PDA

View Full Version : Official Commercial License Input Thread



jack.slocum
12 Feb 2007, 10:54 AM
Here's what I am currently leaning towards for 1.0:

Ext Core:

Continues BSD as before. This allows for sharing key pieces like DomQuery, DomHelper and YUI extensions with the javascript community.

Ext Widgets:

- Free for non-commercial use

- Free licenses to open source projects which request one (does that help dfenwick?)

- Single developer license $299
Includes 1 year email support, source code
The single developer license is good for 1 developer, one machine.
Added (as per KimH) - 1 yr minor and major upgrades

- 5 developer license $799
Includes 1 year email support, 1 yr free minor and major upgrades, source code

- Site license (unlimited developers) $1299
Includes 1 year priority support, 1 yr free minor and major upgrades, source code

There will be no charge for distributing Ext with an application, no per server or per cpu fees and no royalties!

Does this sound fair? It is much less than some other libraries are charging. For example, the nitobi Grid (just a grid!) is $3700 for a 10 developer license (and that is the only license that includes the source).

One other note, I plan to do something like if you purchase a license during the 1.0 beta period, the price will be 25% off. The revenue generated will be used to get help with documentation, so it will be well spent!

This is a rough draft. Please provide your thoughts on how we can make this work.

dewd
12 Feb 2007, 11:07 AM
Seems good to me. Maybe you could start with 1-2 developers for the first fee, which is closer to what small shops might need. :-) The values sound good, even though a mercenary could try to get more out of it. :-)

Well, I am glad there's such a good plan.

jack.slocum
12 Feb 2007, 11:40 AM
dewd, hopefully the 25% will help? I just added that note, I forgot to put it in the original post. :)

KimH
12 Feb 2007, 11:40 AM
Ext Widgets:

- Free for non-commercial use

- Free licenses to open source projects which request one (does that help dfenwick?)
Sounds great!


- Single developer license $299
Includes 1 year email support, source code
The single developer license is good for 1 developer, one machine.
Could you add "1 yr free minor and major upgrades" to the single developer license as well. As you know yourself the web and probably this API will be better and better each day. It would not be attractive for a single developer to buy this kind of license without at least the upgrade too.

jack.slocum
12 Feb 2007, 11:49 AM
KimH - I added it.

KimH
12 Feb 2007, 12:00 PM
KimH - I added it.
Thanks! You're the greatest. It's nice that you listen to "us" :wink:

greyknght1
12 Feb 2007, 12:14 PM
Can you update your blog with this information so I can start working with procurement to get you paid?

Thanks for all the really hard work!

Michael

jack.slocum
12 Feb 2007, 12:18 PM
Of course I listen! I want this to work for everyone. In many ways, even if it doesn't seem like it, I think this will be a good thing for everyone.

Some positives that I think can come from it:
- More documentation
- Improved support
- Better organization
- Happier wife - that's my benefit ;)

lstroud
12 Feb 2007, 12:19 PM
Have you considered a support centric model? I suppose this would be more of the JBoss approach.

I will gladly pay to use your work on my project. However, my personal opinion is that the commercial options on license agreements sometimes prevent internal adoption. The problem is that you need popularity to drive either model. So, to me it would make more sense to keep nonprofit free (like you are suggesting) and make corporate dev free, but offer a support arrangement. Then include a certain size commercial deployment for free and force pay for licensing on deployments larger than that. I would bet that with your design and coding skills, that there is more money to be made by using license fees for large commercial deployments to drive consulting engagements. In the mean time it would allow you to build a support org/consulting company that does ajax ui design.

Just my thoughts,
LES

jack.slocum
12 Feb 2007, 12:23 PM
Michael, there is a new Ext site in the works. It will have the finalized license info. I will have it up in the next week.

greyknght1
12 Feb 2007, 12:26 PM
Thanks Jack! Your work has been critical in a project I have been working on. You make me look good and I thank you for it!

Michael

PortraitPainter
12 Feb 2007, 12:57 PM
Please provide your thoughts on how we can make this work.
Looks about right to me and I have been on the purchasing side in recent years. When the reputation of Ext becomes more widespread in corporate IT you could bump up the site developer licence by 2 or 3 times. The early bird beta phase licence discount needs to be a little more generous in view of the state of docs and tutorial material.

Very pleased to see no per server deployment charges as this would have killed off my interest and planned use of Ext.

Open source is not a religion in my book, I see it as a mechanism that allows a talented individual to muscle into the space previously tied up by big boys with their marketing budgets. Your licence structure keeps the open source ideal going for many but provides an income allowing you to focus on the project full time which is the key objective.

I would also recommend layering a "premier support" product on top of your standard licence structure. The company I work for pays a 6 figure sum per year for this type of support from one large IT company, it is mainly telephone consulting but they do fly in trouble shooting experts from time to time.

Put yourself in the position of an IT executive charged with a $ ½ million conversion of a web application to an Ajax style UI. Your smart developers are suggesting Ext but you have never heard of it before. The $1300 site licence is trivial and Ext still feels too risky. Given an option to spend another $5k to $10k for extra telephone consulting from the author himself and the Executive might be swayed.

schmidetzki
12 Feb 2007, 1:01 PM
I see not problems with you licences and I will surely buy some for my company.
I would also be happy to licence the 1.0 beta as well (and save 25% :-))

But why don't you post this (and other news like this: http://www.yui-ext.com/forum/viewtopic.php?t=2675) in your bog?
I added you blog to my newsreader - but manualy have to scan the forum on a daily basis for important new news like this. Isn't RSS much better to keep your "customers" up to date?

jon.whitcraft
12 Feb 2007, 1:26 PM
I like your licensing. I only see one problem. Where would this example fit.

currently as my job i use the Ext Widgets for my admin tools (posting news stories, photo management, etc...) and I use the Basic Dialog to show a bigger picture of a thumbnail on my public site but 95% of my development goes into the admin tools. Which license would i need?

