0ea85e7@mail.gmail.com> <58C6107D-21A3-4248-A20C-C1FAC9D892C9@ericgorr.net> Message-ID: On Aug 15, 2007, at 4:51 PM, Eric Gorr wrote: > > On Aug 15, 2007, at 4:28 PM, Tom Harrington wrote: >> On 8/15/07, Eric Gorr wrote: >>> The only thing I really need to know is whether or not VRAM is being >>> shared with the 'real' ram like it is on a MacMini or MacBook. If >>> there is a way to determine this, that would be the best solution >>> for >>> the problem I am attempting to solve. >> >> I don't have a mini or MacBook to test on, but I'd guess that Apple's >> VideoHardwareDemo app at >> would >> demonstrate how to find that. Among other things it looks up the >> amount of video RAM-- maybe that'd be zero on those Macs? > > Thanks for the reply. > > I have already checked that previously. > Unfortunately, it is not zero. Although, the model does say 'GMA 950' which I believe is referring to http://en.wikipedia.org/wiki/Intel_GMA While checking against the model name would probably work, it would not seem to be any better then checking against the result of 'sysctl hw.model'. From kusterer at gmail.com Wed Aug 15 14:46:05 2007 From: kusterer at gmail.com (Uli Kusterer) Date: Wed Aug 15 14:46:19 2007 Subject: VRAM ( was Re: sysctl hw.model ) In-Reply-To: References: <6cf7ded20708141951w243bfa87tff726145948cffbf@mail.gmail.com> <68162e550708142008y18b1678sbfc5b71736931626@mail.gmail.com> <1C413769-5740-43AC-9C37-8CDEC29D08A0@ericgorr.net> <68162e550708151328l7c749a4cr24131eb8c0ea85e7@mail.gmail.com> <58C6107D-21A3-4248-A20C-C1FAC9D892C9@ericgorr.net> Message-ID: <5ACD5AFA-3A0E-4CCF-A689-FB721DD2FD11@gmail.com> On 15.08.2007, at 23:18, Eric Gorr wrote: > Although, the model does say 'GMA 950' which I believe is referring to > > http://en.wikipedia.org/wiki/Intel_GMA > > While checking against the model name would probably work, it would > not seem to be any better then checking against the result of > 'sysctl hw.model'. Actually, it would be better. Because any Mac that contains a GMA950 GPU will trigger this, even ones that Apple may not have released yet. So, it would potentially catch some cases for which the Mac name would still need to be updated. Cheers, -- M. Uli Kusterer http://www.zathras.de From cmhofman at gmail.com Thu Aug 16 06:57:08 2007 From: cmhofman at gmail.com (Christiaan Hofman) Date: Thu Aug 16 06:57:16 2007 Subject: Screen frames and fullscreen windows Message-ID: <6D78EACA-54AC-4A2F-BDE7-855D852A9945@gmail.com> I am struggling at the moment with properly placing fullscreen windows on arbitrary screens. In particular when the fullscreen window is not on the primary screen, the fullscreen window sometimes gets offset, and I've no idea what causes this. I use [screen frame] for the frame of the fullscreen window and place some secondary windows relative to this frame. But they sometimes seem to get offset relative to the primary screen. First of all, I need to understand the coordinates that are used for windows and screens. Unfortunately the docs are vague and inconsistent on this point, so I'm not sure if my interpretation is correct. They mention in several place "screen coordinates". E.g. window frames and content frames and screen frames are relative to this. My main question is : what are screen coordinates (as this is not defined anywhere in the docs). I can think of two distinct interpretations: 1. relative to the lower right corner of the primary screen 2. relative to the lower right corner of the current screen (the one that is referred to). I've made 1 my working hypothesis. But it is inconsistent with what the docs say about the window's contentRect (which implies 2 should be correct). However 2 seems utterly useless to me as you would not be able to refer to the frame of any screen in an unambiguous manner (it always has origin NSZeroPoint relative to itself) and would make it impossible to move windows to another screen. And how are different resolutions oif the different screens taken into account in the screen coordinates? I've no idea, frankly. It would be helpful is someone could enlighten me a bit. Thanks, Christiaan From jswitte at bloomington.in.us Thu Aug 16 08:28:23 2007 From: jswitte at bloomington.in.us (Jim Witte) Date: Thu Aug 16 08:21:30 2007 Subject: Is Safari 'find' function (visual) public API yet? Message-ID: <862A35A3-CD7D-4413-80D9-DBF9B763F527@bloomington.in.us> Is there a public API for having a find function like in Safari 3 with the colored highlighting of terms and such? I'd *love* to see this in other programs, but nothing else seems to - is this going to be yet another 'really neat and useful technology we'll keep all to ourselves' from Apple's POV? Jim From cmhofman at gmail.com Thu Aug 16 08:29:21 2007 From: cmhofman at gmail.com (Christiaan Hofman) Date: Thu Aug 16 08:29:27 2007 Subject: Is Safari 'find' function (visual) public API yet? In-Reply-To: <862A35A3-CD7D-4413-80D9-DBF9B763F527@bloomington.in.us> References: <862A35A3-CD7D-4413-80D9-DBF9B763F527@bloomington.in.us> Message-ID: Pretty sure it's custom. It's available on Tiger, so it's not a leopard feature, and there definitely is no documentation 9read public API) for this. Christiaan On 16 Aug 2007, at 5:28 PM, Jim Witte wrote: > Is there a public API for having a find function like in Safari 3 > with the colored highlighting of terms and