language localization
Richard Gaskin wrote: More importantly, this may be a healthy wake-up call for all of us: There are other cultures out there, and not all of them speak English. How many of you are getting requests for localized versions of your MC/Rev-based software? How do you approach localization with a tool like MetaCard where we commonly embed user feedback directly in the code (the ask and answer commands for example)? I got my wakeup call recently. I created a large educational project without ever having the faintest thought about anyone in another country or speaking another language using it. After I posted my software for download, it was very entertaining to count how many different countries there were in the download record. Then a prof in Brazil asked if I would incorporate his translation of all the text in my project into Portuguese. It was easy to say sure. I was astounded to go to the web and find that there are 50 universities in Brazil with chemical engineering departments. I started hoping someone would volunteer to do a Spanish translation. Then I got the Portuguese translation in a big Word document. Turns out translation is the easy part (ignoring the effort of learning a second language). The hard part is that now I am committed to - and am in the middle of - replacing many hundreds of strings in hundreds of objects. So I learned: Whenever I start any new project, I am going to design the project to make translation and language localization easy. Jacqueline suggested putting text in the custom properties of each object, answer text in the custom props of a button, for example. Adding a new language then would mean accessing all the individual objects. Maybe put the text strings at the stack (or higher level), e.g., in a custom property set of the stack whose keys are formed from the name of the object using the string plus a string number (or other identifier) for objects using multiple strings... Rich Herz [EMAIL PROTECTED] Archives: http://www.mail-archive.com/metacard@lists.runrev.com/ Info: http://www.xworlds.com/metacard/mailinglist.htm Please send bug reports to [EMAIL PROTECTED], not this list.
Re: libURL
-- MESSAGE metacard.v004.n394.10 --- From: Scott Rossi [EMAIL PROTECTED] Subject: Re: libURL Date: Mon, 16 Jul 2001 12:19:36 -0700 MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit In-Reply-To: [EMAIL PROTECTED] I'm just now trying to get up to speed with the new URL implementation in 2.4B3 and can't seem to get the POST method to work (this works fine in 2.3.2). My form looks likes this: set httpheaders to specialHeaderInforeturn put http://www.myURL.com/special.cfm?method=postaction=action; into HREF put data= fld data into tData start using stack libURL post tData to url HREF answer it Problem = I'm not getting any data back in it; in fact, I'm not getting the answer dialog at all. I've tried removing the libURL line (I assume you need to use this with a POST command) with the same results. The simple loading of a URL via the libURL stack works fine, and the server that is handling the CFM action responds as expected using MC 2.3.2. Am I missing something about POST in MC 2.4? On a related note, is it the case that you need to include the libURL stack as a substack in your own stack if you distribute your stack as a standalone? If this is the case, why isn't the libURL stack included as an option in the resource mover? FWIW, I've tried cloning the libURL stack on MacOS 9.1 three times, each time resulting in a frozen crash. Regards, Scott The problem is that http headers are controled by the liburl stack properties. Your changes are not taken in account, unless you customize the liburl stack. My impression is that liburl is experimental (and promising). H. Chaudet -- - Herve Chaudet e-mail : [EMAIL PROTECTED] mobile : 33-(0)-608-05-98-09 Unité UPRES EA 2672 Faculte de Medecine - 27, Bd Jean Moulin - 13385 Marseille cedex 5 - France Tel 33-(0)-491-79-19-10 ; Fax 33-(0)-491-79-40-13 Service de l'Information Medicale Hopital Nord - Chemin des Bourrelly - 13326 Marseille cedex 15 - France Tel 33-(0)-491-96-80-20 - Archives: http://www.mail-archive.com/metacard@lists.runrev.com/ Info: http://www.xworlds.com/metacard/mailinglist.htm Please send bug reports to [EMAIL PROTECTED], not this list.
