EQEmulator Forums

EQEmulator Forums (http://www.eqemulator.org/forums/index.php)
-   General::News (http://www.eqemulator.org/forums/forumdisplay.php?f=594)
-   -   EQEmu Considering Deprecating Clients (http://www.eqemulator.org/forums/showthread.php?t=40853)

Akkadius 08-24-2016 02:31 PM

EQEmu Considering Deprecating Clients
 
Due to a variety reasons, we've been considering deprecating all game clients up to the Rain of Fear clients.

This means that the following clients would no longer be maintained in the main server code:
  • Titanium
  • SoF
  • SoD
  • UF

Maintainability being one of the biggest reasons, another large reason is because of having to hack server code heavily which introduced a large amount of complexity in trying to implement new inventory code.

This also would forces players on all servers to be on a newer version of code.

We would branch our code into a Github branch to a frozen state where all clients work, but further features and implementations would only support RoF+ clients.

I would like to extend this discussion into a thread for players and server operators to chime in on.

Again, this is nothing final, but bringing the discussion public for all to chime in on.

We've briefly discussed this in coders and would like to see others feedback.

N0ctrnl 08-24-2016 02:33 PM

I have only supported the RoF2 client for some time on Vegarlson. This would definitely allow for some good things to happen much easier.

Mortykins 08-24-2016 02:34 PM

On Raid Addicts everyone uses UF - ROF2 , none of my player base uses those old clients to my knowledge.

Morty

demonstar55 08-24-2016 02:49 PM

I don't really see dropping UF as a possibility, which means we still need to support the older inventory system, which means there is no reason to drop tit :(

SoF we might consider dropping since that client is all kinds of screwy though :P

noudess 08-24-2016 04:49 PM

My server is Titanium based. What would I be giving up moving all my clients forward?

I run old versions of many zones (nektulos, lava, tox, kerra, etc.) if that matters. I went wit Titanium because it felt more like the older game, and my server is an original only world for the most part. And, all my players use Titanium, as they came over from P1999.

DanCanDo 08-24-2016 05:36 PM

Although this wouldn't effect the server project I am working on, because the editing is
being done to support both UF and RoF clients but, like noudness, I would wonder what
kind of effect this would have on existing servers built around older clients, that are in
the habit of updating their code from gitpull all the time. Servers like P99 have nothing
to be concerned about since, (I'm guessing) they do their own code updating.
As far as using a client, as a player, my preferences will always stick with the UF client,
just because it's a "like the feeling" thing, so it would be a sad day for me, if support was
dropped for that. But I'm guessing everyone out there has a client they love more than
than the rest.

Mortow 08-24-2016 05:57 PM

I, personally prefer the RoF2 client. All the players on my server use only that client as most of them like the newer features. I would love to see things move forward, especially in the inventory department.

DanCanDo 08-24-2016 06:22 PM

Quote:

Originally Posted by Mortow (Post 250703)
I, personally prefer the RoF2 client. All the players on my server use only that client as most of them like the newer features. I would love to see things move forward, especially in the inventory department.

Yes, the RoF2 client seems to be popular because of the features that come
with it. While some servers develop, they work to "roll back" EQ, and others like
to see it roll forward. I've read so many opinions from players ranging from the
"old school vets" who miss the original game, to the modern day players who
love things like in-bag clickies. (chuckle)

ghanja 08-24-2016 06:58 PM

Does the "frozen state" denote what it suggests, in that, no more commits what-so-ever? Sounds a little extreme, which is why I'm leaning towards believing that what is meant by it is that, any -active- development would be frozen (i.e. no concentration on the branch/fork? that supports the older clients).

I'm far from as knowledgeable of C++ let alone the EQEMU source which, I feel many would admit, it's a mess (no complaints mind you, but strictly from a technical stand point I am speaking). If being able to work around that source to offer improvements both of functionality and structure/organization of the code while doing so without the fear of "shit, adding that RoF feature now broke this with the older clients", by all means, it would seem it's a dire necessity. If for nothing else but to help keep your sanity. Maybe too, it would allow for cleaning up parts of the code, like qglobal rewrite, who (GD thats a mess), etc. because for all intents and purposes only one client would be supported.

Heck, the inquiry of the public is appreciated, but, in the end, we're at the mercy of those intimately familiar with the source (or those daring enough and with enough time in their lives to become intimate with it), so go with what ensures longevity both as a whole AND of the participating developers. However, I will say, I hope that if a developer doesn't find it to lead to getting burnt out, if a bug fix here or there on the "old master" branch (UF, SoF, etc.) could be done, that would be more ideal to most, I believe?

TL;DR = Game for whatever makes it easier on the project and the developers, albeit, I do hope that the "old master" isn't totally abandoned if a dev could toss it a bug fix or two time to time.

blackdragonsdg 08-24-2016 09:25 PM

From a coding point of view I would imagine it is a great idea to deprecate the older clients. But what about new players or even some existing players that can not acquire RoF/RoF2 because it is not readily available. A time frozen state is fine but I guess a lack of RoF/RoF2 client availability would be my biggest concern.

Maze_EQ 08-24-2016 09:44 PM

This is a godsend.

I have added so much bullshit for older clients.

Rof2 for ever.

rudeboy88 08-25-2016 11:10 PM

I wouldn't mind seeing as ghanja said an "old master" version that may get the occasional update or so. I am excited to see how the single-client or 2 client server possibility pans out as when I've ran servers I frequently have restricted things to only 1 to level the playing ground for people (ex. titanium lacking extended target).

With that said I'd understand this choice- I've used the same client to play 80% of all servers I played here for almost 10 years now, it seems like the appropriate time to look to better things.

One point I'm curious about is how this might effect other projects like the Unity client one? Haven't read up on that enough lately but seemed worth mentioning given that it's been around a while. In any case if eqemu's really moving this way I see where RoF/2 would be among the most ideal, and for consistency's sake in the future, the only client(s).

Secrets 08-25-2016 11:57 PM

<iframe width="560" height="315" src="https://www.youtube.com/embed/ZXsQAXx_ao0" frameborder="0" allowfullscreen></iframe>

Tyen05 08-26-2016 03:22 AM

dropping all client support except for Unity & Unreal.

imo

daerath 08-26-2016 08:35 AM

Quote:

Originally Posted by Akkadius (Post 250694)
Due to a variety reasons, we've been considering deprecating all game clients up to the Rain of Fear clients.

This means that the following clients would no longer be maintained in the main server code:
  • Titanium
  • SoF
  • SoD
  • UF

Maintainability being one of the biggest reasons, another large reason is because of having to hack server code heavily which introduced a large amount of complexity in trying to implement new inventory code.

This also would forces players on all servers to be on a newer version of code.

We would branch our code into a Github branch to a frozen state where all clients work, but further features and implementations would only support RoF+ clients.

I would like to extend this discussion into a thread for players and server operators to chime in on.

Again, this is nothing final, but bringing the discussion public for all to chime in on.

We've briefly discussed this in coders and would like to see others feedback.

I love this idea. The sheer quantity of legacy code that could be removed in the RoF+ branch would vastly simplify things and may even speed up introduction of new features if only because the backwards compatibility issues would be reduced (or outright eliminated).

And for those people still on the older clients, it's not like they couldn't branch the source if they wanted to maintain it for their own servers.

bakajikara 08-27-2016 02:04 PM

I am all for this it time move forward not stay in past go for it.

RinoaV 08-27-2016 03:17 PM

I could be wrong here, but as far as I know the steam client has been upgraded and the new version no-longer works, if that is in fact the case where will my players(most of which use Titanium) get the Rof client? (Amazon and Ebgames/Gamestop don't carry it), I mean If there is no legal way to obtain the Rof client then we will be stuck like we were when the 6.2 client was the only working client and impossible for new players to get, unless of course your planning to chase the steam/live client opcodes each time they are changed like was done in the past?

