“This is true for all free software projects, not just ones that do dual-licensing.”
See comment that is in “awaiting moderation” state.
grin I hate to break up the fun here, but wouldn’t the time spent arguing here be better spent improving each’s favorite XMPP server? And curse you Daniel for having my same name! ;D I am the only one who’s allowed to be called Daniel . . . just like EST is the only true time zone!
(just trying to lighten the mood )
forks the name “Daniel”
Take that!
Hahahah open standards, open names!
Wonder what the source code to the name Daniel looks like.
Hopefully it’s not written in visual basic.
All in all, and on a more serious note, I am always glad to see a bunch of good well liked XMPP servers out there with their own followings. What fun would it be if there were only one server implementation? We can all learn from each other and since we are forced to approach scenarios differently by the sheer nature of the languages we are using or the architecture we have chosen, we can see problems approached in different ways. All of this is great for making the XMPP standards rock solid if you ask me.
So everyone raise a glass to your favorite server and ALL of the XMPP servers out there. wpjabber, jabberd1, jabberd2, openim, openfire, ejabberd, djabberd, tigase, and whatever else I’m missing! Kudos to all of ya!
Hi Sander! Just saw your comment about the python transports stuff. Clearly you thought I was Daniel from above. Either way, the python transports were losing a bit of interest from me before Jive wanted me to do this. On that front I have to implement both the protocol and the transport. While I did intend to keep working on them, I was getting “bored” with them. I -want- some folk to help with the project. If I could get a team of people working on them instead of just me, that would be awesome! The problem appears to be though that there’s not a lot of Python expertise out there. Sucks.
Originally when Jive approached me about this I balked and fought them. Tried to test under Jyphon but that crap sucks. I acknowledge that Python is a confusing monkey to deal with for a lot of admins. That sucks a lot but I can’t argue with it too much. Some of the documentation that badlop put up for ejabberd for installing the Py*Transports is some of the greatest docs I’ve seen for it.
Unfortunately it looks like James (PyMSNt) is also not having time for the Py*Transports either. Norman (PyYIMt, PyIRCt) was doing a lot with his but I haven’t seen him all that active lately.
I’d wanted an “excuse” to learn Java for a long time. I’m kind of a programming language whore. =) And this turned out to be a great opportunity to dive into it. Plus I don’t have to implement the freakin’ protocols from nearly scratch! I also hoped that by working with Openfire that I might be able to discover new and interesting ways to improve concepts that the PyTransports use. Some day I would like to quite frankly ride on the coat tails of Jive to push for some good XEP’s to be proposed to the standards committee based off work I’ve done with them that I think will aid transports in general. The world of transports -sucks- IMO. Everything I hate about the PyTransports I am able to cheat and use internal APIs with Openfire to get around. I believe some of the things I’ve learned I can use to improve other transports later. =) We’ll see though. I have to find like… free time to dive into writing XEP’s to make this come to fruition.
It’s also nice to be able to work with these java library implementations because they’ve implement things I haven’t yet in PyAIMt/PyICQt. I think I can borrow concepts and such at some point. =D (kinda like I’m already borrowing concepts from libpurple and such)
Anyway, don’t fault Jive for “stealing me away”. I wouldn’t have done this if I didn’t want it. I was bored with the Pys. I still sorta am. Sucks for the projects but really, it’s not a good thing that I was the sole developer anyway. Those projects -really- need some assistance. If you know of people that are interested and have python expertise please send them my way. I’d love to see a “team” working on them. =) I do at least have a helper for PyAIMt, but I don’t think he’s had a lot of time lately either.
And hell, if someone can write up transports in erlang, kudos to them! That language drives me insane to try to write for it. =) (you may recall I tried to throw my hat into writing some features for ejabberd a couple of years ago) Alexey has my envy for his grasp of it’s concepts. =)
I’ll sorta re-reply to the one post, since it wasn’t all there when I replied before.
"Though, this is technically true, businesses that want to do that have several major disadvantages over MySQL (the same is true for Jive) that makes competition with MySQL difficult:
-
They cannot provide a closed source version as MySQL can do (from their website: ???With that said, we recommend the commercial license to all commercial and government organizations. This frees you from the broad and strict requirements of the GPL license.???)
-
They cannot influence the roadmap without forking (which is made extremely costly, see further), so they have no control and cannot adapt to all their customers needs
-
There is a big threat for them when another company, which isn???t that friendly to them, acquires MySQL.
Regarding the last point, I found next interesting article: http://www.eweek.com/article2/0,1895,1930346,00.asp Just imagine what would happen when Microsoft???s Live departement acquires Jive Software???"
