[ Updated threads · New messages · Members · Forum rules · Search · RSS ]
  • Page 1 of 1
  • 1
Forum » [FHSW Europ] FHSW 0.61 Server: » Suggestions: » Reverting back to 0.61 from 0.611
Reverting back to 0.61 from 0.611
bubuDate: Monday, 2020-08-17, 12:51 PM | Message # 1
Colonel
Group: Admins
Messages: 381
Awards: 28
Reputation: 11
Status: Offline
Yesterday's Pakfront with approx. 20 disconnected players (cumulative value in total) was a last nail to my patience about this version.

Next issues:
- Map loading screen wrong for Berlin-45 and Midway-42
- Similar disconnect issues in Tarawa (even though that was already happening in 0.6 but not to such level)
- Westminster-42 might have the same issues (iirc)
- Appeninnes might have the cap-flag issue, Jan and Torquemada disconnected during the map test, me/e-3/gameplayer did not disconnect

Main question: Shall we roll-back to FHSW0.6? From my view point, I expect our only loss would be compatibility with Japanese server, and possibly Super Yamato on 1 map, as well as initial loss of player base till they realize we are going to roll back.
 
mardukDate: Monday, 2020-08-17, 1:02 PM | Message # 2
General
Group: Admins
Messages: 1592
Awards: 42
Reputation: 12
Status: Offline
Although I haven't been playing for three weeks I didn't witness comparable issues like with version 0.55 in 2014 ( http://www.fhsw-europe.com/forum/44-574-1 ) and for example the wrong loading screens aren't a real issue.

This is a gaming community. People come here to relax. If you start an argument you will get kicked.

Quid pro quo.
 
bubuDate: Monday, 2020-08-17, 1:43 PM | Message # 3
Colonel
Group: Admins
Messages: 381
Awards: 28
Reputation: 11
Status: Offline
Sorry, yes, indeed, I forgot to explain the map loading screen issues are just a symptom (as far as I understand), both Berlin-45 and Midway-42 load as Savoy1940, pointing to the lexiconall.dat issue (as Eye pointed out). The problem stems from that - are other maps ok or which other map has similar issues because of this file/are there any more issues stemming from this file?
 
starking018Date: Monday, 2020-08-17, 6:54 PM | Message # 4
Colonel
Group: Friends
Messages: 359
Awards: 21
Reputation: 9
Status: Offline
TL;DR: terrible idea

For a second I thought you were parodying someone or something.
I don't see a single issue which didn't already exist in 0.61 and in 0.6:
- Pakfront has been disconnecting me several times if I joined after the start of the round as far ago as I can remember. I see a couple of SSM files for it, but I don't know if they have any effect on this.
- Tarawa disconnects many players at only one event during the course of the map (IIRC it was the capture of the northernmost flag, which can't be recaptured). Maybe it can be fixed via a SSM: e.g. try making the flag allied from the start, or tweak whatever spawns from it.
- The issue with lexiconall.dat existed in all old versions. In some it was worse - actually causing crashes on maps (it turned out it didn't cause crashes in 0.611, as far as I could test). The usual effects include wrong loading screens and texts, and missing maps in the map lists for singleplayer or starting a new server from the menu (http://www.fhsw-europe.com/forum/21-438-1).
- Let's try Westminster-1942 again without the old SSMs which were removed at some point.
- Apennines also has an old SSM, which may or may not have any bad effects (the map is unchanged since FHSW 0.6).

If a map is too buggy I'd rather lose that map until it is fixed than losing other improvements and additions which came with new versions. Going back to 0.6 would mean you'd have to re-test all maps which we didn't have back then, including Fall_of_Berlin-1945 and all the latest maps in our map pack, and replacing any objects which didn't exist in 0.6.

I'd rather test and evaluate some of the many custom maps which we've never tested, and have a new mappack version.


