Feb 01 - Initial testing / installation

A forum for chatting about in-development game features.
mharmless
Posts: 235
Joined: Sun Feb 01, 2015 11:11 pm
Location: Washington State

Re: Feb 01 - Initial testing / installation

Postby mharmless » Mon Feb 02, 2015 8:04 pm

On lunch, gathered the logs. I had deleted the game from D: on the tablet PC, reinstalled it and produced the same error as before, producing a new log for your use.
Attachments
03 - Tablet PC - logs.zip
(3.16 KiB) Downloaded 645 times
02 - Primary PC - logs.rar
(5.93 KiB) Downloaded 635 times
01 - HP Laptop - logs.rar
(5.8 KiB) Downloaded 715 times

mharmless
Posts: 235
Joined: Sun Feb 01, 2015 11:11 pm
Location: Washington State

Re: Feb 01 - Initial testing / installation

Postby mharmless » Mon Feb 02, 2015 8:04 pm

and the fourth one. Looks like cap of 3 attachments per post.
Attachments
04 - Media PC - logs.rar
(6.6 KiB) Downloaded 638 times

User avatar
sven
Site Admin
Posts: 1620
Joined: Sat Jan 31, 2015 10:24 pm
Location: British Columbia, Canada
Contact:

Re: Feb 01 - Initial testing / installation

Postby sven » Mon Feb 02, 2015 8:12 pm

Many thanks for those. I'll take a look at them once I've finished replacing the current terrible end turn button :)

mharmless
Posts: 235
Joined: Sun Feb 01, 2015 11:11 pm
Location: Washington State

Re: Feb 01 - Initial testing / installation

Postby mharmless » Mon Feb 02, 2015 9:47 pm

Device 06, the Oracle VirtualBox, alt port gives

Code: Select all

updater\connect_to_server.lua:141: connection refused
updater\connect_to_server.lua:141:connect_to_server:
 updater\client_update.lua:508:
  updater\client_update.lua:558


Disabled eset temporarily, no difference (as expected).

Added to firewall exempt devices, update started correctly. Closed updater.

Removed firewall exemption for the device, added 54.201.202.107 (update.stars-in-shadow.com) to the SSL exceptions list (no ssl inspection). Same behavior as before.

Removed SSL exception. Added to list of trusted HTTP sites (no packet inspection). Updater was still on alt port, still failed same message. Switching to default port, download proceeded as expected. Canceled/closed updater, removed trusted site status.

So, looks like definitely something to do with watchguard's packet inspection. I'm not seeing block messages, so looks our firewall doesn't see anything wrong with the traffic. The packet inspection is interfering somehow?

mharmless
Posts: 235
Joined: Sun Feb 01, 2015 11:11 pm
Location: Washington State

Re: Feb 01 - Initial testing / installation

Postby mharmless » Mon Feb 02, 2015 9:58 pm

6) Oracle VirtualBox on PC 5
6.OS) Windows 7 Professional SP 1, 32 bit
6.Spec) Intel i7-4770 processor, 2gb of RAM, 32mb VirtualBox Graphics without Acceleration, 40gb VDI
6.Res) 1 adjustable display, tested at 1032x777
6.Inst) Copied the installed game from HP Elitedesk onto a mapped network drive. "E:\Stars in Shadow". Location maps to the host machine.
6.Notes) Game fails to run, probably expected given the graphics above.
VM without acceleration error.PNG
VM without acceleration error.PNG (9 KiB) Viewed 23640 times


Next time I have to wipe this VM I'll step it up to 2d acceleration, and later to 3d.

User avatar
sven
Site Admin
Posts: 1620
Joined: Sat Jan 31, 2015 10:24 pm
Location: British Columbia, Canada
Contact:

Re: Feb 01 - Initial testing / installation

Postby sven » Mon Feb 02, 2015 10:06 pm

mharmless wrote:So, looks like definitely something to do with watchguard's packet inspection. I'm not seeing block messages, so looks our firewall doesn't see anything wrong with the traffic. The packet inspection is interfering somehow?


Yeah -- I've seen this in alpha testing. The client uses a raw TLS protocol for data transfer, but, a relatively recent protocol version of TLS (1.2). I suspect some security software may mistake this traffic for the older SSL protocol, or even just a less current version of TLS, and thus mangles the data in ways that lead to errors of various sorts.

This is a hairy issue to try and solve well. Indeed, fundamentally, handing over responsibility for secure content delivery is one of the most attractive things we might get by signing up with a main stream publisher.

But unless or until that happens, the best we can do is likely to just document the more common issues, and supply people with workarounds as we figure them out.

A virtual machine behind a corporate VPN is enough of a rare case that I'm happy enough to leave this one be, call it a "known issue", and move on to more pressing bugs :)

User avatar
sven
Site Admin
Posts: 1620
Joined: Sat Jan 31, 2015 10:24 pm
Location: British Columbia, Canada
Contact:

Re: Feb 01 - Initial testing / installation

Postby sven » Mon Feb 02, 2015 10:13 pm

mharmless wrote:6.Notes) Game fails to run, probably expected given the graphics above.
VM without acceleration error.PNG



Yup -- that's a known issue. You've managed to create a virtual machine that has less than the effective minimum requirements for the game.

