open All Channels
seplocked Test Server Feedback
blankseplocked Mass-Testing 2.0 - Results
 
This thread is older than 90 days and has been locked due to inactivity.


 
Author Topic

CCP Tanis

Posted - 2010.03.03 19:53:00 - [1]
 

Howdy folks,

Saturday, February 20th

Summary:
  • Average FPS: 21.8

  • Average Player rating: 5.2 (1=fail, 5=same as TQ, 10=win)


During the course of this test; the following issues were reported, which we will be following up on:

  • Players report very-low FPS in some cases, even on high-end machines

  • Fleet-swapping may add load to nodes

  • Fleets, or being in one, seems to cause the most noticeable lag/FPS hit for players


More details:

We were able to get around 400 players for this test. The test involved two fleets, ‘campers’ and ‘attackers’, one jumping into the other. We were able to use FleetFinder ads to bring players into fleets much more easily compared to previous test events. We were able to test jumping between 2 dedicated nodes as well as from a dedicated node to a shared node. Almost no lag was reported when jumping a 200-man fleet into another of roughly equal size; however, when jumping 400 players at once and if combat is happening on the destination node, there came reports of “no-load” issues. Combat occurring caused the most noticeable difference, by far. In addition to this, we also ran a brief test of fleet-swapping; meaning having an entire fleet drop and join a new fleet en-masse. This last test showed that there is a little extra load when large numbers of people join a new fleet, we will be following up on this internally.


CCP Tanis

Posted - 2010.04.12 15:11:00 - [2]
 

Howdy folks,

Saturday, March 20th

Summary:
  • Average FPS: 17.5

  • Average Player rating: 5.1 (1=fail, 5=same as TQ, 10=win)


During the course of this test; the following issues were reported, which we will be following up on:

  • Still seeing low FPS reports, seems to be getting worse

  • Fleet and overview windows cause the most reports of “lag” from players

  • Being in a fleet still seems to result in noticeably decreased performance

  • Internal tools need improvement


More details:

We were able to get around 500 players for this test. The test involved two fleets, ‘campers’ and ‘attackers’, one jumping into the other. We tested jumping between 2 dedicated nodes and also jumping between 2 systems on the same node. We did not have time to get to the starbase siege portion of the test. Most lag was reported while forming the fleets, or while combat was occurring on the destination node. Most players reported the overview causing issues for them, too frequent updates making it hard/impossible to pick targets and also clearly impacting the client’s performance. FPS reports this round were terrible, many players were reporting below 1 FPS. Fortunately, developers were able to get some good data in chasing down the load-lag issue further.

CCP Tanis

Posted - 2010.04.12 15:14:00 - [3]
 

Howdy folks,

Saturday, April 3rd

Summary:
  • Average FPS: 13.1

  • Average Player rating: 4.7 (1=fail, 5=same as TQ, 10=win)


During the course of this test; the following issues were reported, which we will be following up on:

  • Overview updating too often/causing lag is the #1 issue reported by players

  • Avg FPS reported by players is consistently dropping

  • Internal tools need improvement


More details:

This round of testing saw the exact same test cases as the previous 2 runs, we were able to get ~400 players this time. During the first jump the grid loaded quickly and smoothly, as expected. Players reported quite smooth performance from the client and server. During the second jump we began seeing quite a few reports of grid-load-lag, incredibly low FPS, dysyncs, and a few crashes. The crashes can likely be accounted for by players not running logserver in server mode, based on discussions with players. This is the first time we’ve seen reports of players getting ‘traffic control’ type messages when initiating a jump. The message they get tells them that they’re in a queue to jump and must wait 2m 30s (was the exact same time reported by all players who got the message). That said, Loading into the system was a bit smoother this time around than in previous tests.