Air Troll a.k.a. starking018

Message edited by starking018 - Monday, 2020-08-17, 6:55 PM
 
waldhurzakDate: Monday, 2020-08-17, 7:59 PM | Message # 5
Lieutenant
Group: Trusted
Messages: 65
Awards: 3
Reputation: 2
Status: Offline
if 0.611 already exist its very bad idea for whole community, to roll back any previous version

Best Bf1942 version for FHSW:
http://www.mediafire.com/file/818jqqez94ka80h/BF1942-HD.rar

rafal
 
E-3Date: Tuesday, 2020-08-18, 9:33 PM | Message # 6
Lieutenant Colonel
Group: Bronze Donator
Messages: 185
Awards: 8
Reputation: 9
Status: Offline
I am somewhere in between at this point I don't like the idea of running a different version than the Japanese. However I would prefer
map, server, and mod,  stability overall. but at this time we have too little data to
assume that 0.611 is unstable time will tell.


I'm an old bf1942 player I go far back all the way to 2002
and I also play cnc generals.
 
mardukDate: Tuesday, 2020-08-18, 9:41 PM | Message # 7
General
Group: Admins
Messages: 1592
Awards: 42
Reputation: 12
Status: Offline
Quote E-3: " but at this time we have too little data to assume that 0.611 is unstable time will tell."

Exactly. Today for example I've been setting maps (Suomussalmi, Britain, Edsons, Bizerte, Gela) and we had no problems.


This is a gaming community. People come here to relax. If you start an argument you will get kicked.

Quid pro quo.
 
bubuDate: Wednesday, 2020-08-19, 8:12 AM | Message # 8
Colonel
Group: Admins
Messages: 381
Awards: 28
Reputation: 11
Status: Offline
Quote starking018 ()
Going back to 0.6 would mean you'd have to re-test all maps which we didn't have back then, including Fall_of_Berlin-1945 and all the latest maps in our map pack, and replacing any objects which didn't exist in 0.6.
Ah, damn, sorry I am talking about 0.61. These version numbers are killing me. So rolling from 0.611 to 0.61. I renamed the thread, sorry for confusion! sad

Quote starking018 ()
The issue with lexiconall.dat existed in all old versions. In some it was worse - actually causing crashes on maps (it turned out it didn't cause crashes in 0.611, as far as I could test).
Did we have any issues with it in 0.61?

Maps performance
I suggest of having approx. 20 people at server and running the following maps:
- Midway42 - No ssm
- Westminster42 - No ssm
- Appeninnes43 - ssm from E3 (freshly done)
- Tarawa44 - No ssm
- Berlin-45 - No ssm

- Pakfront: Has 2 SSMs applied: 001.rfa from 19/11/2017 and 002.rfa from 13/05/2018. Shall I remove them?

Alternatively, I am keen to removing all SSMs except those lately made as attempt to improve balance, what do you think about this suggestion?

Hardware performance
Next thing, the main server currently runs on Intel Core i7-3770, 3.40GHz, with 16GB RAM, 64b WinServer2012R2.
The currently significant processes running are:
- BF1918 server
- FHSW main server
- FHSW test server
- TS3 server
- Virtualized server on VirtualBox for FHSW Stats page (reserved 1 CPU core, 1GB of RAM, 300 GB HDD)

Performance wise, the VirtualBox uses 10-13% of the CPU, while each instance of BF1942 dedicated server uses 0.4-1% of CPU. However the test server was using 1.6% of CPU when it was running Midway-42, which I find interesting. Too many objects on the Midway42 map, maybe? Check these:







So from my understanding we should try modifying VM's params? Currently it seems to use 12% of the 1 core, if I understand resource monitor output correctly.

 
bubuDate: Wednesday, 2020-09-02, 8:35 AM | Message # 9
Colonel
Group: Admins
Messages: 381
Awards: 28
Reputation: 11
Status: Offline
Now operation Varsity needs to be removed....
 
Forum » [FHSW Europ] FHSW 0.61 Server: » Suggestions: » Reverting back to 0.61 from 0.611
  • Page 1 of 1
  • 1
Search: