Darwinbots Forum
Code center => Suggestions => Topic started by: Griz on December 16, 2005, 08:56:57 AM
-
could we sorta, kinda work out the stuff that doesn't work first?
like day/night, and the leagues, ensuring all the data/settings are
saved and reloaded properly ... or whatever ...
you know ... get a version that works as advertised before adding
still more new stuff?
just a thought.
-
This isn't exactly new stuff.
Ties have always been kind of buggy and have never done what they were designed to do. Overhauling them is more in the vein of getting the existing system working.
Other than that I agree. We need to focus on defects in the existing version(s)
-
This isn't exactly new stuff.
Ties have always been kind of buggy and have never done what they were designed to do. Overhauling them is more in the vein of getting the existing system working.
Other than that I agree. We need to focus on defects in the existing version(s)
yeah, ok.
I know Nums is wired up much the same as I ...
getting off on discovering how to do something new ...
and then moving on to the next big thing.
I understand that very well, believe me. ;)
at the same time ...
we do also have to actually get the stuff discovered
to work in conjunction with what we already have ...
or we don't have an end product that is on any use
to anyone ... ie ... the User.
now I'm just beginning to play with VB ...
and trying to figure out the DB code as well ...
and I wish I could be of more help in resolving some
of the existing 'bugs' or in implimenting the ideas
discovered into the existing program.
I'm afraid it will be a while ...
esp when it keeps changing.
also ... I can see that if/when I start to get a handle on
VB ... it will then be in C++. ;)
I just happen to think that part of this project is being
overlooked or not given the priority it needs to have
to make DB something that people can actually use
and get excited about ...
and out of that coming the support for DB as a real group
effort with lots of folks willing to contribute.
that's all.
-
How bout we not derail a topic about ties with a discussion of priority? (That can be another topic if you want to talk about it).
The most basic test bot would be some sort of jet propulsion bot like I tried to make months ago that moves by a butterfly kick. I think that's what would be cool looking.
how about you actually address it somewhere?
-
Basically everything in 2.4 should be more or less functioning. If something isn't, you should point to it so I can fix it. However, unless something is a real deal breaker, I'd prefer to work out issues with the C++ port and get that finished up so I can start on metabolism issues, env grid, etc.
Also keep in mind that programming takes time. Alot of time. Lots and lots of time. I devoted about every waking free moment I had to Darwinbots from about the beginning of October to the end of November, to give you the idea of time to progress ratio.
So if something takes a while to happen, relax, and either help me do it or sit back and wait.
Since I'm the one donating my time I get final say on what I work on. That's not to say that I won't take input from others. I just have a clear vision of what I want to work on and that's what I follow. Only if you change my vision can you change my priorities. If you're unhappy with either the pace or direction of upgrades, feel free to suggest new changes in suggestions forum to get discussed and program them in yourself. It's open source for a reason.
Basically, if you pardon my French, the mass generalizations used to bitch are getting old. Things like "we do also have to actually get the stuff discovered
to work in conjunction with what we already have ...
or we don't have an end product that is on any use
to anyone ... ie ... the User." do not tell me what it is specifically you're at issue with, so I don't know how to correct it myself. And you're either unwilling or unable to do it yourself. Hence hurt feelings and alot of tire spinning.
-
see I had the mistaken impression that DB was a group project ...
kinda like some open source project where people could brainstorm
and share their ideas and contribute to where it was going.
that's where I went wrong.
I do understand that control of information is power ...
and I haven't seen you exibit much interest in what anyone else
has to say or any willingness to allow anyone but yourself to
determine the course of DB.
you have your own vision/agenda and that's all that matters.
you the man ...
it's your show.
good luck
-
Your damn passive aggressive attitude is hella annoying.
I am not stopping anyone from adding something to DB. If you want it to do something it doesn't, tell us and either do it yourself or wait and shut the Hell up. That's what it boils down to.
And so far I've done every specific thing you've requested. "Fix the bugs" you said. I did.
Darwinbots is not a Democracy. It's an incredibly fascinating governing system based on interest. People who are more interested in the program have more say in how it develops. I guess it's Interestocracy. It's how all collaborative efforts sort themselves out. People who have the talent, time and interest to be hardcore form the top power users, and users of less interest form the basic user base. The hardcore users are the ones who shape the project.
