A clue for using Valentina by XCMD
Can anyone give me some clues about using Valentina from Rev via the XCMD? I've been able to use it via the revdb functions with some limited success, but had some crashes (sadly not been able to pin down) and I think not everything can be done via the SQL interface - for example, I've not been able to create a new database from scratch. So I'd like to dig into the direct interface. I've looked at the documentation, both Rev's and Valentina's, but not got anywhere. As I understand it, Rev ships with the Valentina XCMD already integrated it. Is that right, or should I need to do install anything else to get working? My first - perhaps it's my only - problem is the name of the function. The documentation from Rev is nowhere (that I can find) explicit about this. The documentation from Paradigma suggests that there should be a function called Valentina, eg that there should be possible to execute put Valentina( Version ) When I attempt the above, I get error in function handler. (I also tried it as a handler rather than a function - just get the equivalent, no such handler.) Do I need to do something else to make the Valentina functions available? Or is just a slightly different name? Any clues gratefully accepted. Ben Rubinstein | Email: [EMAIL PROTECTED] Cognitive Applications Ltd | Phone: +44 (0)1273-821600 http://www.cogapp.com| Fax : +44 (0)1273-728866 ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: A clue for using Valentina by XCMD
Hi Ben, Can anyone give me some clues about using Valentina from Rev via the XCMD? I've been able to use it via the revdb functions with some limited success, but had some crashes (sadly not been able to pin down) and I think not everything can be done via the SQL interface - for example, I've not been able to create a new database from scratch. So I'd like to dig into the direct interface. I've looked at the documentation, both Rev's and Valentina's, but not got anywhere. As I understand it, Rev ships with the Valentina XCMD already integrated it. Is that right, or should I need to do install anything else to get working? Yes, but it is not always the latest version (Ruslan is extremely busy in releasing updates, so this is impossible per se :-)... And it is in this subfolder: /Applications/Revolution 2.1/components/global environment/database_drivers/MacOSX/ So you better to put it into the Rev folder to make it work in your stacks. ;-) (This behaviour of Rev has always confused me, since sometimes the sample stacks seem to loose the connection to the appropriate DLLs/External...) Download the latest package here: http://www.paradigmasoft.com/files/vxcmd.html My first - perhaps it's my only - problem is the name of the function. The documentation from Rev is nowhere (that I can find) explicit about this. The documentation from Paradigma suggests that there should be a function called Valentina, eg that there should be possible to execute put Valentina( Version ) See below... When I attempt the above, I get error in function handler. (I also tried it as a handler rather than a function - just get the equivalent, no such handler.) Do I need to do something else to make the Valentina functions available? Or is just a slightly different name? Any clues gratefully accepted. did you take a look at the VXCMD Docs? http://www.paradigmasoft.com/docs/VXCMDReference.pdf I am sure you wil find the answers here... Ben Rubinstein | Email: [EMAIL PROTECTED] Cognitive Applications Ltd | Phone: +44 (0)1273-821600 http://www.cogapp.com| Fax : +44 (0)1273-728866 Hope that helps... Regards Klaus Major [EMAIL PROTECTED] http://www.major-k.de ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Euro symbol woes
Put the htmlText of field 1 into field 2: ยค = curren; OK, what I needed to test was set the htmlText of field 2 to the htmlText of field 1 This, rtfText, and unicodeText all seem to be working now. The original script was get the htmlText of char firstChar to (firstChar+messageLength) of field 1 set the htmlText of field 2 to it and I was getting blanks for the euro in field 2. That must have been due to the font specifier included when I cut pasted the euro symbol from an eMail. Rob Cozens CCW, Serendipity Software Company And I, which was two fooles, do so grow three; Who are a little wise, the best fooles bee. from The Triple Foole by John Donne (1572-1631) ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: FTP queries
Sarah, Should work fine. Just upload the new file with .new at the end of the name, check the length, and if it looks good, do the rename. Here's my rename code. The PWD is there from the debugging I was doing. put libURLftpCommand(PWD, LFtpServer, LFtpUserName, LFtpUserPassword) into theResult1 put libURLftpCommand(DELE renameToFile, LFtpServer, LFtpUserName, LFtpUserPassword) into theResult2 put libURLftpCommand(RNFR renameFromFile, LFtpServer, LFtpUserName, LFtpUserPassword) into theResult3 put libURLftpCommand(RNTO renameToFile, LFtpServer, LFtpUserName, LFtpUserPassword) into theResult4 put libURLftpCommand(DELE renameFromFile, LFtpServer, LFtpUserName, LFtpUserPassword) into theResult5 -- Frank On Jun 21, 2004, at 2:55 PM, [EMAIL PROTECTED] wrote: From: Sarah Reichelt [EMAIL PROTECTED] Subject: FTP queries To: [EMAIL PROTECTED] Message-ID: [EMAIL PROTECTED] Content-Type: text/plain; charset=US-ASCII; format=flowed Hi All, Pursuing my remote notification requirements, I am creating an HTML file which gets re-generated and uploaded to the server every few minutes. However FTP uploads don't always work, but I don't want to be left with a situation where I upload half a file, leaving nothing that is readable by a web browser. A good solution would seem to be to upload the file using a temporary file name, check that the file length is the same as expected, then re-name the temporary file to the display file name. Is this a good way to go - perhaps people with more experience at FTP could suggest something better? If this is a good plan, how do I do the renaming? Will I have to delete the original file first, or will the renaming automatically replace it? ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Splash screen in standalone - too long to show
