| *** tpb has joined #tp | 00:00 | |
| *** ChanServ sets mode: +o tpb | 00:00 | |
| *** llnz has joined #tp | 00:12 | |
| llnz | damn! | 00:12 |
|---|---|---|
| llnz | tpserver-cpp sigabrt'ed | 00:12 |
| llnz | humm.... glibc detected double free or corruption | 00:13 |
| llnz | bbs, dinner | 00:28 |
| nash | Anywa... night5 all - heading home | 01:39 |
| *** nash has quit IRC | 01:41 | |
| *** greywhind has joined #tp | 01:45 | |
| *** greywhind_ has quit IRC | 01:45 | |
| mithro | llnz: so what killed tpserver-cpp? | 02:28 |
| llnz | looks like a double free | 02:35 |
| *** llnz2 has joined #tp | 02:46 | |
| *** llnz has quit IRC | 02:47 | |
| *** peres has joined #tp | 02:50 | |
| *** llnz2_64 has joined #tp | 02:50 | |
| peres | oh somebody has connection problems :P | 02:51 |
| *** llnz2_64 is now known as llnz | 02:52 | |
| llnz | always | 02:52 |
| *** llnz2 has quit IRC | 03:04 | |
| *** mithro has quit IRC | 03:05 | |
| llnz | starmapper.... | 03:34 |
| llnz | anyone have jars handy? | 03:35 |
| *** llnz2 has joined #tp | 04:05 | |
| *** llnz has quit IRC | 04:05 | |
| *** llnz2 is now known as llnz | 04:05 | |
| llnz | jotham: are you around? | 04:35 |
| jotham | yeah | 04:35 |
| llnz | i'm going to be in Wellington on Monday | 04:36 |
| jotham | uh oh | 04:36 |
| jotham | where will you be | 04:36 |
| jotham | i work by the council/civic square | 04:36 |
| llnz | City | 04:36 |
| llnz | that's close | 04:36 |
| jotham | if you want to get lunch or something you can txt me or catch me on here | 04:37 |
| llnz | i will probably be free from about 3pm till 5pm | 04:38 |
| jotham | cool | 04:38 |
| llnz | txt would be fine | 04:38 |
| *** TBBle_ is now known as TBBle | 05:01 | |
| *** llnz2 has joined #tp | 05:20 | |
| *** llnz has quit IRC | 05:21 | |
| *** llnz2 is now known as llnz | 05:21 | |
| JLP | morning all | 05:30 |
| llnz | hi JLP | 05:30 |
| JLP | damn i ilmost forgot that i have a presentation to make today, if it wasn't for korganizer... | 05:32 |
| *** greywhind_ has joined #tp | 05:45 | |
| *** greywhind has quit IRC | 05:45 | |
| JLP | lectures await, see you later | 05:49 |
| *** JLP has quit IRC | 05:49 | |
| * llnz wanders off | 06:44 | |
| llnz | later all | 06:44 |
| *** llnz has quit IRC | 06:44 | |
| *** peres has quit IRC | 06:46 | |
| *** mithro has joined #tp | 08:09 | |
| mithro | ~seen llnz | 08:12 |
| tpb | mithro: llnz was last seen in #tp 1 hour, 27 minutes, and 35 seconds ago: <llnz> later all | 08:12 |
| *** peres has joined #tp | 08:32 | |
| *** mithro has quit IRC | 09:13 | |
| *** greywhind has joined #tp | 09:45 | |
| *** greywhind_ has quit IRC | 09:45 | |
| *** peres has quit IRC | 12:33 | |
| *** Erroneous has joined #tp | 12:54 | |
| *** greywhind_ has joined #tp | 13:45 | |
| *** greywhind has quit IRC | 13:45 | |
| *** peres has joined #tp | 14:40 | |
| *** nash has joined #tp | 16:29 | |
| *** greywhind_ has quit IRC | 17:35 | |
| *** DTRemenak has quit IRC | 17:35 | |
| *** kandinski has quit IRC | 17:35 | |
| *** kandinski has joined #tp | 17:37 | |
| *** greywhind_ has joined #tp | 17:38 | |
| *** DTRemenak has joined #tp | 17:38 | |
| *** peres has quit IRC | 17:38 | |
| *** greywhind has joined #tp | 17:42 | |
| *** greywhind_ has quit IRC | 17:43 | |
| *** DTRemenak has quit IRC | 17:43 | |
| *** DTRemenak has joined #tp | 17:43 | |
| *** greywhind_ has joined #tp | 17:45 | |
| *** greywhind has quit IRC | 17:45 | |
| *** mithro has joined #tp | 17:49 | |
| *** jotham has quit IRC | 17:56 | |
| *** mithro has quit IRC | 17:56 | |
| *** TBBle has quit IRC | 17:56 | |