These are all risks to using free software, whether there are closed-source versions of the same products or not. If Microsoft bought Jive, then we’d still have the last version of OpenFire that was released as free software, and could build on it from there. Is it a risk? Perhaps. But in my opinion, not in the way you portray it.
“What do you mean with this? Can you explain? If I understand you right, I???m afraid to tell you that your employer started with posting something that is more excessive than Process-One???s last post: http://web.archive.org/web/20050810235429/www.jivesoftware.org/messenger/status_ report_2_2.jsp”
Jive isn’t my employer, but I think you know that now. I look at that post, and I don’t see anything that’s particularly excessive as far as self-congratulation is concerned. But, a few tidbits from the Process One blog:
-
http://www.process-one.net/en/blogs/article/according_to_ohloh_ejabberd_the_most _popular_open_source_xmpp_server/
-
http://www.process-one.net/en/blogs/article/ejabberd_passes_the_120000_download_ mark/
-
http://www.process-one.net/en/blogs/article/angie_millions_users_instant_messagi ng/
-
http://www.process-one.net/en/blogs/article/ejabberd_4th_birthday/
Not everything in the aforementioned posts is bad. I agree with many of the things that Process One says. But, seemingly arbitrary milestones and benchmarks, assertations that ejabberd is the “best” or “fastest” or “most scalable” at various things at various times. Various XMPP mailing lists have had many similar posts, and my personal opinion is that it gets old.
Does Jive say similar things about OpenFire? Yes, they do at times. But normally, from what I’ve seen, it’s not often as “we’re the best”, “we’re number one”, “zomg we pwnz j00”, but rather pointing out specific areas where their software has filled a need very well, and proclaimed that they have been successful in those areas. There’s a difference.
Perhaps it’s something lost in translation, or the fact that most of the OpenFire/Jive things are written from more of a business perspective. I can only take so much of somebody tooting their own horn and pretending that everyone else is inferior to them. It’s tiresome, and that’s where I take issue with it.
“* Fragmentation: It???s just a matter of time until Erlang-based transports will arrive. In fact I know some individuals that started with such transports. Fragmentation in transport projects is IMO a problem because it is a waste of developer resources in the community; these transport projects are temporary (until these networks become XMPP compatible, which already happened with Sametime, if I am right!)”
I don’t see the big networks becoming XMPP compatible in the near future. But, you mention that fragmentation in transports is a problem, then go on to talk about people writing transports in erlang. Why? Why not contribute to the existing Py* transports? Or is it so they can be ejabberd specific, like the OpenFire transport plugin?
“These consequences were predictable, so I guess this was done by intention. That???s why I get so angry when Jive Software???s PR shouts about its care for the Jabber community :-s It also explains my passionate posts over here, because I DO CARE about the Jabber community.”
To my knowledge, Jive has been fairly active on the various XMPP mailing lists, has worked on a number of XEPs, has been very open to ideas from the community, and has given a lot away to the community as well when they really didn’t have to. I don’t know how much support they get from outside developers, but Jive doesn’t appear to me to have any problems being successful on their own. Their forum software is used by a multitude of large corporations all across the Internet. I think they’re interested in real-time communication and the flexibility of XMPP, and honestly want to see it succeed. Any extra push XMPP gets can’t do anything but benefit Jive in the long run.
I don’t work on any XMPP software, nor do I work for any XMPP-related company. I’ve been a Jabber user for years and years, and have been running a public server for a fairly long time as well (off and on, on mostly the past two years or so). I know what works for me and what doesn’t, and OpenFire has been the biggest blessing to me and many other server administrators. Of course it has flaws, and things that could be done better, but it’s a work in progress and so far has done nothing but improve. Yes, Jive sells a commercial version of the server too. What does this do? It gets more companies involved in XMPP than they would normally be, because they have a company to work with for support, customization, etc. It’s not a bad thing, and my opinion is that it strengthens the community as a whole.
Hey all,
First, let me express deep thanks for the supportive comments in this thread. It’s clear that Sander’s viewpoints are not mainstream, which I’ve certainly found to be true in my interactions with the larger XMPP/Jabber community. Anytime I get together with Process One employees (and other community members), we buy one another beers and have a good time. That kind of camaraderie is pretty necessary since we all have to cooperate on building open standards together.
Hopefully many have seen the announcement for the XMPP devcon event in July (http://blog.xmpp.org/?p=25). In the spirit of XMPP goodwill, a beer is on me for all that can end up making it.
Hi,
really funny how people care about page ranking within yahoo/ohloh. As the last ejabberd release with new features (1.1.0) is a year old one may really wonder how active the project is and if new features will be added. This may be much more interesting than a ranking based on links or a feature comparison. Having said this I should add that I don’t like projects which release new versions every three weeks as no one has the time to do 20 software updates within a year, four releases within a year are usually enough.