CCP Tanis

Posted - 2010.06.09 17:32:00 - [4]
 

Howdy folks,

Saturday, April 17th

Summary:
Average FPS: 11.1
Average Player rating: 5.1 (1=fail, 5=same as TQ, 10=win)


During the course of this test; the following issues were reported, which we will be following up on:

  • Overview updating too often/causing lag is the #1 issue reported by players

  • Avg FPS reported by players is consistently dropping

  • Client memory usage is unusually high

  • Players find the PI UI hard to use


More details:

For this test event we ran the standard jump-test in addition to a special test of the Planetary Interaction feature set. In total, we had ~500 players participate, with ~1000 players on the server (assuming many of these folks were also playing with the new PI feature, outside of the test system). The PI feature testing went fairly smoothly, little to no impact on the PI nodes and no reports of lag. There was plenty of great usability feedback though. During the jump testing, we saw a noticeable improvement in playability from previous tests, especially for those players who disabled in-space brackets. We saw again the 2m 30s “jump queue” messages for many players, though all did eventually load into the target system.

CCP Tanis

Posted - 2010.06.09 17:37:00 - [5]
 

Howdy folks,

Saturday, May 15th

Summary:
Average FPS: 18.3
Average Player rating: 4.5 (1=fail, 5=same as TQ, 10=win)


During the course of this test; the following issues were reported, which we will be following up on:

  • Major memory leaks in the client

  • Guns wont cycle/do damage properly in high-load situations

  • PI UI could use a little more love

  • EVE Gate login is teh slow

  • Overview updating too often/causing lag is the #1 issue reported by players



More details:

For this test event we ran the standard jump-test in addition to two special tests of the PI feature (, as well as testing of EVE Gate. In total, we had ~350 players participate, though we only had ~200 players for the EVE Gate portion of the testing. The PI feature testing went fairly smoothly, little to no impact on the PI service nodes and the only ‘lag’ that was reported was on hitting the ‘submit’ button. The only real load that was placed on the PI services was when we used scripts to pre-configure planets for the test. EVE Gate testing was also quite smooth, with the only major complaint being that initial login was very slow. Overall, most players are quite happy with EVE Gate and seem very keen to use it. During the jump testing, we saw a noticeable very dramatic upturn in reported FPS from players, especially for those players who disabled in-space brackets. We saw again the 2m 30s “jump queue” messages for many players, though all did eventually load into the target system

CCP Tanis

Posted - 2010.06.09 17:47:00 - [6]
 

Howdy folks,

This is a special double-header report. This covers special mass-tests run to address specific client performance issues that cropped up during Tyrannis development.

Thursday, May 20th & Friday, May 21st

Summary:
Average FPS: 19.7
Average Player rating: 6.5 (1=fail, 5=same as TQ, 10=win)


During this mass-test, we tested a few potential fixes for the following issues:

  • Negative Resource cache (i.e. Memory leaks)

  • Overview stutter

  • Fleet window stutter



More details:

During the course of Tyrannis development, we found a few major issues which we needed to correct: client memory leak (Trinity), overview stutter, fleet window stutter. To test fixes for these issues, we ran two mass-tests on the 20th and 21st. During the first test, which had roughly 450 pilots, it was immediately apparent that we’d not isolated the root cause as memory usage was still astronomical and not releasing properly. The test on the 21st, which had only about 250 pilots, yielded far better results. In the second test; we saw no reports or indication of the “negative cache” issue, resources appeared to be released on time, and everything seemed to run much smoother, by all accounts. That having been said, the 2nd test was with fairly low numbers of players, thus cannot give us 100% confidence that these issues have been completely resolved. A later test, with better numbers, will be required to achieve optimal confidence.

CCP Tanis

Posted - 2010.06.09 17:55:00 - [7]
 

Howdy folks,

Saturday, May 29th

Summary:
Average FPS: 14.9
Average Player rating: 4.8 (1=fail, 5=same as TQ, 10=win)


During this mass-test, we tested a few potential fixes for the following issues:

  • Overview updating too often/causing 'lag' is the #1 issue reported by players

  • Overview presets and configuration is problematic for players



More details:

For this test event we ran only the standard jump-test. The During the jump testing, we saw again a decrease in FPS reported by players. That said, we are not into dangerous territory yet. This test we also see some reports of the load-lag issue again; though much more minor than before. We also see more complaints about the overview, this time that it’s not updating often enough. We saw again the 2m 30s “jump queue” messages for many players, though all did eventually load into the target system.