Re: Multi-column list workarounds
Richard, In my last project I opted to readjust the width of each column as needed. The downside is the calculation time this requires for a list with many columns, but otherwise it works quite well. Perhaps you can improve on my code to gain some speed (please let me know if you can!). Here's what I've been doing: on reformatTabStops set the itemDel to tab set the numberFormat to 000.## put the margins of fld text into tMargins put the number of lines of fld text into tNum put the number of items of line 1 of fld text into tItems repeat with i = 1 to tItems set cursor to busy repeat with x = 1 to tNum put (the formattedWidth of item i of line x of fld text) comma after tLengths end repeat if last char of tLengths is comma then delete last char of tLengths put max(tLengths) after aColWidths[item i of fld column labels] put max(aColWidths[item i of fld column labels]) comma after tColumnWidths put empty into tLengths -- prep for next column end repeat delete last char of tColumnWidths -- remove final comma set the itemDel to comma put 0 into tTabStop repeat for each item y in tColumnWidths -- calc tabstops add (y + tMargins) + 2 to tTabStop -- 2 seems to work best, esp. for smaller sizes -- each tabstop = current width + prev value put tTabStop comma after tTabList end repeat delete last char of tTabList -- remove final comma set the tabStops of fld text to tTabList set the scroll of fld text to 0 end reformatTabStops The other alternative I've played with is to use a single MC field, but to algorithmically truncate each line of the displated text if it exceeds the column width. I found this solution to have its own issues: - Performance of the truncation bites. - Requires a monospaced font in order to calculate where to truncate, or an even slower algorithm which queries the formattedwidth. - You need to maintain two sets of data, the complete set and then a copy in the displayed field with some lines truncated. Sometimes this is a RAM issue, and you have to remember as you're coding that everytime you touch the data you need to refresh the display version. So for displaying database columns, how do you folks handle that? -- Richard Gaskin Fourth World Media Corporation Multimedia Design and Development for Mac, Windows, UNIX, and the Web -- /\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\ \/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/ Craig Spooner Instructional Technology Training and Development WebCT Administrator Office of Instructional Services Colorado State University (970) 491-2516 Archives: http://www.mail-archive.com/metacard@lists.runrev.com/ Info: http://www.xworlds.com/metacard/mailinglist.htm Please send bug reports to [EMAIL PROTECTED], not this list.
Re: language localization
On Mon, 16 Jul 2001 Richard Herz [EMAIL PROTECTED] wrote: (snip) So I learned: Whenever I start any new project, I am going to design the project to make translation and language localization easy. Jacqueline suggested putting text in the custom properties of each object, answer text in the custom props of a button, for example. Adding a new language then would mean accessing all the individual objects. Maybe put the text strings at the stack (or higher level), e.g., in a custom property set of the stack whose keys are formed from the name of the object using the string plus a string number (or other identifier) for objects using multiple strings... Unfortunately another common requirement for large multilanguage products like this is that the actual translation must be done by people who only use word processors. The above architecture is efficient and will work fine if you're willing to reimport all of the text each time any of it changes. I've also seen projects that store the text for each object in a separate file in a language-specific directory that's opened and read in as a card is opened. That way you just hand the whole directory structure over to the translator and they just go through the files one at a time. This architecture is also pretty much a requirement if your product relies on a lot of text in images (importing lots of images is something you have to be careful about even in single-language products, and with multiple languages it pretty much becomes a necessity). Of course you could combine the two approaches and instead of loading the text at run time you write a script that goes through the whole stack and loads all of the text at once. Then you just have to push a button when you get a new translation. Regards, Scott Rich Herz [EMAIL PROTECTED] Scott Raney [EMAIL PROTECTED] http://www.metacard.com MetaCard: You know, there's an easier way to do that... Archives: http://www.mail-archive.com/metacard@lists.runrev.com/ Info: http://www.xworlds.com/metacard/mailinglist.htm Please send bug reports to [EMAIL PROTECTED], not this list.