LG
“Either way, the python transports were losing a bit of interest from me before Jive wanted me to do this.”
What I saw, is that after you started working on a nearly duplicate thing for Jive, you development dropped even much less.
“On that front I have to implement both the protocol and the transport. While I did intend to keep working on them, I was getting ???bored??? with them. I -want- some folk to help with the project. If I could get a team of people working on them instead of just me, that would be awesome! The problem appears to be though that there???s not a lot of Python expertise out there. Sucks.”
From my experience, to get a project of the ground, you need at least 1 passionated person. This passion then inspires other people and spreads like a virus. After a while you have built a community that is sustainable and which will not die after the first passionated person stops with the project (cfr. The Gimp!). You were on the right way with the Pytransports to replace the dead C-based transports. After Jive’s move, I’m afraid the same will happen with the Pytransports as what happened with the C-based transports. Too sad
“The world of transports -sucks- IMO.”
I agree, that’s why Jabber should become more popular so that the proprietary networks become XMPP compatible. At least, we don’t need a Sametime transport anymore AFAIR
“Anyway, don???t fault Jive for ???stealing me away???. I wouldn???t have done this if I didn???t want it. I was bored with the Pys. I still sorta am.”
Well, I do fault them for this. They could have at least made the transports you make for Openfire also accessible for other Jabber servers. In that way that would be consistent with their claim that they support to general Jabber community instead of harming it.
“Sucks for the projects but really, it???s not a good thing that I was the sole developer anyway. Those projects -really- need some assistance. If you know of people that are interested and have python expertise please send them my way. I???d love to see a ???team??? working on them. =)”
There are 2 options to form a team:
-
You can pay people
-
You can inspire people
“And hell, if someone can write up transports in erlang, kudos to them!”
Actually there were even 2 duplicate efforts to write transports in Erlang…maybe because these people ALSO (*) felt bad about Jive’s move.
(*) cfr: “It???s clear that Sander???s viewpoints are not mainstream, which I???ve certainly found to be true in my interactions with the larger XMPP/Jabber community.”
“I look at that post, and I don???t see anything that???s particularly excessive as far as self-congratulation is concerned.”
Besides the mentioning of the feature score on jabber.org without telling that this is based on subjective feature measurements, in particular this: “The discussion forums have become very busy. The Jive Messenger community is much more active than any of the other XMPP/Jabber Open Source projects. Below are some statistics generated by the Jive Forums instance at jivesoftware.org between May 1, 2005 and July 24, 2005.”
At the time I read this, I was pissed of because this claim was (and is?) for sure unfair to ejabberd for next reasons:
-
How can they know it is more active? They don’t have any stats of other projects.
-
They forget to tell that the Jive forums were mostly attended by customers from before the open-sourcing of Jive Messenger; the forums already were busy, this had nothing to do with the open-source community as probably most of these people asking questions at that time in the forums bought the closed-source version and were asking questions about that.
-
Jive does forget to tell that other open-source projects are more decentralised regarding support (different support websites in different languages, different mailing lists where you could find help, multiple chat rooms that were always open,…)
"But, a few tidbits from the Process One blog:
But, seemingly arbitrary milestones and benchmarks, assertations that ejabberd is the ???best??? or ???fastest??? or ???most scalable??? at various things at various times. Various XMPP mailing lists have had many similar posts, and my personal opinion is that it gets old."
Well, what is wrong with telling the things anybody can verify using benchmarks? At the time of these benchmarks (and probably still), ejabberd was (and is still?) the fastest and most scalable open-source Jabber server (not “open-source”, as it only was tested with open-source servers!). Regarding “best”, I don’t see any problem with that; any company will claim it has the best product/service… (and then you know at least some of this is probably true, at least in their eyes) So: “fastest” and “most scalable” can be measured, while “best” is an objective thing.
btw: a fun thing that is true for some of the Process-One posts you mention, is that after a few days you could see on the Jive blog a post regarding some number of forum posts, number of downloads of Jive software, or other statistical information. There’s nothing wrong with replies to blog posts, but Jive does not seem to know that it is common use in the blog world to mention and link to the blog article you are replying to :o)
“I can only take so much of somebody tooting their own horn and pretending that everyone else is inferior to them.”
See “Jive Messenger Project Status (August 2, 2005)” and in particular the note I made about it. For sure at that time Jive Messenger had not the most activity with an open-source origin, it was not even 3th in the rankings I guess.
“I don???t see the big networks becoming XMPP compatible in the near future.”