CCP Habakuk

Posted - 2010.07.08 17:57:00 - [8]
 

Hi,

Tuesday, June 22nd

Summary:
Average FPS: 18.7
Average Player rating: 6.6 (1=fail, 5=same as TQ, 10=win)
Number of participants: ~250


  • In this mass-test we concentrated on testing performance improvements for the overview, the fleet window and brackets.

  • The performance was very good (partially also caused by the lower number of participants).

  • Some minor issues (specially for the fleet window) were found (and mostly already fixed on the next day).


More details:

For this test event we ran the standard jump-test, modified with some extra warp-tests. During the test we concentrated on the overview, the fleet window and brackets. For the overview and the fleet window we were specially in the behavior while other players joined or left the fleet. Several ISD bughunters helped us with this by leaving and joining the fleets as often as possible. We found small issues, but the overall results were very good. About 15 minutes after the test several server-nodes crashed, after trying to gather server-side data.

CCP Habakuk

Posted - 2010.07.08 18:30:00 - [9]
 

Hi,

Saturday, July 3rd

Summary:
Average FPS: 13.0
Average Player rating: 5.0 (1=fail, 5=same as TQ, 10=win)
Number of participants: ~310


  • In this test we concentrated on server-side improvements for jumping between systems.

  • The tested changes looked very good, although the overall performance in the test was worse than in the last test (partially caused by the higher number of participants


More details:

For this test event we ran the standard jump-test, but we stayed in the systems X-BV98 and Poitot to test the same cross-node jump several times. Server-side options were changed between the jumps to analyze their impact, specially regarding traffic control. Several issues were observed during the test like high module lag or problems with drones. Both the client as the server performance was worse compared to the last test, but this was expected.

CCP Tanis

Posted - 2010.07.21 18:18:00 - [10]
 

Howdy folks,

Thursday, July 15

Summary:
Average FPS: 21.6
Average Player rating: 3.8 (1=fail, 5=same as TQ, 10=win)
Number of participants: ~150



  • In this test we focused more on server-side improvements for fleets jumping as well as modules cycling.

  • Most of the changes looked good, while others clearly made things worse. Testing was incredibly laggy as a result of us changing various server-side toggles for the new fixes



More details:

For this test event we ran the standard jump-test, jumping first from X-BV98 into Poitot, en-masse and then jumping fleet X from F67E-Q into Fleet W's gate-camp in MHC-R3. Combat occurred in the MHC system. Server-side options were changed between the jumps to analyze their impact, again focused on jump queues. In addition, during the combat portion of testing, we also tested various new server-side settings to affect module activation delays. Results of the new toggles were mixed, the jumping ones looked good, the module ones not so much. More work will be needed on the module activation issues in large fleets. Both the client as the server performance was worse compared to the last test, but this was entirely expected.

CCP Habakuk

Posted - 2010.08.25 15:11:00 - [11]
 

Hi!

Thursday, August 5th

Summary:
Average FPS: 19.3
Average Player rating: 3.8 (1=fail, 5=same as TQ, 10=win)
Number of participants: ~550


  • Changes to module cycling

  • Awesome participation

  • good results, but the changes needed more refinement



More details:
Check out this DevBlog from CCP Tanis, he talks about MassTesting in general, but also gives specifics to this mass test.


