Anope IRC Services

Please login or register.

Login with username, password and session length
Advanced search  


Pages: [1]   Go Down

Author Topic: I would like to but I can't update...  (Read 4124 times)

0 Members and 1 Guest are viewing this topic.


  • Anope User
  • Offline Offline
  • Posts: 158
I would like to but I can't update...
« on: November 07, 2008, 05:34:38 PM »


I recognized a lack of updated modules. I know: You (devs) already have to finalize 1.8 but I officially want to ask what will happen with all these nice modules that were made for the 1.7 unstable branch that are currently not compatible with 1.8? Will you ask/please the developers of that modules to update them (immediately)?
And in this case I want to beef you a bit (in a polite way): Why did you set a feature freeze long time ago so that the 3rd-party-module-devs and me, one part of the thankful irc admins, could think this modules will also work in the stable branch but still made a "huge" change shortly before the RC?

I know I have no claim to get updated modules but I know I can tell you that I cannot update to the newest version of anope because some of my favourite modules won't work on there. So what can I do? What will you do? And I guess also some other irc-admins can't update because they won't make their users saddened.



  • Guest
Re: I would like to but I can't update...
« Reply #1 on: November 07, 2008, 06:54:30 PM »

This point has already been stressed in a less friendly mannor by myself. It's being looked into and the incompatability of the most popular modules, such as bs_fantasy_ext is being considered. Module authors are never forced to maintain compatability, however I'm sure the Anope Team recongizes the importance of allowing authors to "catch up" before support is completely lost. The last thing Anope wants or needs is to drop support for versions of Anope that are used by people that can't simply upgrade. chaz and I have been constantly communicating on this topic and I feel that in the end, this issue would be resoved prior to support is lost. The Anope Team has already converted their orginal 1.8 forum layout.
« Last Edit: November 07, 2008, 06:58:10 PM by katsklaw »

Jan Milants

  • Team
  • *
  • Offline Offline
  • Gender: Male
  • Posts: 1372
Re: I would like to but I can't update...
« Reply #2 on: November 07, 2008, 09:59:28 PM »

It s up to the individual module developers to decide when/whether to update their modules.
There have always been modules that became out of date at all stages in development.. many modules for 1.6 and or early 1.7 don't work on 1.8 any more either. As for the "huge" changes made shortly before releasing the RC, they were needed to properly fix anope bugs. Updating a normal sized module to work with the newer versions shouldn't even take so long or be so much work, however many of the module authors have gone inactive.. So it doesn't really matter how big the changes are, the main issue is that it would seem modules are no longer being actively maintained. When that happens problems are bound to arise.

As for dropping support on older versions.. 1.6 will probably be supported until somewhere around march 2009, which is still quite soon for a stable branch, but it s soo deprecated... No more updates will be made to the  branch however .
1.7, being a development branch for 1.8 should go end of life much sooner: when people had issues with an 1.7 release the first thing we told them to do was always update to the latest release, in this case this is very true. 1.7 is/was a development branch. Personally I am inclined to say I wouldn't support any version of a development branch for more then one month after an update to that branch is released...
The fact that module authors need to catch up is - sorry if I offend anyone with this - largely on their end.. The .21 release was simply around for wayyy too long.. once the release cycle was put into gear again it would seem none of the module authors were following.. even though those "huge" changes were available as a patch on the bug tracker and later on svn with plenty advance before new releases were published. I wouldn't call the changerate in anope of the last months so high that it s impossible to keep up with..  :-X

And I suspect you were probably referring to bs_fantasy_ext as well, which hasn't exactly been following anopes updates either. However after joining the ranks of anope devs I had to change my priorities.. my 3rd party modules are less important then anope 1.8 (and soon 1.9) and I have to divide my time over all of these.. I will admit that most of the changes in anope that broke compatibility are to blame on me, yet I would make the sacrifice again.. I will update all  of my popular modules, however it may take some time.. I won't publish them before 1.8.0 goes gold anyways as 1.8 will be a requirement for all of them.
If you like me donate coins to 1FBmZVT4J8WAUMHKqpWhgNVj3XXnRN1cCk :)


  • Guest
Re: I would like to but I can't update...
« Reply #3 on: November 08, 2008, 04:13:12 PM »

IMHO, viper, your personal definition of "plenty of advance notice" is way too short. Case in point is adding CIDR support. That patch wasn't in svn for more than a few days. We all know that real life interferes with IRC. Some people don't get on everyday, every hour .. many don't do an svn check out everyday. Even as a QA I didn't have the time to checkout CIDR before the next dev release and I _do_ watch the bug tracker, I just simply hadn't the time. Thankfully none of my modules should be affected by changes after .21. However, this thread is proof that my more privately expressed concerns are right on target.

