cancel
Showing results for 
Search instead for 
Did you mean: 

Is Alfresco a real Open Source?

edless
Champ in-the-making
Champ in-the-making
Certified Alfresco Partners only offer Level 1 Support, Consulting, Integration or Training on either the Alfresco Enterprise.

Open source was born to give people the possibility to share knowledge and expertise.

Why Alfresco is giving just to big company the possibility to be supported  on the product? Why I can't pay a Alfresco Partner just to support me on a community version, without asking any guaranty?

ED
102 REPLIES 102

dmatejka
Champ in-the-making
Champ in-the-making
Sorry, I don't understand your "more time" argument. I think it's a little naive to think that Alfresco is trying hard to do different. I think they know exactly what they are doing (and they've been right so far!).

As I have mentioned… they have changed a lot. not only "in code" but also "in the community strategy". The way from a closed source (e.g. Documentum) business model to an open source is not easy and some decisions are really hard to be made. Maybe you feel it as Alfresco's way to threat the community. I see it as a way how to find acceptable business model for the product. Anyway, loosing the community in a long run is not what they need.

vsuarez
Champ in-the-making
Champ in-the-making
Thank you for your information Luis, but I must say that my experience the last three months with Alfresco Labs (3A, 3B and various nightly builds) was with fresh installs, and finally I downgraded configuration files from Alfresco 3 Labs to my old and loved 2.1 CE. I have some issues, but as you said, it was a 30 minutes job.

Surely I am very much more skilled in 2.1, but when I am testing and showing features to a client, I cannot get failures adding groups or users to a space because of a javascript problem, or getting regression bugs in the SOAP clients.

Note 1. You made me search in the dictionary the "Howdy" word  Smiley Surprisedops:  :lol:
Note 2. This thread is going more and more interesting but… here it's 00:52 and I must wake up at 06:00!!! Tomorrow I won't be "fresco" at all.

mjasay
Champ in-the-making
Champ in-the-making
1. Source code is CLOSED. We have only dirty HEAD code. Fixes from community as long as fixes from alfresco team is not in HEAD. Community assist to make ENTERPRISE version more stable but this stability inaccessible for community. It's stealing towards us (Community). We help make Alfresco better ECM but only for commecial users.

I'm sorry, but this is not true, on several levels.  (And no, don't worry about your English - it is much better than my Russian.  🙂

First, the source code for Labs is 100% open, and it's also open in Enterprise for our customers.

Second, it's not "dirty HEAD code."  All bug fixes that go into Enterprise also go into Labs.  Some immediately, some later.  If you prefer to have immediate service on all bug fixes, I have a very good idea: fork the code, start a rival project, and fix bugs to your heart's content.  There is absolutely nothing stopping you or anyone else from doing this.  It's open source, and the right to fork is the fundamental freedom.  You have that with Alfresco.  I invite you to exercise it.

Third, if, in fact, you have contributed code to Labs (and, hence, to Enterprise, as you say), can you please show me what code you have contributed?  I hear a lot about Alfresco taking from you, but the only thing I see is Alfresco giving you thousands and thousands of lines of excellent code.  I have yet to see what you are prepared to contribute back….

I have colorful example on statments above: in one of my projtcts I've use Alfresco CE 2.1. After clean install on my portal (liferay 4.2.2) I try to upload some content - exception. 100% reproducable. I search in JIRA and WOW - bug is fixed ONLY for ENTERPRISE (http://issues.alfresco.com/browse/AWC-1493).
After that case I've many similar troubls…

Wait a bit, and those bug fixes will make their way into Labs.  I'm sorry if you would like to have 100% of the Alfresco code 100% of the time, but we have families to feed, just like you, and we think it's a fair tradeoff to provide immediate bug fixes to our paying customers and to have them delayed a little for non-paid users. 

Of course, if you don't like this there's a very easy solution, which I mention above: write the bug fixes yourself and contribute them, either to Alfresco or to a forked project.  I'd be very happy to discuss with the rest of the management team a plan to keep all community-written bug fixes available in Labs immediately, but first I need your commitment that you're actually going to contribute code.[/block]

2. Developer documentation is inaccessible. See at Spring, Hibernate, Apache projects (all used in Alfresco) - developer documentation is open and absolutly free. In case of Alfresco we have some small wiki articles - not more. Many of them is out of date.