Thanks Richard...that explains some things. I also have a few substacks in this stack, so it might help to break this up a bit. My expectation was that I have control over the subatcks that are loaded. I suppose the 'code' for the substacks (which is embedded in my splash screen) is being loaded with the splash screen whether it is used or not...about 2 megabytes worth of stuff. When my splash screen opens, I have a progress bar that advances as I manually open each substack, getting it ready for the environment. If I don't do this, the stack is not available...I wonder if there is some other initialization going on for substacks regardless of whether they are used or not? Is the code just being loaded, or are they being opened 'internally', but not open to the user? That would also explain my overhead! Maybe an option to NOT load substack code until it is called for? That would allow me to keep my single file application intact...which is still a unique feature to RunRev AFAIK. JR Date: Sun, 20 Jun 2004 11:46:11 -0700 From: Richard Gaskin [EMAIL PROTECTED] Subject: Re: Splash screen in standalone - too long to show To: How to use Revolution [EMAIL PROTECTED] Message-ID: [EMAIL PROTECTED] Content-Type: text/plain; charset=us-ascii; format=flowed John Rule wrote: I am finding that the fastest I can get my splash screen to show is about 8 seconds...that seems a little long compared to any 'real' app. Unless it's made by Adobe or Macromedia, where the launch times will commonly run between 30 and 60 seconds. ;) I know the RunRev engine has to load, and do it's internal stuff, but this is just on the border of being annoying to the user (me) in my opionion. I'm a big fan of snappy launch times. Made myself somewhat unpopular on a Big Corporate Software Vendor list by raising the issue there, asking if maybe it was time for some optimization (note to self: never mention optimization to engineers forced to succumb to a marketing VP's unbridled lust for featuritis). Seems it's kind of a sore spot with developers stuck using bloated frameworks (ah, the special hell that is developing in a Big Corporate environment; not everyone has it as good as we Rev developers. :) Are there any tricks to getting a stack to display right away in a standalone? Maybe an option in the RunRev engine to do this? A small raw stack (no initialization scripts of any kind) turned into a standalone opens in well under half a second on my modest 1GHz single-processor Mac (PB 4G; Dock set to not animate launches; 768MB RAM). So the engine load time itself seems pretty darn quick. Things that can affect launch time include: - The size of the standalone stack: objects need to be read from disk and unpacked in memory, so bigger stacks will take longer. - Other people's initialization: Are you using any Rev libraries? If not, does removing them from the build process result in a faster launch? - Your own initialization: What does your app do before the line that shows the splash screen? -- Richard Gaskin Fourth World Media Corporation ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: A clue for using Valentina by XCMD
Are you using Windows or OSX. If you are I think you might need to type in the message box of the stack you are using set the externals of this stack to ./VXCMD_Macho_MC so that the stack knows to use the Xcmd (p 4 of the VXCMD docs). You have to quit Rev or purge the stack from memory and then start afresh before the property is registered. If your file path or Xcmd is different then modify the set command accordingly. I think the macho version doesn't ship with OSX Rev so you will need to download it from Paridigmasoft.com. I don't know about windows but I think you need VXCMD_Win_MC.dll. That is what the docs say. If you get the externals set to your stack you should be able to type put valentina(version) into the message box and have the version returned. Michael On Jun 21, 2004, at 7:54 AM, Ben Rubinstein wrote: Can anyone give me some clues about using Valentina from Rev via the XCMD? I've been able to use it via the revdb functions with some limited success, but had some crashes (sadly not been able to pin down) and I think not everything can be done via the SQL interface - for example, I've not been able to create a new database from scratch. So I'd like to dig into the direct interface. I've looked at the documentation, both Rev's and Valentina's, but not got anywhere. As I understand it, Rev ships with the Valentina XCMD already integrated it. Is that right, or should I need to do install anything else to get working? My first - perhaps it's my only - problem is the name of the function. The documentation from Rev is nowhere (that I can find) explicit about this. The documentation from Paradigma suggests that there should be a function called Valentina, eg that there should be possible to execute put Valentina( Version ) When I attempt the above, I get error in function handler. (I also tried it as a handler rather than a function - just get the equivalent, no such handler.) Do I need to do something else to make the Valentina functions available? Or is just a slightly different name? ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
What happens with igame3d?