Thanks,

Jon

jack.slocum
12 Feb 2007, 1:32 PM
PortraitPainter - Before the 1.0 release the docs will be there. The idea is to use the early beta funds to get help with tutorial type documentation. I can write good example code and I can document the API but when it comes to entry level tutorials, I don't do very well.

If you can provide any more input on how to structure a premier support options I would definitely add it. I am more than happy to provide paid phone support.

schmidetzki - Historically, I usually post tutorials, examples and releases to the blog and discuss other stuff in here. I agree with you though. What is needed is an Ext blog, and my blog can again be a little more informal for discussions like this.

jack.slocum
12 Feb 2007, 1:36 PM
sid, if I understand correctly you could get the single dev license. There's no per deployment fees. :D

Bobafart
12 Feb 2007, 1:40 PM
...when it comes to entry level tutorials, I don't do very well.


Sir I am glad you can identify this weakness (probably the ONLY weakness of yui-ext). I really wish there were more basic tutorials. I think yui-ext is missing a large market of users to frameworks with smaller learning curves (mootools, dojo). yui-ext is obviously superior to these but you have to think of marketing and word-of-mouth popularity which the "smaller learning curve" have.

It would be excellent if you could get a JS noob or intermediate JS tinkler to write some basic tutorials for the novice JS programmers to get addicted to yui-ext. I am sure a lot of the dojo/mootool-types would migrate to yui-ext if that were the case.

I would like to write basic tutorials for yui-ext if you would like -- basic, dumb-down, vanilla tutorials for noobs.

jason
12 Feb 2007, 1:52 PM
hey jack, i think your commercial license w/ free for non-commercial is excellent. i will personally have no problem paying $299 for a developer license.

i do know some guys who might have problem scrounging up $299 at one time. If it fits your model, you might consider a per-application license as well that people could buy. that way they can tack it onto their clients bill for small projects. I know there are a ton of $1,000 website projects out there and they could pitch an extra $99 fee or whatever to the client for "advanced GUI" or whatever.

It may be that it's better to write that category off because they might never pay the license fee anyway no matter how low it is. And perhaps it's better to keep your officially supported clients down to a smaller number...? I don't know the answer! But, if you want to go for the freelancers and small-time guys, i think some type of price option under $100 would be a way.

jack.slocum
12 Feb 2007, 2:12 PM
For sure Bob - tutorial help would be great. I will let you know when I have some funds and we can get started.

jason - That why I chose the simple license. The developer buys 1 license, and can use it to make as many sites as they want with no additional costs.

lstroud
12 Feb 2007, 2:37 PM
How does the 'developer buys one license, but develops many sites' plan apply in a contract situation? If you are contracting and that contractor own the license, are they then allowed to develop sites at each of these companies under the one license? I suppose each company could get an independant support plan.

LES

Bobafart
12 Feb 2007, 2:39 PM
... I will let you know when I have some funds and we can get started.


funds? No, I am not requesting to be paid. I have enough money. I would be happy to write and maintain tutorials for free -- consider it me giving back to you after all the giving you have provided for us.

I have already posted a couple of basic tutorials.. I just need the CSS that you would like me. we can talk more in private message if you would like

kjordan
12 Feb 2007, 4:01 PM
How does the per developer license work? There's two developers where I work and we each work off a mutual library directory on the server.

JeffHowden
12 Feb 2007, 4:12 PM
We're in a similar situation, but we have 3-4 developers. I think that equates to needing 4 developer licenses. In that case, we'd be better off with either the 5 developer license or the $1299 site license.

jason
12 Feb 2007, 4:15 PM
kjordan - per developer means that each developer on your team needs a license. but you can produce as many applications as you want. so in your case, you would need two licenses. If the other dev on your team will never touch the actual yui-ext code, but works on other parts, then you might be able to say that person doesn't require a license and only buy one license.

basically, if you are a consultant and you release a lot of apps, this is a cheap license type. but, if you are a large team working on one single application, then the price is going to be really high even though you only have one application. A large team usually = corporation. small team usual = freelance or small company. so this type of pricing usually is considered to be fair in that the larger the company, the more they pay.

kjordan
12 Feb 2007, 4:19 PM
kjordan - per developer means that each developer on your team needs a license. but you can produce as many applications as you want. so in your case, you would need two licenses. If the other dev on your team will never touch the actual yui-ext code, but works on other parts, then you might be able to say that person doesn't require a license and only buy one license.

basically, if you are a consultant and you release a lot of apps, this is a cheap license type. but, if you are a large team working on one single application, then the price is going to be really high even though you only have one application. A large team usually = corporation. small team usual = freelance or small company. so this type of pricing usually is considered to be fair in that the larger the company, the more they pay.
I guess I'm more curious as to how the license is enforced. It would have to be pretty much on the honor system that you're not misrepresenting your developer count, correct?

vmorale4
12 Feb 2007, 6:51 PM
Does free for non-commercial use includes non-profit organizations? I work for a University, and we have been using your excellent grid in an upcoming project...

ilazarte
12 Feb 2007, 8:15 PM
unfortunately this type of licensing would probably kick me back out into widget wild. i'd use the core happily, but end up rebuilding the items i need as my company and the type i like working at prefer "free" techologies.

i second the vote for support based model, but you gotta do what you gotta do. like someone noted earlier, support is really where you build the confidence and the billable hours. consider talking to someone like Tod Liebeck from Echo2, it seems they are up to their ears in consulting work to the point where they can't put out any new releases :).

