| 00:00.13 | Patlabor221 | I have mine adjusted to my eyes pretty good |
| 00:00.30 | Patlabor221 | it is more tiring to use for long times |
| 00:00.48 | Patlabor221 | but I take it off, put my glasses back on and I'm generaly fine |
| 00:01.12 | Patlabor221 | optimaly I'd have the inner lenses ground for my astigmatism |
| 00:03.26 | Patlabor221 | it would probably be better if it had higher resolution |
| 00:03.49 | Patlabor221 | I am able to see the inter-pixel spaces and sometimes my eye's focus on those gaps |
| 00:04.15 | Patlabor221 | but back to the Joyhat thing, the analog joyhat his VERY rare these days |
| 00:04.29 | Patlabor221 | most use 5 or 9 position |
| 00:04.37 | Patlabor221 | some APIs return an angle, some enumerations |
| 00:05.41 | Patlabor221 | I have never seen a hat that identifies it self as a set of axes |
| 00:05.47 | Patlabor221 | that is usualy a Dpad |
| 00:25.58 | *** join/#bzflag Notify (~notify@66-118-151-70.static.sagonet.net) |
| 00:27.06 | *** join/#bzflag brlcad (~sean@66-118-151-70.static.sagonet.net) |
| 00:53.12 | *** join/#bzflag tofu_ (~sean@66-118-151-70.static.sagonet.net) |
| 01:42.44 | *** join/#bzflag Snake12534 (329c524c@gateway/web/cgi-irc/kiwiirc.com/ip.50.156.82.76) |
| 01:48.10 | *** join/#bzflag khonkhortisan (~Khonkhort@2601:8:8600:9b0:4a5b:39ff:fe94:be23) |
| 02:32.28 | *** join/#bzflag Snake12534 (635cd32e@gateway/web/cgi-irc/kiwiirc.com/ip.99.92.211.46) |
| 02:54.33 | *** join/#bzflag Snake12534 (635cd32e@gateway/web/cgi-irc/kiwiirc.com/ip.99.92.211.46) |
| 03:21.11 | *** join/#bzflag Snake12534 (635cd32e@gateway/web/cgi-irc/kiwiirc.com/ip.99.92.211.46) |
| 04:06.57 | *** join/#bzflag Patlabor221 (uid48618@gateway/web/irccloud.com/x-kwppeahnijxsglva) |
| 04:32.23 | Flash | ~bzgit |
| 04:32.23 | infobot | i guess bzgit is git clone --branch v2_4_x https://github.com/BZFlag-Dev/bzflag-import-3.git bzflag |
| 04:42.26 | Flash | Can someone explain how we are/want to be using git ... for example, why am I seeing notifications from Jeff for his personal fork? |
| 04:42.45 | Flash | From the discussion last night, it seems that personal forks are the way to go... |
| 04:43.02 | Flash | but it doesn't seem like everyone should be spammed from my doodling |
| 04:45.04 | Flash | and why are we considering removing geno from 2.4? |
| 04:55.25 | Flash | and if I want to try out Jeff's pull request, how do I get it to my repository without committing it everywhere? |
| 05:02.35 | *** join/#bzflag Snake12534 (635cd32e@gateway/web/cgi-irc/kiwiirc.com/ip.99.92.211.46) |
| 05:26.55 | *** join/#bzflag Snake12534 (635cd32e@gateway/web/cgi-irc/kiwiirc.com/ip.99.92.211.46) |
| 05:27.08 | alezakos | Flash: your personal fork won't trigger any notifications, Jeff has just set them up for his fork on this channel |
| 05:27.08 | Snake12534 | baby, I just want to not do this |
| 05:27.23 | Snake12534 | wrong channel |
| 05:27.28 | Snake12534 | good morning, alezakos! |
| 05:27.53 | alezakos | food morning :) |
| 05:28.04 | alezakos | good* |
| 05:30.19 | alezakos | while geno is going to be removed, the BZFlag codebase will contain a plugin to exactly replicate its behaviour (specifically for Mofo, I don't think their geno plugin will be affected) |
| 05:32.57 | BulletCatcher | Even if it is protocol compatible, I think that moving the geno flag into a plugin is too big a change for the 2.4 branch. |
| 05:32.59 | BulletCatcher | I would prefer to see it in v2_6_x, and with a more recognizable name. |
| 05:38.01 | Patlabor221 | Yeah notify should not be holed up to my fork and bz fork, probably screwed up when I set it up |
| 05:39.32 | Patlabor221 | I know other forks don't do that |
| 05:39.35 | allejo | I'd agree with bc on this, would seem fitting for it to be something done in a major release. and the alezakos is correct, mofo's geno functionality is entirely separate from the geno flag |
| 05:40.06 | Patlabor221 | I am probably weird because I am me and bzflag-dev |