Does anyone knows happens to http://www.igame3d.com The OpenGL programming system that they implemented shows a lot of possibilities, but it has vanished from their site. I was expecting for detailed documentations but looks like they do not publish it. When we could have a OpenGL canvas in RR? Thanks in advance. al = Visit my site: http://www.geocities.com/capellan2000/ Search the mail list: http://mindlube.com/cgi-bin/search-use-rev.cgi __ Do you Yahoo!? New and Improved Yahoo! Mail - Send 10MB messages! http://promotions.yahoo.com/new_mail ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Revolution gui wrapper
on Sun, 20 Jun 2004 20:54:35 -0300 Andre Garzia wrote: we can always build a python command line tool, a Revolution gui wrapper... this would work on any MacOS X, *nix system (they come with python) and in win32 boxes with python installed. I'll like to read a tutorial about making a GUI wrapper for DOS command line utilities. Compact Pro, an archive utility for macintosh offers a DOS command line tool for decompressing cpt files. Could you show me how to create a gui for this DOS utility? Thanks in advance. al program? = Visit my site: http://www.geocities.com/capellan2000/ Search the mail list: http://mindlube.com/cgi-bin/search-use-rev.cgi __ Do you Yahoo!? New and Improved Yahoo! Mail - Send 10MB messages! http://promotions.yahoo.com/new_mail ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Rulers
RunRev allows one to display horizontal and vertical rulers--when the pointer tool is selected. Has anyone developed rulers that one could display in the stand-alone? Can the scale and/or origin be changed? Looking to benefit from someone else's efforts, Jim ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Splash screen in standalone - too long to show
One thing I just experimented with is an extremely basic hub standalone--that works pretty well My Title Page has a fairly long embedded piece of music, so it takes 10-15 seconds to load off HD, longer off HD. Instead of using this stack as my standalone hub, if I create a minimalist hub consisting simply of a window that says Loading. Please wait. the launch time is minimal off HD (not tested from CD yet) and the user at least has something on his screen letting him know why he's waiting. The other advantage of this is that one does not have to recompile the Title Page each time it is revised. One very basic hub standalone and then everything else, including the title page, can remain in the form of non-standalone Rev files. Marian On Jun 21, 2004, at 10:26 AM, John Rule wrote: Thanks Richard...that explains some things. I also have a few substacks in this stack, so it might help to break this up a bit. My expectation was that I have control over the subatcks that are loaded. I suppose the 'code' for the substacks (which is embedded in my splash screen) is being loaded with the splash screen whether it is used or not...about 2 megabytes worth of stuff. When my splash screen opens, I have a progress bar that advances as I manually open each substack, getting it ready for the environment. If I don't do this, the stack is not available...I wonder if there is some other initialization going on for substacks regardless of whether they are used or not? Is the code just being loaded, or are they being opened 'internally', but not open to the user? That would also explain my overhead! Maybe an option to NOT load substack code until it is called for? That would allow me to keep my single file application intact...which is still a unique feature to RunRev AFAIK. JR Date: Sun, 20 Jun 2004 11:46:11 -0700 From: Richard Gaskin [EMAIL PROTECTED] Subject: Re: Splash screen in standalone - too long to show To: How to use Revolution [EMAIL PROTECTED] Message-ID: [EMAIL PROTECTED] Content-Type: text/plain; charset=us-ascii; format=flowed John Rule wrote: I am finding that the fastest I can get my splash screen to show is about 8 seconds...that seems a little long compared to any 'real' app. Unless it's made by Adobe or Macromedia, where the launch times will commonly run between 30 and 60 seconds. ;) I know the RunRev engine has to load, and do it's internal stuff, but this is just on the border of being annoying to the user (me) in my opionion. I'm a big fan of snappy launch times. Made myself somewhat unpopular on a Big Corporate Software Vendor list by raising the issue there, asking if maybe it was time for some optimization (note to self: never mention optimization to engineers forced to succumb to a marketing VP's unbridled lust for featuritis). Seems it's kind of a sore spot with developers stuck using bloated frameworks (ah, the special hell that is developing in a Big Corporate environment; not everyone has it as good as we Rev developers. :) Are there any tricks to getting a stack to display right away in a standalone? Maybe an option in the RunRev engine to do this? A small raw stack (no initialization scripts of any kind) turned into a standalone opens in well under half a second on my modest 1GHz single-processor Mac (PB 4G; Dock set to not animate launches; 768MB RAM). So the engine load time itself seems pretty darn quick. Things that can affect launch time include: - The size of the standalone stack: objects need to be read from disk and unpacked in memory, so bigger stacks will take longer. - Other people's initialization: Are you using any Rev libraries? If not, does removing them from the build process result in a faster launch? - Your own initialization: What does your app do before the line that shows the splash screen? -- Richard Gaskin Fourth World Media Corporation ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
RE: Revolution gui wrapper