Your interest does not even need to extend so far as either writing bots or editing the code. shvarz has had probably a larger impact on how DB developed than most other users.
He's done this by working with us on shaping and refining new features. Giving us a clear biological base we can collaborate with.
So far all you've done is bitch. Alot.
I could see alot of potential use coming from your wiki knowledge. If you'd spend half the time working on developing the wiki (specifically DNA pages, etc.) that you spend on bitching, you'd be more respected in the community and your voice would carry more weight.
"But I don't know that stuff" I can hear you moan. Guess what, most of them are outlined in the DArwinbots included help and readme files. And what you don't know you can ask.
-
I don't see this as anything to do with vision or agenda.
Like Nums said, it is about time and priorities.
For instance, I can only devote significant time to fixing problems with 2.37 when somebody reports a specific problem. Vague issues with file saves and stuff, while annoying, are not really worth my time to do when I have so many other things going on and when a whole new DB version will be coming out soon anyway. I will get around to it eventually though as I would really like to see the VB side of the project kept alive even after the C++ version is released.
You guys wanted a stable version.
As far as I know, 2.36.7 is now that stable version since nobody has reported a crash in weeks.
I have not worked on 2.4 at all and I don't really think Nums will be doing anything significant with it in the near future.
The C++ version is what we are all dying to see. It should be faster and more stable than any other to date.
DB IS open source. and IS a group project. However Nums and myself are the only people who have actually done any of the programming since Carlo bowed out after 2.11 (mostly Nums lately). Others have helped on the wiki and that is greatly appreciated. Others have contributed knowledge that has helped to shape DB into something potentially ground breaking.
Anybody can have an input. Just try to make it constructive. please.
-
People who have the talent, time and interest to be hardcore form the top power users
yes. that comes thru Numbs, loud and clear.
but you see ...
there is also this idea of bringing others along ...
not excluding those who are not as talented as yourself ...
those who make up the base of support for the success of any program ...
the end Users. the general public, the 'commoners'.
if you want this to be anything other than a pet project for your own
programming amusement ... you have to include them.
[how many users do you figure DB really has?]
and I tell you my friend ...
you aren't worth a shit when it comes to communicationg/interacting
with those folks. that is my own direct experience anyway.
now my talent lies in being a 'troubleshooter' ...
in finding 'problems' and 'shortcomings' ... in any system.
I'm a systems control dude ... that's what I know how to do ...
what I get paid very well for doing in the real world ...
and I'm pretty good at it.
part of that is the writing of procedures to guide those who are not
familiar through the steps they need to take to make something work.
ie ... the User/Operator ... to bring them up to speed, so to speak.
now in the world I live in, I also know enough about the details that
I can also come up with the solutions to the problems I sus out.
in this case, with VB, I don't yet have the expertise to go in and isolate
or tell you preciesly what needs to be changed ...
I don't yet know how to make sense of your code.
you wrote the code, know what it means, what it does, and where it is.
I do not.
and therefore you find it easy to blow me off because I can't point
directly at where it is messed up ...
as if there are no problems that actually need to be addressed.
I get that. believe me, I get that.
you, PY ... a few others, including Bots, btw ...
do have the expertise to actually address and make the changes ...
but you first have to be open to seeing that they exist ...
and willing to do something about it.
and I'm sorry ... I just don't see much interest in this area rom you.
you have your own agenda/priorities.
so tell you what I'll do ...
I'll report in as much detail as possible ...
the bugs if/when/where I find them ...
and do what I can at the wiki to help those coming in cold to at least
see if they can get a version working so they don't throw up their
hands in frustration and give up.
that's not easy, as I still have a lot of trouble myself but ...
will at least pass on what I find as I find and resolve it ...
so not everyone has to reinvent the wheel to get DB happening.
am working on getting the full version of PY's Bot tutorial there ...
and Endy is working on documenting the DNA commands ...
and whatever else a 'new user' just coming in might find to be of use.
so you go ahead and program your ass off ...
and don't worry about the rest of us ...
except to please address what specific problems you can should I
or other bring them up.
we aren't making this stuff up you know.
for a start:L
seems Testlund has reported, more than once, that Day/Night doesn't work
in 2.4.A ...
and I can't get Leagues to load without crashing.
I can't tell you 'why' ...
just that it crashes.
if I can get it to write an error.sim when it does ...
will pass it along.
btw ...
'true virtue fears no critique' ;)
-
Can you get leagues to work in 2.37?