| 05:41.17 | alezakos | all hail the great alezakos |
| 05:41.38 | Patlabor221 | I did not make the plugin for mofo |
| 05:43.11 | allejo | never said you did, was simply stating alezakos was correct :p |
| 05:44.29 | Patlabor221 | I know, I am just clarifying |
| 05:45.12 | Patlabor221 | It would be a change for server configs, for the better imho, but that's why it's just a pull request now |
| 05:47.07 | Patlabor221 | goes to find a real computer to see about notify |
| 05:48.05 | *** join/#bzflag Patlabor (~JeffM@2605:e000:7d02:c700:a9e9:de64:4f66:316d) |
| 05:48.13 | Patlabor | there we go |
| 05:48.21 | Patlabor | what website deals with notify again? |
| 05:48.26 | Flash | had to get on a mac? |
| 05:48.45 | Patlabor | nope, had to get off an apple product |
| 05:48.55 | Patlabor | ipad is not awesome for IRC |
| 05:49.04 | allejo | lol |
| 05:49.24 | Patlabor | tho had my macbook not died it would have been the next closest |
| 05:50.15 | Patlabor | I think I may have bound all my projects that say "bzflag" to the IRC bot |
| 05:50.24 | Patlabor | so no, your fork would not report here |
| 05:50.38 | Patlabor | it's just an artifiact of the fact that I setup the account for bzflag on the bot |
| 05:50.52 | Flash | ok, so how do I pull your pull request to my personal fork? |
| 05:50.54 | allejo | and isn't notify brclad's bot? |
| 05:51.08 | allejo | brlcad* |
| 05:51.26 | Patlabor | you don't get my pull request, you merge my fork into your fork |
| 05:51.42 | Patlabor | my pull request is a request for bzflag-dev |
| 05:51.45 | alezakos | Patlabor: http://n.tkte.ch and github |
| 05:51.46 | Patlabor | no it is no |
| 05:51.48 | Patlabor | it's a service |
| 05:51.59 | Patlabor | that service :) |
| 05:52.15 | allejo | ohhh |
| 05:52.21 | allejo | that's an interesting domain |
| 05:52.40 | Patlabor | so yeah flash, you'd have a personal for, and then could merge in from any other fork, including the master fork |
| 05:52.47 | Patlabor | that's how you pick up other commited changes |
| 05:53.05 | Flash | is definitely old... svn made sense... |
| 05:53.13 | Patlabor | it's a different method |
| 05:53.26 | BulletCatcher | Flash, the simplest and least disruptive (to your own development work) way to try Patlabor's changes is to clone his repo. |
| 05:53.28 | Patlabor | some of it I don't get but the fork stuff I made myself understand |
| 05:53.40 | Patlabor | yeah, you can clone my repo and build it |
| 05:54.01 | Patlabor | there Should be a way to pull from my repo to yours but that is more invasive |
| 05:54.16 | Patlabor | if you are just trying, you just build my code, since it's public |
| 05:54.34 | Patlabor | Hey, I rememberd the notifico username :) |
| 05:54.51 | allejo | jeff, did you remove roguegeno from the xcode project as well? (not sure if that's needed) |
| 05:54.56 | Patlabor | too many things named bzflag :) |
| 05:55.07 | Patlabor | allejo, I did not, I have no mac to do that with |
| 05:55.15 | Patlabor | thanks to a cat and a bathtub |
| 05:55.47 | Flash | my brain is rotted by decades of using ClearCase ... this copy stuff is mind numbing |
| 05:56.26 | Patlabor | the simplest thing I find to do is the pull requests, it gives me a nice single place to put a coment for a list of commeits that implemnt one thing, that part of it I like since it's more closely matchs SVN |
| 05:56.27 | allejo | you actually called it UnfairMurderMachine?! i thought that was a joke |
| 05:56.37 | Patlabor | I wanted to be honest |
| 05:59.16 | Patlabor | ahh I see why notify got odd |
| 05:59.21 | Patlabor | I shall fix |
| 05:59.28 | BulletCatcher | A plugin name like genocideFlag will more meaningful to everyone but you, Patlabor. |
| 05:59.48 | Patlabor | it's defentaly a mess up on my part, not a feature of forks |
| 05:59.51 | allejo | meaningful! that's the word I was looking for |
| 05:59.58 | Patlabor | BulletCatcher, of course, I never expected the pull request to be accepted ;) |