I’m more ambitious Do not forget Google has already one of the semi-major networks…
“But, you mention that fragmentation in transports is a problem, then go on to talk about people writing transports in erlang. Why? Why not contribute to the existing Py* transports?”
Because it is more easy to start new transports than it is to revive semi-dead projects; we’ve seen the same with the C-based projects, do you remember?
“Or is it so they can be ejabberd specific, like the OpenFire transport plugin?”
http://www.process-one.net/en/blogs/article/from_jabber_to_xmpp_im_standard_conv ergence/ --> @Matt: this was not written by me and it seems straightforward that there is a hidden complaint to Jive’s way of “helping” the community in it (cfr: ???It???s clear that Sander???s viewpoints are not mainstream, which I???ve certainly found to be true in my interactions with the larger XMPP/Jabber community.???) So, please do not do this: http://en.wikipedia.org/wiki/Ad_hominem I try to give some - yes, negative - feedback.
http://www.process-one.net/en/blogs/article/epeios_write_a_module_for_an_xmpp_se rver_once_run_it_everywhere/ --> “Epeios allows to run an ejabberd module on any XMPP server supporting the component protocol.” (so, this is IMO a major difference with Jive: giving back to the broader Jabber community instead of harming the Jabber community nd binding users to Openfire)
“To my knowledge, Jive has been fairly active on the various XMPP mailing lists, has worked on a number of XEPs, has been very open to ideas from the community, and has given a lot away to the community as well when they really didn???t have to.”
I do not try to convince people that Jive does nothing good, I just try to make some things more visible to people that seems to be very conflicting with Jive’s claim that they do not want to harm the broader Jabber community. (btw: this makes me think about another example: the unfair promotion of Spark by only making the automatically network deployment using Openfire available for Spark instead of making this feature client independent)
chuckle Sander, believe what you like man. =) I’ve been working on the python transport for a long time now and no one has stepped up to help me out really. All I got was complaints and wishes and occasionally a patch or two. I started getting “real help” from folk the moment I started with the IM Gateway plugin. It’s nice to have people actually interested in helping. =) Hell if you were so interested in it, why weren’t you offering to help join the team? If the projects die, so be it. If they don’t great. Chris seems to be interested in working with them but doesn’t appear to have time. If you know of so many people who you think would be interested in helping with the PyTransports, then by all means point them towards me and I’ll talk with them about getting them access to the repository and the ability to do real work. Amusingly enough, you just made me even -less- interested in working with the PyTransports with this last little comment by reminding me just how much actual community support I get with this project that I don’t get with those projects. You did inspire me to post something on blathersource.org’s news section asking for developers though!
Hey Sander. Not sure if you remember me, but we’ve chatted in both the ejabberd forums and on some MUC sessions in the past. (I’ve run ejabberd on Windows for some time now.)
Anyway, I think most folks involved with the Jabber/XMPP community at large know who you are, as you’re quite prolific on both forums and mailing lists. You’re obviously very passionate about ejabberd, and that’s great. And I don’t doubt your intentions are honorable. But having said that, if you read this thread not from your own eyes, but as a person who just visited the www.igniterealtime.org website or as a Jive/Wildfire/Openfire user/admin, let’s just say you’re doing both yourself and the ejabberd community a disservice.
Being a loyal fan of some software project is one thing. But doing so to the detriment of others is another. Regarding Daniel (jadestorm) being “lured away” (my expression here) by money…no! Say it isn’t so! A programmer who would like to be paid so they can eat and have a reasonable living? The insanity of it all! Even Linus Torvalds has (gasp) taken a job where they pay him money. Crazy. Seriously, though, what Daniel chose to do is HIS choice and should be respected. Let’s face it. Open-source is great, but it won’t by itself put food on the table. And whether we like it or not, we still gotta eat.
And regarding Jive making plugins that only work with OpenFire and not for all XMPP servers…I don’t even know where to begin. Besides the fact that each XMPP server implementation uses a different gateway interface (assuming they have one), making doing so quite the challenge, this is a bit over the top. Jive is a business. That it uses the MySQL business model is their choice. That they pay people to write software for their products is, quite frankly, only logical. Does it suck for the Py* community? Sure. Is it “fair”? Maybe not. But life’s not fair. Get over it. Do what you can for the things in life that matter to you.
I mean, some folks might consider the fact that you’re sitting here, on a “competitor’s” website writing blog entries, to be a form of trolling. Not sure if Jive/OpenFire folks do the same on the ejabberd forums/mailing lists/etc., but honestly, it does no one any good.