such? I'd *love* to see > this in other programs, but nothing else seems to - is this going > to be yet another 'really neat and useful technology we'll keep all > to ourselves' from Apple's POV? > > Jim > _______________________________________________ > MacOSX-dev mailing list > MacOSX-dev@omnigroup.com > http://www.omnigroup.com/mailman/listinfo/macosx-dev From kusterer at gmail.com Thu Aug 16 14:42:23 2007 From: kusterer at gmail.com (Uli Kusterer) Date: Thu Aug 16 14:48:42 2007 Subject: Is Safari 'find' function (visual) public API yet? In-Reply-To: <862A35A3-CD7D-4413-80D9-DBF9B763F527@bloomington.in.us> References: <862A35A3-CD7D-4413-80D9-DBF9B763F527@bloomington.in.us> Message-ID: <52282AC9-47A1-4F04-B0E6-949BFB6CBEBC@gmail.com> On 16.08.2007, at 17:28, Jim Witte wrote: > Is there a public API for having a find function like in Safari 3 > with the colored highlighting of terms and such? I'd *love* to see > this in other programs, but nothing else seems to - is this going > to be yet another 'really neat and useful technology we'll keep all > to ourselves' from Apple's POV? It probably won't be too hard to do with CoreAnimation. Cheers, -- M. Uli Kusterer http://www.zathras.de From danhd123 at mac.com Thu Aug 16 15:05:51 2007 From: danhd123 at mac.com (Daniel DeCovnick) Date: Thu Aug 16 15:06:04 2007 Subject: Is Safari 'find' function (visual) public API yet? In-Reply-To: <52282AC9-47A1-4F04-B0E6-949BFB6CBEBC@gmail.com> References: <862A35A3-CD7D-4413-80D9-DBF9B763F527@bloomington.in.us> <52282AC9-47A1-4F04-B0E6-949BFB6CBEBC@gmail.com> Message-ID: <9D03A6AF-321B-4A8C-B90B-F4AAB82CB290@mac.com> I also imagine you can find the same thing (if not as pretty) in the Firefox source code. -Dan On Aug 16, 2007, at 2:42 PM, Uli Kusterer wrote: > On 16.08.2007, at 17:28, Jim Witte wrote: >> Is there a public API for having a find function like in Safari 3 >> with the colored highlighting of terms and such? I'd *love* to >> see this in other programs, but nothing else seems to - is this >> going to be yet another 'really neat and useful technology we'll >> keep all to ourselves' from Apple's POV? > > It probably won't be too hard to do with CoreAnimation. > > Cheers, > -- M. Uli Kusterer > http://www.zathras.de > > > > _______________________________________________ > MacOSX-dev mailing list > MacOSX-dev@omnigroup.com > http://www.omnigroup.com/mailman/listinfo/macosx-dev From jswitte at bloomington.in.us Thu Aug 16 18:07:03 2007 From: jswitte at bloomington.in.us (Jim Witte) Date: Thu Aug 16 18:00:11 2007 Subject: Way to make 'modifiable-zip' file-system? (DiskDoubler for OSX?) Message-ID: Hi, I'm not sure whether this should go on this list, on darwin-dev, or no-where at all, but it's rather odd. I know that disk space today is ridiculously cheap (I remember when a 256 *mb* HD was considered "big" and 1GB was "huge".. Not that long ago either :)) But I look at a lot of the sample codes (I sort of collect sample code :), which often is something like 1.1 MB expanded, but only 200K when zipped (or DMG'd), and always think it would be nice if there were some way I could work with it, bit keep it stored in a compressed format. I can imagine 2 ways to do this, assuming the DMG is not R/W (can they be, and still be compressed?) One would be to decompress the file to a ram disk (*aheh* memory resident FS I think), and then recompress. The other would save memory and maybe space at the expense of (maybe) time and a lot of complexity (I think). That would make an FS that would LOOK like the contents of the compressed files, and actually read file files and uncompress them on-the-fly (is it possible to uncompress just part of a zip archive, or must it be un-compressed 'in sequence'?), but then any changes you made would either make a NEW file that would be compressed (onto a separate 'diff' archive, either compressing the entire changed file, or just compressing a diff version of it. Then you'd have TWO files, the original and the 'diff' archive, which would probably be bundled together. Is this feasible to do (or is there something to do it already)? I'd imagine the first idea (uncompress to memory disk and then re- compress) would be easiest, but would also mean rewriting the entire archive, and not just the changes.) I guess this is also kinda like what I've read about how Time Machine will work - except that it's not for backup only, it's a main FS. Aside: This kinda similar to how the Newton did it's memory management: a "frame" accessed from a package was RO and still in virtual memory on the slow RAM cards, but if it was modified, it was copied copy in DRAM was RW format. The system transparently did the copying and access of the correct part of the memory - DRAM space or the high-address virtual package store space. Jim Witte From cmhofman at gmail.com Fri Aug 17 12:14:51 2007 From: cmhofman at gmail.com (Christiaan Hofman) Date: Fri Aug 17 12:14:58 2007 Subject: Weird shortcut messup Message-ID: <22498F29-775E-47DC-B1DD-1A09BA0761C0@gmail.com> On my system I'm having some very weird shortcut messup. When I hit Option-Command-P, the action bound to Command-, (Command-Comma) is activated (usually showing the preferences panel). However this only happens after I have hit a few unmodified character keys. I can reproduce it with every Cocoa app on my system that