I believe also people like IBM are banking big on selling application lifecycle, less on software releases as open source makes second tier products very attractive...

dewd
12 Feb 2007, 8:31 PM
unfortunately this type of licensing would probably kick me back out into widget wild. i'd use the core happily, but end up rebuilding the items i need as my company and the type i like working at prefer "free" techologies.


I am in the same boat. I wish all the best to YUI-ext and Jack!

For now, I am going to keep YUI and jQuery as the building tools for me. Eventually I might need YUI-ext as well.

The main difficulty of YUI-ext will be the relicensing issues.

dfenwick
12 Feb 2007, 8:45 PM
Here's what I am currently leaning towards for 1.0:


- Free licenses to open source projects which request one (does that help dfenwick?)



I just sent a message to Eben Moglen (one of the FSF's law advisors) about it. Can you provide any details on what the license content will be? You're going to end up with a weird balancing act with the license and hopefully you have a lawyer to straighten the license out. Hopefully Eben will get back to me about it sometime soon.

jratcliff
12 Feb 2007, 9:33 PM
@ilazarte and @dewd,

ilazarte wrote:

unfortunately this type of licensing would probably kick me back out into widget wild. i'd use the core happily, but end up rebuilding the items i need as my company and the type i like working at prefer "free" techologies.

dewd wrote:

I am in the same boat. I wish all the best to YUI-ext and Jack!

So you will be able to reproduce the widgets and/or mix and match other free widgets for less than $299 of your own time's worth? I've been doing the mix and match thing for years and it just seems like so much work that $299/year seems like a bargain to me. :wink:

I guess you just have to do a real price comparison and compare $299/year to what your hourly rate is (or what your company pays you) and multiple that by the number of hours it takes to find and support and integrate with all of your free widgets. Good luck! :)

jack.slocum
12 Feb 2007, 10:19 PM
Sorry guys, had to get some sleep. :)

@ilazarte and @dewd,

I believe the license fees are reasonable in comparison with other licenses. Also, in particular, they allow for royalty free usage.

Can you please define "this type of licensing". What in particular is bad?

@dfenwick
Please let me know the results of your query. You are welcome to email me, jack.slocum@yahoo.com.

@vmorale4
I have modified it to include non-profit

@kjordan
Yes, that is correct.

@Bobafart
Please send me a pm with IM id and we can discuss further. Thanks!

manugoel2003
12 Feb 2007, 10:46 PM
Hi Jack, its nice to know that you finally came up with a licensing policy.... but it is sort of bad news for me coz we are a very low budget firm and cant afford $1299... things are very different here in India.... we are over a dozen developers but the setup cost is not much so it is relatively easy for us to manage our operations with a low budget... but certainly cant afford this much :(.... it would be nice though if the support only was paid and not the library itself (or any prt of it).... moreover, assuming that u stick to the current draft, would there be an evaluation version??

Stephan
13 Feb 2007, 4:27 AM
- Single developer license $299
Includes 1 year email support, source code
The single developer license is good for 1 developer, one machine.
Added (as per KimH) - 1 yr minor and major upgrades

How do you get new updates after one year ? Do we have to buy another license ?

Does a license allow me to use the code as long as I wish or is it limited to 1 year and after that period I have to renew it ?

lstroud
13 Feb 2007, 5:12 AM
My previous post flowed off page before it was seen, but I still think that you would be better off with a support agreement financial model. I think that you need wide spread adoption to drive ext. So, you need startups and such to deploy it. A startup will always choose free first. So, it gives an advantage to your competition. Mid to large size corporations always by maintenance agreements. They think of it as an insurance policy. I think that with a support agreement and corporate adoption that you could drive a consulting org. That would stand to make a lot more money.

JBoss has made more money that just about any other commercial open source venture (short of linux). Their model may prove to be the most successful by far. I think you should consider that before you go with the typical gui component developer seat license model.

Your example of Nitobi is actually a pretty good one. They do use dev seats. However, they are also providing their components so that they plug into tools (JSF, ASF, ColdFusion, etc). To me, that seems like a slightly different market. They also provide UI design consulting services (which is where they are making a lot of their money right now).

Just wanted to provide a counter case for thought.

digital.alterego
13 Feb 2007, 5:51 AM
Hi Jack,

I think the prices are very moderate, I suggest that you rise the prices after the beta / early adopters period.

Additionally I would suggest that you offer also a subscription model, after purchasing. For a yearly fee of 100$ per developer the customer gets support and new versions of the library. The subscription extends itself every year unless it's canceled.

The benefit of this approach is, that you can calculate with a constant income without the pressure of needing new sales.
For our company it would be okay to pay the subscription, but if we bought the lib a year ago and we would need to rebuy it now, we only would buy it if we gain big advantages.

Webnet
13 Feb 2007, 5:51 AM
Jack, I can't wait for you to steady your income with licensing. Even now the quality of your library is insane, the features you offer are off the charts. And as you have said in your posts, tutorials are a key element that need to be stronger, more simplified, and easier to understand. In the same note, you also need to make sure you market this library to webmasters, who have no idea how the code works, but the features themselves. I know PHP/SQL/HTML/CSS and minor Javascript. When I learned PHP there weren't many Javascript resources out there, at least none that I could understand very quickly.

In the end I didn't learn too much javascript, when I look at your code and documentation none of it makes sense to me. However when a friend of mine, who uses your library, looks at your documentation he's blown away at it. All I see is the demos and examples you have set up.

You currently have a demo/Example section setup, but it doesn't nearly do your library justice. In most cases it's not up to the programmers on whether the money can be spent to purchase a license, it's up to the webmasters. On this note, I think your new site needs to seriously consider this and be geared towards this market. Examples and Demos need to be taken a step further, when a webmaster comes to your site, he doesn't want to see code, he wants to see results. The current examples (which I'm sure you just threw together geared towards programmers initially to save time) aren't sufficient to market to webmasters. It's difficult to understand just by viewing your demos which items can be used together.