| *** CIA-10 has quit IRC | 17:56 | |
| *** nash has quit IRC | 17:56 | |
| *** greywhind_ has quit IRC | 17:56 | |
| *** DTRemenak has quit IRC | 17:56 | |
| *** Erroneous has quit IRC | 17:56 | |
| *** kandinski has quit IRC | 17:56 | |
| *** mithro has joined #tp | 17:57 | |
| *** greywhind has joined #tp | 17:57 | |
| *** DTRemenak has joined #tp | 17:57 | |
| *** nash has joined #tp | 17:57 | |
| *** Erroneous has joined #tp | 17:57 | |
| *** TBBle has joined #tp | 17:57 | |
| *** jotham has joined #tp | 17:57 | |
| *** CIA-10 has joined #tp | 17:57 | |
| *** kandinski has joined #tp | 17:57 | |
| mithro | nash: no, I'm just being slow to respond | 17:58 |
| greywhind | mithro: fine, but i still need to figure out the cause of the lack of drawing of pie charts | 17:58 |
| mithro | greywhind: I would download the wxFloatCanvas from the floatcanvas website and try some of the piechart demos | 17:58 |
| greywhind | mithro: hmm... not yet. still have some other things to check. | 17:58 |
| *** tuna has quit IRC | 18:00 | |
| *** tuna has joined #tp | 18:02 | |
| greywhind | mithro: should a line like self[oid] = PieChart.PieChart(<arguments>) be enough to get a pie chart to be created and then drawn? | 18:10 |
| mithro | greywhind: yes | 18:11 |
| mithro | (if that code is in an overlay) | 18:11 |
| mithro | oh | 18:11 |
| greywhind | it's in Proportional.py, so yes. | 18:11 |
| mithro | you may need to call self.Canvas.Draw() | 18:11 |
| greywhind | well, when it is done with a FloatCanvas.Point, it works | 18:11 |
| greywhind | but i notice that Point has a def _Draw(self, dc , WorldToPixel, ScaleWorldToPixel, HTdc=None): method | 18:14 |
| greywhind | does PieChart need one? | 18:14 |
| greywhind | hmm... it doesn't since it extends Group, right? | 18:15 |
| mithro | greywhind: right | 18:19 |
| greywhind | mithro: hmm.... | 18:24 |
| greywhind | mithro: if i create Point objects before creating PieChart objects, the Point objects don't seem to be drawn | 18:24 |
| greywhind | mithro: but if i create them after the PieChart objects, they are drawn | 18:24 |
| mithro | you can only assign one item to the value | 18:24 |
| mithro | self[oid] = a | 18:24 |
| mithro | self[oid] = b | 18:24 |
| greywhind | mithro: ah, makes sense | 18:25 |
| greywhind | of course | 18:25 |
| mithro | if you want both | 18:25 |
| mithro | do a | 18:25 |
| mithro | self[oid] = (a, b) | 18:25 |
| greywhind | mithro: so, on a hunch, i replaced the ObjectList.append(ArcPoint(EndXY, StartXY, CenterXY, **args)) call in PieChart with an ObjectList.append(FloatCanvas.Point(XY, 'White', Diameter)), and it draws the points correctly | 18:32 |
| greywhind | so i guess it has something to do with the ArcPoints not drawing correctly | 18:32 |
| mithro | greywhind: what size Diameter are you using? tried fixing it to something like 100 pixels? | 18:33 |
| mithro | you also need to give it a list of more then a single slice I believe | 18:33 |
| greywhind | some of the diameters were about 16.6666666667 | 18:33 |
| greywhind | some 8.33333 | 18:34 |
| greywhind | i gave it two slices. | 18:34 |
| mithro | greywhind: I suggest trying to demo from wxFloatCanvas | 18:42 |
| greywhind | mithro: i guess i will | 18:42 |
| mithro | it could be that PieChart got horrible broken sometime ago | 18:42 |
| mithro | nash: ping? | 18:42 |
| nash | pong | 18:43 |
| nash | mithro: ? | 18:44 |
| mithro | nash: opps sorry, I'll reply via email | 18:44 |
| nash | mithro: cool - meeting in a mninute anyway | 18:45 |
| greywhind | mithro: err... when i download FloatCanvas-0.9.18 rev 25 zip from the FloatCanvas trac, the folders that it says should have stuff in them are empty | 18:46 |