If there is no way for my players to obtain the Rof client then I will be forced to use the frozen code, those of us that have no C/C++ knowledge will be *bleep* out of luck because there will be no way to fix any bugs we run across.

In my opinion *take it for what you feel it's worth*, removing viable clients and freezing development on all non Rof server code to add an inventory upgrade for a single client that may no longer be available is not a good solution, there has to be some kind of alternative method which won't alienate players and servers.

-Rin

Kingly_Krab 08-27-2016 06:17 PM

I believe that deprecating old clients is a good idea. Think about it this way, we have to port every single bit of functionality backwards, map opcodes from older clients when we find them in the new client and hope they maintain the same information and size. Honestly there is nothing but good in doing away with the older clients. It would cut development time down a lot since we wouldn't have to worry about, "Will this break Titanium?" Yes, freezing development on older clients may come as a detriment to some, but with the majority of our servers being on the new source code and having the ability to utilize newer clients, there's no need to use something with 1/10 of the functionality. ROF2 has direct augmentation, 10 inventory slots, massive bag-sizes, more spell slots, more spell book slots, more everything. If we were to continue development on ROF2 singularly, the benefits would far outweigh any shortcomings.

TL;DR: I support the deprecation of older clients, as I believe it would allow for better future development of systems we simply cannot have due to backwards-compatibility constraints.

Addendum: If you want to see how bad we try to support older clients, take a look at common/patches/Titanium.cpp or any of the other clients. We have several thousand lines of translation code. Unifying into one client we could remove a lot of legacy code and standardize everything.

Uleat 08-27-2016 06:34 PM

I honestly think if we're going to go the singular client route, then now is the time to develop our own client.


We can't focus on one client and hope to add a new one when it comes out.

There are already changes in the available steam client that are as drastic as the UF to HoT inventory models..and I hear live spells are completely incompatible now.

EQEMU2010 08-27-2016 06:35 PM

Myself and tons of others use UF that I know of I would like for it to be supported still if thats possible but thats up to the admins who run this project.

I have the free steam version of ROF2 but dont use it on most servers.Unless they are really basic ones.

Uleat 08-27-2016 06:41 PM

Strictly using the inventory model as a standard for client focusing..

..the UF client would not be supported since it uses the old 'mapped range' design, where RoF+ uses a pseudo-4D matrix one.

N0ctrnl 08-27-2016 07:28 PM

Quote:

Originally Posted by RinoaV (Post 250745)
In my opinion *take it for what you feel it's worth*, removing viable clients and freezing development on all non Rof server code to add an inventory upgrade for a single client that may no longer be available is not a good solution, there has to be some kind of alternative method which won't alienate players and servers.

To be fair, that's quite an oversimplification. "Viable" is a relative term. When you're talking in development time and effort, it's really not. That's why this conversation is taking place :)

Proxeeus 08-28-2016 11:00 AM

Probably a stupid question but anyway:

What would deprecating older clients mean regarding the use of old/classic models, like the old skeletons (white & brown flavors), some qeynos citizens NPCs, and classic spell effects? Would that force me to use newer models & effects? Or could I keep this level of customization?

demonstar55 08-28-2016 11:04 AM

Love actually supports those now. I'm unsure if rof2 does, something I've been meaning to find out.

Maze_EQ 08-28-2016 01:19 PM

People are only mad about RoF2 because they are still using the free UF active hack compile.

demonstar55 08-28-2016 02:50 PM

Quote:

Originally Posted by Maze_EQ (Post 250780)
People are only mad about RoF2 because they are still using the free UF active hack compile.

So you're saying I should make use of all the cheat detection packets the client sends to us?

daerath 08-28-2016 02:54 PM

Any idea as to the approximate reduction in size of the Emu source if all of the non-RoF2 code was removed?