Alejandro DOS GUI wrappers are easy and very useful. I manage a whole NT domain using the NT Resource kit and metacard with these. It sure avoid a lot of dos typing... Open a DOS box, type your command with '/?' to display the help switches. Then for any interesting switch, you build a parameter field or menu. The execute button will assemble the command to type with the parameters do a script like on mouseup get fld server if \\ is not in char 1 to 2 of server then put \\ before server get shell(rmtshare servername \ fld sharename) -- clean and parse the output after... Firsteable, you have to know that the shell function will block MC or RR until the result is returned (if it comes back...) - a solution to this problem is to use (if available) the start command line (try start /? for the many options with this cmd) and you can always dump the output using the piping command character at the end of the command line. or (one to replace the output, 2 to append). for example dir c: myoutput.txt Second, you have to be careful about different commands running on different system (W98, nt4, 2000, xp, 2003 and their corresponding reskits) as they have different inputs and outputs! I use many GUI kits combining different of these IO's to do many many things like user, group, share management much better than MS's own administrative tools... So if you are interested in more let me know. cheers Xavier -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Alejandro Tejada Sent: Monday, June 21, 2004 17:23 To: [EMAIL PROTECTED] Subject: Revolution gui wrapper on Sun, 20 Jun 2004 20:54:35 -0300 Andre Garzia wrote: we can always build a python command line tool, a Revolution gui wrapper... this would work on any MacOS X, *nix system (they come with python) and in win32 boxes with python installed. I'll like to read a tutorial about making a GUI wrapper for DOS command line utilities. Compact Pro, an archive utility for macintosh offers a DOS command line tool for decompressing cpt files. Could you show me how to create a gui for this DOS utility? Thanks in advance. al program? = Visit my site: http://www.geocities.com/capellan2000/ Search the mail list: http://mindlube.com/cgi-bin/search-use-rev.cgi __ Do you Yahoo!? New and Improved Yahoo! Mail - Send 10MB messages! http://promotions.yahoo.com/new_mail ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Splash screen in standalone - too long to show
Hi John, Date: Mon, 21 Jun 2004 07:26:36 -0700 From: John Rule [EMAIL PROTECTED] Subject: Re: Splash screen in standalone - too long to show Thanks Richard...that explains some things. I also have a few substacks in this stack, so it might help to break this up a bit. My expectation was that I have control over the subatcks that are loaded. I suppose the 'code' for the substacks (which is embedded in my splash screen) is being loaded with the splash screen whether it is used or not...about 2 megabytes worth of stuff. That, by itself, doesn't explain it unless there's something you're not saying. If you run init routines in a preOpenStack handler, then I think it's waiting until they are done to show the splash window. If that's the case, move the init routines into an openStack handler of the splash. That way the splash window shows _first_, and _then_ your init routines run, which is what you want. HTH, Ken N. ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: A clue for using Valentina by XCMD
Thanks Andre, Klaus, and Robert (offlist) for your swift replies. For the benefit of anyone else tracking this in the list archives; Rev does ship with the VXCMD (although not necessarily the latest version); but in order to make use of it in your own stacks, you need to set the externals property of the stack and then re-open it. So to get started, set the externals property of a stack to a path to the VXCMD, ie to start with to something along the lines of /Applications/Revolution 2.1/components/global environment/database_drivers/MacOSX/VXCMD_macho Save this stack, close+remove from memory, open it, and either with it as the top stack, or after 'start using' it, one can then at last call the function Valentina, eg put Valentina(version) which on Rev 2.2 returned 1.9.8b2. Hurrah! Secondly, one needs to know that Valentina doesn't seem to do much in the way of error checking; if you make mistakes, you are liable to be punished by Rev immediately crashing. So save early and often. For example, although you can get the version without initialising Valentina, attempting to create a database before initialising will result in immediate crash. Thirdly - the VXCMD_Reference.pdf does not contain a reference to the error codes. These can be found in ValentinaKernel.pdf. Well, that's got me started. Thanks very much. Here we go... Ben Rubinstein | Email: [EMAIL PROTECTED] Cognitive Applications Ltd | Phone: +44 (0)1273-821600 http://www.cogapp.com| Fax : +44 (0)1273-728866 ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Playing quicktimes while pinging
On Jun 21, 2004, at 12:58 PM, [EMAIL PROTECTED] wrote: I'm attempting to use the Shell command in Mac OS X to ping some servers and check their availability. My client wants to play a small looping Quicktime animation while this is going on to mask the latency. Unfortunatly, shell and ping working together freezes Revolution for a few seconds as it's doing it, which in turn, causes the animation to freeze. Is there any way to do these operations at the same time without the temporary freezes? Maybe an animated gif does not have this problem. Maybe you can ping a different way. Perhaps you can make a test open on the service you need. Dar Scott ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Crazy imageData
On Jun 21, 2004, at 3:34 PM, Dar Scott wrote: You say you import image ID 1018. When it fails it is pure black? Do you mean image ID 1018? Then what does image ThumbNail have to do with it? When setting the imageData, the height and width must match the original. (Strictly, the size of the imageData must fit the product of the height and width.) I usually clear the destination image (put empty into the text of image ThumbNail) and set the height and width, first. This will reset the alphaData/maskData, too. I also clear before setting the filename property in some situations. The source image is always fine. It is an 80x60 imported JPG (currently, I've tried other formats.) Image Thumbnail is the target. Basically, the source image is a default state. Thumbnail can receive other data placed into it at various times, but gets returned to the default at other times. There are an arbitrary amount of these thumbnail images, one per data card. So, just setting the ID to the source is not acceptable, I need an independent copy of the image data there. I'll try the emptying trick, using the text of the image. I hadn't gone that route yet. When you dump in your troubleshooting, be sure and include height/width and formattedHeight/formattedWidth. Also alphaData (and maybe maskData). You can also save the image itself. And you can export the image to whatever form you want, including the human readable PBM. OK. Reset the dimensions. Got it. I'll try that too. At this point, I'll try anything. And when it works, I won't trust it, because I've had it working flawlessly before... until it doesn't. *Any* other ideas are still very welcomed. -- Troy RPSystems, Ltd. http://www.rpsystems.net ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Crazy imageData
