[quake3-bugzilla] [Bug 4962] Protocol version 69 + legacy support

[bugzilla-daemon at icculus.org]

https://bugzilla.icculus.org/show_bug.cgi?id=4962

Tim Angus <tim at ngus.net> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |tim at ngus.net

--- Comment #2 from Tim Angus <tim at ngus.net> 2011-04-27 16:38:51 EDT ---
I'm lazy:

<Timbo> so is the main point only to fix
http://aluigi.altervista.org/papers/q3noclient.txt ?
<Thilo> not only
<Timbo> but the main reason?
<Thilo> one of the majore reasons
<Timbo> because really that seems like a non-bug
<Thilo> the other reason is to get rid of the VoIP hacks
<Thilo> it's not a non-bug
<Timbo> it's only really an issue when you know the client address
<Timbo> which is almost never
<Thilo> skype, teamspeak...
<Thilo> forums
<Thilo> LAN connections
<Timbo> feh
<Timbo> tenuous
<Timbo> I mean I don't see it being used any where
<Timbo> the changes are waaaay too extensive for just addressing that
<Thilo> i told you
<Thilo> it's not only that
<Timbo> the #ifdef stuff is horrid
<Timbo> #ifdefs are a last resort
<Timbo> every time you add one you double your test cases
<Thilo> well, you can make your points on the bugtracker, i have to go now ;)(
<Timbo> and your failure modes
<Timbo> ok
<Thilo> night everyone :)
<Undeference> also being able to disconnect everyone on your network is kinda
neat
<Timbo> haha
<Timbo> also, supporting two protocols just sounds like pain
<Lakitu7> "surely someone on my campus is playing quake 3 right now"
<Timbo> and that's ignoring the #ifdeffery
<Timbo> it really seems more like a patches page type thing to me that
something we want in trunk
<Thilo> well i can get rid of the ifdeffery
<Timbo> it's massively disruptive to anyone using ioq3 as an engine
<Lakitu7> changes to "if" :]
<Timbo> the STANDALONE stuff was bad enough
<-- Henkan (~Henkan at h70n2fls31o1105.telia.com) has quit (Quit: Lämnar)
<Thilo> thats because your personal dislike of ifdefs
<Timbo> it's not a personal dislike
<Thilo> but i for one am not content to be stuck with a version number that id
can change on a whim
<Undeference> it seems like a bad idea for a server to respond on a different
address, since there's a good chance the client will never see the response
<Timbo> it's engineering sense
<Thilo> look at the hacks we're doing to get around the old protocol
<Thilo> seriously
<Thilo> reading past message end?
<Thilo> that's just as ugly
<Timbo> but it works, and isn't nearly as disruptive
<Thilo> disruptive?
<Thilo> you mean during an svn merge?
<Timbo> yeah, and future maintenance wise
<Thilo> the two voip bug reports would benefit from this new protocol
<Thilo> the changed code looks more than it really is
<Thilo> most of the patch size is the changed readme
<Timbo> alright, but this doesn't fix those things
<Thilo> we fix a bug that is exploitable and can disconnect any client on a
whim
<Timbo> so if you did, that means another protocol version?
<Timbo> so we support three protocols then?
<Thilo> no
<Thilo> we only need to maintain the one ID legacy protocol
<Thilo> that was our stated goal
<Thilo> if we increase the version number we need not support the other
protocol numbers
<Thilo> users using ioq3 can upgrade to newest version
<Thilo> those stuck with idq3 can still play together with ioq3 just minus our
features
<Undeference> problem comes in with different distros maintaining different
versions with incompatible protocols. not much of a problem as long as they can
figure it out and fallback to the old protocol
<Timbo> so if an arbitrary ioq3 version connects, and it doesn't speak the new
protocol, what happens?
<Thilo> Timbo: falls back to idq3
<Timbo> ok
<Timbo> I still hate the idea of multiple protocols :)
<Thilo> well alright, let's see how this plays out ;)

-- 
Configure bugmail: https://bugzilla.icculus.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the QA contact for the bug.