------


Thursday, August 19th

Summary:
Average FPS: 12.9
Average Player rating: 5.4 (1=fail, 5=same as TQ, 10=win)
Number of participants: ~380


  • More changes to module cycling

  • Improvement for bombs and smartbombs

  • Testing EVE-Voice



More details:
While we had not as many participants as last time, the test was still quite successful. A previous change to module cycling was improved (read more about this in the DevBlog from CCP Veritas) and a new change for module cycling was tested. We also tested successfully an improvement to the performance of area of effect weapons (like smartbombs).
This time we tried to use EVE-Voice for all voice communication in the fleets and while it performed quite well, once the voice channel was joined, there were issues with joining the channel. We figured out what the problem in this case was and have a solution for it ready. In one of the next mass tests we need to check if this solution really works and if there are other problems.

CCP Habakuk

Posted - 2010.08.30 17:25:00 - [12]
 

Thursday, August 26th

Summary:
Average FPS: 27.3
Average Player rating: 6.1 (1=fail, 5=same as TQ, 10=win)
Number of participants: ~200


  • Planetary Interaction - everything fine

  • Spamming jump - problems found

  • standard test - not enough participants for useful data



More details:
Only 200 players participated in this test, which was much less than expected, but we still got some useful results.
As first part we had a short test of Planetary Interaction to go sure that the last code improvements do not have bad consequences for the server load. The server load was very low and we could not find any major issues.
As second part we tested a potential improvement to prevent increased server-load from spamming the "jump" button. We identified two problems, one of them, which was unrelated to this improvement, caused major confusion as cloaking and uncloaking was not working correctly. We will test this again in the next mass-test to get real data.
The last part was jumping into a gate-camp. We noticed again the cloaking bug, as most attackers were not visible after uncloaking. The fight had nearly no lag, but the metrics of this test are not really usable due to the low number of participants.

CCP Habakuk

Posted - 2010.09.07 19:43:00 - [13]
 

Thursday, September 2nd

Summary:
Average FPS: 14.6
Average Player rating: 4.1 (1=fail, 5=same as TQ, 10=win)
Number of participants: ~310


  • The "grid not loaded after jumping" issue was experienced again when spamming the jump button and we got useful data.

  • Fleet-fight worked without major lag or problems

  • Some problems with the UI like e-war icons, which were not removed



More details:
This test was mostly concentrated on the consequences of spam-clicking buttons like the jump button. Several users (both players, bughunters and Devs) experienced the issue of being unable to load the grid after jumping to another system. The logs were very helpful and we will test a theory on what might causing it in the next mass test (on the 9th September).


CCP Habakuk

Posted - 2010.09.15 13:54:00 - [14]
 

Thursday, September 9th

Summary:
Average FPS: 12.3
Average Player rating: 5.2 (1=fail, 5=same as TQ, 10=win)
Number of participants: ~590

  • Very high participation - awesome!

  • Gathered data for failing to load grid when spam-clicking the jump button

  • The fleet-fight was quite lagged out (as to be expected with nearly 600 players)


More details:
We had about 590 players at the same time in this test, the total number of participants was about 660. For the first steps we jumped between MHC-R3 and F67E-Q and spam-clicked the jump button as much as possible. This was to test a improvement to the jump-process and to gather more data regarding the issue of failing to load the grid. As comparison we had a jump without spam-clicking into Poitot.
Then we had two fights at planets in Poitot to test an fix to stuck modules.
As last step we had a standard jump + fleet-fight test in X-BV98. This fight caused high load on both the server-node and the clients, as is to be expected with nearly 600 players. There were several reports of problems with exploding ships, but beside this there were no "major" problems, beside the lag.

CCP Habakuk

Posted - 2010.10.13 16:53:00 - [15]
 

Thursday, September 16th

Summary:
Average FPS: 11.2
Average Player rating: 3.4 (1=fail, 5=same as TQ, 10=win)
Number of participants: >800

  • Too high participation - some of the logs were useless due to noise

  • lots of LAG

  • Duplicate entries on the overview with a specific server-setting

  • Good data from the smartbomb test, which verified data from thin clients


More details:
We had lots of participants (about 800 in the fleets, more than 1000 on the server), which caused a long list of problems. This included creating additional fleets in the last minute, lots of waiting time, undisciplined participants, noise in server logs and client logs, ...
On the good side the smartbomb test (last step) resulted in useful data and we have a better feeling now about the capability of the server hardware.

