01:32.48 | *** join/#bzflag katyusha (~quassel@unaffiliated/katyusha) |
03:04.08 | *** join/#bzflag mebigfatguy (~dave@128.177.74.185) |
03:24.20 | *** join/#bzflag Constitution (~josh@bzflag/developer/Constitution) |
03:37.24 | Constitution | I want to start working on a native mac platform library |
03:37.43 | Constitution | is the new git repo ready to receive in it's current state? and what would be the best place to work on that as far as branches, etc.? |
03:41.53 | mebigfatguy | one would think you would clone the repo and work on it there |
03:43.12 | Constitution | I'm not sure working on it in a private repo and then dumping it back into the main repo is the best idea, considering the size of the project |
03:45.50 | *** join/#bzflag Slinkwyde (6363e107@gateway/web/freenode/ip.99.99.225.7) |
03:47.14 | mebigfatguy | ?? really?? thats pretty standard github workmoel |
03:47.19 | mebigfatguy | model |
03:47.29 | *** join/#bzflag katyusha (~quassel@unaffiliated/katyusha) |
03:49.32 | Slinkwyde | Constitution, are you there? |
03:51.54 | blast007 | Constitution: the 'bzflag-import-3' repo with the v2_4_x branch, I'd imagine |
03:56.40 | Slinkwyde | I'm one of the people who's been having trouble running BZFlag on OS X 10.9 Mavericks. Constitution has been making test builds blindly because he can't reproduce the issue on his own hardware. I'm trying to contact him so that he can remote into my computer screen using TeamViewer. |
03:57.58 | blast007 | just prior to you joining, he had said "I want to start working on a native mac platform library" |
03:58.15 | Slinkwyde | I saw him post on the forum about 35 minutes ago, but I guess he's afk now. |
03:58.44 | Slinkwyde | Yes, I saw him mention that in the forums and I think that is an awesome idea. |
04:00.23 | blast007 | I'd also be curious if SDL 2 would solve the issue |
04:00.40 | JeffM_ | <PROTECTED> |
04:01.28 | blast007 | certainly less involved (with clear migration documentation) and with benefits for other platforms as well |
04:02.16 | JeffM_ | yeah |
04:03.14 | JeffM_ | there is the old native mac code in history, so there is a starting point there, but sdl2 is not a bad idea |
04:03.33 | blast007 | I'm sure it's less of a trivial migration for the native mac code |
04:03.39 | JeffM_ | yeah it would be |
04:03.49 | JeffM_ | SDL2 should only take a few hours |
04:11.58 | *** part/#bzflag katyusha (~quassel@unaffiliated/katyusha) |
04:21.07 | Constitution | so just right into the mainline code? |
04:21.39 | JeffM_ | I'd do SDL2 on a local fork then merge it when it works |
04:21.53 | JeffM_ | it should not require major changes |
04:23.09 | JeffM_ | if you want/need help from others we can pull your fork and work on it with you |
04:23.28 | Constitution | SDL 2 would, of course, be a multi-platform effort |
04:23.50 | JeffM_ | yeah, get the stuff working on OSX first |
04:24.06 | JeffM_ | to at least see if it fixes the problem |
04:26.35 | JeffM_ | you can do it in the main repository on github if you want, but you can still do it in a branch |
04:26.50 | JeffM_ | making then merging forks is the DVCS way |
04:36.48 | Constitution | there are more issues to resolve than just the mavericks crash? the current mac binary uses a patched SDL to avoid a crash on exit, and iconify still doesn't work, and there was some strange FPS/input issue when I tried to update to a more recent SDL 1.2 |
04:37.43 | Constitution | I was a little worried about older OS support for SDL 2, but it looks like it supports all the way back to 10.5 |
04:38.01 | JeffM_ | that's less of a concern for us |
04:38.16 | JeffM_ | do you think SDL 2 has the same problems? |
04:38.24 | JeffM_ | or are you just weary of SDL? |
04:38.45 | Constitution | more the latter? but I have to agree it's probably worth a shot |