I'd also like to get something perfectly clear. Just because a specific release was out for longer than expected, it's not automaticlly obsolete. So what if .21 was out a year? Appearently it was stable enough as it was, but nope .. since it's old it must be replaced. .22 and forward has caused 10 times the headaches that .21 ever did. That should be a clue.

Here's an idea. Instead of just blindly going by age of the last stable release, why not take a poll and see what Anope's users want. Ask them if they want life sql in 1.8. Ask them id they want a fully functional BotServ with included, CORE bs fantasy commands, ask them if they want os ignores to be saved to db instead of only cached, ask if they'd like the option to have vHosts in NickServ instead of HostServ, ask them if they would like a HelpServ that is actually helpful that runs on trouble tickets instead of the pathetic waste of resources that it currently is (which by the way the current helpserv style was invented in 1994, it's now 2008 .. that's 14 years), ask them if they'd like to be able to mlock modes like +f and +j in 1.8. Ask them if they'd like to have all Serv bots optional, ask if they would like the choice of choosing AuthServ vs NickServ, Ask them if they are willing to wait while all this is added for 1.8 or if they'd rather have 1.8 as-is and wait for Anope 2.0 which could easily be another 2 years before getting such features that already exist in other services packages today.

The fact that there are few/no active module coders proves that anope can't sit around and wait for mod authors to pick up the slack. Additionally, your top 10 modules should be considered a list of desired features. Applications are designed with the USER in mind, not the desires of the development team in hopes the users agree.

I'm seriously not trying to be difficult, I'm speaking as an experienced Net Admin that likes and usually uses Anope speaking his mind about what he'd like to see in Anope. I also think that the Anope Team should ask themselves why the anope-ng team exists. Is it an attept as a fork? is it because they feel Anope is slacking? .. why? Appearently there is enough interest in Anope to evolve. I've said before that the anope-ng, Anope and denora team sohuld join forces into 1 team and fix Anope so it's more state of the art and not something left over from 1996. There is alot of talent in those 3 groups and I see no reason why they can't team up and build something useful that doesn't rely on talanted users to expand upon via modules.

Anope 1.7's cmd set is not vastly different than Anope 1.6 nor the version of Epona that it was forked from nor current Epona for that fact. Which isn't much different than IRC Services 1.4 which is similar to DALnet Services cmd set from the late 90's. All this time has been spent on back-end and modularity, which I agree is important .. but don't you think that the users would like to see and feel a more modern cmd set from todays services packages? I know I would.

« Last Edit: November 08, 2008, 05:47:23 PM by katsklaw »

Trystan Scott Lee

  • Contributor
  • *
  • Offline Offline
  • Gender: Male
  • Posts: 343
Re: I would like to but I can't update...
« Reply #4 on: November 08, 2008, 11:21:48 PM »

Viper is correct, its up the module authors to catch up with the times. I remember when I moved all of the defines from the code and gave anope part of its ircd independence. SGR was so pissed at me that I had to add depricated.h so that he could build some of his modules. Over time I think this idea got lost as we moved more forward and figured if they really wanted to keep up they would.

Kat, there was a time when anope had 2 module authors and one of them was an active dev member. The fact that today anope has a 3 or 4 public and probably many more none public coders is testament to the fact that its moving forward. Granted I believe that its seen now as a serious problem to over come with the changes made by the dev team. In time the mod coders will catch up for will fade away. And new ones will take their place.

In general, rushing a release is never the way to do it. If .21 was up there for a year, make sure there are no known segfaults and release it don't add to it, and wait till you got the next chance to work on it, his is not a race. Adding a large feature IMHO should wait at least in svn for a month with request on the forum for user feed back to make sure its working nicely.

As for the ideas put out by kat, yes all these features are a great idea. But I would like to point out a few features that have never made it into anope in the last five years (yes 5 years cause it was 2003 that we started this all)

1. Oper Flags - the debate is long, and code samples numerous, to date of the features that died and never got far.
2. MySQL - in July 2003, phase 1 was done, phase 2 wasn't that far behind, now its still a pipe dream.
3. Fixing static code problems - much of why news functions are not modularized is cause of this

So 1.7.x has had a long rough life. Its seen its share of project leaders, coders, and supports come and go. This has changed anope in many ways. I hope that chaz has the will and energy to keep the project moving forward.

As for -ng I think it was the logic off shot of the failed attempt of DrStein and GD with their project. Making -ng the next anope core again is where anope has wanted to go for some time. If you think Viper broke stuff then you have not seen what -ng will do.
my God my tourniquet, return to me salvation
Pages: [1]   Go Up