---

Saturday, September 25th

Summary:
Average FPS: 10.1
Average Player rating: 5.8 (1=fail, 5=same as TQ, 10=win)
Number of participants: ~580

  • Optimal participation - awesome!

  • Still problems with loading the grid after spam-clicking the "jump" button

  • Warping one fleet into another fleet was much better with server-side optimizations.

  • low FPS - caused by overview?


More details:
Great test! A problem with the overview was fixed on the next day, which probably caused the really bad average FPS in this test.

---

Thursday, October 7th

Summary:
Average FPS: 13.6
Average Player rating: 6.8 (1=fail, 5=same as TQ, 10=win)
Number of participants: ~410

  • First test with the code for the Incursion expansion: no major performance problems found

  • Performance for the Incursion feature tests was slightly better than expected


More details:
In this test we had a slightly lower participation, but the results were still very good - mostly for the Incursion feature.

CCP Habakuk

Posted - 2010.10.21 13:38:00 - [16]
 

Thursday, October 14th

Summary:
Average FPS, zoomed in: 9.8 (9.9 for players with AA enabled)
Average FPS, zoomed out: 24.4 (23.9 for players with AA enabled)
Average Player rating: 5.1 (1=fail, 5=same as TQ, 10=win)
Number of participants: ~400

  • Incursion test gave good results

  • Memory leak detected

  • Many client crashes


More details:
The first part of the test consisted in jumping into a system (Harroule) with an active Incursion (which changes several ship-attributes). In this system the Incursion influence level was changed (which also changes the ship-attributes of all ships in the system). Then the fleets jumped back to MHC-R3. We gathered very useful data and found that especially the process of jumping out of the system caused a high server-load, which is being addressed by improving how the ship-attributes are being changed (the result should be the same).
After the next jump a small group of Incursion NPCs was waiting - the server load was as expected.
The test was concluded with a fight in Poitot, where one fleet was camping and the 2nd was jumping in. The server performance did not show any surprises, but some players noticed a very bad memory performance of the client, which we could reproduce as major memory leak.
The biggest problem through all the test were a high count of client crashes, specially after jumping to another system. We could identify one problem, but future test will be necessary to show if the problems are gone.

CCP Habakuk

Posted - 2010.10.27 17:03:00 - [17]
 

Thursday, October 21st

Summary:
Average FPS, zoomed in: 11.5
Average FPS, zoomed out: 29.0
Average Player rating: 5.8 (1=fail, 5=same as TQ, 10=win)
Number of participants: ~280

  • Low player count, but it was ok for this type of test

  • Fix for duplicate entries on the overview confirmed

  • No major issues found for 64bit Inventory during the test, some server errors recorded

  • One server-node crashed during shooting the POS - caused by the new performance profiling (but it gathered very valuable data).



More details:
This test was basically a standard fleet-test, using the 64bit inventory code, but test-steps were added for verifying the overview fix and for testing performance profiling improvements.
The jumps and the fleet-fight at the gate did not show much lag, mostly due to the low amount of participants. For testing the overview we added a test-step to dock at a station and undock all at the same time. The overview did not show any issues, but we noticed a quite high server-load while undocking, which will be investigated.
During shooting a control tower in X-BV98 a new profiling tool was activated, which crashed the server-node - but good data could be salvaged in the time between activation and crash of the node.


CCP Habakuk

Posted - 2010.11.17 17:37:00 - [18]
 

Thursday, October 28th

Summary:
Average FPS, zoomed in: 9.2
Average FPS, zoomed out: 17.2
Average Player rating: 5.1 (1=fail, 5=same as TQ, 10=win)
Number of participants: ~480

  • good results for the destiny optimization, but one bug found, which caused node-crashes before the test

  • Low client-side FPS



More details:
This test was mostly for testing the first part of destiny optimizations aimed at reducing the impact of missiles on the server load. The gathered data was very promising, but we found one bug, which was able to crash server-nodes (which caused for examples crashes of the FD-MLJ-node earlier to the test).
In this test we were specifically interested in client crashes and while there were reports about single crashes it was much better than in the test on the 14th October.
In the first half hour several players got moved to FD-MLJ by the moveme script, which was caused by a bug in the script. This should be fixed now.