Groups again
Today I've been playing with groups in MC and found it quite frustrating, for 2 main reasons (if I missed anything please MC gurus enlighten me) : - the only way to group controls is to select them first, which makes grouping/ungrouping on the fly via script quite cumbersome : you need to set the selected of each control to true, then group, then set the selected of all controls back to false. And if you don't wznt your controls to be selected on screen during the process, you have to perform a lock screen / unlock screen... Frankly, I miss the groups in OMO where you could group any number of objects by simply listing them in a variable (the same kind of list that you get with the selectedobjects)... The good side of it was that the same object could be part of different groups at the same time... - it looks like grouping controls changes the numbering of those controls : for instance I have 50 images on screen and the user can select any of them (like image 12 to 25 and 37 to 42) and they're supposed to get automatically grouped. What happens is that all images being in a group are re-numbered ! Therefore I had no choice but to change my scripts and use the id instead... Is there any *good* reason for that kind of behavior ? Thanks. JB Archives: http://www.mail-archive.com/metacard@lists.runrev.com/ Info: http://www.xworlds.com/metacard/mailinglist.htm Please send bug reports to [EMAIL PROTECTED], not this list.
Re: how do I turn off the focusColor?
On Sat, 14 Jul 2001 Richard Gaskin [EMAIL PROTECTED] wrote: Karl Becker wrote: Actually, if we're really adhering to the HIG, shouldn't the focusColor not be able to be chosen by MC developers, and only by the user? ;-) Maybe so, but at least this is the way it works by default unless you specifically override the focus color by setting it. That's the beauty of MC: It's not just HIG-compliant, it's better than HIG-compliant, in that it often lets us decide when the HIG applies. Please, Scott, allow another option for the focusColor which turns it off. In the big picture, this is the most HIG-compliant option. I guess I've overestimated the number of people developing for (or even testing for) UNIX systems. This has been the way the Motif look and feel has worked for almost 10 years now and we've yet to get a single complaint about it or request for a way to turn it off. The only other thing I can come up with is just conservatism (it *changed* on the Mac, and many people just don't like change, even if it's for the better). As for ways to implement this, a new field-only property is unfortunately not an option because it would require reworking the file format. So that leaves us with the following options: 1) Tie drawing the focus border to the showBorder property. This means you'd have to put another object (e.g. a group) around a field you want to have a border but no focus border. More work, but then, breaking the rules should require more work. 2) Make it apply to all objects (i.e., buttons and players also get a focus border in the Motif look and feel, and it might be nice to be able to turn this off). 3) Make it global or an inheritable stack property. 4) Hack the focusColor property to have some non-existing color be a cue to not draw the focus border at all. Preferences? Scott -- Richard Gaskin Fourth World Media Corporation Multimedia Design and Development for Mac, Windows, UNIX, and the Web _ [EMAIL PROTECTED] http://www.FourthWorld.com Tel: 323-225-3717 ICQ#60248349Fax: 323-225-0716 Scott Raney [EMAIL PROTECTED] http://www.metacard.com MetaCard: You know, there's an easier way to do that... Archives: http://www.mail-archive.com/metacard@lists.runrev.com/ Info: http://www.xworlds.com/metacard/mailinglist.htm Please send bug reports to [EMAIL PROTECTED], not this list.
Re: language localization
At 9:08 AM -0600 7/17/01, Scott Raney wrote: Of course you could combine the two approaches and instead of loading the text at run time you write a script that goes through the whole stack and loads all of the text at once. Then you just have to push a button when you get a new translation. That's the solution I was thinking about.It seems like it would be fairly straightforward to write two scripts: 1. Go through a project and pull out all displayed text and build a list of it. 2. Take a returned list with the original text and the translated text, and go through the project and replace each piece of text with the translated version. gc -- Archives: http://www.mail-archive.com/metacard@lists.runrev.com/ Info: http://www.xworlds.com/metacard/mailinglist.htm Please send bug reports to [EMAIL PROTECTED], not this list.
Re: language localization
At 9:08 AM -0600 7/17/01, Scott Raney wrote: Of course you could combine the two approaches and instead of loading the text at run time you write a script that goes through the whole stack and loads all of the text at once. Then you just have to push a button when you get a new translation. That's the solution I was thinking about.It seems like it would be fairly straightforward to write two scripts: 1. Go through a project and pull out all displayed text and build a list of it. 2. Take a returned list with the original text and the translated text, and go through the project and replace each piece of text with the translated version. gc -- Archives: http://www.mail-archive.com/metacard@lists.runrev.com/ Info: http://www.xworlds.com/metacard/mailinglist.htm Please send bug reports to [EMAIL PROTECTED], not this list.