In the beginning I thought that the dialog and tabs had to go hand-in-hand, I hadn't realized the full power and flexability of your library. These are elements, that must be considered. Perhaps you should open up submittable demos and examples? Let the community show what they've done with your library, that'll save you time and also help you sell your lisences.

And I also agree with subscription services. Purely sales aren't going to be frequent enough in the long run. Subscription is the route to take by far. I've used IPB (Invision Power Board) in the past, one of the key elements with their hosting, is that they also provide tech support for software you purchase from them, if you need help doing something.

Maybe you should start a "Community Support" forum for current subscribers, they can go and discuss ideas and things they've achieved with your code, and of course, help each other.

BernardChhun
13 Feb 2007, 6:59 AM
sounds good to me Jack. I will definitely pay for that license.

moraes
13 Feb 2007, 7:18 AM
sounds good to me Jack. I will definitely pay for that license.
Me too. However, as I'm working on a open source publishing tool that makes extensive use of Ext widgets, and I don't understand a lot about the possible mixes of commercial and free licenses, I have some doubts that maybe somebody here can answer.


Free licenses to open source projects which request one
- Does this mean that the source code will be distributable in a open source project which requested a free license?
- Will it allow anyone to use my (BSD licensed) tool in a commercial site without paying for a license?

I wish you the best with your plans, Jack, and although I think my project would fit in the "open source projects" category, I intend to pay for a dev license as I think this work deserves much more. :D

PortraitPainter
13 Feb 2007, 8:53 AM
@vmorale4
I have modified it to include non-profit
Jack I wanted to give you a European perspective. Over here the public sector represents 50% of the economy, Governments should be paying you to incorporate Ext in their projects that often run to $10’s millions.

A report in the UK recently highlighted that the top 50 charities paid their CEO a salary around $150k per year, and I know some big charities use contract IT developers at $100k p/a rates. They too can afford your licence.

So why my money grabbing outlook?

Within a year I think you will need to recruit additional developers to sustain the Ext development and stay in the top league of Web 2.0 UI frameworks.

gordon
13 Feb 2007, 10:13 AM
Jack,

As you know, we've made productive use of your libraries and we are happy to support you in this endeavour.

If a commercial license is the right way to go we'll subscribe to that.

If there are ways we can help in addition to cold hard cash let me know.

Best of luck with Ext 1.0

Gordon

glassy
13 Feb 2007, 10:20 AM
Having used the previous version of yui-ext, it has really given me the confidence that Ext 1.0 for $299 is a great deal :)

Congratulations on going commercial with it!

And having no per server / installation fees is great! Otherwise I wouldn't be able to use it.

Anyways, I just can't wait to start using it ... (I've stopped development on my app for a while now to wait for Ext 1.0!)

SwellJoe
13 Feb 2007, 11:02 AM
Hey Jack,

Welcome to the world of Open Source vs. Proprietary madness, where everyone wants to consult attorneys and wring their hands and worry about your legal safety any time you want to modify the license on code you hold the copyright on. (I've traveled in these waters several times over the past few years.) ;-)

You, as the copyright holder, have a lot of flexibility in what you're doing...don't let anyone convince you otherwise (even when they talk about contacting the lawyers of the FSF...actually, especially then...I can assure you that the FSF doesn't and can't care two sh*ts about software that isn't Copyright FSF).

As an example, our product is dual-licensed and available in commercial and proprietary variants. We use the GPL on our Open Source version (a slightly less powerful and slightly less easy to use version, though the gap is about to shrink some), which protects us from competitors "borrowing" our code--if they do so, they have to give it back to us with any improvements they make. It is entirely sound, from a legal perspective, to license your code in as many different ways as you choose. As long as you are the copyright holder, no amount of gnashing of teeth among the Open Source fanatics can alter that.

That said, your proprietary modules are unlikely to see widespread adoption in Open Source projects, because they are then integrating proprietary software into their codebase, and if their code is GPL, the integrated code would fall under that license--effectively changing the license on your code (which they don't have a license to do). You may find that adoption of yui-ext slows dramatically. When it comes to development tools, being Open Source is pretty much a killer feature.

I suspect your proprietary license will work fine for us, though, as you've described the terms above. Since we'll happily pay the license fees for our commercial variant, and distribution of an Open Source version would fall under the free license (our UI is modular, so it can have a different license from the underlying code, even one that is incompatible with the GPL).

I can think of a lot of ways for things to work out well for everyone (even the "Open Source or die" ninnies), as long as the code quality and scope keeps improving. So, try out the proprietary model on some of the components to keep your bills paid, and see how things shake out over the next few months.

rmesser
13 Feb 2007, 11:48 AM
I'd suggest that you keep the whole library open source and charge more for support, to hopefully make more money for yourself and speed adoption. I manage a small company (16 people), and we use YUI for our web-based apps. I've been watching your Ext library closely as it has some great features that we'd like to use also.

If we could use Ext under a standard license like the BSD license that would give us great comfort, since so many other products use BSD licenses. If we are going to spend the time and money to integrate your library into our products, we need to know that we can absolutely positively no matter what be able to continue to use the code. It would be disastrous to use some library, then later find out that for some legal reason we have to rip it out. It scares people of to even think about losing access to code that they are counting on.

So your proposal here:

- Site license (unlimited developers) $1299
Includes 1 year priority support, 1 yr free minor and major upgrades, source code

may be ok, but I'd have to read the license. Then I might also feel compelled to pay our lawyer $250/hour to read it and render an opinion. What happens if somebody buys you out? What if we no longer need support and we paid the first year? Are we compelled to continue to pay your then-current site license fee in perpetuity?