| mithro | ? | 18:47 |
| greywhind | there's nothing in the demos, floatcanvas, etc. folders | 18:47 |
| greywhind | the file is a total of about 24 kb | 18:47 |
| mithro | always choose a tar.gz over a zip :P | 18:49 |
| greywhind | mithro: i did... but that one just creates a symlink to nowhere | 18:49 |
| mithro | greywhind: looks like you are right | 18:49 |
| mithro | the tarballs/zips are screwed | 18:49 |
| mithro | just checkout the code using subversion | 18:49 |
| greywhind | will do | 18:49 |
| greywhind | that works | 18:51 |
| greywhind | mithro: hmm... PieChart demo in FloatCanvas gives this: | 18:52 |
| greywhind | http://rafb.net/p/nDzjta15.html (8 lines) | 18:52 |
| tpb | Title: Nopaste - No description (at rafb.net) | 18:52 |
| greywhind | and there's no Colors.py in FloatCanvas/Utilities | 18:53 |
| mithro | - DefaultColorList = Colors.CategoricalColor1 | 18:55 |
| mithro | - #["Red", "Green", "Blue", "Purple", "Yellow", "Cyan"] | 18:56 |
| mithro | + DefaultColorList = ["Red", "Green", "Blue", "Purple", "Yellow", "Cyan"] | 18:56 |
| mithro | + #Colors.CategoricalColor1 | 18:56 |
| mithro | -from floatcanvas.Utilities import Colors | 18:56 |
| mithro | +#from floatcanvas.Utilities import Colors | 18:56 |
| mithro | dunno why there is no Colors.py | 18:56 |
| greywhind | mmmhmmm.... well, it draws... shapes | 18:58 |
| greywhind | they're not pie charts. | 18:58 |
| greywhind | want a screenshot? | 18:59 |
| nash | greywhind: Everyone wants screenshots - all the time | 19:00 |
| greywhind | nash: yes, but i was hoping for, "no - don't bother, I can see it myself by running the demo." :P | 19:01 |
| nash | greywhind: I can't ;-) | 19:02 |
| greywhind | mithro/nash: http://free.prohosting.com/whind/stuff/piechartsproblem.png | 19:02 |
| tpb | <http://ln-s.net/1:5-> (at free.prohosting.com) | 19:02 |
| mithro | greywhind: :P | 19:03 |
| mithro | looks like the code is borked | 19:03 |
| greywhind | mithro: looks like it | 19:03 |
| mithro | (under MacOS X that is) | 19:03 |
| mithro | works fine here | 19:03 |
| greywhind | but note that the PieChart class in FloatCanvas is totally different from the one in ThousandParsec | 19:04 |
| greywhind | they don't even remotely resemble each otehr | 19:07 |
| greywhind | *other | 19:07 |
| greywhind | well, i'm going to stop for today | 19:09 |
| mithro | so much to do, so little time | 20:02 |
| nash | mithro: See my reply? | 20:04 |
| nash | mithro: Essentially, you are right, but I think it makes as many changes, and is really not that elegant. | 20:06 |
| mithro | nash: yeah | 20:06 |
| mithro | nash: what if you wanted to say "any object which has a move order" ? | 20:07 |
| nash | And I'm not sure if it works. Basically the rule is: You need to supply an object which is of type referred too by this list of references | 20:07 |
| mithro | i'm not sure it works either.... | 20:07 |
| nash | mithro: There is a limit to how far you can go - else you need the server to say "one of these objects" | 20:07 |
| mithro | or "any object which has resource x"? | 20:07 |
| mithro | I would like the ability to click on the starmap with these options - how can we get that? | 20:08 |
| nash | Maybe we should walk before we fly... | 20:10 |
| nash | I think we need to go to code to do such things | 20:10 |
| nash | I thought about the do everything solution as well | 20:10 |
| mithro | well, we already have the "non-starmap click" option - the server could just use the list order argument | 20:10 |