| 06:00.27 | Flash | swears at git |
| 06:01.09 | Flash | so I cloned https://github.com/JeffM2501/bzflag-import-3.git and got bupkiss |
| 06:01.22 | Patlabor | did you use the right branch? |
| 06:01.29 | Patlabor | or did you clone mastre? |
| 06:01.34 | BulletCatcher | Now do a "git checkout v2_4_x" command. |
| 06:01.43 | Flash | it's from https://github.com/JeffM2501/bzflag-import-3/tree/v2_4_x |
| 06:01.44 | Patlabor | yeah that will get the bracch |
| 06:01.55 | Flash | I clicked on the "Clone in desktop" button |
| 06:01.58 | Patlabor | it doesn't work like SVN |
| 06:02.03 | Patlabor | yeah that cloned the entire REPO |
| 06:02.08 | Patlabor | you did not pick a branch |
| 06:02.10 | Flash | not reallyy |
| 06:02.17 | Flash | very little actually came down |
| 06:02.23 | Patlabor | exactly |
| 06:02.30 | Patlabor | it cloned the entire MASTER branch |
| 06:02.40 | Patlabor | and gave yuo the info to checkout other branches |
| 06:02.48 | Flash | github isn't helping me much :@ |
| 06:02.49 | Patlabor | the branches are not stored in sub URLS like svn |
| 06:03.01 | Patlabor | not if you expect it to work like SVN |
| 06:03.08 | Patlabor | are you using the mac github client? |
| 06:03.14 | Patlabor | or command line? |
| 06:03.36 | Flash | I was using the web page/mac client, but I can revert to the command line as required |
| 06:03.56 | Patlabor | clone in desktop should try to run the mac github client |
| 06:04.00 | Patlabor | did it do that? |
| 06:04.05 | Flash | yes |
| 06:04.21 | Patlabor | ok, then in that mac github client is an option to change branches and do a checkout |
| 06:04.29 | Patlabor | that will change your working tree to the branch |
| 06:05.13 | Patlabor | forking a repo, forks the entire repo, not just a single branch. so what you did was the same as getting the master of bzflag-dev/import-3 |
| 06:05.34 | Patlabor | and I did not make any changes in that branch so you get the same thing as the upstream origin |
| 06:06.03 | Patlabor | when I do a clone, I usualy use clone --branch BRANCH_NAME https://REPO_URL |
| 06:06.24 | Flash | ok.... looks like I got actual code now .... |
| 06:06.29 | Flash | bangs head on desk |
| 06:06.40 | Patlabor | I do most of my clones using TortoiseGit |
| 06:06.49 | Patlabor | then commit managment from te github client |
| 06:06.58 | *** join/#bzflag zuii (~akos@BC06BD4B.catv.pool.telekom.hu) |
| 06:07.03 | Patlabor | the github client really assumes your master is the only branch you want to ever use |
| 06:07.14 | Patlabor | and that is not the case with import-3 |
| 06:07.35 | Patlabor | clones are rare so I don't mind using a different tool |
| 06:08.02 | Patlabor | tho for some things I've had good luck using the built in git support in VC12 |
| 06:08.19 | Patlabor | ok, notifico should be fixed now |
| 06:08.23 | Patlabor | no more spam from me |
| 06:08.34 | Patlabor | I didn't realize it, I thought it was just dupes |
| 06:16.26 | BulletCatcher | I can think of two reasons a BZFlag developer would want a personal GitHub fork. |
| 06:16.30 | BulletCatcher | The first is to have "cloud" storage of their (semi) private repo, which may be convenient when development is done on multiple computers. |
| 06:16.35 | BulletCatcher | I created my private repo with "git clone --bare --mirror" on my sole development system, and all of my actual development is done in clones of that. |
| 06:16.41 | BulletCatcher | This makes it easy to discard experiments, and the two-step push process (first to my private repo, then to GitHub) makes it more likely that I will catch mistakes and be able to rebase rather than create an unnecessary tiny branch. |
| 06:16.45 | BulletCatcher | The second reason for a personal GitHub fork is to allow review of commits (via pull requests) before they are pulled into the main repo. |
| 06:17.08 | Flash | so as an experiment, I'd like to offer a change to Patlabor's soon to be rejected pull request that removes the RogueGeno from the XCode project. |