Re: how do I turn off the focusColor?
Scott Raney wrote: On Sat, 14 Jul 2001 Richard Gaskin [EMAIL PROTECTED] wrote: Karl Becker wrote: Actually, if we're really adhering to the HIG, shouldn't the focusColor not be able to be chosen by MC developers, and only by the user? ;-) Maybe so, but at least this is the way it works by default unless you specifically override the focus color by setting it. That's the beauty of MC: It's not just HIG-compliant, it's better than HIG-compliant, in that it often lets us decide when the HIG applies. Please, Scott, allow another option for the focusColor which turns it off. In the big picture, this is the most HIG-compliant option. I guess I've overestimated the number of people developing for (or even testing for) UNIX systems. This has been the way the Motif look and feel has worked for almost 10 years now and we've yet to get a single complaint about it or request for a way to turn it off. The only other thing I can come up with is just conservatism (it *changed* on the Mac, and many people just don't like change, even if it's for the better). As for ways to implement this, a new field-only property is unfortunately not an option because it would require reworking the file format. So that leaves us with the following options: 1) Tie drawing the focus border to the showBorder property. This means you'd have to put another object (e.g. a group) around a field you want to have a border but no focus border. More work, but then, breaking the rules should require more work. 2) Make it apply to all objects (i.e., buttons and players also get a focus border in the Motif look and feel, and it might be nice to be able to turn this off). 3) Make it global or an inheritable stack property. 4) Hack the focusColor property to have some non-existing color be a cue to not draw the focus border at all. I'd go for either 2 or 4, most likely 4 if there is something like transparent color. Preferences? Scott -- Regards, Andu ___ [EMAIL PROTECTED] Archives: http://www.mail-archive.com/metacard@lists.runrev.com/ Info: http://www.xworlds.com/metacard/mailinglist.htm Please send bug reports to [EMAIL PROTECTED], not this list.
Re: how do I turn off the focusColor?
On 17/7/01 6:34 pm, Scott Raney [EMAIL PROTECTED] wrote: 2) Make it apply to all objects (i.e., buttons and players also get a focus border in the Motif look and feel, and it might be nice to be able to turn this off). 3) Make it global or an inheritable stack property. 4) Hack the focusColor property to have some non-existing color be a cue to not draw the focus border at all. It does need to apply object by object - as you do generally want this on. Turning it off happens on a case-by-case basis where you're doing something cool that the field object probably wasn't originally designed for (such as a multi-column list). In such a case, no way do you want to have an extra object there to draw the border with. As such it has to be 2 or 4, and I would go for 4, as 2 is going to be a real hassle from the cross-platform point of view (you have to turn this property off on Mac for buttons, but then turn it on again on UNIX). Regards, Kevin Kevin Miller [EMAIL PROTECTED] http://www.runrev.com/ Runtime Revolution Limited - Power to the Developer! Tel: +44 (0)131 718 4333. Fax: +44 (0)1639 830 707. Archives: http://www.mail-archive.com/metacard@lists.runrev.com/ Info: http://www.xworlds.com/metacard/mailinglist.htm Please send bug reports to [EMAIL PROTECTED], not this list.
Re: how do I turn off the focusColor?
on 07.17.01 12:34 PM, Scott Raney wrote: 2) Make it apply to all objects (i.e., buttons and players also get a focus border in the Motif look and feel, and it might be nice to be able to turn this off). 3) Make it global or an inheritable stack property. In my projects, either I want focus borders everywhere (a conventional text-entry app) or I don't want them anywhere (an educational quasi-game). So I guess I'd back method 2 and/or 3. t + Tereza Snyder +Software Developer + Attainment Company, Inc. + www.attainmentcompany.com + 608.845.7880 Archives: http://www.mail-archive.com/metacard@lists.runrev.com/ Info: http://www.xworlds.com/metacard/mailinglist.htm Please send bug reports to [EMAIL PROTECTED], not this list.