I’m sure you’ll have some rebuttal, but honestly, take a look at this thread. Think of the combined time it took every reader who (like myself) took the time to read ALL the comments to get a fair picture of the conversation. Add in the time it took others to respond. Factor in the time it took you alone to both read and respond to these comments. Wouldn’t that time have been better served helping folks on the ejabberd forums? Or writing docs for ejabberd? Or maybe even helping code in Erlang?
(Heck, the only reason I’m typing this response is because it’s Sunday and I’m stuck at the office waiting for some hardware diagnostics to finish on a misbehaving server so that I can go home. So I came online and got caught up on my reading and hit this thread.)
Now I happen to know you don’t know (at least as of last year) any programming languages. You told me so in a MUC session. Why not take the time and energy you have and go learn Erlang so that Alexey isn’t doing all the work? I downloaded the tutorials on Erlang and gave it a shot, but it’s just not my language. Like some folks find one human language easier to learn than another, it’s the same with programming languages. (I’ll never understand my one friend who found Japanese easier to learn than German or Spanish, but go figure.) Erlang is quite powerful, but I’d be lying if I said it came as easy as C-like languages (C, C , Java, etc.) for me.
Anyway, it’s clear that even within our community (and with this I mean the ENTIRE Jabber/XMPP community, from Jabber Inc. to Google to Jive to ejabberd and all the others involved in any and all capacities, whether commercial, open-source, or some combination of the two) that things get tense at times. It would be nice that instead of sniping at one another, each focuses on the projects/things that they want to see succeed. That way we all win.
And honestly, before this initial post, I’d never even heard of ohloh, so I wonder how much of an impact, if any, any such ratings have in the grand scheme of things. I think all projects would be better served focusing on their code, documentation, etc., than seeing if they’re the flavor of the week.
P.S. In case you’re wondering, Sander, I am on this site because I like to keep tabs on the various XMPP/Jabber projects out there. And Wildfire/Openfire is such a project, and I think they’ve done an incredible job making an XMPP server that’s easy to setup/administer. It’s not perfect (no project is), but with luck, all the projects will incorporate each other’s best ideas and the world at large will be better for it.
Thank you Frank =) Very well put! In case you are wondering, I’m not getting paid a salary or anything. =) I started on the project and then when I got a first release out they rewarded me (sponsored my project, a common thing done in the open source world apparently) with something that turned out to aid development for all of my projects. =D Later they rewarded me something else, but generally I’m doing this on my own with no real compensation. If they highly need some sort of functionality out of it though, they can always “bribe” me to focus on what they need. To be honest, I was a tad jealous of James (PyMSNt) for a while as he was actually getting paid to work on PyMSNt for a time.
What I -wanted- to happen, originally, is for Jython not to suck and for me to be able to just make PyAIMt and PyICQt work with it so it could be easily packaged alongside Java server implementations. But I did some research on that and well… it didn’t work out at all. (Twisted doesn’t work … lots of things don’t work … rather disappointing)
If I recall correctly, joscar is what lured me to the world of java more than anything. It’s a very thorough AIM/ICQ library and at the time it was going to be the base of Adium X’s support for AIM/ICQ, so I knew it would have a solid developer set and I wouldn’t have to keep reinvesting the wheel. =) Hasn’t turned out exactly that way on the ICQ front, but hey…
“Hell if you were so interested in it, why weren???t you offering to help join the team?”
As already should be known, I’m no coder. Though, several of the improvements of the Py*transports tutorials on the ejabberd website are mine.
“If you know of so many people who you think would be interested in helping with the PyTransports, then by all means point them towards me and I???ll talk with them about getting them access to the repository and the ability to do real work. You did inspire me to post something on blathersource.org???s news section asking for developers though!”
This will probably not help I’m afraid; to attract new enthusiastic contributors, you should give the good example by alo being enthusiastic and ambitious. Otherwise, they’ll prefer to start their own project from scratch or join another project that has enthusiastic people in it.
“You???re obviously very passionate about ejabberd”
Actually, I’m not anymore until some issues are fixed in the ejabberd community. But I still do like Process-One more than Jive Software as they never harmed the Jabber community whilst claiming they support it in all ways.
“A programmer who would like to be paid so they can eat and have a reasonable living?”
Read my whole reasoning about this problem; I BELIEVE IT IS GREAT THAT JADESTORM IS PAID BY JIVE SOFTWARE; read my whole post about this issue what is the problem.
“But life???s not fair. Get over it.”
Yes, I know. But the key problem is that Jive Software CLAIMS to admire the Jabber community. I just want to open the eyes of people that DO admore the Jabber community…free speech you know. What even makes me more passionate is that Jive Software tries to prohibit me to open people’s eyes. Jive Software unfortunately is a shareholders-only focussed company it seems
“I mean, some folks might consider the fact that you???re sitting here, on a ???competitor???s??? website writing blog entries, to be a form of trolling.”