| nash | But then a) we destroy associations in the UI - which sucks - both for the GUI, and for things like AIs | 20:11 |
| nash | b) The list may be too long, and then we need to fragment order frames... | 20:11 |
| mithro | so is there a way to preserve the association in a list? | 20:12 |
| mithro | on a side note, colonise no longer takes an argument on tpserver-cpp's Minisec | 20:12 |
| nash | mithro: Sure... but then what to do with the object list? And if we do that we may as turn coordinates into lists that can only contain 3 items et al... | 20:14 |
| mithro | you are effectively adding a "filter" to the object list right? | 20:15 |
| nash | Yeah, with the default being no filter | 20:15 |
| nash | Basically I think it solves 90% of the problems easily | 20:16 |
| mithro | nash: well, I don't think "no filter" ever actually makes sense | 20:16 |
| nash | General solutions are ugly | 20:16 |
| nash | mithro: A game where I can rename objects? | 20:16 |
| mithro | nash: what does renaming objects have to do with no filter? | 20:16 |
| nash | mithro: Game engines who haven't done any further work | 20:16 |
| nash | How about a move_to_object order... it would take any object as an arg | 20:17 |
| nash | Or at least all but maybe the universe... and that would just move to the center of the universe if you wanted | 20:18 |
| mithro | okay move to/intercept would be the possible time when it makes sense to have no filter | 20:18 |
| nash | In any case, the reason I suggested an empty list meaning "any object" is to make the upgrade path easier... a dumb implementation on the server just appends an extra int, and a simple client just ignroes it (and presents too many choices) | 20:20 |
| mithro | nash: would lists which preserve data about what the items refer to be a better option? - do we need to add that anyway? | 20:20 |
| nash | Basically make it as simple as possible - not for a specific usability case | 20:20 |
| nash | mithro: Only if you want to make lists be the only type for everything. | 20:21 |
| nash | I'd like to keep _some_ context in the protocol - at least for now | 20:21 |
| mithro | okay this is off on a tangent - for example AI's have decent way of telling which options in a "Build Fleet" order refer to which designs right? | 20:22 |
| mithro | bah | 20:23 |
| mithro | let me try again | 20:23 |
| mithro | okay this is off on a tangent - for example AI's **don't** have decent way of telling which options in a "Build Fleet" order refer to which designs right? | 20:24 |
| nash | Yes. | 20:24 |
| nash | So yes, lists which refer to other things would be useful (and in the UI I can click on 'scout' and see the design details for scout) | 20:25 |
| nash | mithro: But that sounds like a massive change - a TP05 level one ;-) | 20:27 |
| mithro | nash: so in the case where an order is going to have only a small set of objects options the list (with references) could be useful? | 20:29 |
| mithro | nash: so in the case where an order is going to have only a small set of objects options the list (with references) could be used? | 20:29 |
| nash | mithro: Yes... | 20:30 |
| nash | And it could even switch between arg types if the list got too long... | 20:30 |
| mithro | hence we don't need a generic system for filtering objects - a simple type based filter would be sufficent for border cases | 20:30 |
| nash | s/border/most/? | 20:31 |
| mithro | s/border/broader/ | 20:31 |
| nash | Exactly my reasoning | 20:31 |
| nash | Well... except for maybe a 'now' - or lets wait and see if we need a generic system for filtering shit later | 20:32 |
