This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.
Messages - KermMartian
Pages: 1 ... 20 21 [22] 23 24 ... 34
316
« on: June 01, 2010, 05:04:16 pm »
Oooh awesome trick! Nice job on doing this 
Thanks! By the way, since I released Doors CS 6.4 beta, I found that FileOpen/FileSaveAs bug that was making weird hashnoise appear on the screen when you move the mouse, ztrumpet. XLib/CelticIII integration is moving along, so I anticipate making or beating my deadline of (*checks sig*) 18 days 6 hours.
317
« on: June 01, 2010, 12:16:50 pm »
Btw Kerm will the custom Garbage College message work on every calc OS? I am a bit worried because I now remember the following post: http://ourl.ca/4228/78222
Yes, it works perfectly, because it doesn't actually call the Garbage Collect routine itself. Whenever there's a danger of the GC message triggering (for example, when archiving a program), Doors CS installs a GetCSCHook. If the GC menu triggers, then the GetCSCHook fires, which detects that the OS is in the GC context, and forces [2] (perform Garbage Collect) to be pressed. Before the GetCSCHook presses [2], though, it uninstalls itself and installs a LocalizationHook. The LocalizationHook intercepts the OS displaying "Garbage" "Collecting...", and aborts both. On the "Collecting", it draws the window and puts it up, and then uninstalls itself. Therefore, the OS still does all the Garbage Collecting itself, Doors CS just forces Garbage Collecting to occur and prettifies it.
318
« on: May 31, 2010, 04:19:06 pm »
Oooh, I found a bug I believe. try and run a program that has errors, then press quit, and then clear. Then try and turn your calc off.
Yup, this is a known problem, which I fixed in the version that I have. Thanks for that.  Would you mind posting the same in the Cemetech thread, so I remember to make a note of it?
319
« on: May 30, 2010, 11:36:52 pm »
I read over the first few tutorials, and kudos for trying to keep it informative and lighthearted. I gotta say though, I am quite concerned at your avoidance of hex, such as the confusing .org 40339 instead of the universally-accepted .org $9D93 or .org $9d95-2 or .org progstart or .org 9d95h-2. I also take issue with B_CALL_STUFF; I understand if you prefer that form, but you should at least mention that most people use b_call(_stuff) or bcall(_stuff). Miotatsu was having troubles assembling a program after reading your tutorials, though luckily he stopped by and gave me a shout, because he was using the standard include file that uses bcall(_stuff). I'll let you know if I have any further comments when I get to look through the tutorials more thoroughly.
Before I say anything, I really appreciate your comments.
Glad to hear it.  Hexadecimal does come later. However, universally accepted does not necessarily mean "easy." I'm not defending myself as much as asking, "does it really make a difference?" Oh good, I only got as far as 4 so far. No, technically it doesn't make a difference, but I for one filled in a lot of my confusion by staring at existing z80 ASM programs, examining them, toying with them, and reassembling them. I'm afraid if you err to far from the accepted form, your loyal pupils may be confused looking at existing source. I also understand my error with the wrong include file, and thank you for pointing out that it's an error I need to fix. However, in terms of B_CALL with parenthesis, I'm afraid I lack knowledge on other compilers, which is part of the reason why I say in the 4th tutorial that Spasm is the only compiler I will be refering to in the lessons. After all, there's many difference besides simply B_CALL with parenthesis.
Definitely, and you're absolutely within your rights to use SPASM form. I think it might be helpful to acknowledge the standard form that is used with [Brass, TASM, and a few others].
320
« on: May 30, 2010, 11:10:36 pm »
I read over the first few tutorials, and kudos for trying to keep it informative and lighthearted. I gotta say though, I am quite concerned at your avoidance of hex, such as the confusing .org 40339 instead of the universally-accepted .org $9D93 or .org $9d95-2 or .org progstart or .org 9d95h-2. I also take issue with B_CALL_STUFF; I understand if you prefer that form, but you should at least mention that most people use b_call(_stuff) or bcall(_stuff). Miotatsu was having troubles assembling a program after reading your tutorials, though luckily he stopped by and gave me a shout, because he was using the standard include file that uses bcall(_stuff). I'll let you know if I have any further comments when I get to look through the tutorials more thoroughly.