But if you release the product under the BSD license, I have no worries. I don't need to read the license, I've already read it. And I don't have to pay my lawyer anything either. And I know that if for some reason you go away or get bought out, I'll always be able to use the code I've integrated, so I won't be living under a legal cloud.

Maybe I'm overstating it and your license will be fine, but my initial take is that I'd rather pay $2500 for annual support for a product with a BSD license than $1299 for support for a product with a more restrictive license. In fact, I think you should have higher tiers of support. Go up to $5000, $10,000 or whatever for premium or super premium support and you will probably get some takers. Maybe throw in some proprietary customizations at those points but keep the core and widget stuff open source.

Anyway, it's a tough decision for you but that's my 2 cents.

ilazarte
13 Feb 2007, 1:26 PM
the code that powers the widgets is a convenience layer for us, not a necessity. also, the widgets are not a one to one match for what we need, as we focus on context that needs to be crawlable first, interactive second. that combined with the various upcoming frameworks just tells me to begin investigating again and keep the options open while we build core competency in dynamic web apps. better said, if asked by the bean counters, "do you really need this", i would answer with "well, i can look again".

yui itself has a strong community and at least a starting set of widgets... it makes for a good cost free alternative...

Wolfgang
13 Feb 2007, 2:39 PM
Hello,
I feel that the licence fees and are more than moderate. Anyone doing serious web development should be able to pay 299$ per developer. Also the site licence is very very moderate, as it covers an entire dev-team.

Given that there are no additional licence fees involved for deployment, what else does one want?

I second that there should be a sort of premium support offer on top. But, i am sure that Jack is aware of this.

So one step first.

Regards

Wolfgang

Herm
13 Feb 2007, 2:47 PM
I'm happy to pay a license/dev fee (and stop donating :D ).

But... do you want to be the man behind a wave or a product flogger/controls vendor?

I agree with the folks who say that not open sourcing the code will have a large impact on its takeup in the world. I think an open sourced Ext.fx will have 1000 or 10,000 times more takeup than a paid license version and would expose the library to be considered for use by any larger framework, project, product or vendor (e.g IBM will use dojo libs in announced products, shipping Q2). As some people have said, a closed Ext.fx could NOT be included in most OSS because of their license (not yours).

I'm sure productizing the premium support + premium consulting is the way to go. I would expect to pay US$2000 per day for "Jack Slocum" and US$200/hr for a principal consultant on your team (if you choose to get one together). I would expect $1K, $5K and $10K support packs (that last a year).

I really believe if the whole lib goes fully open and you have the support and consulting products lined up, the revenue will flood in (esp support packs). It is totally a numbers game - a set percentage of the downloads will ALWAYS buy support and a set percentage of the bigger players will always need consulting on a big/new/timeshort project.

If you're seriously concerned about short term revenue issues (like the next month), get an $xK support pack out asap, release the beta fx libs to them and when you're 'safe' open it up to all.

This is hopefully just grist for the mill - it's your call and I'll support you financially which ever way you go.

All the best.
Peter

dfenwick
13 Feb 2007, 2:58 PM
Hey Jack,

You, as the copyright holder, have a lot of flexibility in what you're doing...don't let anyone convince you otherwise (even when they talk about contacting the lawyers of the FSF...actually, especially then...I can assure you that the FSF doesn't and can't care two sh*ts about software that isn't Copyright FSF).



I'm sorry you've had bad experiences with FSF legal advice. My experience with them has been extremely insightful and helpful regardless of the actual license that's applied to a piece of code. As you alluded to later in your post, there are definitely legal issues that need to be ironed out, particularly when one splits a license into three.

Developers need to understand the impacts and ramifications of using the Ext within their code and need to prepare for that. Since Open Source covers a wide variety of licenses (BSD, free commons, Mozilla, GPL v1, GPL v2, GPL v3, LGPL, etc.) it's important to understand the difference between usage and distribution. This will need to be made clear in the license.

As I said in a previous case, I'm not directly tied to GPL. I actually prefer the less restrictive BSD license. The model MySQL uses is they dual license it. It's provided under the GPL, but they also sell their copyrighted code to allow proprietary vendors to embed it in their products. Fortunately the GPL allows the original authors to negotiate terms with authors of derived works, which is what MySQL does.

For the projects I'm looking at writing (all open source), I foresee something along these lines:

- Create a new branch of code with a fully extensible and compatible license as a new project (perhaps Ext-Ext!) that does not include Ext, but simply states it as a requirement for Ext-Ext.

- Each Open Source project I creates has a requirements section that says "You need to download the open source version of Ext as well as the open source version of Ext-Ext".

This will allow me to license my project under whichever license I see fit. Because I won't be redistributing Ext, there is no license incompatibility regardless of what license I choose. Unfortunately, the only downfall to this is that there will end up being at least 2 different projects revolving around Ext; Ext itself and the community extensions to Ext (Ext-Ext).

In the end, I believe the majority of the extensions to Ext will be written by Jack himself because, well, that's what he does. I'm not sure how derived works of the commercial license will even work. Can they sell those works? Can Ext be redistributed in a derived work under the commercial license? We'll have AnimalExt, TryanExt, etc. This is why I'm assuming Jack is consulting a lawyer on the licensing. I just told him I'd get some perspective from the FSF law advisors (since I know them) and see what they have to say.

Incidentally, when 1.0 is released, I will be buying a license for myself. I probably won't even use the license for anything, but I consider it almost obligatory for solving one of the larger UI problems currently out there for web interfaces, which is the grid. His grid is actually akin to the ListView control for Microsoft Windows in report mode. With very little extending I'm sure it could support the other modes (icon, tiles, list). As I've said before several times, the grid alone is a huge enabler for more robust web applications.

SwellJoe
13 Feb 2007, 4:39 PM
I'm sorry you've had bad experiences with FSF legal advice. My experience with them has been extremely insightful and helpful regardless of the actual license that's applied to a piece of code. As you alluded to later in your post, there are definitely legal issues that need to be ironed out, particularly when one splits a license into three.

Who says I had a bad experience with them?

I was merely stating a fact about copyright law, and how involved the FSF can legally become in anything that that they don't hold the copyright on. Being an Open Source developer on several large projects for over eight years I've seen way too many people expect the FSF to become involved in things that they simply have no interest in, and no legally sound basis for, being involved in. I'm just pointing out that the FSF has no interest in Jack's libraries, and simply can't have anything to say about them.

License discussions are always entertaining in the Open Source community, because folks love to wave around "Be careful! Consult an attorney before making this decision!", etc., as though a license has magical powers to ruin lives. Copyright is strong and not easily broken (i.e. the grantor of the license is in a very strong position, while the receipient of the license is generally in a weaker position, by virtue of who created the work).

Jack's problem is basically a stock standard software library license situation, suitable for boilerplate language--licensing libraries isn't new territory. Nobody needs to spend a bunch of time discussing it with attorneys and wasting all that money (Jack's trying to make a little money, not spend a lot). Jack is protected by Copyright...now he just needs to pick license terms that gives away what he wants to give away, to the folks who send him money. He can run those terms by an attorney, just for safety, if he's feeling serious...but the fact is, if someone wants to screw him, he's gonna have to sue no matter what the terms, and if Jack wanted to be an and sue his users, he could do so no matter what the license. The assumption in most situations like this is that all parties are going to behave in good faith, and if they don't it will be egregious enough for a court to see it too (or there's nothing anyone can do about it anyway). The folks worth suing are the folks who aren't paying, and they never enter into the license terms, anyway.

In short, I'm not saying the FSF aren't a great lot of folks (our Open Source product is GPL licensed, against the preference of my co-founder). I'm just saying it's just silly to ask them to chime in on software they didn't write. If Jack wants to produce proprietary code, don't point to the FSF for advice...their only advice to proprietary software vendors is to go forth and sin no more.

In short, I guess what I'm saying is: Let's not waste our time talking to the preacher about how to spend Friday night. ;-)

As I said, I believe Open Source is the number one killer feature of development tools. But it's damned hard to sell something that is free--and nobody has bought the "consulting and support" line for about 7 years (roughly coincident to when there was too much money flowing into technology--everybody floated back then, even the folks with a non-supportable business model). Seriously, come on out here to Silicon Valley and try to sell a "consulting and support" Open Source business to the angels, VCs, etc. I may love Open Source software, and enjoy working on it, but it aint paying the bills.

brian.moeskau
13 Feb 2007, 4:46 PM
Several people are advocating a support-only model. I personally do not think this is a good model for this project -- not because it can't make money, but because I personally would not want to become a "consultant." I suspect Jack doesn't either. Put another way, if someone is paying Jack big $$$ to be available to fly somewhere (or otherwise drop what he's doing) at a moment's notice, then Jack is spending his time doing support and not actually building the product. Yes, Jack could potentially build a support organization so he's not doing all the flying around himself, but that's a whole new set of challenges. Does Jack want to be a consultant "manager"? That sounds even worse. Is Jack even interested in or capable of running a full-fledged consulting organization? Well, only he can tell that for certain, but again, my guess is the answer to some or all of that is "no." Yes, support has to be part of the equation, but predicating the entire business model on it makes actually building Ext the lowest priority.

I think at the end of the day, it's also a lifestyle choice. Some people are happy consulting, some supporting. Others (not naming any names here) are at their best when sitting at home, alone, fueled by Red Bull, and cranking out new stuff left and right, 24x7. Unfortunately, you can't force someone to adopt a lifestyle not suited to them, regardless of how that may affect the Open Source world. :)

ilazarte
13 Feb 2007, 6:58 PM
Several people are advocating a support-only model. I personally do not think this is a good model for this project -- not because it can't make money, but because I personally would not want to become a "consultant." I suspect Jack doesn't either. Put another way, if someone is paying Jack big $$$ to be available to fly somewhere (or otherwise drop what he's doing) at a moment's notice, then Jack is spending his time doing support and not actually building the product. Yes, Jack could potentially build a support organization so he's not doing all the flying around himself, but that's a whole new set of challenges. Does Jack want to be a consultant "manager"? That sounds even worse. Is Jack even interested in or capable of running a full-fledged consulting organization? Well, only he can tell that for certain, but again, my guess is the answer to some or all of that is "no." Yes, support has to be part of the equation, but predicating the entire business model on it makes actually building Ext the lowest priority.

I think at the end of the day, it's also a lifestyle choice. Some people are happy consulting, some supporting. Others (not naming any names here) are at their best when sitting at home, alone, fueled by Red Bull, and cranking out new stuff left and right, 24x7. Unfortunately, you can't force someone to adopt a lifestyle not suited to them, regardless of how that may affect the Open Source world. :)

we also have the project in mind. i didn't come back here and watch this product obsessively because it was a commercial lib. there were other successful projects that had trees and grids before even dojo became popular...

http://www.scbr.com/docs/products.shtml
http://www.componentgo.com/
http://www.milonic.com/

are a few, yet you never read about these guys on digg, or see them enjoy high integration into products... why is that?

you are right in that it really comes down to what Jack wants to do, and as a member of the fascinated and grateful community I'll be happy for him if commercializing his lib can support him and his family. I'm just vocalizing an opinion that I know is probably not unique to me.

brian.moeskau
13 Feb 2007, 7:31 PM
are a few, yet you never read about these guys on digg, or see them enjoy high integration into products... why is that?
Um, so does it then follow that if those other libs simply adopted a GPL + support model, that suddenly they would become more popular on Digg, etc.? I think it might also have something to do with the amazing level of work that Jack produces and the quality/comprehensiveness of the library that has contributed to it's rise.

I'm not arguing that open source is bad (we all know it's great!). What I'm arguing is simply that the support model is not the best answer to how Jack can pay his bills. If anyone has any other ideas (aside frrom strictly support/consulting) I'm sure everyone would love to hear them.

dfenwick
13 Feb 2007, 8:14 PM
License discussions are always entertaining in the Open Source community, because folks love to wave around "Be careful! Consult an attorney before making this decision!", etc., as though a license has magical powers to ruin lives. Copyright is strong and not easily broken (i.e. the grantor of the license is in a very strong position, while the receipient of the license is generally in a weaker position, by virtue of who created the work).

Since I've been doing this for 25+ years now both professionally and as a hobby, I can tell you that consulting a lawyer about licensing issues is smart and prudent for your own protection. I am not, by any means, an FSF zealot. I've written both proprietary and open source code. The point I was trying to get across, which apparently I'm failing at, is not the licensing from Jack's perspective, but the licensing from the consumer's perspective. It's important to understand, as a consumer, what you can and can't do with this library.

A license can't ruin lives, but it can certainly affect the direction a project has to take. I'd be just as happy if Jack uses the BSD license for non-commercial and open source work. That would be a compatible license, provided I don't extend his widgets.

I guess most of this will shake out when the official license is released. It's funny, we're probably both on the same side of the fence on this licensing. I just offered to get a little advice from someone I know will give proper (albeit slightly slanted toward openness) copyright advice. I've also asked another friend of mine, who specializes in proprietary copyright law, about the topic. He's much more corporate about things. I'm really getting opinions from both sides of the fence on the topic so I understand what I may or may not be able to do with it. Until there's an official license, I don't think either one of them will give me any advice. Standard lawyer stuff. "Send me the license and I'll have a look."

justheatingup
13 Feb 2007, 9:03 PM
I don't think that this project would have taken off as quickly or with as much enthusiasm if it not for two things: 1) it's open source 2) Jack slaving away.

There's a fork in the road.... Jack can continue to put in sleepless hours, reap the rewards with the current honor system, and go broke. The alternative, try to commercialize the product and hope that corporations will pore money into buying licenses.

My personal feelings are both of these solutions are not in Jack's best interest. With all this Web 2.0 stuff going on.. SOA enabled SaaS solutions taking over the web, Jack's building widgets?? I'd give the library away, get traction to the site and build a working product that has a pay per drink model.

37 signals does this with prototype and basecamp.

garyR
13 Feb 2007, 9:43 PM
- Site license (unlimited developers) $1299
Includes 1 year priority support, 1 yr free minor and major upgrades, source code

Can you please define "Site license" is this a single physical development site?

Can you please clarify exactly what "priority support" is?

Thanks

ilazarte
13 Feb 2007, 9:46 PM
Brian,

A possible alternative is aligning with a framework that is making money off of it's own consultancy. Spring is a popular one (supported by Interface21 consulting) and another popular option is JBoss Seam (supported by JBoss). Both already have models in place to support their open source products and both are lacking strong web 2.0 technologies, and both have very strong Java community mindshare (read: customers with BIG bucks willing to pay top dollar for quality).

I would approach it from the perspective of integration and new development in exchange for financial backing. That way, they stay with the consultancy and Jack could code to his hearts content while empowering their consultants so they could deploy these gorgeous front ends. I would think it'd be a highly symbiotic relationship as customers would quickly come back with what features they need added to the toolkit.

Seam would probably express more interest quickly as it seems to be going after the Rails audience with it's feature set, and thus is probably looking for a strong JS toolkit to integrate with; plus as the late-comer they're looking for feature differentiation to put them ahead of Spring. FWIW I'm a Spring guy.

I dunno, just something to consider as a potentially happy medium.

Ivan

ilazarte
13 Feb 2007, 10:31 PM
- Site license (unlimited developers) $1299
Includes 1 year priority support, 1 yr free minor and major upgrades, source code

Can you please define "Site license" is this a single physical development site?

Can you please clarify exactly what "priority support" is?

Thanks

The former is important for me as well. What would be the cheapest option for:

1. 15 domains worked on by 12 developers (2/3 leave per year)
2. Team of 6 work on an internal administrative application which powers a commercial site but is never viewable by a consumer.

mmcmahon
14 Feb 2007, 7:33 AM
I agree with ilazarte on this topic. I wouldn't be tracking this lib with the same amount of interest if it were commercial. Should Jack be compensated, yes of course.

I've been actively tracking the progress to ver. .40 and patiently waiting for the break from YUI to model a solution for review at work. If accepted... sure I would get them to pay for the lib if it remains reasonable.

Going from $0/donate --> $1300 + support or whatever sales model is a sign of things to come. I didn't figure on Ext joining Telerik and the others... but's that seems to be where it's heading so I'll have to begin evaluating them as well.

I know I could never approach the manager and say we should donate to this lib developer as that is non-professional. A mechanism to for 'pay what it's worth' with a capability of generating a quote page for the boss to cut a cheque to would be great. I think if Ext measured up to the other libs/controls (which it does), I would happily have the boss cut you a cheque for a few $K. I guess my perception and I think your popularity up to this point is the quality of your solution and that it is free for review/use. Altering it from a *donate* to a 'pay what it's worth to you' are two entirely different scenarios.

Those are my thoughts, take them for what their worth (similar model :) ). No reply necessary.