On Jun 21, 2004, at 3:51 PM, Troy Rollins wrote: I'll try the emptying trick, using the text of the image. I hadn't gone that route yet. The empty trick doesn't work. When you dump in your troubleshooting, be sure and include height/width and formattedHeight/formattedWidth. Also alphaData (and maybe maskData). You can also save the image itself. And you can export the image to whatever form you want, including the human readable PBM. OK. Reset the dimensions. Got it. I'll try that too. At this point, I'll try anything. And when it works, I won't trust it, because I've had it working flawlessly before... until it doesn't. The resize doesn't work. The combination of empty and resize doesn't work. Just to be clear, the imageData of the source NEVER changes. It is an imported image. It sits on a resource card in a one card resource substack of the main application. I must have over 8 hours into troubleshooting this. This is just the sort of thing that will drive you crazy in using Revolution. -- Troy RPSystems, Ltd. http://www.rpsystems.net ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: A clue for using Valentina by XCMD
Secondly, one needs to know that Valentina doesn't seem to do much in the way of error checking; if you make mistakes, you are liable to be punished by Rev immediately crashing. So save early and often. For example, although you can get the version without initialising Valentina, attempting to create a database before initialising will result in immediate crash. Yes, Valentina is optimized for max performance at run time, so there is some crashing when starting to develop. You can set the debugging level before calling init (I think it is the first or second function described in the vxcmd manual). This function allows you to turn on additional error checking in kernel during development (at the expense of performance), although it won't necessarily stop crashing if things are called out of order. The errors are passed back as the result in format error nnn but a few functions return true/false so watch out. Robert Brenstein ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
RE: Crazy imageData
That would avoid using the imagedata and not require you have exactly the same dimensions (the destination image will adjust to fit the incoming image unless the lockLoc is true). Doesn't this technique actually create a link to the source, rather than an independent copy of the image data? No, it copies the data to the destination image, AFAIK. Ken Ray Sons of Thunder Software Email: [EMAIL PROTECTED] Web Site: http://www.sonsothunder.com/ ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
The Case of the Vanishing Insertion Point
Hello everyone, I have a stack where I've layered text entry fields so that I can tab through them in order. The trouble is, when I've tabbed out of the last field in the sequence, the insertion point does not go back into the first. Where did it go? Greg ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
RE: A clue for using Valentina by XCMD
Actually, I've been using and trapping errors in Valentina with MetaCard for a long time and very rarely will it crash MetaCard; perhaps it's an IDE interation thing with Rev? Ken Ray Sons of Thunder Software Email: [EMAIL PROTECTED] Web Site: http://www.sonsothunder.com/ -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Robert Brenstein Sent: Monday, June 21, 2004 2:57 PM To: How to use Revolution Subject: Re: A clue for using Valentina by XCMD Secondly, one needs to know that Valentina doesn't seem to do much in the way of error checking; if you make mistakes, you are liable to be punished by Rev immediately crashing. So save early and often. For example, although you can get the version without initialising Valentina, attempting to create a database before initialising will result in immediate crash. Yes, Valentina is optimized for max performance at run time, so there is some crashing when starting to develop. You can set the debugging level before calling init (I think it is the first or second function described in the vxcmd manual). This function allows you to turn on additional error checking in kernel during development (at the expense of performance), although it won't necessarily stop crashing if things are called out of order. The errors are passed back as the result in format error nnn but a few functions return true/false so watch out. Robert Brenstein ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: A clue for using Valentina by XCMD
For debugging: get Valentina(SetDebugLevel, 3 ) The 3 parameter tells Valentina to log messages from the Valentina Xcmd to the Console program. When Rev crashes, look over at Console and you will see what was happening when you crashed or just to see if the xcmd is doing what you think it should be doing. See page 5 of the VXCMD docs. Michael On Jun 21, 2004, at 11:45 AM, Ben Rubinstein wrote: Thirdly - the VXCMD_Reference.pdf does not contain a reference to the error codes. These can be found in ValentinaKernel.pdf. Well, that's got me started. Thanks very much. Here we go... ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Crazy imageData
On Jun 21, 2004, at 2:10 PM, Troy Rollins wrote: Just to be clear, the imageData of the source NEVER changes. It is an imported image. It sits on a resource card in a one card resource substack of the main application. What do you mean by imported? Do you mean you set the filename? Do you mean you set the text property of the image? Key Ray's idea of copying the image (the text) should work. I wouldn't consider setting the height and width a trick. The imageData has no height and width info. Dar Scott ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
RE: A clue for using Valentina by XCMD
Actually, I've been using and trapping errors in Valentina with MetaCard for a long time and very rarely will it crash MetaCard; perhaps it's an IDE interation thing with Rev? Ken Ray Sons of Thunder Software Email: [EMAIL PROTECTED] Web Site: http://www.sonsothunder.com/ Ken, as you say, you have been using Valentina for a long time. So did I. Just like you, I do not remember when was the last time I crashed it. The thing is that as long-time users, we know the order of calls well. However, it is known that calling certain functions out of order results in crashes and invariably new users fall into that trap. Using a different version of xfcn and the valentina library under Classic will also crash reliably. Robert Brenstein ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Crazy imageData