04:39.12 | JeffM_ | it should take less time |
04:39.28 | JeffM_ | and if we learn that SDL2 is not as stable as we need, then we should not use it for other platforms ether |
04:39.38 | Constitution | a native platform library would be more challenging in several ways? primarily because cocoa apps don't really have a main application loop |
04:39.46 | JeffM_ | yeah |
04:39.52 | JeffM_ | it'd require much larger code changes |
04:40.24 | JeffM_ | at that point we should probalb just refactor the entire platform system to make it easy to do all modern platforms |
04:41.05 | JeffM_ | cocoa apps should be able to fire up a message pump |
04:41.06 | Constitution | windows doesn't use SDL already, correct? |
04:41.19 | JeffM_ | yeah mostly just because of joystickness |
04:41.41 | Constitution | force feedback and such? |
04:41.49 | JeffM_ | yeah |
04:42.08 | JeffM_ | breaking platform up a bit more would let all platforms use SDL for window and sound |
04:42.49 | JeffM_ | and letting windows use native for input and the like |
04:43.40 | JeffM_ | but I think an easy thing to do, at least for a 2.4.x compatable change would be to try updating to SDL2 and see how it works |
04:43.58 | JeffM_ | or at the very least build a test app using SDL 2 that you can use to see if it has problems |
04:44.22 | Constitution | ok |
04:45.06 | JeffM_ | the main problem I see with going right to native is that there are fewer people to maintain it then there would be with SDL2 |
04:47.30 | Constitution | SDL on mac still has it's intricacies? and I've really been solo on mac development for several years |
04:47.38 | Constitution | but yes, it could be easier |
04:48.55 | JeffM_ | start a test app usng SDL2 to see how different it is then normal stuff |
04:49.10 | BulletCatcher | An SDL2 branch off of the v2_4_x head in the bzflag-import-3 repo would be a fine way to go on this. |
04:49.12 | JeffM_ | the idea behind SDL is that the same code runs on all platforms |
04:49.49 | JeffM_ | I think SDL 1.x on the mac was a bit of a hack, I'd hope they redid it propery for SDL2 |
04:52.20 | Constitution | it is different in a few regards? I know you don't need extra files that replace the main function anymore |
04:53.20 | JeffM_ | try it and see |
04:53.23 | JeffM_ | I' don't know |
04:53.32 | JeffM_ | I can barely use xcode :) |
04:53.52 | Constitution | I already went that far and discovered the difference a few weeks ago? so, we'll see |
04:53.55 | JeffM_ | you fear SDL2 so much you haven't even looked at it? |
04:54.26 | Constitution | I did look at it? but it's not a drop-in replacement |
04:54.46 | JeffM_ | yeah I know that you have to make changes |
04:55.06 | JeffM_ | but it would not suprize me if you need a different main, apple and ms are wiered that way |
04:55.18 | JeffM_ | at least MS lets your main be in C |
04:57.30 | Constitution | so 2.4-compatible code goes into v2_4_x branch, and development goes into master? |
04:57.47 | JeffM_ | dev goes into 2.6.x branch |
04:57.52 | JeffM_ | master is empty |
04:58.01 | BulletCatcher | 2.4 compatible goes into 2_4_x, and breaking stuff goes into 2_6_x. |
04:59.18 | Constitution | is it still at https://github.com/BZFlag-Dev/bzflag/tree/master ? I don't see a v2_6_x, and master does not appear empty |
04:59.33 | JeffM_ | no it's the import-3 one |
04:59.35 | BulletCatcher | Go up to the bzflag-import-3 repo. |
04:59.49 | JeffM_ | I should probalby just lock the normal one |
05:00.17 | JeffM_ | BulletCatcher, is there any reason to keep the bzflag one? |
05:00.55 | BulletCatcher | For now, let's keep it for historical purposes. |
05:01.05 | BulletCatcher | Maybe we could rename it to bzflag-import-1 |
05:01.44 | BulletCatcher | If we do that now it will be easier to name the final version just "bzflag". |
05:01.53 | JeffM_ | done |