| 06:17.13 | Flash | How do I do that? |
| 06:17.45 | Patlabor | you shoud be able to make a pull request on my branch |
| 06:17.48 | Patlabor | fork sory |
| 06:18.14 | Patlabor | I know I could give you commit access to my fork |
| 06:18.28 | Patlabor | or you can offer me a push I think |
| 06:18.36 | Patlabor | this is the part I am not super clear on |
| 06:18.57 | Patlabor | I have only done it back to the upstream origin, but you are supposed to be able to do it anywhere |
| 06:19.04 | BulletCatcher | For him to offer a pull request I think he has to clone your repo on GitHub. |
| 06:19.13 | Patlabor | fork the fork |
| 06:19.21 | BulletCatcher | yeah |
| 06:19.30 | Patlabor | that's a lot of forkin forks |
| 06:20.16 | Flash | stumbles around in the dark |
| 06:20.22 | Not-311d | [02bzflag-import-3] 07JeffM2501 pushed 032 commits to 03v2_4_x [+0/-0/±19] 13http://git.io/RAXS7w |
| 06:20.23 | Not-311d | [02bzflag-import-3] 07JeffM2501 03499265f - add increment functions for score for convenience |
| 06:20.24 | Not-311d | [02bzflag-import-3] 07JeffM2501 03eb3446a - Merge branch 'v2_4_x' of https://github.com/BZFlag-Dev/bzflag-import-3 into v2_4_x |
| 06:20.30 | Patlabor | damnit! |
| 06:20.36 | Patlabor | where did it get that from |
| 06:20.48 | Patlabor | oh wait no I just screwed up *sigh* |
| 06:20.51 | Flash | So in my GitHub client, I "sync"d my commit ... where did that go? |
| 06:21.12 | Patlabor | I was using the legit fork not my fork |
| 06:21.39 | Patlabor | you synced the commit you made on my fork? |
| 06:21.44 | Flash | yes |
| 06:22.33 | Patlabor | check your repos in github do you have one with my name in it now? |
| 06:23.24 | Flash | no |
| 06:24.06 | Flash | the black hole of source "control" |
| 06:24.18 | Patlabor | yeah that is distressing |
| 06:24.37 | Patlabor | does it show in your history? |
| 06:24.40 | Flash | so I should fork your fork and then create a pull request? |
| 06:24.54 | Flash | which history? |
| 06:24.55 | Patlabor | I am fairly sure that would work yes |
| 06:25.00 | Flash | the local repo is good |
| 06:25.02 | Patlabor | your commit history |
| 06:25.08 | Patlabor | in the local |
| 06:25.57 | Patlabor | one thing I don't like about the github client is it doesn't let you name local repos based on where the come from |
| 06:26.09 | Patlabor | that's the main reason I don't clone using it |
| 06:26.25 | Flash | it doesn't show in the repo clone of your fork |
| 06:26.50 | Patlabor | so it failed to push, then synced with my repo |
| 06:26.52 | Flash | I have no idea where it went |
| 06:26.53 | Patlabor | did you commit forst? |
| 06:26.57 | Patlabor | or just sync? |
| 06:27.17 | Flash | I committed first |
| 06:27.28 | Patlabor | and it doesn't show as an unsynced commit? |
| 06:28.18 | Flash | no |
| 06:28.35 | Flash | Hmmm. I don't know if I can fork your fork, since I already have a BZFlag fork |
| 06:28.50 | Patlabor | exactly why the github client sucks |
| 06:29.02 | Flash | this is actually the web page |
| 06:29.04 | Patlabor | but yeah I dunno |
| 06:29.21 | Patlabor | there should be a way for us to trade commits across forks |
| 06:29.28 | Patlabor | I just dont know how to do that |
| 06:30.05 | Patlabor | you probalby have to set up my fork as a different remote |
| 06:30.12 | Patlabor | sync to that to get my changes |
| 06:30.23 | Patlabor | then make your changes, and do a pull request from there |
| 06:30.36 | BulletCatcher | I am confident there is a way for multiple people to contribute to a pull request, but it's too late for me to play with that right now. |
| 06:30.52 | Patlabor | yeah I know it can be done too, I just don't know how |
| 06:31.07 | Patlabor | I need to throw away all my repos and name them properly |
| 06:31.19 | Flash | so if I try to push, I get an authentication error, as expected. |