has a Cmd-, shortcut. I've tried to turn off every deamon, APE, input manager I could find and re-login, without success. I'm completely lost. If anyone on this list could tell me what could cause this, I would be very grateful. Thanks, Christiaan From kusterer at gmail.com Fri Aug 17 15:02:05 2007 From: kusterer at gmail.com (Uli Kusterer) Date: Fri Aug 17 15:02:34 2007 Subject: Weird shortcut messup In-Reply-To: <22498F29-775E-47DC-B1DD-1A09BA0761C0@gmail.com> References: <22498F29-775E-47DC-B1DD-1A09BA0761C0@gmail.com> Message-ID: <9FB922F0-81FE-4A7D-A5DC-994A0D83E099@gmail.com> On 17.08.2007, at 21:14, Christiaan Hofman wrote: > On my system I'm having some very weird shortcut messup. When I hit > Option-Command-P, the action bound to Command-, (Command-Comma) is > activated (usually showing the preferences panel). Just a few guesses? Is Cmd-Opt-P defined anywhere? Or is the OS maybe not finding that shortcut and trying to find a best match? Have you perhaps changed the keyboard mapping? Show the keyboard layouts menu and see whether the flag is different than what you want. There's also a setting that causes the keyboard mapping to be picked anew for each text field, which can reset the mapping to something undesired. Finally, have you checked in the "Keyboard" system preference whether there's any shortcut defined there that may be the cause? You can redefine them there, globally or for a particular app. Cheers, -- M. Uli Kusterer http://www.zathras.de From cmhofman at gmail.com Fri Aug 17 15:54:31 2007 From: cmhofman at gmail.com (Christiaan Hofman) Date: Fri Aug 17 15:54:41 2007 Subject: Weird shortcut messup In-Reply-To: <9FB922F0-81FE-4A7D-A5DC-994A0D83E099@gmail.com> References: <22498F29-775E-47DC-B1DD-1A09BA0761C0@gmail.com> <9FB922F0-81FE-4A7D-A5DC-994A0D83E099@gmail.com> Message-ID: On 18 Aug 2007, at 12:02 AM, Uli Kusterer wrote: > On 17.08.2007, at 21:14, Christiaan Hofman wrote: >> On my system I'm having some very weird shortcut messup. When I >> hit Option-Command-P, the action bound to Command-, (Command- >> Comma) is activated (usually showing the preferences panel). > > Just a few guesses? Is Cmd-Opt-P defined anywhere? Or is the OS > maybe not finding that shortcut and trying to find a best match? > It doesn't matter, it happens either when the shortcut is defined and when it isn't. > Have you perhaps changed the keyboard mapping? Show the keyboard > layouts menu and see whether the flag is different than what you > want. There's also a setting that causes the keyboard mapping to be > picked anew for each text field, which can reset the mapping to > something undesired. > I'm not sure what you mean here; you mean the list in the International preferences? This actually may explain it. In fact I just noticed that it happens only if I use the US Extended mapping, which somehow is included in the menu, and not with e.g. the US mapping. So is there a problem with that mapping? Can it be that textfield input can switch the input mapping somehow? I never understood what triggers changes of the mapping, I would have expected it to change > Finally, have you checked in the "Keyboard" system preference > whether there's any shortcut defined there that may be the cause? > You can redefine them there, globally or for a particular app. > I checked that, there's nothing even close. > Cheers, > -- M. Uli Kusterer > http://www.zathras.de > Thanks, Christiaan From dunham at mac.com Fri Aug 17 16:28:06 2007 From: dunham at mac.com (David Dunham) Date: Fri Aug 17 16:28:11 2007 Subject: Weird shortcut messup In-Reply-To: References: <22498F29-775E-47DC-B1DD-1A09BA0761C0@gmail.com> <9FB922F0-81FE-4A7D-A5DC-994A0D83E099@gmail.com> Message-ID: On 17 Aug 2007, at 15:54, Christiaan Hofman wrote: >> Have you perhaps changed the keyboard mapping? Show the keyboard >> layouts menu and see whether the flag is different than what you >> want. There's also a setting that causes the keyboard mapping to >> be picked anew for each text field, which can reset the mapping to >> something undesired. >> > > I'm not sure what you mean here; you mean the list in the > International preferences? > > This actually may explain it. In fact I just noticed that it > happens only if I use the US Extended mapping, which somehow is > included in the menu, and not with e.g. the US mapping. So is there > a problem with that mapping? Can it be that textfield input can > switch the input mapping somehow? I never understood what triggers > changes of the mapping, I would have expected it to change If I turn on US Extended (which is not the default, which is why your new account must have worked), I can reproduce this bug with Opal (a Cocoa app). Pretzel+option+P opens my Preferences dialog (normally activated by pretzel+,). David Dunham A Sharp, LLC Voice/Fax: 206 783 7404 http://a-sharp.com Efficiency is intelligent laziness. From toddransom at mac.com Sat Aug 18 09:53:52 2007 From: toddransom at mac.com (Todd Ransom) Date: Sat Aug 18 09:54:03 2007 Subject: Screen frames and fullscreen windows In-Reply-To: <6D78EACA-54AC-4A2F-BDE7-855D852A9945@gmail.com> References: <6D78EACA-54AC-4A2F-BDE7-855D852A9945@gmail.com> Message-ID: Have you checked - (BOOL)shouldCascadeWindows in the window controller? Todd Ransom Return Self Software http://returnself.com On Aug 16, 2007, at 9:57 AM, Christiaan