If people see me as a competitor, they in fact see the Jabber community as a competitor as I am only complaining about issues that concern the Jabber community. Did I said anything about ejabberd? I don’t think so. Also, as I already said, I’m also not so kind anymore about Process-One, and currently I am focussed on Coccinella. But my main focus is the Jabber community, the project doesn’t matter for me.
"Wouldn???t that time have been better served helping folks on… "
Nope, protecting the Jabber community is IMO a more worthful task than working on a specific project that uses Jabber.
“Anyway, it???s clear that even within our community that things get tense at times. It would be nice that instead of sniping at one another, each focuses on the projects/things that they want to see succeed. That way we all win.”
I don’t agree. Criticism can help to get things right again. E.g. do you remember the complaints from the Jabber community when Google Talk was not able to federate with the bigger Jabber community. Instead of denying that this was something that harmed the Jabber community they gave some arguments why they did not yet federated with other servers (they had concerns about spam) and after a while (probably when they investigated it was only a minor threat) they actually listened to this criticism and started federating with the broader community.
To make things more clear, I am NOT complaining about the software produced by Jive Software in this thread. Instead, I have some concerns relates to Jive Software strategic actions and the way they harm the Jabber community (and not only ejabberd or Coccinella, but also Tigase, jabberd14, jabberd2, Jabber XCP, Jabber libraries, other Jabber clients, amongst others). IMO Jive Software exploited the whole Jabber community in some ways to please their shareholders at the expense of their stakeholders, and I don’t like this at all as I am one of these stakeholders. HTH to understand that I am not comparing software in this thread.
“To be honest, I was a tad jealous of James (PyMSNt) for a while as he was actually getting paid to work on PyMSNt for a time.”
Yes, what SAPO did was a great example of support the Jabber community.
Sander – your accusations of us harming the community seem a bit nutty to me. But I think we’ve hashed through all of that several times over already, so it’s not worth arguing further.
To everyone else – you probably think I’m a bit nutty for letting these posts from Sander through moderation. I’m just hoping that letting him have his say will let this topic of conversation reach its end. I’m certainly ready to move on…
“To everyone else ??? you probably think I???m a bit nutty for letting these posts from Sander through moderation. I???m just hoping that letting him have his say will let this topic of conversation reach its end. I???m certainly ready to move on???”
I don’t think it’s nutty. To do otherwise would just embolden his cause further. Granted, you don’t have to allow them, but I think the consensus on this blog is that his opinions are in the minority (at least, as far as Jive, OpenFire, and related users are concerned).
But, in response to a few more of Sander’s posts… (sorry, I can’t help it!)
“This will probably not help I???m afraid; to attract new enthusiastic contributors, you should give the good example by alo being enthusiastic and ambitious. Otherwise, they???ll prefer to start their own project from scratch or join another project that has enthusiastic people in it.”
Jadestorm worked rather enthusiastically on PyAIMt and PyICQt for years. During that time, I’m pretty certain he received few offers of help. If people want to start their own projects now, rather than use the code he’s painstakingly written because he’s not “enthusiastic and ambitious” about his own project after years of solitary development, I’d say that’s their loss and their mistake. I won’t presume to speak for him, but if the comment was directed at me I’d be rather insulted.
“Yes, I know. But the key problem is that Jive Software CLAIMS to admire the Jabber community. I just want to open the eyes of people that DO admore the Jabber community???free speech you know. What even makes me more passionate is that Jive Software tries to prohibit me to open people???s eyes. Jive Software unfortunately is a shareholders-only focussed company it seems”
Let’s see, GPLed server, GPLed client, GPLed plug-ins, GPLed client libraries, GPLed component libraries, free community based support, and participation in the JSF (or is it XSF now?). Yeah, sounds a lot like “claiming” to me. All because they require the ability to dual license software so they can sell commercial versions to support themselves. How evil. But, of course, maybe those developers who would refuse to use Jadestorm’s python transport code because he’s not “enthusiastic and ambitious” about it would also refuse to use any of Jive’s code because they actually have a business agenda that runs parallel to the XMPP agenda.
Jive is a company. They want to make money and turn a profit. That’s the whole point of going into business. You don’t go into business so you can go broke. What would be the point? They’ve chosen to involve the community, like many companies do nowadays, and not only receive the benefits of the community and open source development, but they also give back to the community. If you say they don’t, you’re just being disingenuous. You can say you dislike the fact that they dual license code (and require copyrights be signed over to them for that purpose), and that’s fine. I’m sure a lot of people, free software proponents especially, don’t like that. But to say that, because of that, they don’t care about the XMPP community and do nothing to support it is a bald faced lie.