AppleScript
Has anyone created scripts for retrieving data from an application via AppleScript? Specifically, I'm trying to use MC to pull addresses/content from email addresses in Outlook Express. Using MC, I can send AppleEvents to OE, but I'm having a hard time figuring out how to get anything back from OE, via request or the appleEvent message. Thanks for any suggestions. Regards, Scott _ Scott Rossi Tactile Media - Multimedia Design Creative Director Email: [EMAIL PROTECTED] Web: www.tactilemedia.com Archives: http://www.mail-archive.com/metacard@lists.runrev.com/ Info: http://www.xworlds.com/metacard/mailinglist.htm Please send bug reports to [EMAIL PROTECTED], not this list.
Re: AppleScript
Scott Rossi wrote: Has anyone created scripts for retrieving data from an application via AppleScript? Specifically, I'm trying to use MC to pull addresses/content from email addresses in Outlook Express. Using MC, I can send AppleEvents to OE, but I'm having a hard time figuring out how to get anything back from OE, via request or the appleEvent message. Check the data types returned from OE. If they're anything other than type 'TEXT' (such as lists), forget it, as MC, like most xTalks, only understands TEXT data. You're using the request command to obtain the data portion from the incoming message, yes? -- Richard Gaskin Fourth World Media Corporation Multimedia Design and Development for Mac, Windows, UNIX, and the Web _ [EMAIL PROTECTED] http://www.FourthWorld.com Tel: 323-225-3717 ICQ#60248349Fax: 323-225-0716 Archives: http://www.mail-archive.com/metacard@lists.runrev.com/ Info: http://www.xworlds.com/metacard/mailinglist.htm Please send bug reports to [EMAIL PROTECTED], not this list.
Re: Groups again
jbv wrote: - it looks like grouping controls changes the numbering of those controls : for instance I have 50 images on screen and the user can select any of them (like image 12 to 25 and 37 to 42) and they're supposed to get automatically grouped. What happens is that all images being in a group are re-numbered ! Therefore I had no choice but to change my scripts and use the id instead... Is there any *good* reason for that kind of behavior ? In my experience, grouped objects maintain their relative order, but their positions relative to objects outside the group will change, since the group is placed on top of other objects by default when it is created. If you address objects relative to the group things should be fine, using something like: get the long id of image 1 of grp MyGroup One of the amazing niceties about MC is that when you ungroup objects to work on them and then regroup them again, most of the time the original group name is retained -- one less step to worry about. -- Richard Gaskin Fourth World Media Corporation Multimedia Design and Development for Mac, Windows, UNIX, and the Web _ [EMAIL PROTECTED] http://www.FourthWorld.com Tel: 323-225-3717 ICQ#60248349Fax: 323-225-0716 Archives: http://www.mail-archive.com/metacard@lists.runrev.com/ Info: http://www.xworlds.com/metacard/mailinglist.htm Please send bug reports to [EMAIL PROTECTED], not this list.
Re: how do I turn off the focusColor?
Kevin Miller wrote: On 17/7/01 6:34 pm, Scott Raney [EMAIL PROTECTED] wrote: 2) Make it apply to all objects (i.e., buttons and players also get a focus border in the Motif look and feel, and it might be nice to be able to turn this off). 3) Make it global or an inheritable stack property. 4) Hack the focusColor property to have some non-existing color be a cue to not draw the focus border at all. It does need to apply object by object - as you do generally want this on. Turning it off happens on a case-by-case basis where you're doing something cool that the field object probably wasn't originally designed for (such as a multi-column list). In such a case, no way do you want to have an extra object there to draw the border with. As such it has to be 2 or 4, and I would go for 4, as 2 is going to be a real hassle from the cross-platform point of view (you have to turn this property off on Mac for buttons, but then turn it on again on UNIX). My vote: I agree with Kevin's comments. -- Richard Gaskin Fourth World Media Corporation Multimedia Design and Development for Mac, Windows, UNIX, and the Web _ [EMAIL PROTECTED] http://www.FourthWorld.com Tel: 323-225-3717 ICQ#60248349Fax: 323-225-0716 Archives: http://www.mail-archive.com/metacard@lists.runrev.com/ Info: http://www.xworlds.com/metacard/mailinglist.htm Please send bug reports to [EMAIL PROTECTED], not this list.