On Jun 21, 2004, at 3:33 PM, Ken Ray wrote: Can you solve it with a simple put: put img id 1018 of stack Button Icons into image \ Thumbnail of card tClipSelected of stack CurrentP Holy moly. This in fact DOES seem to be working. Ugh. I would have thought this would repoint the ID to use as the image, but then, what do I know? Now... we'll see just how long it works for. ;-) THANKS!! -- Troy RPSystems, Ltd. http://www.rpsystems.net ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Crazy imageData
On Jun 21, 2004, at 9:41 PM, [EMAIL PROTECTED] wrote: From: Troy Rollins [EMAIL PROTECTED] Subject: Re: Crazy imageData To: How to use Revolution [EMAIL PROTECTED] Message-ID: [EMAIL PROTECTED] Content-Type: text/plain; charset=US-ASCII; format=flowed Anything else? This is driving me nuts, as it will appear to be resolved... work for a LONG time (many data transactions), and then suddenly start failing. But basically, it either will work in a run of the program, or not. Once it doesn't, it never seems to work again, until I change something about the code. Then, it works for a while, and then... well, you get the... picture. -- You might consider not using imageData -- you could import/export using temp files instead -- I do it that way, and it's pretty fast, and much less error prone. But best of all you won't be munging with imageData. Fooling around with imageData is a bit like playing with fire -- I'm a bit surprised that they support imageData at all, it's really the one place in RR that you can touch 1's and 0's directly. I would guess that the imageData is a binary data structure (probably a QuickDraw CGrafPort on the Mac), and I can definitely see how it could get screwed up (imagine the height and width not matching the grafPort data), and that after it's been screwed up (however that happens), it stays screwed up. Actually I'm surprised you haven't seen any crashes. -- Frank ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: The Case of the Vanishing Insertion Point
Hi Greg, The focus probably changed to a button or other object on the page.. best, Chipp Gregory Lypny wrote: Hello everyone, I have a stack where I've layered text entry fields so that I can tab through them in order. The trouble is, when I've tabbed out of the last field in the sequence, the insertion point does not go back into the first. Where did it go? Greg ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Crazy imageData
On Jun 21, 2004, at 3:15 PM, Frank Leahy wrote: But best of all you won't be munging with imageData. Fooling around with imageData is a bit like playing with fire One has to do it if one wants to compute images without the bother of building BNP or JPEG. You can export BPM, a much simpler format, but I get nothing when I try to set the image with it. (A candidate for an enhancement?) Dar Scott ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Crazy imageData
On Jun 21, 2004, at 3:18 PM, Troy Rollins wrote: put img id 1018 of stack Button Icons into image \ Thumbnail of card tClipSelected of stack CurrentP This in fact DOES seem to be working. Bugzilla the one that does not work, unless you think one of the imageData bugs already covers it. Dar Scott ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Saving a stack on OSX creates a 'classic' stack
When I save a stack in OSX it is saved as a classic stack. Do I have something configured wrong? Michael ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: A clue for using Valentina by XCMD
I am having trouble getting an application to work. I have a test stack which only has the mouseUp script in a button: put valentina(version) into fld 1. As a stack, when I click the button the version is put into the fld. When I create the stack nothing happens. I think I have the application settings correct. I see the xcmd is in the package (./VXCMD_Macho_MC Untitled 2.app/Contents/MacOS/externals/revdb.bundle). Any clues as to what I am doing wrong? Thanks, Michael On Jun 21, 2004, at 3:42 PM, Ken Ray wrote: Actually, I've been using and trapping errors in Valentina with MetaCard for a long time and very rarely will it crash MetaCard; perhaps it's an IDE interation thing with Rev? ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Crazy imageData
On Jun 21, 2004, at 5:39 PM, Dar Scott wrote: Bugzilla the one that does not work, unless you think one of the imageData bugs already covers it. Looks like #1409 to me. It claims to be fixed, how do I tell which version it is supposed to be fixed for? -- Troy RPSystems, Ltd. http://www.rpsystems.net ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Crazy imageData
On Jun 21, 2004, at 4:09 PM, Troy Rollins wrote: Looks like #1409 to me. It claims to be fixed, how do I tell which version it is supposed to be fixed for? The procedure requires a target milestone to be set, but that sometimes falls in the crack. Based on the date, I would expect this to be in 2.3. I hope the missing targets get fixed soon. Dar Scott ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Revolution gui wrapper
on Mon, 21 Jun 2004 Xavier Bury wrote: DOS GUI wrappers are easy and very useful. This means that every dos program has the potential to become an auxiliar to RR/MC. Right? Open a DOS box, type your command with '/?' to display the help switches. [snip] you can always dump the output using the piping command character at the end of the command line. or (one to replace the output, 2 to append). for example dir c: myoutput.txt I'll have to study carefully the dos shell commands. I'm not a DOS wizard! ;-) Second, you have to be careful about different commands running on different system (W98, nt4, 2000, xp, 2003 and their corresponding reskits) as they have different inputs and outputs! Could you explain us the ways in which these i/o differs? I use many GUI kits combining different of these IO's to do many many things like user, group, share management much better than MS's own administrative tools... This is extraordinary! :-) So if you are interested in more let me know. Of course i'm interested. But i'll like to read a small tutorial with a simple dos application. Understanding the inputs and outputs is crucial to make this works. So you explanations are welcome! Thanks in advance. al = Visit my site: http://www.geocities.com/capellan2000/ Search the mail list: http://mindlube.com/cgi-bin/search-use-rev.cgi __ Do you Yahoo!? New and Improved Yahoo! Mail - 100MB free storage! http://promotions.yahoo.com/new_mail ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Revolution gui wrapper