| 06:31.57 | BulletCatcher | The commits in a pull request are just branches, so it should be possible to just add further commits to them and push. |
| 06:31.58 | Patlabor | I've seen options to email patches from clients, that's probalby what a lot of people do, lame |
| 06:33.45 | Patlabor | I'd also like to know how to remove/rollback local commits that have not been synced |
| 06:33.51 | BulletCatcher | I find the "gitk" program to be very useful for browsing a local repo. |
| 06:35.44 | BulletCatcher | There are a couple of ways to do remove an unpushed commit, but "git reset --soft HEAD~" is the most direct. |
| 06:36.13 | Flash | ok, just using the github client, commit works; sync fails. That's what I expected |
| 06:37.26 | Patlabor | what is "Squash" |
| 06:37.52 | Flash | it's a game kind of like racquet ball |
| 06:37.57 | allejo | a fruit? |
| 06:38.05 | Patlabor | I can "rebase" with "Squash all" and it shows my unsynced commits |
| 06:38.49 | allejo | squash merges all the commits into one, iirc |
| 06:38.55 | Patlabor | ahh |
| 06:39.32 | allejo | well not "all" of them, which ever ones you specify |
| 06:39.54 | allejo | only works if you haven't pushed to remote though |
| 06:40.43 | Patlabor | I have not |
| 06:40.58 | Patlabor | but I just deleted the repo, was easier :) |
| 06:45.47 | *** join/#bzflag zuii (~akos@BC06BD4B.catv.pool.telekom.hu) |
| 06:52.25 | Patlabor | there we go, 4 clearly labeled folders |
| 06:55.22 | allejo | i'm assuming the self destruct time is client side, any chance that can be a bzdb variable in 2.6? |
| 06:56.12 | Flash | anything is possible :) |
| 06:56.25 | Patlabor | the entire self destruct should be server side |
| 06:56.26 | Flash | IIRC, the client uses bzdb variables ... |
| 06:56.43 | Patlabor | secured with 2 private keys that have to be turned at the same time |
| 06:57.05 | Flash | see? you are warming to the PKI notion! |
| 06:57.13 | Patlabor | no, real plastic ones |
| 06:57.19 | Flash | lol |
| 06:57.37 | Patlabor | then it overloads the matter intermix chamber |
| 06:58.03 | Flash | what about the flux capacitor? |
| 06:58.45 | Patlabor | we don't use those anymore, those are so 80's |
| 07:01.04 | Patlabor | ok cool, no notification.. at least that was fixed |
| 07:11.19 | allejo | oh hmm... |
| 07:11.42 | allejo | once we do move to import 4 or the finalized one. how are we gonna migrate issues/pull requests? |
| 07:12.01 | Patlabor | I sugest we merge or close them all beforehand |
| 07:12.55 | allejo | what if issues are more long term... like the ones i'm creating |
| 07:13.09 | Patlabor | so I reverted my geno suff in my fork after making a branch, then made a new pull rewuest with another change, but still shows all the history items of geno removal, but only one file changed, not related to geno removal |
| 07:13.10 | Flash | be a part of the solution, not the problem |
| 07:13.31 | Patlabor | allejo, hence why I want to get our repo setup sooner rather then later! |
| 07:13.59 | allejo | would be nice... |
| 07:13.59 | Patlabor | I'm guessing we'll have to redo any that are open |
| 07:14.19 | Patlabor | well it's moving at bzflag speed, years instead of weeks |
| 07:16.12 | allejo | woo light speed! |
| 07:20.16 | allejo | can the readme file be converted to markdown so it renders nicely on github? |
| 07:21.02 | Patlabor | it's included with our builds, will it still render correctly ouside of github? |
| 07:21.18 | Flash | markdown is pretty readable as is |
| 07:21.34 | Patlabor | if it's acceptable in a text editor then sure |
| 07:21.42 | Patlabor | but I'd like to see what it'd look like |
| 07:21.42 | Flash | yup |
| 07:25.20 | allejo | writes it up |
| 07:28.17 | Patlabor | just remember that the vast majority of users arn't developers, don't use Git, or Github, so anything that we hae they goes out with binary or package installs needs to look good outside of the context of "developer" |
| 07:31.05 | allejo | wonders |
| 07:31.29 | Flash | allejo: wonder not. MD is very human readable |