05:02.29 | JeffM_ | I put a comment in the description too |
05:05.39 | BulletCatcher | https://github.com/BZFlag-Dev/bzflag-import-1/wiki/SDL-2.0-and-Platform-Rework has some notes relevant to recent discussion. |
05:05.53 | Constitution | thinks we need a higher-quality icon for the github page |
05:06.17 | JeffM_ | yeah, for an inital pass we can just do what we do now |
05:07.14 | JeffM_ | sadly you can't just set the icon, it's tied to gravatar |
05:18.20 | *** part/#bzflag Slinkwyde (6363e107@gateway/web/freenode/ip.99.99.225.7) |
05:22.55 | *** join/#bzflag delusional_ (~delusiona@unaffiliated/delusional) |
05:25.00 | *** join/#bzflag Gnurdux (~gnurdux@pool-108-48-227-29.washdc.east.verizon.net) |
06:35.03 | *** join/#bzflag DTRemenak (~DTRemenak@76-227-151-154.lightspeed.irvnca.sbcglobal.net) |
06:35.03 | *** join/#bzflag DTRemenak (~DTRemenak@about/essy/CrazyCoder/DTRemenak) |
06:35.03 | *** mode/#bzflag [+v DTRemenak] by ChanServ |
06:40.25 | Constitution | looks like SDL 2 has much better joystick support on all platforms |
06:56.25 | JeffM_ | yeah it's supposed to be better |
07:04.40 | *** join/#bzflag ashvala (~ashvala@unaffiliated/ashvala) |
08:07.13 | *** join/#bzflag ashvala (~ashvala@unaffiliated/ashvala) |
11:34.18 | *** join/#bzflag Pimpinella (~frank@gondolin.pimpi.org) |
12:11.34 | *** join/#bzflag ashvala (~ashvala@unaffiliated/ashvala) |
16:10.06 | *** join/#bzflag ashvala (~textual@106.51.128.187) |
16:10.38 | *** join/#bzflag ashvala (~textual@unaffiliated/ashvala) |
16:37.59 | *** join/#bzflag TimRiker (~TimRiker@bzflag/projectlead/TimRiker) |
16:37.59 | *** mode/#bzflag [+o TimRiker] by ChanServ |
16:49.57 | *** join/#bzflag ashvala (~textual@unaffiliated/ashvala) |
17:32.48 | *** join/#bzflag JeffM (~JeffM@107-209-61-105.lightspeed.irvnca.sbcglobal.net) |
17:49.44 | JeffM | BulletCatcher, is there any reason to keep import 2? |
18:07.43 | BulletCatcher | Not at this point. |
18:07.51 | BulletCatcher | I have a clone of it, just in case. |
18:11.53 | JeffM | time to head into the danger zone then |
18:14.33 | JeffM | what do we want to do with the other stuff in the old SVN trunk, bzedit, bzauthd, and the like |
18:15.09 | JeffM | leave it on SF or make a github repo called "archive" and keep them in there as branches? |
18:25.56 | blast007 | JeffM: those may be trivial to convert to individual git repos as I'd imagine most of them have been fairly self-contained |
18:26.07 | JeffM | that's an option too |
18:26.46 | JeffM | It'd be nice for the bzworldGen to get moved to it's own, I think it's valuable code |
18:36.07 | BulletCatcher | It isn't trivial to put the other stuff into their own repos, but that's already a feature of the svn->git conversion work I have done. |
18:37.51 | BulletCatcher | One of my goals is to be able to account for all of the Subversion revisions in the Git repos. |
18:39.29 | blast007 | BulletCatcher: trivial compared to the crazyness that is the 'bzflag' repo |
18:39.53 | BulletCatcher | :-) It's all done by the same script. |
18:40.11 | BulletCatcher | It turns out that completeness isn't necessarily compatible with a clean set of Git branches, but there aren't too many Subversion revisions that translate poorly. |
18:40.46 | *** join/#bzflag mebigfatguy (~dave@128.177.74.185) |
18:42.19 | BulletCatcher | I suggest putting everything in the following Git repos: bzflag (game code), bzworkbench, bzflag-web (trunk/web + trunk/admin), and bzflag-misc (everything else). |
18:42.54 | JeffM | sounds reaslonable |
18:43.33 | BulletCatcher | Everything else is: bzauthd, bzedit, bzeditw32, bzstats, bzwgen, custom_plugins, pybzflag, BZStatCollector. |
18:43.33 | blast007 | and then I'll convert the bzflag-web one to mercurial and put it in bitbucket ;) |
18:43.35 | JeffM | or do bzflag-game bzflag-editor, bzflag-web, bzflag-misc for consistentcy |