-
There you go, that's much better.
I'm pretty sure I fixed Day/Night cycles for 2.4, which is what became 2.4.A, but it's been so long since I released a version I don't remember to be honest. The version I have works, but it may be a reworked 2.4.A. Not sure.
Leagues seem to not crash in my version. I'm not saying they work 100% mind you, but they don't seem to crash.
-
Day/night stuff works fine in 2.37.
The only exception is that I don't think it is recorded in the save file.
-
I think the leagues have some problem with the ranking order changing. Haven't used them much latly so can't say for sure.
Do like the league table's background, I've been using it as my computer's background picture. ^_^
-
Me too actually ;)
-
It is probably well and good if you make this C++ version, Nums, but what concerns me is are you just porting over everything from 2.4 as it is, or will you fix all the shortcomings we've been nagging about here? I managed to run a pretty interesting sim the other night, but because it was night the whole time (wich I suspect it was) the vegs never reproduced. And it annoyed me too much that the cycle counter and moon resets whenever I click 'change' in the GUI. It worked perfectly in 2.37, why did you mess with THAT code? Couldn't you just copy over that code from 2.37.6 into 2.4.A? If you could just FIX this little tidbit we will have a version that runs acceptible, something to play with until you release the C++ version. PRETTY PLEASE!!! :pray:
-
Everything's not as cut and dry as you would make it seem. I didn't mess with day/night cycles. If it works in 2.37 and not in 2.4 it's probably a case of one part of the code affecting another part adversely when it shouldn't. The code is really complicated.
I'm certain that the day/night cycles thing is a GUI bug only. Well, I don't know what happens when you change something in options, that might reset it. I thought we went over this already? It takes one full cycle before you see the sun, remember?
I'm rewriting the code as I go, making the program somewhat object oriented, which serves to isolate parts of the program from effecting other parts of the program so directly, which should help with alot of bugs.
-
I am certain it is ALLWAYS NIGHT in 2.4.A. When the moon is up, it is night. When the sun is up, it is night, because of the behavior of the veggies. I've followed the suggestions to make the veggies reproduce, but they behave like they are sleeping and waiting for the day to appear, the day that never comes. It is an atomic winter!
-
Have you checked if they're getting energy directly by looking at their robot info screen?
I redownloaded 2.4.A from the FTP, ran a sim with day/night cycles, and lo and behold veggies get energy during the day but not the night.
I just honestly don't understand why you're saying this is broken. It works fine for me.
-
try making a veggie that behaves different;y in day and night.
use *.daytime to detect if it is day (day returns 1 and night returns 0)
I have a veggie for pond mode that floats up to the surface in the day and descends to a depth of about 5000 at night. Can't remember if I posted it to the bestiary ever. If not then I will try to remember to do so tomorrow as I don't have it on this PC.
-
I'm jellous that you manage to develop such advanced bots! Grrr! Anyway...I think I have found a workaround for my problem. Increasing the energy/cycle for the veggies seem to make them reproduce better. I never thought of that before, because in earlier versions it whould crash the program with lots of energy for the veggies. In 2.37.6 the veggies reproduce like hell with energy set to 10, in 2.4.A I have to set it to 100 to make them stand a chance. Later though when the bots develops to a larger population I need to increase the energy even more. I'm just wondering how much I can increase it before the program crash in this version.
So... I GUESS that if I have day/night cycles turned OFF, then it will INDEED be daytime all the time, but the MOON will show! Turning day/night cycles ON will reset to NIGHT and then the SUN will show after the set amount of cycles. I must say this program is not user friendly at all! You should put a sticky in the newbie section that warns people from using this if they are not programmers. This is very complicated. I think we've been through before that it whould be nice with a DEFAULT settings file for THIS version so one doesn't have to rip the hair from the skull in frustration trying to figure out how to run this program. :pokey:
Anyway, at least I can finely run a simulation that works. I've been running a simulation through the whole night and day with no crash and the bots behave nicely. :)
Pfew!
-
I'm jellous that you manage to develop such advanced bots!
Advanced? :blink:
Alga Stratificus is a tiny little thing that I threw together in about a minute. it only has 3 simple little genes and does nothing but float up and down, and reproduce.
Did you check it out in the bestiary yet?
-
If you spend a bit of time learnign the program you can manage to create some fairly sophisticated bots.