| 07:31.58 | allejo | oh i know it is, i write a lot of my stuff in MD. i'm just curious as to how some of this should be laid out so it makes sense |
| 07:32.23 | Patlabor | and yes, getting a player record and not freeing it can leak ;) |
| 07:33.07 | allejo | why not std::shared_ptr? |
| 07:33.35 | Patlabor | they didn't exist when we made it, and they are template classes that can't be passed across a DLL boundry |
| 07:33.41 | Patlabor | so we don't use them |
| 07:33.56 | Patlabor | just like we can't use std::strin |
| 07:33.59 | Patlabor | string |
| 07:34.27 | allejo | oh gross... |
| 07:34.33 | Patlabor | if bzfs and the plug-in were built using different C++ runtimes the allocation and the free may not be compatable |
| 07:35.21 | Patlabor | it's one of the things you have to deal with when you want a truely cross platform API |
| 07:35.33 | Patlabor | not everyone can use your fancy features ;) |
| 07:36.01 | allejo | but fanciness! :( |
| 07:36.16 | Patlabor | don't mean Jack when you have to ensure binary compatability |
| 07:36.32 | Patlabor | this is one reason you see most "plug-ins" done in non native languages |
| 07:36.40 | Patlabor | python, lua, etc.. |
| 07:36.52 | Patlabor | things that are interperted or JIT compiled |
| 07:37.00 | Patlabor | then you can use fancyness |
| 07:37.13 | Patlabor | but all we got is bits and assembly codes :) |
| 07:37.22 | allejo | oh |
| 07:37.41 | Patlabor | as a simple example, lets say we have template code that allocates and deallocates |
| 07:37.55 | Patlabor | bzfs uses that template code and is built debug |
| 07:38.11 | Patlabor | because it's debug it allocates some extra bytes with each structure for debug symbols |
| 07:38.27 | Patlabor | then we make a plugin that uses the same template but is used release |
| 07:38.34 | Patlabor | it dosn't know about those extra bytes |
| 07:39.10 | Patlabor | if bzfs were to allocate some memory, but then in he course of runnig the plugin has to deallocate it (like to resize a string), it won't free those extra debyg bytes since it didn't know about them |
| 07:39.17 | Patlabor | now you are leaking memory |
| 07:39.23 | Patlabor | or even worse, overwriting memory |
| 07:39.29 | Patlabor | then bad things happen |
| 07:40.35 | Patlabor | or te modules could use different versions or implemtations of the STL, ones that maybe incompatabile |
| 07:41.42 | allejo | well that's inconvenient... |
| 07:41.47 | allejo | likes the STL |
| 07:41.52 | Patlabor | yeah |
| 07:42.05 | Patlabor | if you an be sure everyone is using the same thing it can work |
| 07:42.23 | Patlabor | but that's the deal with a public API, you may not know what they are doing |
| 07:42.43 | Patlabor | so yeah the API can't pass STL or templates across the function boundry |
| 07:42.56 | Patlabor | we can wrap them up, but not pass them over |
| 07:43.20 | Patlabor | the bzAPI string is just a wraper around std::string to ensure that the calls to std::string all happen inside bzfs |
| 07:43.35 | Patlabor | thta pushes it to one side of the boundry so it's safe |
| 07:44.15 | allejo | wait |
| 07:44.17 | Patlabor | most managed langauges don't have the same issue, since the same runtime always does allocations and garbage collections |
| 07:44.31 | allejo | i meant std::shared_ptr instead of using bz_BasePlayerRecord in the plug-in |
| 07:44.57 | Flash | templates are a code-writing convenience. We COULD explicitly write our own "smart pointer" if needed. |
| 07:44.57 | allejo | or am I entirely blanking out and misreading what you're saying? |
| 07:45.07 | Patlabor | yeah the API can't return a std::shared_ptr to the Record |
| 07:45.17 | Patlabor | the record would be allocated inside bzfs |
| 07:45.28 | Patlabor | but then it would go out of scope inside the plug=in |
| 07:45.46 | Patlabor | so the destructor for the shared_ptr would try to deallocate the memory inside the plug-in |