321
« on: May 30, 2010, 05:56:24 pm »
(crosspost from Cemetech) One or two days ahead of schedule, depending on how you figure it, Doors CS 6.4 beta has been completed. This release contains both new features and a host of bug fixed based on all of your helpful reports on Doors CS 6.3 beta. Among the most notable new features is the completed "Homerun" parser hook, which allows BASIC and ASM programs, whether nostub or written for a shell, in ROM or in RAM, and even Doors CS AP files such as Document DE files, to be run directly from the homescreen. This is made possible by a complete overhaul of the program execution system (RunProg) within Doors CS, which some of you may have been following over the last several weeks. Another new feature is much more powerful FileOpen, FileSave, and FileSaveAs routines, which will be a boon for programmers writing programs for DCS7 that work with files. With some heroic assistance from the indomitable Brandon Wilson, I traced some very difficult bugs, and patch quite a number of more minor bugs. As always, I worked on several small tweaks to improve the user experience. It is recommended that you don't use Document DE 6 with this release, as FileOpen and FileSaveAs are unstable under TI-OS 2.41 and 2.53MP. Please give this release a try, and continue to let me know about bugs that you find, features that you'd like, and comments that you have. Doors CS 6.5 beta, planned for June 20th or earlier, will contain a full set of XLib and Celtic III compatibility routines. For more information about my ongoing progress on the project, please feel free to browse around the Doors CS wiki, particularly the Doors CS 7 Scratchwork page. Relevant to this release is information on the Runprog Chaining page. Edit: Updated the zip with a new .8xk, with several high-profile bug fixed based on feedback by Qazz42 and TIFreak8x, as well as an early beta of Document DE 7. Download: Doors CS 6.4 beta
322
« on: May 29, 2010, 08:04:57 pm »
awsome on a side note are you making it so it can support more files on the calc i forgto and i know when i have a large game like dlq on it dcs laggs so bad that its nigh impossible to use X.x
(dlq is about 150 programs i believe)
Yup, if you take a look at this page you'll see it's a pending item on my To-Do list. I'm going to wait until after I have XLib/CelticIII support ironed out before I stare at that though. http://dcs.cemetech.net/index.php?title=Doors_CS_7_Scratchwork
323
« on: May 24, 2010, 02:32:47 am »
(crosspost from Cemetech) Just so no one think I've been slacking, I've been working very VERY heavily over the past week or so, here's what I've achieved (in reverse order) :: Initial successes running ASM programs and AP files from the homescreen. (5/24/2010) :: Repaired cleanup of RunProg chain when no handler is found for an AP file. (5/23/2010) :: Fixed responsiveness issue with shortcut keys ([1] through [6]). (5/23/2010) :: Fixed a minor bug where an unnecessary temporary file would be created and left on the DCS desktop under certain FileOpen conditions. (5/22/2010) :: Repaired bug in canceling out of a FileOpen prompt, properly removes GUI items added for opening. (5/22/2010) :: Fixed the Launch DCS on Poweron bug - one instance of "DoorsCS6" had not been changed to "DoorsCS7". (5/22/2010) :: Fixed three bugs related to custom mouse cursors, including editing the cursor and resetting the default. (5/21/2010) :: Renamed and copied files now end up in the same folder as their original. (5/21/2010) :: Repaired renaming folders. (5/21/2010) :: Repaired two bugs with folders: both hiding folders and pasting a folder into itself are now detected and prevented. (5/21/2010) :: Found final (?) bug in FileOpen/SaveAs, Save remains untested. (5/21/2010) :: Continued debugging FileOpen/Save/SaveAs, also solved an unrelated graphics bug. (5/20/2010) :: Worked on FileOpen/Save/SaveAs, introduced RelocatablePtr1 and 2. (5/19/2010) :: Wrote and debugged routine to move gui7 appvar to highmem. Removed Power menu to save 280 bytes. (5/18/2010) :: Debugging ASM RunProg Chaining functionality, began work on AP GUIs. (5/17/2010) :: Upgraded ASM sections, writeback, and AP sections of RunProg Chaining, performed testing and debugging, found and more items to upgrade. (5/16/2010) :: Upgraded BASIC sections of RunProg, tested, and debugged as per RunProg Chaining. (5/16/2010) :: Finished auditing existing ProgNameSpace usage, designed core functionality of new RunProg Chaining system, and implemented four base functions to manage the chain. (5/16/2010) :: With heroic assistance from BrandonW, fixed an issue where the Instant Goto would fail to find what program needed to be edited on error. (05/15/2010) :: Found and repaired problem with Instant Goto'ing to first line of a program. (5/12/2010) Some relevant Wiki pages: :: http://dcs.cemetech.net/index.php?title=Doors_CS_7_Scratchwork :: http://dcs.cemetech.net/index.php?title=RunProg_Chaining