Hofman wrote: > I am struggling at the moment with properly placing fullscreen > windows on arbitrary screens. In particular when the fullscreen > window is not on the primary screen, the fullscreen window > sometimes gets offset, and I've no idea what causes this. I use > [screen frame] for the frame of the fullscreen window and place > some secondary windows relative to this frame. But they sometimes > seem to get offset relative to the primary screen. > > First of all, I need to understand the coordinates that are used > for windows and screens. Unfortunately the docs are vague and > inconsistent on this point, so I'm not sure if my interpretation is > correct. They mention in several place "screen coordinates". E.g. > window frames and content frames and screen frames are relative to > this. My main question is : what are screen coordinates (as this is > not defined anywhere in the docs). > > I can think of two distinct interpretations: > > 1. relative to the lower right corner of the primary screen > 2. relative to the lower right corner of the current screen (the > one that is referred to). > > I've made 1 my working hypothesis. But it is inconsistent with what > the docs say about the window's contentRect (which implies 2 should > be correct). However 2 seems utterly useless to me as you would not > be able to refer to the frame of any screen in an unambiguous > manner (it always has origin NSZeroPoint relative to itself) and > would make it impossible to move windows to another screen. > > And how are different resolutions oif the different screens taken > into account in the screen coordinates? I've no idea, frankly. > > It would be helpful is someone could enlighten me a bit. > > Thanks, > > Christiaan > > _______________________________________________ > MacOSX-dev mailing list > MacOSX-dev@omnigroup.com > http://www.omnigroup.com/mailman/listinfo/macosx-dev From cmhofman at gmail.com Sat Aug 18 10:11:06 2007 From: cmhofman at gmail.com (Christiaan Hofman) Date: Sat Aug 18 10:11:17 2007 Subject: Screen frames and fullscreen windows In-Reply-To: References: <6D78EACA-54AC-4A2F-BDE7-855D852A9945@gmail.com> Message-ID: <40C3CCD9-BC4B-44FC-A7CD-91D1D367D771@gmail.com> That's definitely not the problem, as the fullscreen window is created long after the nib has been loaded. Moreover the offset is far bigger (and the opposite direction) than cascading would do. But I think I have figured out the problem, though I cannot test it at the moment without a secondary screen. The screen of the window seems to be assigned delayed, only when it is actually displayed. Therefore, I may do my calculations using a stale screen variable. (documentation would have been helpful). Christiaan On 18 Aug 2007, at 6:53 PM, Todd Ransom wrote: > Have you checked - (BOOL)shouldCascadeWindows in the window > controller? > > Todd Ransom > Return Self Software > http://returnself.com > > > > On Aug 16, 2007, at 9:57 AM, Christiaan Hofman wrote: > >> I am struggling at the moment with properly placing fullscreen >> windows on arbitrary screens. In particular when the fullscreen >> window is not on the primary screen, the fullscreen window >> sometimes gets offset, and I've no idea what causes this. I use >> [screen frame] for the frame of the fullscreen window and place >> some secondary windows relative to this frame. But they sometimes >> seem to get offset relative to the primary screen. >> >> First of all, I need to understand the coordinates that are used >> for windows and screens. Unfortunately the docs are vague and >> inconsistent on this point, so I'm not sure if my interpretation >> is correct. They mention in several place "screen coordinates". >> E.g. window frames and content frames and screen frames are >> relative to this. My main question is : what are screen >> coordinates (as this is not defined anywhere in the docs). >> >> I can think of two distinct interpretations: >> >> 1. relative to the lower right corner of the primary screen >> 2. relative to the lower right corner of the current screen (the >> one that is referred to). >> >> I've made 1 my working hypothesis. But it is inconsistent with >> what the docs say about the window's contentRect (which implies 2 >> should be correct). However 2 seems utterly useless to me as you >> would not be able to refer to the frame of any screen in an >> unambiguous manner (it always has origin NSZeroPoint relative to >> itself) and would make it impossible to move windows to another >> screen. >> >> And how are different resolutions oif the different screens taken >> into account in the screen coordinates? I've no idea, frankly. >> >> It would be helpful is someone could enlighten me a bit. >> >> Thanks, >> >> Christiaan >> >> _______________________________________________ >> MacOSX-dev mailing list >> MacOSX-dev@omnigroup.com >> http://www.omnigroup.com/mailman/listinfo/macosx-dev > From vlogsnapz at mac.com Sun Aug 19 18:43:11 2007 From: vlogsnapz at mac.com (vlogsnapz support) Date: Sun Aug 19 18:43:24 2007 Subject: Job offer for small company Message-ID: <9964DF97-9238-4F57-82B0-438001ED65F8@mac.com> Hi, Hope I'm not violating any rules by posting a job offer here. Basically, I'm looking for a developer to work with me on a project and to write code while I'm busy at school. The project is a videoblogging