Page 1 of 1

VS-0.5.3 first time(s) playing, bugs/oddities

Posted: Sun Feb 23, 2020 4:57 pm
by wirser
I finally got some play time with new Vegastrike 0.53 from
https://github.com/vegastrike/Vega-Stri ... tree/0.5.3

First maybe a "RC?" sould have been added to the name. :shock:
(I will put issues on GH)
Critical:
During first execution(s) of VS, during first flight, game hangs up hard!
- on linux use {cntrl}{alt}{F1}(for Debian 9, other linux's may vary) to escape xwindows , then login again at prompt and issue this command

Code: Select all

killall -9 vegastrike
Then issue {cntrl}{alt}{F7} (for Debian 9, other linux's may vary) to return to xwindows.
After this Vegastrike has not yet observed to hang on subsiquent executions.
This appears to be a problem with Python compiling a module during first time running. possibly waiting long enough will solve problem - no errors were observed on shell terminal.

Important:
-It appears the database is corrupt, in more than one system: missions are listed as "scan" or "visit" with no apparent location, clues are that all are "real" meaning can be found by cycling "n", using placement in mission list provides usefull clues. All have no names: one miningbase in enyo system, another miningbase in pandey(sp), and in everest a location with no descrition which seems not to be either a base or jumppoint but shows up as a valid taget :shock: .

quirk/annoyance:

-in the ship HUD mission if having multiple mission when say a patrol objective is completed in the first mission it shows up attached the the bottom of the next mission and even the third mission as completed, this can lead to a situation of objective being shown as completed and not completed :?
If an abjective shows not completed, it needs to be :wink:

With this in mind Vegastrike-0.5.3 seems to be "playable" but in need of work.
NOTE: I have not seen these issues in VS-taose.

Re: VS-0.5.3 first time(s) playing, bugs/oddities

Posted: Mon Feb 24, 2020 2:42 pm
by pyramid
wirser wrote:...
There were some changes to the python code after "taose" (though it is not visible from which revision he forked his repo) and those changes broke a lot of stuff combined with the "cleaning" and split into several units*.csv files which probably was never completely tested...

Re: VS-0.5.3 first time(s) playing, bugs/oddities

Posted: Mon Feb 24, 2020 7:40 pm
by wirser
pyramid wrote:There were some changes to the python code...and those changes broke a lot of stuff ... split into several units*.csv files ... probably was never completely tested...
This helps explain the rather uncomfortable sensation from my backside.

ok so what if Vegastrike-0.5.3 is formally called a testing version?

Understanding the current state of the project - resuming after consierable loss of expertize. With all things that occured, this release is not in too bad of shape.

My advice for now , is tahat anyone interested in playing Vegastrike-0.5.3, obtain the source via

Code: Select all

git clone https://github.com/vegastrike/Vega-Strike-Engine-Source.git
This will keep you up to date with any changes unti a stable release candidate is ready

Re: VS-0.5.3 first time(s) playing, bugs/oddities

Posted: Tue Feb 25, 2020 2:29 pm
by pyramid
wirser wrote:... this release is not in too bad of shape ... until a stable release candidate is ready
Keep the ball rolling! Looking out for improvements 8)

Re: VS-0.5.3 first time(s) playing, bugs/oddities

Posted: Wed Feb 26, 2020 10:53 pm
by wirser
pyramid wrote:Keep the ball rolling! Looking out for improvements 8)
It may help if the ball wasn't so square :lol:

Now for some potetially good nnews 8)
I think I have a work around for the lockup issue

THIS IS STILL NEEDING MORE TESTING, I have repeatible results but need to verify cpuopts not the issue (doubtfull)

In tracking this bug I found by chance that compiling with cpu opts changed to "core 2"
and boost set to internal I don.t get locked up on first mission.
In order to compile,your os packages that are needed are boost-dev(boost development), along with packages that provide:<Python.h>, <boost/python.hpp>, <boost/python/class.hpp>, <boost/python/object.hpp>

in top vegastrike diredtory
edit ./engine/CMakeLists.txt
search for "Use system boost"
change option from "OFF" to "ON"
save file
rm -rf build ( just to be sure )
exicute sh/vsbuild.sh

for the more cautious, All saved games and all options set via ./bin/vssetup
will be lost.
run "find . -name *.pyc | xargs /bin/rm -f" ( I doubt this is needed but it doesn't hurt"
and "rm -rf ~/.vegastrike" ( again may not be needed, but just to be sure)

run vegastrike as normal

I hope this is to mangled I just had eye injections :(

Re: VS-0.5.3 first time(s) playing, bugs/oddities

Posted: Mon Mar 02, 2020 7:34 pm
by wirser
Just to review , of the four issues I reported:The compile issue has been fixed by updating /sh/vsbuild.sh to be exclusively bash and not any shell , the crash issue has a workaround which may become permanent ( depending on feedback from Boost devs - I tried IRC but got no response). . This leaves two remaining major issues (so far) Verifing what pyramid related about the database/Python changes, next work will be to attempt to track down the source of the database errors. after that I'll look at the strange hud mission window behavior.