Re: Groups again
At 8:23 PM + 7/17/01, jbv wrote: - the only way to group controls is to select them first, which makes grouping/ungrouping on the fly via script quite cumbersome : you need to set the selected of each control to true, then group, then set the selected of all controls back to false. And if you don't wznt your controls to be selected on screen during the process, you have to perform a lock screen / unlock screen... Depending on how much flexibility you need, relayerGroupedControls can help with this. You probably want pre-existing groups to make it work well, and if I understand correctly, each pre-existing group needs at least one object in it (but the object could be a point, or invisible). Then to add an existing object to a group, you simply set the layer of the object to the layer of the object in the group. - it looks like grouping controls changes the numbering of those controls : for instance I have 50 images on screen and the user can select any of them (like image 12 to 25 and 37 to 42) and they're supposed to get automatically grouped. What happens is that all images being in a group are re-numbered ! Therefore I had no choice but to change my scripts and use the id instead... Is there any *good* reason for that kind of behavior ? Groups are inherently part of the layering scheme of the objects on a card; take a look at the control browser for a card with several objects and groups on it to get a very clear picture of this. The control browser is like an outline, and hence you can't have objects that aren't on consecutive layers be in the same group. When you select items and group them, if necessary they are automatically relayered so that they are next to each other. gc -- Archives: http://www.mail-archive.com/metacard@lists.runrev.com/ Info: http://www.xworlds.com/metacard/mailinglist.htm Please send bug reports to [EMAIL PROTECTED], not this list.
Re: Groups again
At 8:23 PM + 7/17/01, jbv wrote: - the only way to group controls is to select them first, which makes grouping/ungrouping on the fly via script quite cumbersome : you need to set the selected of each control to true, then group, then set the selected of all controls back to false. And if you don't wznt your controls to be selected on screen during the process, you have to perform a lock screen / unlock screen... Depending on how much flexibility you need, relayerGroupedControls can help with this. You probably want pre-existing groups to make it work well, and if I understand correctly, each pre-existing group needs at least one object in it (but the object could be a point, or invisible). Then to add an existing object to a group, you simply set the layer of the object to the layer of the object in the group. - it looks like grouping controls changes the numbering of those controls : for instance I have 50 images on screen and the user can select any of them (like image 12 to 25 and 37 to 42) and they're supposed to get automatically grouped. What happens is that all images being in a group are re-numbered ! Therefore I had no choice but to change my scripts and use the id instead... Is there any *good* reason for that kind of behavior ? Groups are inherently part of the layering scheme of the objects on a card; take a look at the control browser for a card with several objects and groups on it to get a very clear picture of this. The control browser is like an outline, and hence you can't have objects that aren't on consecutive layers be in the same group. When you select items and group them, if necessary they are automatically relayered so that they are next to each other. gc -- Archives: http://www.mail-archive.com/metacard@lists.runrev.com/ Info: http://www.xworlds.com/metacard/mailinglist.htm Please send bug reports to [EMAIL PROTECTED], not this list.
Re: how do I turn off the focusColor?
On Tue, 17 Jul 2001 Kevin Miller [EMAIL PROTECTED] wrote: On 17/7/01 6:34 pm, Scott Raney [EMAIL PROTECTED] wrote: 2) Make it apply to all objects (i.e., buttons and players also get a focus border in the Motif look and feel, and it might be nice to be able to turn this off). 3) Make it global or an inheritable stack property. 4) Hack the focusColor property to have some non-existing color be a cue to not draw the focus border at all. It does need to apply object by object - as you do generally want this on. Turning it off happens on a case-by-case basis where you're doing something cool that the field object probably wasn't originally designed for (such as a multi-column list). In such a case, no way do you want to have an extra object there to draw the border with. As such it has to be 2 or 4, and I would go for 4, as 2 is going to be a real hassle from the cross-platform point of view (you have to turn this property off on Mac for buttons, but then turn it on again on UNIX). Actually I hadn't even considered this tangent: Focus borders for fields on Windows anyone? I was only considering it as a negative option, where setting it to false prevented it from being drawn on Motif and Macintosh look and feel settings, but setting it to true wouldn't cause a focus border to always be drawn for all object types on all platforms. And now that I consider it, I think it's probably not worth considering ;-) So, by your criteria I think option 2 would work too. And I can't say that I'm thrilled with the idea of having a color named transparent which seems confusing to me. Next thing you know, people will be setting the backColor to transparent and complaining when their fields are still opaque... Regards, Scott Regards, Kevin Kevin Miller [EMAIL PROTECTED] http://www.runrev.com/ Runtime Revolution Limited - Power to the Developer! Tel: +44 (0)131 718 4333. Fax: +44 (0)1639 830 707. Scott Raney [EMAIL PROTECTED] http://www.metacard.com MetaCard: You know, there's an easier way to do that... Archives: http://www.mail-archive.com/metacard@lists.runrev.com/ Info: http://www.xworlds.com/metacard/mailinglist.htm Please send bug reports to [EMAIL PROTECTED], not this list.