Alejandro, it's your lucky day! I released a libWrapper some days ago, you can check old threads for the email, is a lib that will help you build simple wrappers for command line tools... I've included two examples, but they are for Unix e Macs, for I know not about win32... I think this lib would help you, I use here for a wrapper for TAR which is a unix archiver, so we can work it's way to Compact Pro! my code is free of charge!!! Cheers Andre On Jun 21, 2004, at 12:22 PM, Alejandro Tejada wrote: on Sun, 20 Jun 2004 20:54:35 -0300 Andre Garzia wrote: we can always build a python command line tool, a Revolution gui wrapper... this would work on any MacOS X, *nix system (they come with python) and in win32 boxes with python installed. I'll like to read a tutorial about making a GUI wrapper for DOS command line utilities. Compact Pro, an archive utility for macintosh offers a DOS command line tool for decompressing cpt files. Could you show me how to create a gui for this DOS utility? Thanks in advance. al program? = Visit my site: http://www.geocities.com/capellan2000/ Search the mail list: http://mindlube.com/cgi-bin/search-use-rev.cgi __ Do you Yahoo!? New and Improved Yahoo! Mail - Send 10MB messages! http://promotions.yahoo.com/new_mail ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution -- Andre Alves Garzia 2004 BRAZIL http://studio.soapdog.org ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
RE: The Case of the Vanishing Insertion Point
Well, if you have a button or other object that has its 'traversalOn' property set to true (as most buttons do when created), it will go to focus on that object. To make it so that ONLY fields can get the focus, turn the 'traversalOn' property of all other objects on the card to false. Ken Ray Sons of Thunder Software Email: [EMAIL PROTECTED] Web Site: http://www.sonsothunder.com/ -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Gregory Lypny Sent: Monday, June 21, 2004 3:43 PM To: Revolution Subject: The Case of the Vanishing Insertion Point Hello everyone, I have a stack where I've layered text entry fields so that I can tab through them in order. The trouble is, when I've tabbed out of the last field in the sequence, the insertion point does not go back into the first. Where did it go? Greg ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: FTP queries
Thanks Frank, this was the sort of thing I needed. Cheers, Sarah Should work fine. Just upload the new file with .new at the end of the name, check the length, and if it looks good, do the rename. Here's my rename code. The PWD is there from the debugging I was doing. put libURLftpCommand(PWD, LFtpServer, LFtpUserName, LFtpUserPassword) into theResult1 put libURLftpCommand(DELE renameToFile, LFtpServer, LFtpUserName, LFtpUserPassword) into theResult2 put libURLftpCommand(RNFR renameFromFile, LFtpServer, LFtpUserName, LFtpUserPassword) into theResult3 put libURLftpCommand(RNTO renameToFile, LFtpServer, LFtpUserName, LFtpUserPassword) into theResult4 put libURLftpCommand(DELE renameFromFile, LFtpServer, LFtpUserName, LFtpUserPassword) into theResult5 -- Frank ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Group anomalies
Howard I can confirm this bug as you describe it. FWIW, I see a fair amount of anomalous behavior with groups in the Rev 2.2 + OS X 10.3.4 combo. I sometimes just try things at random until something works. Quite distressing. Dan On Jun 15, 2004, at 9:56 AM, Howard Bornstein wrote: Using Rev 2.2 and Mac OS X 10.3.4. Create a stack. Make a field and a button. Group the field and button. Select the group. Drag the handles of the group and make it bigger in any direction. The group rectangle changes as you drag it, and the field and button are not affected. How turn on horizontal and vertical scrollbars on the group. Select the group. Drag from the top to extend the border. The group border raises AND the button and field go with it! Now drag the border down. Only the border moves, hiding the button and field. BTW, if you resize from the bottom, the objects are not affected. Call me crazy, but it seems to me it shouldn't work this way. This action only happens with vertical scrollbars on. Same thing happens if you resize the group in the horizontal direction. If you enlarge the group from the left, the object go with the resize. When you move the border back to the right, the objects are not affected. Again, this only occurs with horizontal scrollbars turned on. Can anyone else confirm this before I bugzilla it? Regards, Howard Bornstein --- D E S I G N E Q www.designeq.com ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
RE: Group anomalies
Hmm.. If this is a bug then it's on XP too. What seems to be happening is when the object is resized larger from the top or left the scroll is set to 0 but when resized smaller from the top or left the scroll is incremeted by the amount the group was resized and thus the objects don't appear to move. I guess this begs the question which way do we want it to work? Cheers Monte Howard I can confirm this bug as you describe it. FWIW, I see a fair amount of anomalous behavior with groups in the Rev 2.2 + OS X 10.3.4 combo. I sometimes just try things at random until something works. Quite distressing. Dan On Jun 15, 2004, at 9:56 AM, Howard Bornstein wrote: Using Rev 2.2 and Mac OS X 10.3.4. Create a stack. Make a field and a button. Group the field and button. Select the group. Drag the handles of the group and make it bigger in any direction. The group rectangle changes as you drag it, and the field and button are not affected. How turn on horizontal and vertical scrollbars on the group. Select the group. Drag from the top to extend the border. The group border raises AND the button and field go with it! Now drag the border down. Only the border moves, hiding the button and field. BTW, if you resize from the bottom, the objects are not affected. Call me crazy, but it seems to me it shouldn't work this way. This action only happens with vertical scrollbars on. Same thing happens if you resize the group in the horizontal direction. If you enlarge the group from the left, the object go with the resize. When you move the border back to the right, the objects are not affected. Again, this only occurs with horizontal scrollbars turned on. Can anyone else confirm this before I bugzilla it? Regards, Howard Bornstein --- D E S I G N E Q www.designeq.com ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Group anomalies