app that allows users to apply CoreImage effects while recording movies, then to upload these movies to a number of services. For a full description, see my site at http:// vlogsnapz.com. Basically, you would be joining the team rather than becoming an employee. So, you would get a percent of the profits rather than a salary. Job description is as follows (pulled directly from my site): Looking for a new project? vlogsnapz.com is looking for a talented Mac programmer to join the vmovie development team. The applicant must be trustworthy and reliable. (S)he must be willing to work in a small team (sometimes alone) to develop vmovie. Sometimes, the developer will be provided with detailed instructions describing what to implement. Other times, the developer must evaluate the needs of the market and develop accordingly. Applicants must have great experience with Objective-C and Cocoa. QuickTime experience is must. The applicant must also prove that (s) he has good design skills. A successful applicant may be required to design user interfaces for vmovie or for vlogsnapz.com as part of his/her work. The applicant would receive a percentage of profits as compensation. No other compensation would be provided. To apply, send an email to support@vlogsnapz.com with the subject "developer." Attach a zip archive or dmg of an application you have developed to the email. If you had help with the app, you must indicate how much work you did. The app should be pre-registered or a registration code must be provided in the email. The attached application and any related registration codes will be used only to determine your suitability as an applicant. If we think your app matches our quality standards, we will contact you and ask you to provide additional information. Good luck and we hope you find a place on our team! Eric Brunstad vmovie developer support@vlogsnapz.com http://vlogsnapz.com From hal at mobilegeographics.com Sun Aug 19 19:16:36 2007 From: hal at mobilegeographics.com (Hal Mueller) Date: Sun Aug 19 19:23:01 2007 Subject: Job offer for small company In-Reply-To: <9964DF97-9238-4F57-82B0-438001ED65F8@mac.com> References: <9964DF97-9238-4F57-82B0-438001ED65F8@mac.com> Message-ID: At 9:43 PM -0400 8/19/07, vlogsnapz support wrote (twice): > a percentage of profits as compensation. You can fit an aircraft carrier through that loophole, sideways. Profit after deducting what expenses? Audited by whom? Expenses authorized by whom? The risk is all on the developer, with zero risk to Mr. Brunstad. This posting should by rights be called a partnership request. -- Hal Mueller hal@mobilegeographics.com Mobile Geographics LLC http://www.mobilegeographics.com/ From kusterer at gmail.com Mon Aug 20 03:35:22 2007 From: kusterer at gmail.com (Uli Kusterer) Date: Mon Aug 20 03:35:30 2007 Subject: Mixing C, C++ and exceptions Message-ID: <1FA23FAE-0A5D-4BB9-A88B-D73432F91650@gmail.com> I tried googling for this info, but my Google-fu seems to be out of order today (if anyone wants to share search terms where they found anything, I'd be grateful)... THE QUESTION: If I'm mixing C and C++ code compiled as separate units (i.e. some in .c and some in .cpp files), is it safe for the C+ + code to throw exceptions "through" the C++ code, ABI-wise? Will the stack get unrolled OK and all that? I'm not talking about problems I might be getting with the C code leaking memory or not releasing resources it's acquired before the exception happened, I'm aware that this will happen when an exception happens before clean-up, but apart from that, does it work? I thought I'd heard somewhere it didn't, but couldn't find anything, so I thought I'd ask here if anyone has better memory than I do. And while we're at it, can anyone confirm/deny that the same does NOT apply to Objective C exceptions? Since they use longjmp under the hood, they won't unroll the stack correctly, right? Cheers, -- M. Uli Kusterer http://www.zathras.de From vlogsnapz at mac.com Mon Aug 20 06:08:58 2007 From: vlogsnapz at mac.com (vlogsnapz support) Date: Mon Aug 20 06:09:26 2007 Subject: Partnership request (formerly job offer for small company) In-Reply-To: References: <9964DF97-9238-4F57-82B0-438001ED65F8@mac.com> Message-ID: Hal, Ok, then, this is a partnership request. I'm just an honest person looking for an honest person to help me out. I am not trying to trick anyone. I believe if you live your life in trust and honesty, everything else will work out. The developer would have full access to whatever financial information (s)he needs to ensure (s)he is being paid correctly. And, you are wrong to say I have no risk. I have been working nonstop for the last five months to create this application. It is not released yet, so I have not yet received a single penny for all my efforts. Just looking for an honest developer to help out an honest developer. Thanks, Eric On Aug 19, 2007, at 10:16 PM, Hal Mueller wrote: > At 9:43 PM -0400 8/19/07, vlogsnapz support wrote (twice): >> a percentage of profits as compensation. > > You can fit an aircraft carrier through that loophole, sideways. > Profit after deducting what expenses? Audited by whom? Expenses > authorized by whom? > > The risk is all on the developer, with zero risk to Mr. Brunstad. > This posting should by rights be called a partnership request. > > -- > Hal Mueller