| mithro | greywhind: the Colors.py has been commited to subversion now | 20:32 |
| mithro | so I think the steps are then | 20:32 |
| greywhind | mithro: alright... i still don't think the demo helps though, since PieChart is very different from the one in ThousandParsec | 20:32 |
| mithro | 1. Add a object type field to the object parameter | 20:33 |
| mithro | (should that be a list of types?) | 20:33 |
| mithro | 2. Fix list to refer to in-game objects (probably by adding a GRS to each item selection) | 20:34 |
| nash | I think a list of types | 20:34 |
| nash | Eg: an attack order could be to either fleets or planets | 20:34 |
| nash | (or it could be two different orders) | 20:34 |
| mithro | greywhind: yeah, don't know if it is best to use the FloatCanvas PieChart or not | 20:34 |
| mithro | nash: probably a list would be the best bet | 20:35 |
| mithro | the other thing that needs to be fixed about the list parameter is that it should have a "global max" too | 20:36 |
| mithro | I actually don't think the GRS/List of GRS's (as order parameters) really works | 20:37 |
| nash | mithro: I didn't bother to implement them ;-) | 20:37 |
| mithro | nash: nothing uses em yet, so :P | 20:38 |
| mithro | the GRS works quite well on messages to refer to things | 20:38 |
| nash | Yes | 20:38 |
| nash | I need to expand my support to show more links to references | 20:38 |
| mithro | it seems like it would/should work quite well for explain the meaning of a message to the AI too | 20:39 |
| nash | It would just be good for UI stuff | 20:40 |
| mithro | you don't think an AI would get much from the references? | 20:42 |
| nash | It would, I agree with that | 20:42 |
| mithro | in fact I am seeing a connection between giving the AI more information and giving the clients more ability to be smart | 20:43 |
| mithro | guess they are kind of related | 20:44 |
| nash | Yes they are ;-) | 20:44 |
| mithro | btw, I'm not some generic mad person you seem to think I am :) | 20:49 |
| nash | That's what you think... | 20:51 |
| nash | generically | 20:51 |
| nash | mithro: You seem to be very quick to try and generalise stuff | 20:52 |
| mithro | I like to think generically because it lets me see the flaws and strengths of the non-generic stuff | 20:55 |
| mithro | also lets me better understand what we are actually trying to do | 20:56 |
| mithro | well that is my excuse and I'm sticking to it ;) | 20:59 |
| *** greywhind_ has joined #tp | 21:45 | |
| *** greywhind has quit IRC | 21:45 | |
| *** Erroneous has quit IRC | 22:12 | |
| CIA-10 | nash galaxie * r54c203132721 /ewl/gewl_object.c: Handle most of the main arguments. Need to work on cleanup and data. | 23:04 |
| CIA-10 | nash galaxie * r74279b7dfbb8 /ewl/gewl_object.c: Partially working multiple list drop downs. | 23:04 |
| CIA-10 | mithro tpsai-py * rb0df7bff8aef /tpsai-py: Output the universe status at the begining. | 23:08 |
| CIA-10 | llnz tpsai-py * rb038f0313f2e /tpsai-py: Only do the end of turn processing if the end of turn has happened ie, we have seen a async time remaining frame with time==0 and then one with time != 0 (or anything at all) | 23:08 |
| mithro | arg, I have not been productive today at all :( | 23:25 |
| nash | work or TP? | 23:28 |
| mithro | both | 23:38 |
| mithro | I spent a large portion of the morning writing a peice of code to obsfucate some "proprietary information" | 23:38 |
| nash | Oh lovely | 23:39 |
| *** greywhind has joined #tp | 23:45 | |
| *** greywhind_ has quit IRC | 23:45 | |
Generated by irclog2html.py 2.5 by Marius Gedminas - find it at mg.pov.lt!
