On Mon, Jul 08, 2024 at 07:31:25AM -0400, Thomas Frohwein wrote:
> On Mon, Jul 08, 2024 at 08:58:57AM +0100, Fabien Romano wrote:
> > On 07/07/2024 23:36, Thomas Frohwein wrote:
> > > On Sun, Jul 07, 2024 at 06:24:17PM -0400, Thomas Frohwein wrote:
> > >> On Sun, Jul 07, 2024 at 11:17:06PM +0100, Fabien Romano wrote:
> > >>>
> > >>> On 07/07/2024 22:37, Fabien Romano wrote:
> > >>>> same update including two more things.
> > >>>>
> > >>>> use proper IN_ENGINE_V to match BAR servers and be abale to join lobby.
> > >>>> bring bin/byar-chobby to setup and start BAR.
> > >>>
> > >>>
> > >>>> On 16/06/2024 06:55, Fabien ROMANO wrote:
> > >>>>> update to latest version
> > >>>>>
> > >>>>> https://github.com/beyond-all-reason/spring/compare/spring_bar_%7BBAR105%7D105.1.1-2472-ga5aa45c...spring_bar_%7BBAR105%7D105.1.1-2511-g747f18b
> > >>>>>
> > >>>>> https://github.com/rlcevg/CircuitAI/compare/aded4d13fcec74a3416e7055fcfb8275e49061a1...3dee671176bdae7e6b1dd46a3c6c2c0d03c48755
> > >>>>>
> > >>>>> One more patch merged upstream, thx thfr@.
> > >>>>> Use patch -E to drop the file.
> > >>>>>
> > >>>>> While there, I mute portcheck long lines report by removing the space on url.
> > >>>>> I also noticed other long lines and adjusted indentation (portcheck use awk length, looks like it does not handle tabulation size).
> > >>>>> Only two versions changed in DIST_TUPLE, see distinfo.
> > >>>>>
> > >>>>> Tested on BAR, barbarianIA works again.
> > >>>>>
> > >>> on current.
> > >>>
> > >>> note about --isolation, the official launcher use the flag but it doesn't work atm.
> > >>> diff for testing.
> > >>
> > >> I'm getting this error here:
> > >>
> > >> [thread::error::run] duplicate base content detected:
> > >>         /home/thfr/.local/state/Beyond All Reason/base/spring/
> > >>         /usr/local/share/games/spring/base/spring/
> > >>
> > >> Did you have to work around this?
> > > 
> > > I ended up deleting ~/.local/state/Beyond\ All\ Reason/base, then it
> > > launched. It was running quite slow, and after I rm -rf'd
> > > ~/.local/state/Beyond\ All\ Reason/ and restarted from scratch, I
> > > couldn't launch it because of this confusing error:
> > > 
> > > Dependent archive "byar chobby test-3306-d61c0a2" (resolved to "BYAR Chobby test-3306-d61c0a2") not found
> > > 
> > > Maybe I did something wrong, but I feel like something is still amiss
> > > here...
> > > 
> > 
> > I did not had this kind of errors. Will search to know if there is a tool to
> > check pr-download files, something that may report when a pool is corrupted or
> > when files are missing. I'm still learning those tools.
> > 
> > I just tried from scratch, works well here.
> > rm -rf ~/.local/state/Beyond\ All\ Reason
> > rm -rf ~/.config/spring
> > 
> > Do you also have .config/spring ?
> 
> That's probably it! I'll try it tonight.

I'm still getting this error after deleting all those files. Seems to
me that there is a missing step, but I find pr-downloader stuff very
opaque... do you have any ideas?:

Dependent archive "byar chobby test-2934-d9c546c" (resolved to "BYAR Chobby test-2934-d9c546c") not found
> 
> > 
> > I was able to join a game online, badly 6G wasn't enough (bad_alloc).
> > DATASIZE="8388608" # 8g
> > 
> > I'm not sure where .config/spring came from (previous tests I guess) but it did
> > not create it again after I spectacle a game online (which mean it download at
> > least one map in lobby).
> > 
> > Blind guess, maybe you have two REPO or two version of the same REPO
> > (.config/spring vs .local/state/...). repos-cdn.beyondallreason.dev is not the
> > default one and I think the game exists in official repo also (I'm not sure).
> > Then spring engine pick the wrong one (.config/spring).
> > 
> > If your issue came from ~/.config/spring, then the --isolation flag could help
> > in such situation. Will dig, but we already had two different results in the
> > past so I'm not sure why our two tests were different. If you add the flags it
> > works ? Or you also have an error because of a missing font ?
> > 
> > I also experience missing files issues when I first went into the desync,
> > playing with the previous engine (before updating the port) on top of the latest
> > byar-chobby:test. I was thinking it was because of the interaction with the
> > official launcher using its pre-made data.7z pool (to save bandwidth). This
> > definitly need more test.
> >