I wish I could say that you were correct on this, but we don't hold back documentation.  There is a small but growing body of "best practice" information that we sometimes keep for paid customers, but I believe 99% of our documentation is available to you and to everyone else.  It's possible that the documentation for a given problem isn't very good, but I don't think it's hidden….

Alfresco CE is not intended for production usage! It's my bloody expirience.

But tens of thousands of active deployments of Alfresco Labs suggests that you're wrong.  Some of the biggest companies in the world, like General Electric, Fidelity, and The Hartford, happily run Alfresco Labs in serious production…without paying us a dime.  Perhaps you simply need to talk with them about how they manage to do it so well.

mjasay
Champ in-the-making
Champ in-the-making
But my experience is that companies that run mission-critical will anyway buy support, no matter what quality of code is, and those who can only use free Open Source will never pay (just because they never do that). But they will definitely put demerits on Alfresco for broken releases.

I wish that this were true.  We have 30,000 examples of this not being true, however.  🙂

Regardless, please see my comments on a related thread.  Would the Alfresco community prefer that we focus on a super-stable Labs release…but extend it with a few proprietary extensions (as SugarCRM, Zimbra, Hyperic, etc. do)?  We could do that.  In fact, we used to do that three years ago.  If that model works better for you, please let us know.  We're always trying to make our product and community better, and your feedback will help us do that.

Matt

vsuarez
Champ in-the-making
Champ in-the-making
Third, if, in fact, you have contributed code to Labs (and, hence, to Enterprise, as you say), can you please show me what code you have contributed? I hear a lot about Alfresco taking from you, but the only thing I see is Alfresco giving you thousands and thousands of lines of excellent code. I have yet to see what you are prepared to contribute back….

Wow, mjasay, not cool either. In the same line of argument, you could show us your real contribution to Alfresco and open source world and, if you don't publish any code and Jira snapshots with your name written in the bug resolution or internal OSI reports with your signature, we won't believe you. Several small contributions make open source great, and we cannot start to ask for clues of contribution to everybody that uses and complains about an open source product. Who collaborates with open source tries to offer the best possible product to everybody, and criticism helps to improve the product.

And please, it's better for Alfresco that doesn't continue comparing Alfresco open source policy with Red Hat or MySQL ones. If they were comparable, we would have a stable Alfresco 2.2 CE and an Alfresco 3.0 RC1 simultaneously in the "downloads" page, and all the SVN branches published for everybody, including the Enterprise one.

Regards

cstrom
Champ in-the-making
Champ in-the-making
Luis, mjasay
It is nice to hear about more QA and more frequent releases. We have just started getting experienced enough to be able to provide contributions, but that is definitely our intent. When we do, will our contributions (if accepted) go directly into labs head?

About the pricing: I don't understand why you don't choose to go for volume. Where is the cost? It must just be a matter of defining a SLA at an appropriate level. Company users wants to pay, right now you are missing out on a lot of revenue from smaller projects. At least from my view.

Best Regards,
C-J

mjasay
Champ in-the-making
Champ in-the-making
Well, I used to use Fedora and I dumped it, I am not Linus Torvald and don't need to twinker around with operating systems or anything else. Then I used RedHat Enterprise Linux and I dumped it because I found out that I could get CentOS for free. I don't need any support, so what do I need RedHad Enterprise Linux for. Then I found Ubuntu and dumped even CentOS. I don't need a copy of something if I can have an original. I think that Ubuntu has the right Open Source model. Take the whole thing for free and if you need support then buy it, if not, then no need to. The Ubuntu model will succeed, the other I am not so sure about.

What do you say, my friends?

I think I'd say that the Ubuntu model has so far only worked when a billionaire is willing to pump tens of millions of dollars into propping up the business each year.  I'm glad that you have bought into this system (I know Mark S. and he's trying very hard to find a sustainable model that doesn't depend on the bits, and I'm rooting for him to succeed), but you seem to want a model that has proved to fail in every case except when someone with a lot of cash can subsidize a failed business model.

That's not us, unfortunately, and you've managed to find the only commercial project that can subsist that way.

wbox
Champ in-the-making
Champ in-the-making
My First Open Source Experience

Alfresco's my first foray into open source software - boy have I learned a lot.  It's also my first work with VMs and linux.

Having tested many non-open-source software packages in the past, the typical question I get to with the vendor is "will you let me set up an evaluation version in house?".  Some say yes, some say no.  Ones that say no usually don't get much further with us.  If we can't thoroughly test it before purchasing, I can't ever recommend acquisition.

