ThreadBoard ArchivesSite FeaturesActiveworlds SupportHistoric Archives |
aw_int_set(AW_MY_Y, 0)? (Sdk)
aw_int_set(AW_MY_Y, 0)? // SdkjoshDec 18, 1998, 11:22am
doesn't seem to wkr and I am calling aw_state_change but all other
zeroes in other functuins work just fine. edward sumerfieldDec 18, 1998, 11:31am
<!doctype html public "-//w3c//dtd html 4.0 transitional//en">
<html> Lets see the code Josh. [View Quote] joshDec 18, 1998, 9:20pm
if (strstr (aw_string (AW_CHAT_MESSAGE), "bot-ground")) {
aw_int_set (AW_MY_Y, 0); aw_state_change (); } [View Quote] > Lets see the code Josh. > [View Quote] edward sumerfieldDec 19, 1998, 2:14am
Not a lot of code there bud. Anyway, let me ask some questions.
How do you know that the strstr is working and it is actually executing the aw_int_set? Have you added a printf or something inside the if statement. Check the return codes of both calls. The doc doesn't actually say that it returns a error code. It just says that it returns an int which maybe an error code. [View Quote] > if (strstr (aw_string (AW_CHAT_MESSAGE), "bot-ground")) { > aw_int_set (AW_MY_Y, 0); > aw_state_change (); > } > [View Quote] joshDec 19, 1998, 2:54am
Yes I added a printf after it, and I used the strstr to execute all my other
commands and they came up working fine... (don't really wanna give out much more code than this cause most of it's secretive till I finish up the bot) [View Quote] > Not a lot of code there bud. Anyway, let me ask some questions. > > How do you know that the strstr is working and it is actually executing the > aw_int_set? Have you added a printf or something inside the if statement. > > Check the return codes of both calls. The doc doesn't actually say that it > returns a error code. It just says that it returns an int which maybe an error > code. > [View Quote] edward sumerfieldDec 21, 1998, 1:32pm
<!doctype html public "-//w3c//dtd html 4.0 transitional//en">
<html> There is not much I can do then. <p>One approach is to start to remove code until your program ends up very simple. Once simple it becomes easy to debug. [View Quote] joshDec 21, 1998, 4:53pm
its the AW_MY_Y not anything else.
[View Quote] > There is not much I can do then. > > One approach is to start to remove code until your program ends up > very simple. Once simple it becomes easy to debug. > [View Quote] edward sumerfieldDec 21, 1998, 5:21pm
<!doctype html public "-//w3c//dtd html 4.0 transitional//en">
<html> Are you saying that can right a simple program and get the aw_int_set... not to work? <p>We are all moving our bots around AW using the same mechanisms so I doubt that it is the sdk. <p>So write a little bot program that initially puts your avatar at 10n 10w, then aw_wait a while and try your two lines of code aw_int_set and aw_state_change. If you can prove that this code does not work then we will have something to look at. [View Quote] joshDec 21, 1998, 8:21pm
it accepts all other alts jsut for some reason it wont move back to 0
[View Quote] > Are you saying that can right a simple program and get the > aw_int_set... not to work? > > We are all moving our bots around AW using the same mechanisms so I > doubt that it is the sdk. > > So write a little bot program that initially puts your avatar at 10n > 10w, then aw_wait a while and try your two lines of code aw_int_set > and aw_state_change. If you can prove that this code does not work > then we will have something to look at. > [View Quote] walter knupeDec 21, 1998, 10:04pm
I remember having similar difficulties when changing the altitude of a bot
and nothing else... i think i managed by resetting the other current coords as well, but now looking at my current bot code, it works just like that.. so this posting doesn't really say much :) Walter Josh schrieb in Nachricht <367EC9D1.C97793D2 at usa.net>... >it accepts all other alts jsut for some reason it wont move back to 0 > [View Quote] edward sumerfieldDec 22, 1998, 2:26pm
<!doctype html public "-//w3c//dtd html 4.0 transitional//en">
<html> According to Roland, every aw_state_change sends all information to the server whether the bot set it or not. So if you only set the AW_MY_Y and to a state change then the bot will move to X=0, Z=0, Y=what you set it to. <p>I use AW_MY_Y in AWCPP when you change the avatars pitch and move forward it angles the direction up into the sky. Works just fine. <p>Josh, did you write the little program yet? [View Quote] x@x.com (xelag)Dec 23, 1998, 1:41am
In my experience, setting the altitude in Hambots and SDK bots is a
buggy business. Moving up doesn't seem to cause problems, moving down does. I'm referring to changing *only* the altitude. In many but not all cases, the browser does not apply the change in altitude, until a horizontal coordinated is changed. You can wait half an hour for the bot to descend, but just change the x or z coordinate 1 cm and down it zooms. Which makes me think that the problem might not be at the server end or in the dll, but in the browser, maybe something connected with gravity? It's there, it could be anything, but don't say it's Delphi's fault... [View Quote] >doesn't seem to wkr and I am calling aw_state_change but all other >zeroes in other functuins work just fine. roland vilettDec 23, 1998, 4:42pm
It's always possible that there might be a problem here, but I'm not aware
of any...I will try to look into this when I get a chance. If the problem is in the browser's avatar rendering code, then you would think it could be duplicated simply by having someone stand in one place and move up and down using +/-, after using the '5' key to completely stop all horizontal motion. If the bug is as Xela describes it, then after some point presumably an observer would no longer see the other's change in altitude.... -Roland [View Quote] decastro@cable.a2000.nl (xelag)Dec 23, 1998, 7:00pm
Roland, we just conducted the following experiment with Canopus, in 3