| 07:46.18 | Patlabor | yeah if we could ensure the deallocation happened always in bzfs that'd be fine |
| 07:46.25 | Patlabor | that's wat we do with the API string |
| 07:46.35 | Patlabor | but the standard template dosn't force it to be in any module |
| 07:46.47 | Patlabor | the code defined in the template is compiled into each module |
| 07:47.11 | allejo | wait wait. i don't mean the API returning a template, i mean like: std::shared_ptr<bz_BasePlayerRecord> pr(bz_getPlayerByIndex(...)) |
| 07:47.37 | Patlabor | you don't understand how shared pointers work ;) |
| 07:47.49 | Flash | std::shared_ptr only handles new/delete ... the API requires a different allocate/deallocate pattern |
| 07:47.54 | Patlabor | that would make the plugin try to OWN the pointer |
| 07:47.59 | allejo | ok that's where i'm going wrong |
| 07:48.04 | Patlabor | since there are no other shared pointers to it |
| 07:48.27 | Patlabor | yeah shared pointes work by being able to deallocate the object when all references are out of scope |
| 07:48.35 | Patlabor | there would not be a reference in bzfs |
| 07:48.43 | Patlabor | so the plugin would try to deallocate |
| 07:49.33 | allejo | ohhh ok, now i'm following along |
| 07:49.51 | Patlabor | yeah you can't really mix and match and have it work right |
| 07:50.24 | Patlabor | just like if you use shared ptrs and snap off weak pointers that arn't referene counted, they may go invalid if the shared pointers go out of scope |
| 07:51.20 | Patlabor | if someone is supposed to own the object you use unqiue ptr, if you don't care who "owns" it then use shared_ptr |
| 07:51.45 | Patlabor | but ALWAYS use smart poiners or you can run into problems if your weak ones get used out of scope |
| 07:53.00 | allejo | makes sense |
| 07:53.15 | allejo | what would be an apt use for a weak pointer? |
| 07:53.19 | Patlabor | so we just go all managed, problem solved :) |
| 07:53.24 | Patlabor | they can be faster |
| 07:53.37 | Patlabor | and if you know the scope is cool you can use them |
| 07:54.00 | Patlabor | also some APIs can't use smart pointers so you have to send them a plain old pointer |
| 07:54.07 | Patlabor | as long as you know they won't free it |
| 07:54.26 | Patlabor | but most APIs wont do that for the reasons stated above :) |
| 07:56.42 | allejo | right |
| 07:57.07 | allejo | well i guess speed could be nicer as long as you don't goof with the scope |
| 07:57.31 | Patlabor | sometimes you have to muck with a lot of data and pointers are the way to do that |
| 07:57.48 | Patlabor | and you are going to do math on those pointers to process the data |
| 07:59.36 | allejo | math is always fun |
| 08:00.44 | Patlabor | a lot of data formats have offsets, so you want to get a pointer to a thing that you need to treat as a data structure, but you don't want it to be deallocated when you are done |
| 08:01.11 | Patlabor | so you get a pointer to the middle of your blob and just cast it to your structure |
| 08:02.01 | Patlabor | of you made that as a shared_ptr it'd be wierd if the system went and deallocated the middle of your larger structure when you were done with your pointer :) |
| 08:10.58 | allejo | that would be a problem. not so smart now, smart pointers! |
| 08:11.19 | allejo | (meant as a joke, of course) |
| 11:41.58 | *** join/#bzflag Patlabor (~JeffM@2605:e000:7d02:c700:a57a:5853:c5c0:616c) |
| 11:45.47 | *** join/#bzflag blast007 (~blast007@bzflag/developer/Blast) |
| 14:50.56 | *** join/#bzflag Patlabor (~JeffM@2605:e000:7d02:c700:b473:fd5f:71c5:8eea) |
| 15:24.57 | *** join/#bzflag zlsa (~zlsa@static-205-157-147-198.ded.unwiredbb.net) |
| 15:24.57 | *** join/#bzflag zlsa (~zlsa@unaffiliated/zlsa) |
| 15:32.07 | *** join/#bzflag zuii (~akos@BC06BD4B.catv.pool.telekom.hu) |
| 16:01.44 | brlcad | allejo: notify is not my bot, it's run by ``Erik if you want to get another repo set up, see elfga.com/notify |
| 16:05.40 | *** join/#bzflag I_Died_Once (~I_Died_On@69.136.32.149) |