Uleat 08-28-2016 04:07 PM

Well, if you cut out EQEmu::legacy, you'll probably shrink it to half its current size :D


EDIT:

That's not exactly easy to estimate..I wouldn't expect huge reductions though (probably less than 5% of its current size)

DanCanDo 08-28-2016 05:23 PM

I personally would love to see one client compatible with all emu servers, no matter what
client it is. Yes, the various customization prompt us to download files for the individual
server, but it would still make things easier in the long run.
But as far as the "features" introduced with modern clients, the servers that do the "era
rewind", may have a desire to disable client features (like the in-game maps on P99) or
some of the features that come with RoF2.

RinoaV 08-28-2016 05:23 PM

Would it be possible to get the new free-to-play client from steam looked into?

I know it's asking a lot but my thinking here is both sides would be appeased, we could move forward even further than expected, the old clients could be removed and we could give fair warning to all our players to "Download now or lose out!". The downside to this is of course more work for the devs, but I think if they are willing this may just be a solution?

-Rin

Uleat 08-28-2016 06:03 PM

I had about 20 smart-ass comments pop into my head..but, decided to go with the most appropriate.


Choosing the steam client is our current model..that unfortunately has left us client chasing rather than client focusing.

They tend to update steam days after we get it working...

RinoaV 08-28-2016 06:07 PM

Sorry, I was unaware of how often it was updated.

I was just trying to find a solution to a problem that will effect my players.

Uleat 08-28-2016 06:13 PM

That's just it though..they usually don't update it.


RoF was updated within 2 months and RoF2 was updated less than a week after one of the devs got it working.

I had started on TDS, with a dev branch, and about the time I actually got to the point where I could get into the game, they updated steam :/

RinoaV 08-28-2016 06:32 PM

Ouch, yep I can see how that would get annoying really fast, and why my im-possible solution would be a no-go.

I wish I had the programming knowledge you guys have, I could help you out. unfortunately all I have been able to do to help is add the old cultural armor to the database way back when, and stumble upon the guildmaster crash that you guys fixed in record time. Maybe with time I can figure out C/C++, it just might be my only solution.

rhyotte 08-29-2016 12:49 PM

I truly wish we had our own open source client, but that is highly non-trivial.

Still though,

Thanks to all of the dev's who work so hard on EQEMU!!

rdnck 08-29-2016 07:54 PM

I have no dog in this hunt, save that of a person who has spent as many or more hours on the emus as I have on the prod servers. I will say that we, as players on the emus, need a source for a viable client. Of whatever flavor finally makes the cut. I'm currently playing with a UF client, because the RoF2 client I had has gone bonkers, and I have to reset all my hotbars / socials if I cross a zoneline... o.O AND, I've been singularly unsuccessful at finding a place to get one.

Y'all make a decision, as the devs for the emus, select the client you want us to use, and make it available to us. :cool:

EQEMU2010 08-30-2016 08:16 AM

I like RoF2 tons of great features I prefer it but yeah Maze is right myself and tons of others always use free active hacks and don't want to pay a sub to use RoF2 version, I have compiled my own working version in the past but its haven't taken time to make it again and was on a SSD that went bad. sidenote:dont buy OCZ SSDs.

Up to the Devs, im game to be ROF2 only...

Maze_EQ 08-30-2016 08:57 AM

I'm always right, there's almost 0 reason to use a client other than RoF2.

I haven't used anything since Trevius made RoF2 work.

blackdragonsdg 08-31-2016 09:52 PM

Guess my one concern over this topic is now moot. I was reading the PEQ forums a few minutes ago and the general section was really amusing. I think some people have lost their minds because that one link is a disaster waiting to happen.

brokentechnology 09-01-2016 10:08 AM

I am all for this, my only concern is getting the client legally. I fully own it on Steam, but, that is not an option as the goal is to not chase live.

Obviously I can search the high seas and it's out there, but that is not the wanted approach.

How would we address the client moving forward? Would you update to match F2P (Steam) and say "Update on this day, create backup, don't update" and lock it to whatever client than or a different development approach?


All times are GMT -4. The time now is 02:51 AM.

Powered by vBulletin®, Copyright ©2000 - 2017, Jelsoft Enterprises Ltd.