From an earlier post, you complained that Jive’s transports plugin does not work with other servers. The reason for that, and also why Jadestorm has enjoyed working with the project, is because they make use of internal features of OpenFire itself. This allows them to do things that simply isn’t possible (at least, not in an easy way) with the external transports. If they just made external transports, what would be the point of reinventing the wheel when the python transports would’ve done the same thing? They wanted to create a turnkey solution that allowed people to set up an OpenFire server and transports to other networks. That’s exactly what happened, and it works wonderfully. If OpenFire ceases to exist at some point in the future, yeah, that work might be in vain, but that’s a similar reason to why the C-based transports aren’t around anymore. They only worked with jabberd14, and nothing else (though could sometimes work with JCR), because they were written before the component protocol seemed to gain support.
Anyway. I’ll try to make this my last post on the subject. Matt, if you decide not to allow this one in to let the thread die, that’s okay too. I’ll understand. I just take issue with Sander’s claiming that Jive’s contributions to the open source and XMPP community aren’t genuine, simply because he disagrees with the code contribution terms. I do also find it interesting that he disagrees with them so vehemently, but doesn’t code and doesn’t have any intention of contributing code under those terms or any others.
I’m not bashing people who don’t code, by the way. I don’t code anything for Jive/OpenFire either (I’m a Java n00b). But, I’m also not complaining about the “unfairness” and “wrongness” of their contribtion terms.
“If people want to start their own projects now, rather than use the code he???s painstakingly written because he???s not ???enthusiastic and ambitious??? about his own project after years of solitary development,”
Well, this is no blaim at jadestorm. It is just how open source communities work AFAICS. Look for example at the well-known example of Linux. Linus coded as hell the first years. Compare that with what happened to the GNU Hurd and Minix kernel projects… The first years are critical, I think, and the project is very weak at that stage.
“I???d say that???s their loss and their mistake. I won???t presume to speak for him, but if the comment was directed at me I???d be rather insulted.”
Well, sometimes it is easier to start from scratch if you don’t know how the code works. Note, that I don’t know this as of my experience of coder as I’m no coder, just look at other projects, like jabberd14 and jabberd2, of which the development both lagged for a long time (and still do a bit IMO) because the original creators left the projects.
“All because they require the ability to dual license software so they can sell commercial versions to support themselves. How evil”
I suggest you to carefully read my previous posts in which I explain how Jive Software hurts the JABBER community. I was not talking about the free software community. They don’t even had to GPL their software for me…I have no problems with closed source software like those made by Jabber Inc., Adobe, amongst others. I am no free software fundamentalist (no idea if that exists though) as you seems to think. Please read my reasoning before replying.
"Jive is a company. "
Yes, I known how companies work. In fact, I study economics. Also, to repeat, I was talking about the JABBER community, not about the free software community and its GPL.
“But to say that, because of that, they don???t care about the XMPP community and do nothing to support it is a bald faced lie.”
I didn’t said Jive Software didn’t cared at all and you’re calling my a liar! I’m sure Jive Software also does things that are good for the - again - JABBER community. But my point is that they ALSO do things that contradict with this mission (see my previous posts e.g.).
“From an earlier post, you complained that Jive???s transports plugin does not work with other servers. The reason for that, and also why Jadestorm has enjoyed working with the project, is because they make use of internal features of OpenFire itself. This allows them to do things that simply isn???t possible (at least, not in an easy way) with the external transports.”
You should take a look at the URL to the blog article on Process-One regarding Epeios. Nobody said you cannot create internal transports that also can work as external transports; this exaclty what Epeios does…
“They only worked with jabberd14, and nothing else (though could sometimes work with JCR), because they were written before the component protocol seemed to gain support.”
At least you can get them working with other servers such as ejabberd, Openfire, jabberd2:
http://ejabberd.jabber.ru/install-obsolete-transports
“I just take issue with Sander???s claiming that Jive???s contributions to the open source and XMPP community aren???t genuine, simply because he disagrees with the code contribution terms.”
Well, you’re now quoting the issue that doesn’t matter me that much as it only hurts Jive Software themselves and not the broader JABBER community. The issue with doing this is that it creates a bigger threat of license infringements, results in less people interested in contributing code, results in no community of other companies offering support for Jive’s software, results in less people interested in Jive’s software because of this lock-in. But that is Jive Software’s choice, and I respect that choice, although I think it gives other open-source software an advantage over Jive’s, but so it be.