18:43.58 | JeffM | then bzedit and editw32 can be branches off of workbench |
18:44.02 | JeffM | since they are somewhat related |
18:44.12 | JeffM | but ehter or will work |
18:44.39 | blast007 | the trunk/admin can be ignored.. that's just the masterban file that's pretty pointless anyway |
18:44.41 | JeffM | is there any need to split website from services? |
18:44.47 | BulletCatcher | My thought is that the misc repo would contain stuff that we expect to never touch again. |
18:45.10 | JeffM | call it bzflag-archive then |
18:45.29 | JeffM | oddly people still use bzedit, I don't know why |
18:45.45 | BulletCatcher | It must be better than vi. |
18:45.48 | blast007 | because it actually still runs |
18:45.59 | JeffM | I don't think it can be built anymore |
18:46.07 | JeffM | it's old old MFC |
18:46.14 | JeffM | and as we know that's as evil as closed source |
18:46.20 | blast007 | I'm sure I could find my copy of Visual C++ 6.0 somewhere ;) |
18:46.47 | JeffM | it could be ported cross platform in a day :) |
18:57.24 | BulletCatcher | We might as well toss bzworkbench into the bzflag-archive repo with the other abandoned stuff. |
18:57.26 | BulletCatcher | We can always pull it (or anything else) out into a separate repo for revival. |
18:58.27 | blast007 | is it just as much work to put them into a combined repo as it is putting them into individual repos? |
18:59.31 | BulletCatcher | The process I have right now creates separate repos, so it would be an extra step to combine them. |
18:59.50 | blast007 | I'd say just keep everything separate then from the start |
19:00.04 | blast007 | we could also add -archive to the end of the repo name |
19:00.21 | blast007 | or we could create a bzflag-archive team and assign the repos there instead of bzflag-dev |
19:01.19 | JeffM | all are options |
19:01.38 | BulletCatcher | Does GitHub limit the number of free repos we can have? |
19:01.52 | blast007 | I don't think there is a limit on public repos |
19:02.40 | blast007 | https://github.com/pricing shows unlimited public repos |
19:03.20 | BulletCatcher | Then we can do it whatever way works for us. |
19:04.03 | BulletCatcher | But it's kind of annoying that trunk/bzstats only ever had 1 revision. |
19:04.42 | blast007 | do we even need that..? |
19:05.37 | BulletCatcher | There's a lot of stuff we no longer need, but I think it is worthwhile to maintain the history. |
19:07.09 | blast007 | that one revision we could probably forget ever happened |
19:10.47 | JeffM | I kinda like the idea of one repo per proejct and using a seperate group |
19:11.09 | JeffM | then do a group for third party plugins too |
19:12.59 | BulletCatcher | What sort of group do you mean? |
19:13.58 | blast007 | probably the 'teams' |
19:14.01 | JeffM | like the bzflag-dev team |
19:14.02 | JeffM | yeah |
19:14.36 | JeffM | it's an "organization", it's like a user but can have multiple people |
19:14.44 | JeffM | so it can be the container for repositories |
19:28.15 | *** join/#bzflag Gnurdux (~gnurdux@65.199.61.222) |
19:59.41 | *** join/#bzflag Gnurdux (~gnurdux@65.199.61.222) |
20:10.00 | allejo | weee multiple repos! |
20:32.51 | JeffM | ok, nvidia shadow play is pretty cool, automatic recording of full screen apps |
20:45.43 | *** join/#bzflag mebigfatguy (~dave@128.177.74.185) |
20:47.01 | allejo | oh hey look, the bz icon for github change colors |
20:47.17 | JeffM | yeah I changed it to the higher resoultion one that was red |
20:47.21 | JeffM | gravatar is a pain |
20:47.41 | blast007 | why is it a pain? |
20:47.51 | JeffM | have to log in with wordpress stuff |
20:47.51 | allejo | ^ |
20:48.02 | allejo | errr the arrow was for blast's comment |
20:48.08 | blast007 | well, true... kinda liked it before all that |
20:48.22 | JeffM | yeah it's just the wordpress stuff I have an issue with, I like it as an idea |