----

Thursday, November 11th

Summary:
Average FPS, zoomed in: 13.8
Average FPS, zoomed out: 22.2
Average Player rating: 5.7 (1=fail, 5=same as TQ, 10=win)
Number of participants: ~300

  • Low player count, specially in the beginning - caused by late publication of the patch

  • Many client crashes, many disconnects, even some system crashes

  • Problems detected with the improvement to ship destructions

  • remapping the system did not work properly



More details:
This test showed a lot of problems - which means that the test itself was a success (even if the tested improvements were not really successful). Cool
Low player count: After several internal problems the patch for the mass-test was made public much too late for the size of the patch - nearly 1.5GB. I will make sure that this does not happen again (in this extent) in the future. At start of the test we had about 210 participants, at the end of the test we were at a bit over 300.
Crashes: We got many reports about disconnects, client crashes or even system crashes: While we were able to identify (and hopefully fix) at least one crash issue - more data will be gathered during the next mass test.
Ship destruction: We found several side-effects related to this improvement and we need to work more on this.
System remapping: The remap did not work like intended, an improved version will be tested in the next mass test.

CCP Habakuk

Posted - 2010.12.04 21:27:00 - [19]
 

Thursday, November 18th

Summary:
Average FPS, zoomed in: 7.7
Average FPS, zoomed out: 17.7
Average Player rating: 4.7 (1=fail, 5=same as TQ, 10=win)
Number of participants: ~420

  • Lots of client crashes

  • system remapping worked (but with small issues)

  • missile improvements successful


More details:
In this test we added several steps to our standard test-plan:
A few minutes before the official start we started and ended a Sansha Incursion to check if there are performance issues if an Incursions starts in systems with a high player count. The performance hit was less than expected, but we noticed other issues, for example the UI did not update correctly for some players.
As next part we tested destiny-improvements, specially for reducing the impact of missiles. This included several jumps, a gate-camp and a fight at a starbase. The improvements worked and there were no related problems reported.
At the end of the test the system X-BV98 was moved (=remapped) to another server-node. All players in the system were disconnected (as planned) and could reconnect within a very short time. Some problems were reported, that the position of some ships and the state of the starbase was not exactly like before the remap - this will be investigated further.
Through the whole test we had a very high amount of client crashes (including several clients of Devs). Most of the crashes were graphics related and should have been fixed now (to be tested in the next mass test, which includes Incursions and the new character creator).

----

Thursday, November 25th

Summary:
Average FPS, zoomed in: 5.9
Average FPS, zoomed out: 15.5
Average Player rating: 5.5 (1=fail, 5=same as TQ, 10=win)
Number of participants: ~510

  • The server load of ship death has improved, but more work needed

  • General server performance has improved in the last months

  • The client performance has not changed and causes more problems with higher numbers

More details:
In general this mass test showed a very good server performance, especially for the fact that we had more than 500 participants. The client performance on the other side was not as good, especially the FPS. But the number of client crashes and disconnects was very low this time.
After gate-jumps and during the fights we made some checks for position desync-issues, which showed that we seem to be in a much better position, but it is clear that not everything is fixed.
During the last step (mega smartbomb) the load on the server was less than in previous tests, the server was at 100% CPU for only about 2 minutes. On the bad side many clients took much longer to recover. This will be worked on more (and tested in future mass tests), as we found also some bugs, which need to be fixed.