parts: Present: Canopus, me and my bot Delph. PART 1: Canopus, after stopping all horizontal movement, went repeatedly up and down to 0 altitude using the + and - keys. We didn't observe anything unusual. PART 2: I had my bot move from 0 altitude up 5m: no prob. Then down to 0: althought my UI showed him at 0 altitude (the change had been sent to the dll and aw_state_change() had been called), Canopus and I didn't see him come down. I then moved him forward 1 cm, and yes, both Canopus and I saw him come down to 0 altitude. PART 3: I teleported up to 5m hight, using the menu 'Teleport | To...' option. My bot Delph was following me to my exact position at the time. When going up, no prob: Canopus and I saw me go up and my bot following me and merging into my avatar. When coming down to 0 altitude using the same method, something weird happened: I saw myself coming down to 0, but my bot stayed at 5m altitude. Canopus saw me and my bot still at 5m altitude; he moved around, we stayed at 5m. Then I moved slightly horizontally, causing Delph to move too: my browser showed Delph coming down, Canopus' browser showed both me and my bot coming down to 0 altitude. BTW, this problem was observed while I was not subject to gravity. While subject to gravity, I could not teleport up and stay there: I would inmediately return to ground position. CONCLUSIONS: None really. It's obvious that your own browser updates your position straight away, not through the server. This is also seen when the browser is 'Waiting for server': you can still move around. This explains why I saw myself coming down in PART 3. I also assume the browser sends your new position to the server. Therefore, the problem is either at the server end, or possibly after the browser receives the change from the server. XelaG. On Wed, 23 Dec 1998 10:42:49 -0800, "Roland Vilett" <roland at lmi.net> [View Quote] >It's always possible that there might be a problem here, but I'm not aware >of any...I will try to look into this when I get a chance. If the problem >is in the browser's avatar rendering code, then you would think it could be >duplicated simply by having someone stand in one place and move up and down >using +/-, after using the '5' key to completely stop all horizontal motion. >If the bug is as Xela describes it, then after some point presumably an >observer would no longer see the other's change in altitude.... > >-Roland > [View Quote] joshDec 23, 1998, 7:20pm
well after trying what everyone sai dna dmove it a cm it did reset to 0m
[View Quote] > It's always possible that there might be a problem here, but I'm not aware > of any...I will try to look into this when I get a chance. If the problem > is in the browser's avatar rendering code, then you would think it could be > duplicated simply by having someone stand in one place and move up and down > using +/-, after using the '5' key to completely stop all horizontal motion. > If the bug is as Xela describes it, then after some point presumably an > observer would no longer see the other's change in altitude.... > > -Roland > [View Quote] decastro@cable.a2000.nl (xelag)Dec 24, 1998, 8:41am
Just to confirm Josh's view: it's putting the altitude to 0 which is
buggy, whether thats done upwards or downwards is irrelevant. I hadn't noticed the upward bit before, cause i dont usually send myself or my bot below sea level ;o) {not quite true, I did a lot in Atlantis untill I started S-decaying} I've reconducted the experiments we did yesterday with Canopus and my bot, which confirms what Josh said. XelaG and his XelaGots walter knupeDec 24, 1998, 12:17pm
This explanation fits very nicely to my observations as well...
Walter XelaG schrieb in Nachricht <36821820.70845129 at news.activeworlds.com>... >Just to confirm Josh's view: it's putting the altitude to 0 which is >buggy, whether thats done upwards or downwards is irrelevant. I hadn't >noticed the upward bit before, cause i dont usually send myself or my >bot below sea level ;o) {not quite true, I did a lot in Atlantis >untill I started S-decaying} > >I've reconducted the experiments we did yesterday with Canopus and my >bot, which confirms what Josh said. > >XelaG and his XelaGots > roland vilettDec 24, 1998, 5:47pm
Okay, thanks for all of your hard work on this. From the description of
what you did and what you saw, I'll tentatively conclude that this problem exists somewhere in the browser's avatar rendering code (a nightmarish brew of C++ templates, sigh...) I don't think it is in the server; I looked over all the avatar position handling code in the server and could not find anything that could possibly care whether the Y coordinate is zero or not. -Roland [View Quote] decastro@cable.a2000.nl (xelag)Dec 24, 1998, 11:33pm
Thanks, Roland, for checking this at the server end. I think the bug
and workaround are now clear, and since it does not usually affect the normal performance of bots and other creatures, IMHO it can wait a while, although it should sooner than later be solved. A good holiday and best wishes to all bot makers, users and wishful-thinkers (we're all needed in this universe - and non COF ones as well). On Thu, 24 Dec 1998 11:47:57 -0800, "Roland Vilett" <roland at lmi.net> [View Quote] >Okay, thanks for all of your hard work on this. From the description of >what you did and what you saw, I'll tentatively conclude that this problem >exists somewhere in the browser's avatar rendering code (a nightmarish brew >of C++ templates, sigh...) I don't think it is in the server; I looked >over all the avatar position handling code in the server and could not find >anything that could possibly care whether the Y coordinate is zero or not. > >-Roland > [View Quote] abbot pabisokoJan 4, 1999, 2:18am
I have found problems in my ChattyBot(the deaf guy...*sigh*) also were
pushing the up button and the the down button wont bring the bot to 0 if it was initially on the ground, so I push the down button again and the bot goes underground...but if I do Up, Down, North it goes to the ground [View Quote] |