Gamersgate got back to me and claims that their testing of their server version of the game indicates there is no problem with the source.   Just tried to DL again but still problems, albeit this time an error in a different file.   I don't know much about networking, but I gather that, a very likely factor is that Gamersgate's distribution system is just not good enough to handle a game that large. I have no problem downloading very large games (or other applications) on this rig/network connection, it is only this one Gamersgate game (Arma 2 . . . well to be honest I'm also having some problems with my version of Arma 1 Gold from them). I guess it is just time to write off the larger games I own from their site and procure additional copies from other distributors.  

1.90.145472 new PROFILING branch with PERFORMANCE binaries, v06, server and client, windows 32/64-bit, included linux 32-bit server + sync with 1.90.145471 hotfix 02   important note:
use ``-servermod=`` for mods not available at client, which are used only at server, for mods at both use ``-mod=`` (this will is mandatory for stable 1.90 main branch too)   make sure the server doesn't have obsolete files in-between default data , see example : https://forums.bohemia.net/forums/topic/222192-190-update-troubleshooting/?do=findComment&comment=3344880   note: since 145146 (client, server, headless client) generated `cache.ch` within path ``C:\Users\%windowsusername%\AppData\Local\Arma 3\DataCache\`` unless defined `-profiles=` path via command-line param   when defined then the `cache.ch` generates at `-profiles=` directory into sub-folder `\DataCache\`   temporary workaround (resolved since hotfix 02 (1.90.145471)): sometimes create `cache.ch` fails because of special character in windows username or paths use "-profiles=test" startup command-line to try run game   https://www.dropbox.com/sh/582opsto4mmr8d8/3BSy9PdRGm https://drive.google.com/folderview?id=0B03-H4YIbhkFMUt5RzNqZjFlNGs
available via STEAMclient/STEAMcmd as branch too, read https://community.bistudio.com/wiki/Arma_3_Steam_Branches#Arma_3_Server   BIForum feedback: https://forums.bistudio.com/topic/160288-arma-3-stable-server-184-performance-binary-feedback/?do=getLastComment   Arma anything discord:  https://discord.gg/arma

Then you should know that forwarding range 2302-2306 or just 2302-2305 and 2306 seperately has same end result.

I'm disabling the normal action menu while I'm performing a user interaction (via uiInGameSetEventHandler).  I'm using the same keys and buttons as the action menu for this interaction ("action", "closeContext", "prevAction", "nextAction").  Every step of the interaction is flawless now except for one niggling little problem: if the user clicks the middle mouse button to complete the interaction, the interaction completes, but the action menu pops up.  This is not true of keyboard keys; if the interaction is completed by pressing space, then the action menu does not appear. To be clear, both space and middle mouse button are defined as "action" keys. I'm intercepting KeyDown and MouseButtonDown events, then using inputAction to spot the various actions that I want to override.  I've verified that I'm overriding the mouse click and the space. Has anyone else seen this sort of asymmetry between mouse and keyboard events?  I'm not happy to have the quality of the user interaction damaged by this, but I can live with it.

Nice