hal@mobilegeographics.com > Mobile Geographics LLC http://www.mobilegeographics.com/ From macosxdevlist at personal.fishh2o.com Mon Aug 20 09:19:48 2007 From: macosxdevlist at personal.fishh2o.com (SD) Date: Mon Aug 20 09:20:27 2007 Subject: Partnership request (formerly job offer for small company) In-Reply-To: References: <9964DF97-9238-4F57-82B0-438001ED65F8@mac.com> Message-ID: There is a simple way to ensure one gets paid. Make sure you have all the source code for the application and don't hand it over until you receive some payment. SD ______________________________________________________________________ Previous message on 8/20/07 at 9:08 AM -0400 ********************************************************************** >Hal, > >Ok, then, this is a partnership request. I'm just an honest person >looking for an honest person to help me out. I am not trying to >trick anyone. I believe if you live your life in trust and honesty, >everything else will work out. The developer would have full access >to whatever financial information (s)he needs to ensure (s)he is >being paid correctly. And, you are wrong to say I have no risk. I >have been working nonstop for the last five months to create this >application. It is not released yet, so I have not yet received a >single penny for all my efforts. Just looking for an honest >developer to help out an honest developer. > >Thanks, >Eric > >On Aug 19, 2007, at 10:16 PM, Hal Mueller wrote: > >>At 9:43 PM -0400 8/19/07, vlogsnapz support wrote (twice): >>> a percentage of profits as compensation. >> >>You can fit an aircraft carrier through that loophole, sideways. >>Profit after deducting what expenses? Audited by whom? Expenses >>authorized by whom? >> >>The risk is all on the developer, with zero risk to Mr. Brunstad. >>This posting should by rights be called a partnership request. >> >>-- >>Hal Mueller hal@mobilegeographics.com >>Mobile Geographics LLC http://www.mobilegeographics.com/ > >_______________________________________________ >MacOSX-dev mailing list >MacOSX-dev@omnigroup.com >http://www.omnigroup.com/mailman/listinfo/macosx-dev -- ========================================== SD WARNING: Programming may be habit forming. From alastair at alastairs-place.net Mon Aug 20 17:32:49 2007 From: alastair at alastairs-place.net (Alastair Houghton) Date: Mon Aug 20 17:32:54 2007 Subject: Mixing C, C++ and exceptions In-Reply-To: <1FA23FAE-0A5D-4BB9-A88B-D73432F91650@gmail.com> References: <1FA23FAE-0A5D-4BB9-A88B-D73432F91650@gmail.com> Message-ID: <05F020BC-1243-4DBD-9DE6-14AB245BF076@alastairs-place.net> On 20 Aug 2007, at 11:35, Uli Kusterer wrote: > THE QUESTION: If I'm mixing C and C++ code compiled as separate > units (i.e. some in .c and some in .cpp files), is it safe for the C > ++ code to throw exceptions "through" the C++ code, (I take it you mean to say "'through' the C code" here) > ABI-wise? Will the stack get unrolled OK and all that? Yes, aside from the problems you mention below. However, in the case of C code, there's no real unwinding to do anyway... > I'm not talking about problems I might be getting with the C code > leaking memory or not releasing resources it's acquired before the > exception happened, I'm aware that this will happen when an > exception happens before clean-up, but apart from that, does it work? There are a few other problems besides the ones you mention. The C code could leave something in an invalid state, rather than just leaking memory (in the same way that a C++ function written without RAII and exception handling might do the same if you throw through it). As a result, it's generally regarded as a bad idea to throw through C code unless you know what the C code does. Even then, you'd better be pretty sure that nobody is going to change it in the future in such a way that it will break when you throw through it. > And while we're at it, can anyone confirm/deny that the same does > NOT apply to Objective C exceptions? Since they use longjmp under > the hood, they won't unroll the stack correctly, right? That's not true. They won't interact with C++ exception handling properly (except on Windows, where Microsoft implemented setjmp() and longjmp() in an odd manner), but they unwind the stack to the same extent that C++ code unwinds the stack for C functions. (AFAIK, unwinding the stack for a C function really just means setting the stack pointer to a higher address. It's that simple, and that's why exception handling can be implemented with setjmp() and longjmp(). Indeed, GCC used to do exactly that on some platforms...) Kind regards, Alastair. -- http://alastairs-place.net From jnutting at gmail.com Tue Aug 21 05:17:51 2007 From: jnutting at gmail.com (Jack Nutting) Date: Tue Aug 21 05:17:54 2007 Subject: Has anyone tried SCPlugin lately? Message-ID: I was pretty psyched to see that there's a new version of SCPlugin available (0.7, at http://scplugin.tigris.org/ ). Congrats to Jack Repenning (and any other contributors whose names I'm afraid I don't know) on getting it out the door! I got hooked on the TortoiseSVN style a couple years ago, and based on what's stated in the announcement, it looks like SCPlugin is now approaching that level of functionality. However, I'm kind of superstitious about Finder plugins in general, especially ones that do things like modifying icon displays (which I imagine *must* be relying on undocumented APIs), and therefore somewhat leery of installing SCPlugin at all. Has anyone besides the dev team been using it lately? Any comments on stability/functionality? Compatibility experiences with the revamped Finder in a certain upcoming OS? -- // jack // http://www.nuthole.com From m.vanland at gmail.com Tue Aug 21 13:30:03 2007 From: m.vanland at gmail.com (Michael Vanlandingham) Date: Tue Aug 21 13:35:34 2007 Subject: Has anyone tried SCPlugin lately? In-Reply-To: References: Message-ID: <3E7A49C2-BF4A-4849-8B4C-478371627888@gmail.com> FWIW, I tried it just last week, although I'm hesitant to use it, beyond basic browsing, on active "production' projects. I usually use the cmd-line svn so it's a little odd for me. So far I found it to be still quite rough, although usable for basic ci/co, etc. Some things I noticed were that the "Info" and "Log" menu items didn't seem to be functional, although selecting them would 'unfocus' the Finder's current window (cmd-tilde wouldn't work, etc) -- as if it had brought some hidden process to the foreground. The status window works OK but it'd be nice if it were a panel instead of a big window. Also, the Finder feels slower -- especially after I refresh the icons. I'm getting a few spods when opening some versioned projects in the Finder. As for the icon displays, I'm also wary: I'm not sure I want to run a Finder plugin that's doing mach/dyld hacks -- at least one that I didn't write ;-) I tried it primarily on Tiger, althogh I did install on Leopard as well. It looks like the override stuff for the badge overlay doesn't get away with it on Leopard. It draws, then it looks like the Finder quickly restores the original unmodified icon. -mvl On Aug 21, 2007, at 5:17 AM, Jack Nutting wrote: > I was pretty psyched to see that there's a new version of SCPlugin > available (0.7, at http://scplugin.tigris.org/ ). Congrats to Jack > Repenning (and any other contributors whose names I'm afraid I don't > know) on getting it out the door! I got hooked on the TortoiseSVN > style a couple years ago, and based on what's stated in the > announcement, it looks like SCPlugin is now approaching that level of > functionality. > > However, I'm kind of superstitious about Finder plugins in general, > especially ones that do things like modifying icon displays (which I > imagine *must* be relying on undocumented APIs), and therefore > somewhat leery of installing SCPlugin at all. Has anyone besides the > dev team been using it lately? Any comments on > stability/functionality? Compatibility experiences with the revamped > Finder in a certain upcoming OS? > > -- > // jack > // http://www.nuthole.com > _______________________________________________ > MacOSX-dev mailing list > MacOSX-dev@omnigroup.com > http://www.omnigroup.com/mailman/listinfo/macosx-dev From jswitte at bloomington.in.us Tue Aug 21 16:34:15 2007 From: jswitte at bloomington.in.us (Jim Witte) Date: Tue Aug 21 16:27:27 2007 Subject: Adding modifier key equivs to *contextual* menus (in Safari, so a hack..)? Message-ID: I'm wondering if there is any way that key equivalents can be added to contextual menus (if they are showed on screen with a command click, instead of just temporarily with a real right-click). Furthermore, can anyone think of a way to insert something like this (if possible) into Safari (into the link contextual menu - I want a key-equiv for Download file to..") using an APE. True, it would likely break on the next Safari release, but I am NOT going to release this to the general public.. The contextual menus aren't in Safari's nib files, which makes it much harder, and Safari itself looks like a Carbon app, which might make it harder still to find it in the code. But if it is something as simple as changing a string like 'Download file to' to 'Download file to/D', then it seems doable (but I don't know how the Carbon Menu Manager routines deal with multiple modifiers. Jim From cmhofman at gmail.com Tue Aug 21 17:04:30 2007 From: cmhofman at gmail.com (Christiaan Hofman) Date: Tue Aug 21 17:04:45 2007 Subject: Adding modifier key equivs to *contextual* menus (in Safari, so a hack..)? In-Reply-To: References: Message-ID: On 22 Aug 2007, at 1:34 AM, Jim Witte wrote: > I'm wondering if there is any way that key equivalents can be > added to contextual menus (if they are showed on screen with a > command click, instead of just temporarily with a real right- > click). Furthermore, can anyone think of a way to insert something > like this (if possible) into Safari (into the link contextual menu > - I want a key-equiv for Download file to..") using an APE. True, > it would likely break on the next Safari release, but I am NOT > going to release this to the general public.. > > The contextual menus aren't in Safari's nib files, which makes it > much harder, and Safari itself looks like a Carbon app, which might > make it harder still to find it in the code. But if it is > something as simple as changing a string like 'Download file to' to > 'Download file to/D', then it seems doable (but I don't know how > the Carbon Menu Manager routines deal with multiple modifiers. > > Jim > I don't think it's Carbon, as it is based on WebKit. Anyway, I don't think it is possible to add shortcuts to contextual menu items in Safari, as the menus are generated dynamically from an event rather than being associated to any view, so the menu does not exist to compare for shortcuts. See the WebKit API, in particular the WebUIDelegate protocol. Christiaan From huntc at internode.on.net Thu Aug 23 21:13:09 2007 From: huntc at internode.on.net (Christopher Hunt) Date: Thu Aug 23 21:13:37 2007 Subject: Core Video access to a Digital TV signal Message-ID: <5BFEE47D-ABA8-4B00-838E-FFA121FEA1E3@internode.on.net> Hi there, We wish to receive a free to air Digital Terrestrial Television signal via coaxial cable, select a channel and be able to access a memory buffer of frames. Does anyone know of a hardware product that can do this? Ideally it'd be a internal PCI Express card. I'm wanting this solution for a current model Mac Pro. Once we have a buffer we would then look to present the buffer to Core Video for rendering. Thanks for your advice. Kind regards, Christopher From rah at shipwright.com Thu Aug 23 21:31:21 2007 From: rah at shipwright.com (R.A. Hettinga) Date: Thu Aug 23 21:51:20 2007 Subject: Core Video access to a Digital TV signal In-Reply-To: <5BFEE47D-ABA8-4B00-838E-FFA121FEA1E3@internode.on.net> References: <5BFEE47D-ABA8-4B00-838E-FFA121FEA1E3@internode.on.net> Message-ID: At 2:13 PM +1000 8/24/07, Christopher Hunt wrote: >We wish to receive a free to air Digital Terrestrial Television >signal via coaxial cable, select a channel and be able to access a >memory buffer of frames. Miglia TVMini. Sold by EyeTV -- which I still can't get to work after I installed 10.4... I'll test it for you when you're done, if you like. Cheers, RAH -- ----------------- R. A. Hettinga The Internet Bearer Underwriting Corporation 44 Farquhar Street, Boston, MA 02131 USA "... however it may deserve respect for its usefulness and antiquity, [predicting the end of the world] has not been found agreeable to experience." -- Edward Gibbon, 'Decline and Fall of the Roman Empire' From kusterer at gmail.com Fri Aug 24 03:11:56 2007 From: kusterer at gmail.com (Uli Kusterer) Date: Fri Aug 24 03:12:12 2007 Subject: Core Video access to a Digital TV signal In-Reply-To: References: <5BFEE47D-ABA8-4B00-838E-FFA121FEA1E3@internode.on.net> Message-ID: <0AB980CE-696F-49CA-BF47-141F9DB591C6@gmail.com> On 24.08.2007, at 06:31, R.A. Hettinga wrote: > Miglia TVMini. > > Sold by EyeTV -- which I still can't get to work after I installed > 10.4... Err you're getting some things wrong here: 1) EyeTV is a *product* by a company called Elgato (http:// www.elgato.com), and has nothing to do with this hardware device (well, about half a year ago the manufacturer of the device had a license to bundle EyeTV, but that was revoked), and never sold them. 2) EyeTV works fine for me on 10.4.10. But then again, I have an EyeTV Hybrid stick to go with it. Hybrid devices are cool because you can receive digital TV, but also attach your old VCR or camera to grab videos. 3) the TVMini is not a card, it's a USB device. Other companies sell oodles of USB devices (including Hauppauge, Terratec, Pinnacle, Elgato, ...). I don't think there's any good PCI devices out there anymore (not too many people who own a Mac Pro *and* want to watch TV on it), but there's the HDHomerun (http://www.silicondust.com/), which is a network-attached dual-tuner, you may wanna look into that. 4) AFAIK all of these USB devices come with their own software, and aren't supported by CoreVideo in any way. The HDHomerun however, brings a streaming signal that you can receive e.g. with the free VLC player (http://www.videolan.org/), so it should be possible to write your own code for that open standard. If you don't need the TV tuner software, you can buy it from Silicon Dust themselves and save a few bucks. Disclaimer: I work for Elgato, so I am definitely biased. Cheers, -- M. Uli Kusterer http://www.zathras.de From alanrogersmith at gmail.com Sat Aug 25 09:42:21 2007 From: alanrogersmith at gmail.com (Alan Smith) Date: Sat Aug 25 09:42:24 2007 Subject: [ANN] ArtistKit Message-ID: Hello everyone, I just want to let you all know that I have released my first major accomplishment in the way of giving back to the Cocoa community. Check it out at cocoalatte.info. The third to last post is what it's all about. Even if you don't download it and try it out please leave some feedback, I don't care if it's negative, it all counts. Peace, Alan -- My blog: cocoalatte.info // Things I've said ------------------------- "Maturity resides in the mind." "Silence is the Universe's greatest gift." "When the World realizes that personal beliefs are not something to argue or fight over, it shall evolve." From alanrogersmith at gmail.com Sun Aug 26 10:46:50 2007 From: alanrogersmith at gmail.com (Alan Smith) Date: Sun Aug 26 10:46:57 2007 Subject: [ANN] ArtistKit v1.1 Message-ID: Hi everyone, I have released version 1.1 of ArtistKit. Thanks to all those who tried version 1.0, your support is greatly appreciated. Please try version 1.1 and give me feedback. http://cocoalatte.info/2007/08/26/artistkit-11/ Many thanks. Peace, Alan -- My blog: cocoalatte.info // Things I've said ------------------------- "Maturity resides in the mind." "Silence is the Universe's greatest gift." "When the World realizes that personal beliefs are not something to argue or fight over, it shall evolve."