That's why testing Alfresco seemed so appealing.  It's not so much that the source is available to us, but that I could get a version to test without needing permission, load it up, show it around and then, when I'm convinced all is stable and I get buy in from enough clients, I could go ahead and transition into a support contract.  That was my thought process. 

And I think that's generally what's happening so far.  I've got a Community version of 2.9 running and a Labs version of 3.0 running.  I'm posting in this thread because I'm still not so sure I've got really stable versions of either and it's hard to tell because I'm not sure if its because I'm running a non-enterprise version that maybe doesn't have supported fixes applied, or I happened to pick a version (v. 3.0..) that's not quite baked yet, or are there actually bugs?  Am I running into Open Source issues, or product issues?

Alfresco potential stands out as a standards-based solution for ECM that's easy to develop against.  You'd be amazed at how much interest we've got from various groups about its capabilities - but what we need is some assurance that the 'free' versions to test with are equivalent to what we may eventually buy.  If I'm spending too much time in configuration and debugging because I'm running into free vs. pay issues, then interest will probably peter out.

It sounds like (thanks, Luis and the rest of Alfresco) there's lots of effort being put into getting the free and pay version in sync as the norm.  That's much appreciated and will serve everyone well.

Wally
Nike, Inc.

whoami
Champ in-the-making
Champ in-the-making
Well, I used to use Fedora and I dumped it, I am not Linus Torvald and don't need to twinker around with operating systems or anything else. Then I used RedHat Enterprise Linux and I dumped it because I found out that I could get CentOS for free. I don't need any support, so what do I need RedHad Enterprise Linux for. Then I found Ubuntu and dumped even CentOS. I don't need a copy of something if I can have an original. I think that Ubuntu has the right Open Source model. Take the whole thing for free and if you need support then buy it, if not, then no need to. The Ubuntu model will succeed, the other I am not so sure about.

What do you say, my friends?

I think I'd say that the Ubuntu model has so far only worked when a billionaire is willing to pump tens of millions of dollars into propping up the business each year.  I'm glad that you have bought into this system (I know Mark S. and he's trying very hard to find a sustainable model that doesn't depend on the bits, and I'm rooting for him to succeed), but you seem to want a model that has proved to fail in every case except when someone with a lot of cash can subsidize a failed business model.

That's not us, unfortunately, and you've managed to find the only commercial project that can subsist that way.

Wow, someone answered my post. I can't believe it. Thank you, thank you sooooo much, I am sooooo happy.  Smiley Very Happy

Listen, my dear friends, I understand your situation. You need to make money, just like I need to make money too, and just like you I am a poor guy too. Now, what do we poor guys do? Let's be clever, very clever. Let us create two versions of a product, one version which works, we sell for money and the other version which is total nonsense, we give away for free, call it OpenSource and let stupid people test it and report bugs for free. This way, we don't need to hire beta testers and can reduce our quality assurance staff and save money. And besides, it is OpenSource, the buzzword in the software industry of today, so everyone will want it. Great idea.

Hmmmm, that surely is a great idea, but is it a good idea? Will it create a happy and large community? Will it make the product succeed? Will it spread Alfresco? Will it make Alfresco the Enterprise Content Management System Standard on this planet?

Will it, what do you say my clever friends?

And, by the way, this thread has been going on and on and on and on, and I really don't see that any learning process has taken place. It is a ping pong game, ti tac tic tac tic tac tic tac tic tac. Successful companies usually listen to what people are saying and go through a learning process. I don't really see that taking place over here, at least not in this thread. I see a lot of clever arguments, but clever arguments alone, don't lead to success.

stevewickii
Champ in-the-making
Champ in-the-making
Alfresco is an awesome product!  Lots of people are contributing from both sides of the commercial fence, and even more are benefiting from it.  I'll be interesting to see where it goes.

Most subversion projects I work on have a folder structure like this:

<project-name>/trunk
<project-name>/branches/<branch-name>
<project-name>/tags/<tag-name>

Under the branches folder, you would have a branch for each major version/feature set, and one for each new feature.  So a 2.9, a 3.0, and a surf branch might be found under the branches folder.
Under the tags folder, you would have a tag for each release, like alfresco-2.9.0B, alfresco-2.9.0C, alfresco-3.0b, etc…

Could we get something like this?  I would be glad to help, if you guys need it.

Keep up the great work!!!

Cheers!

Stephen