16 replies to this topic

[huhn_m]

I DID IT!

I just found the bug (or at least the exact location of it) that blocked the release for the Beta4R1.
It is still a problem in the String handling that occurs when converting Integers to strings and when
"adding" strings.

I hope to get this fixed today or tomorrow and then I'll concentrate on adding the keyboard system and the conditional statements (maybe I'll add them before fixing since the bug is not, like I belived before, in the Loop coding). So everything is on course again.

Expect a beta 4 with a first preview of the keyboard system and new conditional statements allowing you to create your first REAL games.

So If everythings works as well as I think you'll have the Beta4R1 by the beginning or the middle of next week!

So happy anticipation!!!

[samy]

congratulations!!!!

You've done a good job.. keep like this during this time i will anticipate!

[SRM]

Great, MLC 2 is reviving !!! Good job. We'll have to make an update of the documentation on the wiki. What will be all commands supported in beta 4R1 ? I ask the question because of the unstability of the 3R1 and because it makes a long time I've not been in (a full sudoku project in basic that isn't even finished (fucking grids generation !!!)).

Did you even corrected the points and lines's bug ?

Don't forget to optimize the code as much as you can.

[huhn_m]

I have nearly corrected the bug. It requires extensive changing of the string handling and the implementation of some kind of "garbage collector". The old model just ate the memory up with each string operation...

About the lines and pixels bug ... I'm not sure. I think they were fixed. If not I'll look into it again. But WHAT pixels bug? I don't remember any?!? Here is my changelog for now:

Added VM Setup menu (AC/On Key)
Added Debugger
Added CodeView to Debugger
Added StackView to Debugger
Added MemoryView to Debugger
Fixed 0 not displaying as white but as transparent
Fixed white printing
Fixed rectangles printing slow/wrong when x0 > x1 or y0 > y1
Fixed mouse bound checking that was missing from previsous release
Fixed mouse_Flag reset that lead to program "remembering" mouse press
Fixed Integer conversion routine not working correctly for 256 - 999
Fixed Bug where pointers were not enforced
Fixed Bug where too many arguments on line were ignored
Fixed Bug where FunctionName could only be the last argument of a command
Fixed Bug with new Color functions
Fixed Bug where color was not correctly restored for an internal handler
Introduced overflow handler, not yet working
Initial support for negative numbers
Marked missing entry check for FOR loops as feature instead of a bug
(especially looking at the neg numbers that would render this check impossible anyways)
Replaced WriteWord with WriteByte in FOR loops. Old command caused wrong addition to be executed resulting in a corrupted stack.
Fixed RET from main routine not leading to termination but needing a prior HALT command.

I noted no other known bug than the string bug ... so maybe it is already solved. After all, the rect drawing relies on the line routine and it works.

As for optimization. ...

I will not optimize the code right now. This would be a waste of resources. Especially the line drawing and rectangle and filled rectangle drawing could be heavily optimized but I don't considder this a priority right now. Main thing is to get a WORKING and USEFULL program. The speed optimisations will definately be in before the final version (1.0-pre0) but before I would not engage in speed comparisions.

And now I'll go for creating the garbage collector ...

[CrimsonCasio]

*rummages for link cable*

[huhn_m]

OK, ... I have a major problem with the string handling (actually only with the joining of strings). There seems to be a complete redesign of the string system necessary.

Has anyone a clue HOW the string joining was solved in the old MLC versions? I would really appreciate any help since this point is really critical to the release. I could also release it without the string joining but this would be useless. I didn't get the garbage collection to really work. I would need another segment for this way to work, but then we have 4 segments and I don't want this because they are not dynamically scalable. And this is not really good since we cound end up with some near empty and some full segments after compiling. This would be a waste of memory or even render the writing of some programs (verry string heavy or integer heavy) nearly impossible.

Please help me. Thanks a lot. (Btw. the debugger rocks )

[samy]

bad luck!!
I was so found of a new release for this week.
I hope you will find an answer to this issue very rapidly, instead of ...
Sorry but I have no skill at all in your problem, but hang on , (as an add says "just do it" )

[Deimos]

Correct me if i'm wrong, but i don't think that strings were even supported by MLC1...

[samy]

is there something new?

[huhn_m]

I'm really sorry for this late reply.
I had a bussy time finding an appartement in dresden to stay at
and organizing some stuff in my life, so I had next to no time during the last
month. This of course delayed MLC2. I hope that I come up with something innovative to fix the problem with the string handling.

However in any case this will require a complete rewrite of the string functions and should take some time. I hope to find this time in the next 2 weeks.

As much as I want to release a beta, I just can not ignore this bug. ...
I really wanted you to see the new perfect debugger and this nice shiny stuff. I coul release a version, if you promise me NOT to mess with string handling. In best case it crashes the calc but in the worst it might reset memories! Tell me if this would be acceptable to you.

I will only look here sparsely so please bear with me if you don't get instant anwsers.

Best regards huhn_m / marcus

[SRM]

You should repair this bug, nobody would like to lose memories.

[huhn_m]

ok. I fixed the static strings with a new model that was quite easy to implement till now.
Also static variables (ones that are not joined etc.) work with the new implementation now.

I hope I get the concat to work this week so that I finally get to release this version. I really can't wait since I did lots of work.

I spend the whole day yesterday and some of the weekend thinking about the new implementation and how to do it best. I hope I don't come over any serious issues because I forgot something ...

[huhn_m]

I know ... the week is over and I DO have the new string model implemented. However, there is still a problem with the internal "concat" function that prevents me from releasing the new beta.

However, this should not be due to a misconception (like in earlier versions) but just a typo or a small bug. I hope to get this fixed soon.

Thanks for your patience

[burntfuse]

Glad to see that this project is still going well! Sorry for my absence, I ended up dropping MLC86 3.0 when it was about 3/4 finished because the size (and lack of speed) of the code was getting way out of hand, and it wouldn't have been able do to half the stuff from the AFX version anyways, so compatibility goes out the window and that made the whole thing sort of pointless. Hope the new programmer does a good job with it... Anyways, it'll be great to see a working version of MLC AFX.

[samy]

what a long time since I came the last time.. oups (the holidays certainly)
good huhm for the new string model, have you managed in correcting the misconception? If it is, you should be happy to get rid of those string...
good continuation

[huhn_m]

hey so many babes, so little time, a friend of mine (called Duke Nukem ) would say. I'm still quite busy in moving my home but once things settle down I have some time left to invest in MLC2.

The string model seems to work fine as testing shows. Only thing from preventing me to release it is that if you add string (+) and the first string is referenced as the last result of a string action the string gets "lost" before it is copied again in the cleaning process. So I need to change the cleaning process a bit but it should not be possible. If you disable cleaning the String memory it will get fragmented extremlely quick (like in the last beta) and will lead to uncontrollable behavious, but it works. So the new string model is fine. So just that you know:

Holidays: 26.8.-9.9.2006 (no internet but maybe I take the notebook with me)
DSL on traffic limit: starting from 28.8. in my old home. my new home will get DSL not before the 1.10. moving to new home 15.9.-26.9.
Uni starting 1.10.2006 (then DSL should be there and everything in my life should be back to normal)

[4nic8]

MLC2 sounds good. Maybe I'll find my afx to test it