On Jun 21, 2004, at 10:18 PM, Monte Goulding wrote: Hmm.. If this is a bug then it's on XP too. What seems to be happening is when the object is resized larger from the top or left the scroll is set to 0 but when resized smaller from the top or left the scroll is incremeted by the amount the group was resized and thus the objects don't appear to move. I guess this begs the question which way do we want it to work? Is this with or without lockLocation of the objects set to true? Dar Scott ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: Group anomalies
On Jun 21, 2004, at 10:18 PM, Monte Goulding wrote: Hmm.. If this is a bug then it's on XP too. What seems to be happening is when the object is resized larger from the top or left the scroll is set to 0 but when resized smaller from the top or left the scroll is incremeted by the amount the group was resized and thus the objects don't appear to move. I guess this begs the question which way do we want it to work? Sometimes I double group to get the behavior I want and lock the inside one and put scroll bars on the outside one. Dar Scott ___ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
Re: iGame3D status 062204
Date: Mon, 21 Jun 2004 08:09:29 -0700 (PDT) From: Alejandro Tejada [EMAIL PROTECTED] Subject: What happens with igame3d? Does anyone knows happens to http://www.igame3d.com The OpenGL programming system that they implemented shows a lot of possibilities, but it has vanished from their site. I was expecting for detailed documentations but looks like they do not publish it. When we could have a OpenGL canvas in RR? Thanks in advance. al Alejandro, Thanks for your interest. I don't have good news. May 18 was the two year anniversary of our work on this project. Tomorrow is the one year anniversary of the first Metacard built release. The website was supposed to be revamped to reflect these events, read along to find out why the site is now a dead zone. For eleven months your name in particular was on my mind to accomplish things with iGame3D that seemed elusive to my mathmetically limited mind, your various vector projects have inspired some interesting iGame3D stacks in the past year. I had my heart set on Judy Perry and her class to be the initial student body that would help develop the roadmap of high level 3D application design education. The licensing logistics of that concept were a bit confusing of course. It was my intent to get a working stack to someone with the ability to educate the masses when the time was right, with the features/UI well formed. At the beginning of May I was in the middle of the fourth GUI rework, seeking the perfect balance of function and form Providing a fully working and user friendly stack was priority number one. I had no intention of licensing out a confusing half finished product. It is now eleven months since we licensed Revolution. This is the iGame3D Feature list as of April 2004 http://www.igame3d.com/features.html Easily a half dozen or more other features were added since that time, including shader support, a method of 3D graphing, a new proprietary model format, a new model animation method, drawing directly into a Rev wind (very slow), keyframe-like timeline abilities, anti-aliasing, alpha rendering I've lost count actually. The iGame3D stack, was going to be the template for educating users in the use of iGame3D's extensive features. Writing documentation is not the driving force of this project, in fact it tends to make us brain dead afterwards. iGame3D exists because we needed a tool for making games that doesn't exist on the Mac. Mac users in general were to be the secondary beneficiaries beneficiaries of the fulfillment of our personal life long goal. Rev users were to get access to a great external in the process of those primary motivators. After eleven months of working non stop on this product, it has been driven into the ground by the faulty IDE we were rented. The iGame3D stack has been pretty much destroyed by Bugzilla Bug 1655 http://www.runrev.com/revolution/developers/bugdatabase/show_bug.cgi? id=1655 Possibly any one of the other 300 or so bugs in the Rev IDE are also culprits. Pop-up buttons stopped working. They sometimes crashed the application on mousedown. Replacing pop-ups with tabbed groups makes an unsightly cluttered interface. Too many buttons was complaint number one of the metacard version. To circumvent that problem, pop ups were removed from stacks, UI re-designed, and extensive menus with submenus were painstakingly developed but these Menu bar menu items vanish. I get menus with random amounts of content, or menus with nothing at all in them, even though they show up 100% in menuBuilder. This breaks opening levels, importing/exporting models, accesing two dozen or more mesh editing commands, and destroys all functionality of the iGame3D console. Revs UI menu items vanish, therefore no saving, no opening stacks, and 90% of the other controls under the REV menus become non existant. This problem has been frustratingly random, several times, across several months I thought the bug had been vanquished only to find myself un-able to save after extensive optimizations and bug fixes. There one minute, gone the next. When reverting to Metacard, the menus functioned perfectly but other elements of the iGame3D external no longer functioned, and eventually Metacard crashed while doing nothing at all. Having switched from Metacard eleven months ago (after only using the demo for less than 2 months) I'm too unfamiliar with it to even begin to guess where in the 3,000+ lines of script to start looking for a fix. Especially if it crashes while trying to figure this out. Instead of getting any help on bug 1655, I received your license has expired, pay up for the latest version. Pay for a version which i've already found out contains the same bugs? Less than twelve months from licensing the first broken version? Are they serious? We don't have an income to feed Runtime, or to tempt others to work on the project with the dedication it
