|
|
Post by mikiex on Feb 24, 2016 4:02:17 GMT -5
On the Mod 2 vectrex, there is code about for converting YM/AY files that are dumps of registers, though that can take quite a bit of memory. Ehm - yup, I know that one :-)... There is some more source stuff, you may or may not of seen, Its "findable" on google but I don't know if its meant to be shared, I will pm you the link. |
|
|
|
Post by mikiex on Feb 24, 2016 4:17:04 GMT -5
On the Mod 2 vectrex, there is code about for converting YM/AY files that are dumps of registers, though that can take quite a bit of memory. Ehm - yup, I know that one :-)... Lol you should do! I didn't realise you wrote it  |
|
|
|
Post by Malban on Feb 28, 2016 14:30:32 GMT -5
Hello to you all,
I know you are all holding your breath for an update on VIDE :-)!
Here the small things I did the last days...
As mentioned in the other thread I did two new "sections":
- "codi" which holds all kinds of sample code, which can be assembled and run directly from VIDE
- "library" (something held me back calling it libbi :-)) - with documents
Both section must get filled out a bit more... but that will happen in time.
I had contact with Ville Krumlinde, and he was Ok with me including his sources to "Thrust" with VIDE, so that is what I did the last couple of days, making assi more or less "Krumlinde" compatible, which was no easy task.
The sources to Thrust are the best and most complete testcases an assembler can ask for. I think every pseudo opcode AS09 has to offer in every possible permutation was used!
- nearly everything needed was now implemented (apart from very, very few things, for which I rather chose to change Ville's sourcecode, but that was only in two cases)
- implemented are now additionally:
- structs and all that goes with them
- "countable" nops
- cmap
- and fixed a couple of bugs in the original sources (if/macros/ds)
(the only thing I did not implement yet, since even Ville didn't use it: "align")
- to free my head a bit of that, I in between translated the mod2Vectrex sources from Pascal to Java, I haven't run the results on a vectrex yet, but the output pas/java are 100% identical, so I expect them to work alright
- I also started to build "vectrex" projects
- added some dissi comands
- expanded cdissi
- support for bankswitching in dissi (debuging and exporting) and "banked" breakpoints, also codescan is supported for banks
- did some editor stuff (pretty print for ASM sources, indent/unindent for selected lines with TAB... the usual)
The next steps will be:
- look at Flex again, to accomodate all the new "syntax" that came with "Krumlinde" style.
- finish the project stuff
- than I will have another look at the emulator, and look if I can't make it any better...
- after that ... veccy
- then.. we'll see
Regards
Malban
PS
Preliminary docs are updated also once in a while.
|
|
|
|
Post by Malban on Mar 13, 2016 12:47:19 GMT -5
Hi, finally! - Emulation wise I get where I want to be - pretty exact! No fiddling, all running in the same "configuration"... (hehe - and all running (pausing) at the same time too...) Regards Malban PS Finaly came around and made a home for VIDE: Vide And in comparisson, the real thing(s):       |
|
sdw
Vector Runner
Posts: 47 
|
Post by sdw on Mar 17, 2016 10:16:42 GMT -5
This looks amazing! If this gets released, I think it might serve as a good replacement for ParaJVD which I run now (which also is pretty damn nice, don't get me wrong!) Don't know if you are looking for suggestions (I know as a coder that it can get quite annoying with people who constantly come up with "good ideas") - but here's one: For the "tracki"-cycle counting part, as an addition to being able to track code between two addresses, have an option to track time between Wait_recal. So for example in the following code main: jsr wait_recal ... some code .. sometimes: jmp main other times: ..run more code.. jmp main Then independent of which code path I run, I want to know how much cycles I used before running the next wait_recal call. Don't know if that made any sense...  |
|
|
|
Post by Malban on Mar 17, 2016 12:41:30 GMT -5
sdw I think tracki is exactly doing that. Allowing it to measure between two addresses is IMHO exactly what you need. The default values are: Start Tracking: $F1A2 End Tracking: $F192 Taking the start of WaitRecal as "end address" of measurement and the end address of WaitRecal as start address of measurement - measure exactly the time between two calls of WaitRecal :-)! Or did I get you wrong? So you want to be able to measure different "parts" - so to say more than one tracki window? Regards Malban PS Apart from that - you can use dissi to measure other things. There is a "cycle" command, which lets you reset a cycle counter - or show its current value. With step and breakpoints you can measure quite exact all you want... PPS Apart from that - I am open to suggestions - and NOT annoyed. Glad someone shows some interest :-)... 