“I do also find it interesting that he disagrees with them so vehemently, but doesn???t code and doesn???t have any intention of contributing code under those terms or any others.”
I don’t see what’s so interesting about telling that I don’t like such a restriction.
“I???m not bashing people who don???t code, by the way.”
Without this sentence I would have thought you bashed people contributing in a non-code way to any open-source project.
chuckle Well for what it’s worth, Daniel-not-me, I just really don’t care about whatever sander says about me or to me. =) It’s just not important. What’s important is the folk who are enjoying the work that I’m enjoying doing!
Sander, your entire argument has been that Jive’s contribution terms are “wrong” or “unacceptable” because they require the contributors to sign their copyright over to Jive, thus allowing Jive to dual license the software. Now you’re saying that’s not the issue, and that Jive hurts the Jabber/XMPP community in other ways.
How? By asking Jadestorm if he’d be interested in writing plugins for them to do transports to third party networks? Are you angry because they wrote their own server, their own client, their own whatever, rather than using existing tools?
The code is there. Anyone can take the things that Jive has done and incorporate them into their own projects if they see fit. Anyone can take all the GPLed software Jive has written to create other Jabber/XMPP implementations. From what I can tell, they’ve not tried to make any proprietary extensions of XMPP, they’ve not tried to pervert the standards so as to keep their products incompatible with the rest of the XMPP community. Their software has become a very popular alternative to jabberd14, jabberd2, and ejabberd, because it works. Looking over the instructions of trying to get the C-based transports working with other servers (which STILL requires jabberd14 to build the stupid things), and knowing the hassles one has to go through to get it working, any sane person has to see why the Jive approach is preferable for most server administrators.
Regarding Epeios for ejabberd:
“Limitations: For now, the ejabberd modules that can be run with epeios must not use any ejabberd hooks or handlers. The module can work, but the feature that are linked to a hook will be ignored. Improving the XEP-0114 to extend the field of this software is however a work in progress.”
In other words, until XEP-0114 is extended, there’s not much benefit to using an ejabberd module on a non-ejabberd server rather than an existing external component. You can use the Python transports with OpenFire just fine, but the integration with everything else is what makes them nice. Perhaps this will change once XEP-0114 is extended to allow all sorts of new interaction, but then the servers will all have to be changed to support it, etc. That’s going to take a long time, and maybe one day it’ll be a reality. But I bet there are ejabberd-specific modules now that don’t work with other XMPP servers, so are you over there yelling at them too?
“I didn???t said Jive Software didn???t cared at all and you???re calling my a liar! I???m sure Jive Software also does things that are good for the - again - JABBER community. But my point is that they ALSO do things that contradict with this mission (see my previous posts e.g.).”
Perhaps in your mind. But, I think you’ve mischaracterized things and you’re wrong. But, you’re entitled to your own opinion, even if it’s wrong.
“Well, you???re now quoting the issue that doesn???t matter me that much as it only hurts Jive Software themselves and not the broader JABBER community. The issue with doing this is that it creates a bigger threat of license infringements, results in less people interested in contributing code, results in no community of other companies offering support for Jive???s software, results in less people interested in Jive???s software because of this lock-in. But that is Jive Software???s choice, and I respect that choice, although I think it gives other open-source software an advantage over Jive???s, but so it be.”
I don’t think this creates any greater threat of license infringements than any other open source / free software project. If somebody contributes code that isn’t under a GPL compatible license, it’s a problem whether they assign Jive the copyright or not. It might scare some people away from contributing, but that’s a personal preference. It seems that it works fairly well from my perspective. As far as lock-in, I’d assume most people prefer to get support from the company who makes a product, rather than someone else who is just watching from the sidelines. You can get support for RHEL from Oracle now, but I’d imagine most people would prefer to just get it from RH. Anyone else can support OpenFire if they want, there just may or may not be a reason to. That’s a business decision they’ll have to make themselves.
Anyway. Your argument is all over the place, and in the end, it boils down to your opinion that Jive doesn’t support the XMPP community. The only reasons I’ve seen to support your claim are that 1) they dual license code and 2) they have a transport plugin that’s specific to OpenFire (even though the plugin is also GPLed!). I’m sorry, but those two facts don’t support your argument.
“I don???t see what???s so interesting about telling that I don???t like such a restriction.”
It’s interesting because it’s a restriction/condition that doesn’t affect you, yet you bring it up as an example of Jive hurting the community. If you don’t like the terms, don’t contribute code. You know the terms up front. It isn’t as if it’s a surprise, and there are other open source projects that do similar things. You’re entitled to your opposition, but it’s apparently not that big of an issue to other code contributors.