On real hardware, I've seen this error happen once before on a clean windows install of a circa-2009 dell laptop -- in that case, updating the graphics driver did fix the issue.

We've also seen it happen on very old (circa 2004) AMD/ATI cards. But anything less than 10 years old, running with updated drivers, will hopefully be immune to particular error.

mharmless
Posts: 235
Joined: Sun Feb 01, 2015 11:11 pm
Location: Washington State

Re: Feb 01 - Initial testing / installation

Postby mharmless » Mon Feb 02, 2015 10:20 pm

Pretty sure I can lay hands on some hardware right at that cusp, might be worthwhile to expose issues that better hardware can mask. I'll poke around tonight.

User avatar
sven
Site Admin
Posts: 1620
Joined: Sat Jan 31, 2015 10:24 pm
Location: British Columbia, Canada
Contact:

Re: Feb 01 - Initial testing / installation

Postby sven » Mon Feb 02, 2015 10:46 pm

mharmless wrote:Pretty sure I can lay hands on some hardware right at that cusp, might be worthwhile to expose issues that better hardware can mask. I'll poke around tonight.


While I certainly appreciate your enthusiasm, we've already got 3 serious AMD support bugs to work through :) And with limited dev-time, those are what I'd like to focus on. If you have time later on tonight, I'll try to get you some patches that may help things on your main gaming machine.

mharmless
Posts: 235
Joined: Sun Feb 01, 2015 11:11 pm
Location: Washington State

Re: Feb 01 - Initial testing / installation

Postby mharmless » Tue Feb 03, 2015 4:54 am

Been somewhat busy tonight, but found time to test this on the girlfriend's computer. I'm replacing machine 6 with hers, seems like there isn't really a good reason to revisit the VM installation. If I do bring a VM into the test configs again, it will be a beefier one with at least some acceleration enabled.

6) Cyberpower PC
6.OS) Windows 8.1, 64 bit
6.Spec) Intel Core i5-4570 processor, 8gb of RAM, NVIDIA GeForce GTX 660 graphics, 1tb conventional drive
6.Res) 1600x1200 primary with 1200x1600 secondary
6.Inst) Installed to the conventional drive, at the location "C:\Program Files (x86)\Stars in Shadow", with elevated privileges.
6.Notes) Attempted installation with standard user permissions, but this resulted in errors opening for writing. Running as an administrator got around that, of course. Perhaps a more elegant failure state when write permissions are a problem? Appears to run correctly. While checking out menus, just noticed that at this resolution the composite image results in option menu text and controls overlapping the game's title logo.
Options menu text overlap.PNG
Options menu text overlap.PNG (524.96 KiB) Viewed 23799 times

Playing around in windowed mode, there are multiple problems at various resolutions, many resulting in the game's logo appearing to be struck through by options menu elements. Was able to get similar behavior out of the load game menu, but not out of the new game menu. No matter how I resize the new game menu, the logo looks like it was meant to be that way, and fits naturally alongside the rest of it.

mharmless
Posts: 235
Joined: Sun Feb 01, 2015 11:11 pm
Location: Washington State

Re: Feb 01 - Initial testing / installation

Postby mharmless » Tue Feb 03, 2015 5:02 am

Last new machine for today, promise. I installed it on our guest PC, a cobbled together collection of parts from past gaming rigs.

7) Guest PC
7.OS) Windows 7 Home Premium SP1, 64 bit
7.Spec) AMD Phenom II x3 720, 4gb of RAM, AMD Radeon HD 7770 graphics, 250gb conventional drive
7.Res) 1440x900
7.Inst) Installed to "c:\games\Stars in Shadow"
7.Notes) Installed and ran fine. The disco ball sun effect is in force here, just like the media PC (4), despite being a dedicated card. Same behavior down to the human star being the only one to look correct. Did not actually start a new game, just observed the effect in the new game menu. Logs attached.

Now then, going to see about some 'ad hoc testing of the software product in an uncontrolled user-driven environment'
Attachments
07 - Guest PC - logs.rar
(4.37 KiB) Downloaded 657 times

mharmless
Posts: 235
Joined: Sun Feb 01, 2015 11:11 pm
Location: Washington State

Re: Feb 01 - Initial testing / installation

Postby mharmless » Tue Feb 03, 2015 6:12 am

Finishing up for the night, found a few more bugs. Pasting my notes in, cleaning up a bit but might still feel disjointed. Logs attached. Bold things I think are clearly bugs and not debatable interface/design choices.


Sven mentioned possible issues on ATI cards around the Phidi ships. Playing on Media PC (4), the unit with disco ball suns, so lets play Phidi. Check on their ships first thing, they look fine. I'm assuming concern was about the water graphical effect?