| 16:05.40 | *** join/#bzflag I_Died_Once (~I_Died_On@unaffiliated/idiedonce/x-1828535) |
| 16:33.04 | *** join/#bzflag Patlabor (~JeffM@2605:e000:7d02:c700:7139:a86:e667:b76c) |
| 16:40.45 | blast007 | ooo, an unlimited number of users in an organization can use Visual Studio Community 2013 for contributing to open source projects |
| 16:41.21 | blast007 | (as well as for classroom learning and academic research) |
| 16:49.41 | alezakos | Patlabor: about the history, github sees the revert commit you made as yet another commit and includes it in the history, since you didn't explicitly delete all the other commits |
| 16:50.02 | alezakos | which is why it's a good idea to make a separate branch for every pull request |
| 16:58.39 | Patlabor | because I could not find a way to delete said history :) |
| 16:58.58 | Patlabor | blast yeah, the 5 user thing is for comerical proejcts |
| 16:59.00 | Patlabor | goes to work |
| 17:15.19 | *** join/#bzflag Patlabor221 (~JeffM@107-209-61-107.lightspeed.irvnca.sbcglobal.net) |
| 17:19.45 | *** join/#bzflag I_Died_Once (~I_Died_On@unaffiliated/idiedonce/x-1828535) |
| 17:22.06 | *** join/#bzflag zlsa (~zlsa@static-205-157-147-198.ded.unwiredbb.net) |
| 17:22.06 | *** join/#bzflag zlsa (~zlsa@unaffiliated/zlsa) |
| 17:28.38 | *** join/#bzflag zlsa (~zlsa@static-205-157-147-198.ded.unwiredbb.net) |
| 17:28.38 | *** join/#bzflag zlsa (~zlsa@unaffiliated/zlsa) |
| 17:36.15 | *** join/#bzflag TimRiker (~TimRiker@bzflag/projectlead/TimRiker) |
| 17:36.15 | *** mode/#bzflag [+o TimRiker] by ChanServ |
| 17:48.15 | Patlabor221 | so how does one remove history? |
| 17:48.27 | Patlabor221 | I guess I should be sure I understand how to make branches too |
| 18:06.49 | *** join/#bzflag zlsa (~zlsa@unaffiliated/zlsa) |
| 18:22.56 | *** join/#bzflag WarPig (~j-mo@207.235.53.82) |
| 18:23.59 | WarPig | global auth down again? |
| 18:24.33 | WarPig | s/?/!!!1!/ |
| 18:33.32 | blast007 | WarPig: is it working now? |
| 19:03.54 | WarPig | Appears to be. |
| 19:04.47 | blast007 | I think the issues from the days earlier was due to a network upgrade that my host was doing |
| 19:21.35 | *** join/#bzflag TimRiker (~TimRiker@bzflag/projectlead/TimRiker) |
| 19:21.36 | *** mode/#bzflag [+o TimRiker] by ChanServ |
| 19:47.16 | WarPig | People are still having problems, it appears.. |
| 19:47.56 | WarPig | I was not able to auth just now |
| 19:48.02 | blast007 | I don't know. |
| 19:48.15 | WarPig | lemme try a different server |
| 19:48.27 | blast007 | are you able to get the server list? |
| 19:48.35 | blast007 | because that *is* the auth server |
| 19:49.00 | WarPig | yes. looks like a mofo problem. I am able to auth elsewhere |
| 19:49.52 | blast007 | there was some talk about DNS being messed up at mofo. looks like zoneedit is borked so they are trying to switch to cloudflare. |
| 19:50.34 | WarPig | It is VERY FAST to tell me I can't auth. Usually, if its a problem with the auth server, there is a short delay. |
| 19:50.59 | WarPig | ...like a timeout expires or something. |
| 19:51.46 | WarPig | Sorry to disturb your nap.. :P |
| 19:52.26 | blast007 | :P |
| 20:34.53 | *** join/#bzflag zuii (~akos@BC06BD4B.catv.pool.telekom.hu) |
| 21:17.17 | *** join/#bzflag TimRiker (~TimRiker@bzflag/projectlead/TimRiker) |
| 21:17.17 | *** mode/#bzflag [+o TimRiker] by ChanServ |
| 21:20.12 | gnu-sense | yeah, mofo still has probs. Other servers are okay. |
| 21:26.58 | *** join/#bzflag Snake12534 (uid45641@gateway/web/irccloud.com/x-gtuqpsivmyxavuxc) |
| 22:50.53 | gnu-sense | hmm. server list seems to be awol now. |
| 22:51.19 | gnu-sense | no, it's back. idk... |
| 22:52.58 | *** join/#bzflag zlsa (~zlsa@static-205-157-147-198.ded.unwiredbb.net) |
| 22:52.58 | *** join/#bzflag zlsa (~zlsa@unaffiliated/zlsa) |
| 23:11.17 | *** join/#bzflag khonkhortisan (~Khonkhort@2601:8:8600:9b0:4a5b:39ff:fe94:be23) |
| 23:33.17 | *** join/#bzflag khonkhortisan (~Khonkhort@c-24-18-42-215.hsd1.wa.comcast.net) |