|
|
|
|
Post by Rapetou33 on Mar 17, 2016 16:32:25 GMT -5
Glad someone shows some interest :-)... Don't worry, a lot among us have a huge interest for what you're making ... I am personnally very interested but as a beginner in coding (no knowledge at all), this remains a bit complex ... :-). I am pretty sure your tools are going to easy my coding but I have to process every info... Please go ahead ! |
|
sdw
Vector Runner
Posts: 47
|
Post by sdw on Mar 17, 2016 17:04:07 GMT -5
I think tracki is exactly doing that. Allowing it to measure between two addresses is IMHO exactly what you need. The default values are: Start Tracking: $F1A2 End Tracking: $F192 Taking the start of WaitRecal as "end address" of measurement and the end address of WaitRecal as start address of measurement - measure exactly the time between two calls of WaitRecal :-)! DOH! Yes, that's exactly how I wanted it to behave, it was just me being stupid and not realizing that the default addresses actually pointed to the ROM stop/start of the waitrecal routine! |
|
|
|
Post by mountaingoat on Mar 19, 2016 11:26:08 GMT -5
Looking better and better, can't wait to try the whole suite.
|
|
|
|
Post by Malban on Mar 29, 2016 17:41:47 GMT -5
Hello again, just another small update... May I remind you, that VIDE also has an "official" homepage, sometimes I do updates even there: Vide HomepageI won't repeat stuff I already posted there... Vide is far from being finished, but I need to get that vaporware feeling of my back and some feedback does not hurt either, so... Here it goes, there is a beta version to download: VideBeta.zip(The zip is password protected, but everything should be well if you remember my name is "Malban") There is no installer in any way, once the zip is unpacked a bunch of directories are created, and "above" all a small jar called "Vide.jar" resides. This is no official release! Coding/documentation and credits are not finished! Use at your own risk!In windows/mac world a double click should start it, if not, than go to the command line of your choice and type: java -jar Vide.jar(an optional -Xmx1g wouldn't hurt though) There is no need for any other binaries or libraries, VIDE is pure Java! Actually I don't know which java version you need to get it working, on every test computer I have Java 1.8, maybe 1.7 does work too. Bug reports, suggestions and so on... post it here, or as pm... or even on the homepage as commentary ('though I might clean that up later). Regards Malban |
|
|
|
Post by thomas on Mar 31, 2016 5:27:00 GMT -5
well, 1.6 is too old  $ java -version java version "1.6.0_65" $ java -jar -Xmx1g VIDE.jar Exception in thread "main" java.lang.UnsupportedClassVersionError: de/malban/VideMain : Unsupported major.minor version 52.0 at java.lang.ClassLoader.defineClass1(Native Method) […] |
|
|
|
Post by mountaingoat on Mar 31, 2016 6:04:24 GMT -5
Malban, this is great, downloading it now.
Seems to be going really slow - if I may suggest, you could probably create a GitHub account which is free if you are sharing your projects and upload it there. They will basically host it for you for free.
Also 1.7 also has the same exception so I think you need 1.8 to run it.
|
|
|
|
Post by mountaingoat on Mar 31, 2016 6:44:47 GMT -5
Ok, downloaded the new Java, tried it.
Malban - this things is FANTASTIC!
The step by step debugging with the symbols loaded from the lst file (which is a very smart idea) is going to make development so much easier.
Also, having all the documentation readily available is very cool.
Register monitoring for the VIA and the AY chip is extremely useful as well.
|
|
|
|
Post by Malban on Mar 31, 2016 16:09:42 GMT -5
Hello - again, just a small notice about two things: - a bug in the current vide beta - cartridge files are saved with absolute paths -> if you don't have my setup, you can't use the starter listing for GCE cartridges... (loading "bin" files with vecxy works ok...) - I created a small first "FAQ", it is now part of the preliminary documentation: VIDE Documentation & FAQRegards Malban |
|
|
|
Post by mountaingoat on Apr 1, 2016 5:55:49 GMT -5
Malban, let me know how you want to go about us reporting bugs. I basically stopped using ParaJVE and using vexci and dissi only.
So I could report any issues I see if you'd like.
|
|