20:48.32 | blast007 | but I already had a wordpress.com account for akismet |
20:48.43 | JeffM | I did not |
20:48.51 | JeffM | and they REALLLLYYYY wanted me to start a blog |
20:48.56 | blast007 | :) |
20:49.09 | allejo | you need a blog, jeff. everyone needs a blog! |
20:49.24 | JeffM | I have several, I never use them and nobody reads them |
20:49.46 | allejo | heck, I think Google may even want you to start a vlog! but you'd need Google+ |
20:50.04 | JeffM | I use google+ a lot |
20:50.43 | JeffM | it's an accepable home page |
20:52.17 | JeffM | youtube did want to post my new video to my g+ stream |
20:52.19 | JeffM | I didn't mind |
20:54.56 | allejo | oh? hmm... would you be the first person who doesn't mind the new youtube comment system? :o |
20:55.27 | JeffM | I dunno, I never comment on youtube videos |
20:55.41 | JeffM | but I don't mind requiring real names |
20:57.01 | JeffM | they were somewhat useless before |
20:57.21 | JeffM | any change to make them even slightly more useful is better |
20:58.04 | allejo | you are like the first person who has ever said that |
20:58.09 | JeffM | heh |
20:58.59 | JeffM | I think that people who bag on the comment system are just mad they can't use derogitory terms with out any consiquence anymore43 |
20:59.05 | JeffM | they'll have to move to XboxLive |
20:59.14 | allejo | oh blast007 who doesn't want Google+ |
20:59.16 | allejo | or* |
20:59.44 | allejo | ahhh Xbox Live... *shivers* |
20:59.47 | JeffM | yeah |
21:01.49 | JeffM | I think the only "social" account I have for blast outside of this stuff is steam |
21:02.03 | JeffM | cus borderlands2 is fun |
21:03.40 | allejo | never played |
21:03.43 | allejo | it |
21:04.14 | JeffM | it's a FPS game with a very good sense of humor |
21:04.21 | JeffM | and 4 player co-op |
21:07.45 | blast007 | JeffM: I think you underestimate how much people don't care about what they say on the Internet even if it's tied to a real name... |
21:07.55 | JeffM | true |
21:08.20 | blast007 | the real name policy won't stop any of that |
21:08.27 | JeffM | and who's to say Dick McButtz isn't a real name? |
21:08.35 | blast007 | :) |
21:09.20 | blast007 | even before Google started that, it was apparent from Facebook comment systems.. |
21:09.33 | JeffM | Doctor Richard McButtz of the New Hampshire McButtz |
21:09.50 | JeffM | yeah |
21:09.58 | JeffM | so I don't mind those changes on youtube |
21:10.04 | JeffM | I never read comments there anyway |
21:10.09 | JeffM | not that I post a lot of videos |
21:10.10 | blast007 | I do... |
21:10.43 | blast007 | it was bad enough that the biggest youtuber turned off youtube comments and told people to comment on reddit instead |
21:10.52 | JeffM | ouch |
21:11.17 | blast007 | he may have turned it back on now as I think the initial issues were with spam comments getting pushed to the top :P |
21:11.21 | JeffM | if I watch something on yourtube it's usualy on a device that isn't a computer |
21:11.42 | blast007 | I almost always watch on a computer so that the ads show up |
21:12.18 | JeffM | I see adds on my TV sometimes |
21:12.26 | JeffM | when I watch youtube there |
21:12.41 | blast007 | I suppose I don't have any devices with a youtube "app" on them, though |
21:13.18 | JeffM | TV and PS3 have em, I usualy watch youtube shows like I watch TV shows |
21:13.40 | JeffM | on the laptop I can just hit "play on TV" and it starts |
21:15.32 | khonkhortisan | I have an edge case bug, confine mouse only confines the first mouse. |
21:16.07 | JeffM | you have multiple cursors? |
21:16.16 | khonkhortisan | and multiple keyboards to go with them |
21:16.28 | JeffM | yeah that wasn't realy part of the design spec :) |
21:16.39 | khonkhortisan | if I move the second mouse into the window, it traps the first one |