CCP Habakuk

Posted - 2011.01.29 21:19:00 - [20]
 

Phhhh, did I really not post any updates since the beginning of December? Well, here are the results of the last three tests:

Thursday, December 9th

Summary:
Average FPS, zoomed in: 4.3
Average FPS, zoomed out: 12.0
Average Player rating: 4.9 (1=fail, 5=same as TQ, 10=win)
Number of participants: ~600

  • High player count (~600) without extreme server lag

  • No problems with ship-death improvements

  • Position desync confirmed




Saturday, December 18th

Summary:
Average FPS, zoomed in: 15.6
Average FPS, zoomed out: 25.4
Average Player rating: 6.1 (1=fail, 5=same as TQ, 10=win)
Number of participants: ~300

  • Low player count, nearly no server lag

  • A few client crashes, but better than before

  • Several defects found for incursions




Thursday, January 13th

Summary:
Average FPS, zoomed in: 9.2
Average FPS, zoomed out: 25
Average Player rating: 5.2 (1=fail, 5=same as TQ, 10=win)
Number of participants: ~350

  • No performance issues found with the Incursion feature

  • Tech 3 issue (already fixed internally) disturbed the test

  • Only a few random disconnects

  • Many client crashes (related to Tech 3 issue?)

  • No problems with drone damage messages


More details:
This mass-test was quite disturbed by the problems with tech 3 ships, which caused some clients to fail to display other objects. They were banned from the test, but some were still present.
The user count was not high enough to cause massive server lag (except during gate jumps).

CCP Habakuk

Posted - 2011.02.05 21:43:00 - [21]
 

Thursday, February 3rd

Summary:
Average FPS, zoomed in: 9.2
Average FPS, zoomed out: 19.5
Average Player rating: 4.5 (1=fail, 5=same as TQ, 10=win)
Number of participants: ~380

  • No real problems found with missile optimizations

  • The POS caused a lot of lag - to be investigated

  • again many client crashes reported, especially at ship-death



In the first part of the test (shooting a POS) a lot of server lag was experienced, which seems to have been caused by the POS and its batteries. This is being investigated. The other parts of the test showed no real lag-related problems. No relevant problems were reported about missiles.

CCP Habakuk

Posted - 2011.03.24 13:32:00 - [22]
 

Thursday, February 10th

Summary:
Average FPS, zoomed in: 13.4
Average FPS, zoomed out: 18.1
Average Player rating: 4.9 (1=fail, 5=same as TQ, 10=win)
Number of participants: ~230

  • Low player count

  • Several major bugs, both server-side and client-side

  • No Mac and EveVoice support

  • No server-side lag (due to low player count)



--
Thursday, February 24th

Summary:
Average FPS, zoomed in:11.2
Average FPS, zoomed out: 23.0
Average Player rating: 5.4 (1=fail, 5=same as TQ, 10=win)
Number of participants: ~380

  • Smooth test without major problems

  • Very good server performance

  • More problems with client performance than with server performance, but varies with user hardware

  • Mac patches brokenSad

  • Still some issues with client crashes



--
Thursday, March 10th

Summary:
Average FPS, zoomed in: 14.5
Average FPS, zoomed out: 29.8
Average Player rating: 5.7 (1=fail, 5=same as TQ, 10=win)
Number of participants: ~140

  • Very low player count, due to:

    • Huge patches

    • Server available too late

    • Many bugs


  • Remapping had to be canceled (because of a bug, which was found shortly before the test)

  • no server lag (not enough players)



--
Thursday, March 17th

Summary:
Average FPS, zoomed in: 17.3
Average FPS, zoomed out: 33.0
Average Player rating: 5.7 (1=fail, 5=same as TQ, 10=win)
Number of participants: ~130

  • Again very low player count (similar reasons as on the 10th)

  • First mass test on the new test-server Duality (this also reduced the player count)

  • Low sec POS bash performed better than the one on the 3rd Feb. (but difficult to compare due to low player count and different server)






 

This thread is older than 90 days and has been locked due to inactivity.


 


The new forums are live

Please adjust your bookmarks to https://forums.eveonline.com

These forums are archived and read-only