Re: Groups again
On Tue, 17 Jul 2001 jbv [EMAIL PROTECTED] wrote: Today I've been playing with groups in MC and found it quite frustrating, for 2 main reasons (if I missed anything please MC gurus enlighten me) : - the only way to group controls is to select them first, which makes grouping/ungrouping on the fly via script quite cumbersome : you need to set the selected of each control to true, then group, then set the selected of all controls back to false. And if you don't wznt your controls to be selected on screen during the process, you have to perform a lock screen / unlock screen... Frankly, I miss the groups in OMO where you could group any number of objects by simply listing them in a variable (the same kind of list that you get with the selectedobjects)... The good side of it was that the same object could be part of different groups at the same time... Frankly I'm a little mystified that it's possible to consider the concept of a group in OMO to be anything other than an ugly hack. You get none of the benefits of groups in MetaCard (e.g., clipping, scrolling, nesting, handler and attribute inheritance, and the ability to place them on different cards like backgrounds), they're also purely extent-based both for selection and manupulation, and not even really objects at all, just (as you've pointed out) some sort of disembodied list of objects that should be moved together. Nevertheless, being able to specify a list of controls to group (as in group button 1 and field 2) is a useful feature, which is why it was implemented in MetaCard years ago. Why it wasn't documented at that time I'm not as sure about, but I'll bug report this so it will be for the 2.4 doc. - it looks like grouping controls changes the numbering of those controls : for instance I have 50 images on screen and the user can select any of them (like image 12 to 25 and 37 to 42) and they're supposed to get automatically grouped. What happens is that all images being in a group are re-numbered ! Therefore I had no choice but to change my scripts and use the id instead... Probably not a bad idea in general: referring to objects by number is exceedingly brittle and can result in maintenance nightmares if you ever have to change the application by adding or deleting a control. Is there any *good* reason for that kind of behavior ? Yes: groups are real objects in MetaCard, so the controls have to be children of the group, which of course means they also have to be contiguous. This also helps with setting up keyboard navigation so you don't have to worry about creating a situation where you tab into a group and then back out when navigating a set of radio buttons. Regards, Scott Thanks. JB Scott Raney [EMAIL PROTECTED] http://www.metacard.com MetaCard: You know, there's an easier way to do that... Archives: http://www.mail-archive.com/metacard@lists.runrev.com/ Info: http://www.xworlds.com/metacard/mailinglist.htm Please send bug reports to [EMAIL PROTECTED], not this list.
Re: set the name of the last image bug?
The following works in new stacks created under MC and RR, but ONLY works under RR if you have the IDENTICAL script in a stack that was originally a HyperCard stack. ie., it fails under MC 2.4b1, 2, and 3 under these circumstances. SCRIPT: on mouseUp answer file Select image: if ((it is empty) or (the result is Cancel)) then exit mouseUp put it into myFileName import paint from file myFileName set the name of last image to My Chosen Name end mouseUp The point of failure is the set the name of line. So... is there a way to turn off the flag metacard uses to tell a stack originally came from hypercard to see if this clears that problem? Cheers, Mark Archives: http://www.mail-archive.com/metacard@lists.runrev.com/ Info: http://www.xworlds.com/metacard/mailinglist.htm Please send bug reports to [EMAIL PROTECTED], not this list.