21:16.46 | khonkhortisan | yeah, it's not in anyone's design |
21:17.00 | JeffM | why would you want to do that? |
21:17.14 | khonkhortisan | dualboxing |
21:17.26 | JeffM | that implies 2 boxes :) |
21:17.35 | JeffM | you are instancing |
21:17.38 | khonkhortisan | oh |
21:17.43 | JeffM | run each in a VM |
21:18.38 | khonkhortisan | I just need the second one in a vm |
21:18.38 | JeffM | but really I belive that the constraint is handled by SDL on your platform and SDL only knows about a single cursor |
21:19.07 | khonkhortisan | My guess was that SDL knew about both, but bzflag was just written for the first one. |
21:19.21 | JeffM | no, there is no API for it IIRC |
21:19.27 | JeffM | at least not in SDL 1.x |
21:19.56 | JeffM | the cursor is attached to the window |
21:20.00 | JeffM | only one of each of those |
21:23.04 | blast007 | multiple mice was a "wishlist" item for SDL 2, but I'm not sure if it was actually implemented |
21:23.52 | khonkhortisan | I only have it because of xinput 2 |
21:26.26 | JeffM | yeah it seems very limited for the cases it'd be useful |
21:27.47 | JeffM | yeah all the SDL APIs are just for THE mouse |
21:27.49 | khonkhortisan | There's a table with a sceen on the surface that can be used by multiple people, but that's for (multi)touch api, which is separate from cursor stuff |
21:27.57 | JeffM | yeah |
21:28.04 | JeffM | multiouch is handled very differently |
21:29.10 | JeffM | you can see the event structure that SDL uses here, http://www.libsdl.org/release/SDL-1.2.15/docs/html/sdlevent.html |
21:29.33 | JeffM | there is nothing that diferentiates multiple mice |
21:29.36 | JeffM | or keyboards |
21:29.54 | JeffM | it's just event type, and the motion data |
21:34.58 | JeffM | I don't see anything in SDL 2 for multiple mice in the basic API |
21:35.16 | JeffM | Uint32 SDL_GetMouseState(int* x, |
21:35.16 | JeffM | int* y) |
21:35.48 | JeffM | and the warp function just takes a window not any kind of cursor identifyer |
21:36.25 | JeffM | ahh looks like the event system has a mouse ID |
21:37.11 | JeffM | but that seems to only give you the option of mouse OR touchscreen |
21:37.34 | JeffM | and they have different "finger touch" events for those kinds of input |
21:38.25 | JeffM | those have a touchscreen ID and a finger ID |
21:39.14 | khonkhortisan | so if someone has a touchscreen and the program doesn't support touchscreens, it just shows up as a regular mouse |
21:45.41 | JeffM | yeah a single fingure can act as a mouse |
21:46.18 | JeffM | the touch event has a lot more info, but if all you need is relative motion you can just use the mouse event |
21:47.04 | JeffM | none of this helps you |
21:47.21 | JeffM | I don't see a constrain function that works with multiple input IDes |
21:47.42 | JeffM | the multiple IDs are for touch, and an application can't keep a finger inside a window :) |
21:49.48 | khonkhortisan | My compiled bzflag links to libSDL-1.2.so.0 anyway |
21:50.07 | JeffM | yeah that was what we discussed yesterday |
21:54.57 | JeffM | we think that someone should do a quicky branch and update to SDL 2 |
21:55.07 | JeffM | it may help some of the OSX problems |
22:01.40 | *** join/#bzflag Gnurdux (~gnurdux@65.199.61.222) |
22:05.43 | khonkhortisan | Is the git repo usable yet? |
22:15.21 | JeffM | you can do stuff in import 3 if you want |
22:15.41 | JeffM | the idea for SDL 2 was to make a branch in import 3 and then merge it back in when it was done |
22:18.38 | *** join/#bzflag mebigfatguy (~dave@128.177.74.185) |
22:23.22 | khonkhortisan | the comments in that script make git-svn look bad |
23:05.34 | *** join/#bzflag mebigfatguy (~dave@128.177.74.185) |
23:20.18 | JeffM | unleashes the power of animated gif for his new avatar |
23:50.46 | *** join/#bzflag Delusional (~delusiona@unaffiliated/delusional) |