1280x800, the new next turn button overlaps content in the planet screen. The bottommost part ends up being above the upper right portion of the planet screen, depicting the race.

New turn button bisects the "SYSTEMS" text on the ship detail screen.

Colonized a new system, moving pop around with the transport I started with. Loading people still feels weird. I click, get an icon of my race, then I click on the planet to take from. Mechanically same as unloading, click button, get race icon, click planet, just feels like having load and unload be the same exact sequence is not quite right. A question mark for when you pick a planet to load from? Don't ask if there is only one planet colonized in this system? I'm not really sure.

Tried embarking and disembarking repeatedly.
Starting from pop in transport, new turn:
Disembark. Embark. Not allowed anymore.
Starting from empty transport, new turn:
Embark. Disembark. Embark. Not allowed anymore.
Getting two embarks in one case and one in the other seems off.

I just realized the bar showing planet population fills in from the center out to represent population growth. Man I'm dense. Also, I like this.

Girlfriend said "the factory and the lab are a lie." What do you mean? "They're a water race." Good point, I guess. I dunno, race customization at that level seems silly. It's clean and clear right now, bogging down in a half dozen graphical versions of the same building with the same job and same basic stats seems like the kind of thing other 4x space games are doing _wrong_.

Fighting another race, using a new scout based design "Escort". In a fight with both, it isn't obvious which is the scout and which is the escort. Have to click the large ship icon to see the name. Perhaps the ship design name under that large icon in the lower left?

So, two vs two fight. My escort has marine bays, so does the scout. Conducted a boarding action on one enemy ship with my scout, doesn't appear to have worked one way or the other. Advance with the escort, I am unable to choose to board. Ship detail screen shows marines. Click over to the scout to compare, crash. Bug report sent. Think it triggered upon examining a component. Restarting game. Lets isolate that.

New game. Go directly south to nearest star with my entire starting fleet. Enter tactical battle, luckily. Click on the scout, go into detail screen, examine warp lane amplifier, crash. Exit game, new game. Different race, human. Scatter fleet in all directions. Found a fight at Castor. Click colony ship, click into details, examine a module, crash. Ok simple reproduction. Uploaded that. Attaching game logs for good measure.
Attachments
04 - Media - logs 02-02-2015.rar
(6.97 KiB) Downloaded 654 times

User avatar
sven
Site Admin
Posts: 1620
Joined: Sat Jan 31, 2015 10:24 pm
Location: British Columbia, Canada
Contact:

Re: Feb 01 - Initial testing / installation

Postby sven » Tue Feb 03, 2015 5:21 pm

mharmless wrote:Playing around in windowed mode, there are multiple problems at various resolutions, many resulting in the game's logo appearing to be struck through by options menu elements. Was able to get similar behavior out of the load game menu, but not out of the new game menu. No matter how I resize the new game menu, the logo looks like it was meant to be that way, and fits naturally alongside the rest of it.


Yup -- this is a classic case of "it's a known issue, and we're working on it".

The Race selection and start menu screens use our newest, prettiest UI styles and conventions. The Options and Load Screens are both shoehorned in from a prior generation of UI assets. We'll upgrade all the UI elements across the whole game eventually -- but as it doesn't seriously affect gameplay, it's not currently a top priority.

The high-priority UI fixes are all involved in the strategic map report system -- there's we've got a lot of very fancy new icons waiting to go in, and when the whole system is up and running, that part of the game should be not only prettier, but also significantly more fun to play.

User avatar
sven
Site Admin
Posts: 1620
Joined: Sat Jan 31, 2015 10:24 pm
Location: British Columbia, Canada
Contact:

Re: Feb 01 - Initial testing / installation

Postby sven » Tue Feb 03, 2015 5:35 pm

mharmless wrote:1280x800, the new next turn button overlaps content in the planet screen. The bottommost part ends up being above the upper right portion of the planet screen, depicting the race.

New turn button bisects the "SYSTEMS" text on the ship detail screen.


This, in a way, is another instance of the same "incomplete ui upgrade" issue. Basically, if you're looking at an ugly interaction between 2 panes, and one pane has glowy borders, and the other doesn't, it's probably safe to assume that the non-glowy pane is scheduled to be upgraded to fit better with the glowy one.

That said -- this is a useful report -- a glitch I hadn't noticed when I pushed the changes last night.

User avatar
sven
Site Admin
Posts: 1620
Joined: Sat Jan 31, 2015 10:24 pm
Location: British Columbia, Canada
Contact:

Re: Feb 01 - Initial testing / installation

Postby sven » Tue Feb 03, 2015 5:48 pm

mharmless wrote:[b Click colony ship, click into details, examine a module, crash. Ok simple reproduction. Uploaded that. Attaching game logs for good measure.[/b]


Bug related to sloppy coding during pre-beta deadline panic. Easy fix. Posted and marked as stable. Thanks for the report.


Return to “Testing”

Who is online

Users browsing this forum: No registered users and 73 guests

cron