324
« on: May 14, 2010, 08:20:49 pm »
@ZTrumpet: I suppose it could be a MathPrint issue, but I try to use only my custom DCS text-output routines that are completely independent from the TI-OS. I'll try it with emulated 2.53, though, thanks. @SilverShadow, yes, I'll be implementing large swaths of Iambian's code, (hopefully) with bugfixes. @Kevin: Thanks for the details on the bugs, that should help. Edit: Not sure on your policy on double-posting, so I'll bump this even though it's over 24 hours old. So to update everyone on recent happenstances with Doors CS development. TIFreak and Reapex helped me find some bugs with Instant Goto and the editor, some of which I fixed, and one of which I was only able to repair after heroic OS disassembly assistance from BrandonW. Qazz42 helped me find a few bugs, which I fixed. Schoolhacker found three bugs for me, which are on my To-Do list, but have been pushed back pending my work on the RunProg Chaining system. I completely reverse-engineered my existing routines that deal with program execution and the four ProgNameSpace areas. Today I designed an AppVar to hold the program execution chain during operation and the basic routines to manage the chain, implemented as a stack, then went ahead and actually wrote said routines in 143 bytes. Next up, I need to painstakingly work my way through runprog.asm, apguis.asm, and ap.asm, and replace everything that uses the static ProgNameSpaces with the new chain. Extensive testing will follow, and then implementation of the hook to perform chaining from the RunProg entry point. That will give me an opportunity to add homescreen execution, after which Doors CS 6.4 beta will be released.
325
« on: May 13, 2010, 02:11:31 pm »
Thanks SilverShadow, I caught your post on Cemetech and added the data to the Nspire Compatibility page. DJ Omnimaga, I will integrate all the functions so you don't need XLib or Celtic on your calculator; what are these display glitches? I'm working from Iambian's source, so I'll try to fix any bugs it has, as per Iambian's request. ZTrumpet, I tried very hard to replicate that bug you got, to no success. I did crosspost your post to Cemetech for my future reference though ( here). Was that bug with 6.2 or the new 6.3 beta?
326
« on: May 13, 2010, 03:32:41 am »
That's definitely one of the bug fixes / system updates that's on my agenda, but it will fall after implementing the two major features, arbitrary program depth chaining (read: NoShell-style program execution from the desktop!) and XLib/Celtic compat libraries. I fixed up a bunch of bugs today, including one that tifreak8x helped me find, in which the Instant Goto feature would freeze if the error was on the first line of a program. And please, if you have an Nspire, try the new beta, especially powering off from the DCS desktop. I have reports of 1.1, 1.3, and 1.6 working, and 1.7 and 2.6 failing miserably.
327
« on: May 12, 2010, 07:49:18 pm »
328
« on: May 12, 2010, 04:59:59 pm »
I hope this will fix the errors I had with 6.2. I'm glad this is progressing! 
I'd appreciate if you would stop by that topic on Cemetech and tell me some of the bugs you've run into.  That way I can add them to my list of bugs I need to find and fix, to create as stable a final product as possible.
329
« on: May 12, 2010, 03:15:02 pm »
330
« on: May 05, 2010, 11:15:25 am »
What I was asking was if it you need the seperate program, or will it be built in, like TabFunctions is now? Cuz, if there's extra space on the third app page when Kerm is done, it would be nice to have it built in, if he didn't intend to originally.
Or, what would actually be cooler, but maybe harder to do, is to have the password program work as a sort of installer script that adds the password to the DoorsCS program, in much the same way that you run installers on a computer, and once the program is installed, you never need the installer again (unless you crash).
That's basically what the program I released on ticalc.org and Cemetech is. It installs a hidden program with your desired password encrypted, which runs at Doors CS startup as a Shell Expansion. There's no point in me incorporating it into Doors CS itself, as I personally feel password protection is annoying and useless on a calculator.
Pages: 1 ... 20 21 [22] 23 24 ... 34
|