Regards, Michael

rmesser
14 Feb 2007, 8:21 AM
One other issue with making Ext more commercial (even if the "core" is open source) -- it will probably be harder to get free help with testing and documentation and such. When a library is commercial, people who can't afford it tend to just go away. With open source, people who can't pay feel compelled to help in other ways. They are contributing to a community resource, so they won't ask to get paid. But with a commercial library, even if it is priced reasonably, people will be much more likely to demand payment for any assistance. Maybe Jack has enough help lined up (paid or unpaid), but just something else to think about.

yogurtearl
14 Feb 2007, 10:15 AM
When you say "free for open source use", are you talking dual-license like QT and MySQL?
Their products are available under a GPL license or a commercial license.

Michael

JeffHowden
14 Feb 2007, 10:42 AM
One other issue with making Ext more commercial (even if the "core" is open source) -- it will probably be harder to get free help with testing and documentation and such. When a library is commercial, people who can't afford it tend to just go away. With open source, people who can't pay feel compelled to help in other ways. They are contributing to a community resource, so they won't ask to get paid. But with a commercial library, even if it is priced reasonably, people will be much more likely to demand payment for any assistance. Maybe Jack has enough help lined up (paid or unpaid), but just something else to think about.

I completely disagree. ColdFusion has never been open source, has always cost a reasonable amount per server license and there is no shortage of help available in the community.

dewd
14 Feb 2007, 11:06 AM
I completely disagree. ColdFusion has never been open source, has always cost a reasonable amount per server license and there is no shortage of help available in the community.

I think that's an important difference between a client side technology and a server side one. YUI-ext, while being client side, can work with any server side technology, but only a little percentage of the folks who use server side technologies will care to go beyond their usual designs and adapt to a whole set of YUI-ext widgets on their own. For example, maybe less than 1% of the Java and ASP.NET guys would be willing to learn and use YUI-ext with their server side techs.

Prototype, meanwhile, by being integrated in the Rails framework by default, is used a lot of the time by the Rails users, just because it's convenient. The same goes for the Java and ASP.NET own technologies.

What interest would the server side technologies have in integrating YUI-ext support, now that every user which is a developer would need a license to customize/develop their sites? Can you imagine Sun, Microsoft, Rails, Zend/PHP, etc, providing YUI-ext out-of-the-box for their users? If not, then YUI-ext will continue to be a niche technology used only by a few users.

I've just subscribed to the jQuery mailing list, as I am trying to learn more about it now, and their community is growing very fast, they have a lot of plugins already, and their future looks bright. jQuery could be integrated with some server side technologies free of charge/licensing issues for their users.

Alright, jQuery does not have the goodness of the YUI-ext designs/widgets, but that's the "Bazaar" way of open source that seems to work OK in the long term.

All that said, I am supportive of Jack trying to find a solution to his financial needs. YUI-ext does not need to be popular for Jack to make it a relative success for himself. :-) We all have our own dreams, right?

rmesser
14 Feb 2007, 12:33 PM
I completely disagree. ColdFusion has never been open source, has always cost a reasonable amount per server license and there is no shortage of help available in the community.

But I'm not referring to help from the community, user to user. Sure, even on closed source commercial products, there can be vibrant and helpful user communities. I'm just saying that if the product is not open source, users are going to be less likely to document stuff for Jack or help him create his yui-ext website for free. He can give them licenses for their efforts, but that will only go so far. People tend to consider open source projects more of a community thing and you get unpaid volunteers much more readily.

That's why you can get help from other users for Cold Fusion, but it's less likely that somebody is going to write a chapter on some new Cold Fusion feature for Adobe and give it to them for nothing.

Anyway, maybe the commercial licensing plan will work out fine for Jack and he can just pay people where he needs help, but IMHO if it's not open source it will make finding help more difficult for him.

Maecenas
14 Feb 2007, 1:14 PM
Prices seem more than fair to me. I hope you are aware that by charging *any* amount of money means that people will become much more demanding. Offering email based support for the cheapest licenses might proof something of a burden.

Besides that, as mentioned before, what is non-profit? Is, for example, the webdevelopper who works for the red-cross (who pays the webdevelopper a competitive loan) non-profit? Or is work non-profit when the webdevelopper makes no profit? I'd go for the latter. But then there still is the issue what if I help a friend whom only pays the webdeveloppers mere expenses. However this seems moot, as in that case the webdevelopper is most likely to also make commercial sites, in which he'll use YUI-EXT, thus he already pays.

What I fail to understand is how people can say that the price is too much? It's not even near the monthly loan of a webdevelopper, i fail to see how any commercial institution cannot afford it. (And in the end, if you think it is cheaper to develop your own lib, no one will stop you.. Just remember it’s no one’s given right to use YUI-EXT)

I think it is unrealistic to assume one person (Jack) is going to spend the better part of his time on this project while not being compensated for it. While it can be argued that a full blown opensource YUI-EXT (as I understand the basic parts will still be opensource) might be more successful.. until now all the work done is by Jack and see what a nice piece of JavaScript he has delivered, i think he deserves enough credit to safely assume the library will continue to be excellent, maybe the success is partly due to the fact that it is a one-man show?!

On the other hand, it is true that it will be more difficult to get YUI-EXT included in OS projects. But is that really needed. YUI-EXT commercial version is solely about js widgets. The ajaxy part is YUI (and some basic parts of EXT), which are/stay opensource. (And can be as it is part of a larger project, it has no dedicated developers that have to make a living on it). If Jack wants to ease the use of his library in other framework, it'll be up to him to provide the tools/binding for that framework.. Not sure if that's going to happen, but it is his choice.

(On a side note, I myself would not pay, thus use, the commercial parts. I’m mostly a hobbyist, my websites, by using banners, generate a tiny amount of income but that’s in no relation to the time spend on them. But that’s ok, I can afford to mess and waste time with the alternatives. If I ever were to take on a project which needed the widgets, I’m sure I could explain the 300$ costs to my client)

jack.slocum
14 Feb 2007, 1:23 PM
This is a bump for those subscribed:

http://www.yui-ext.com/forum/viewtopic.php?t=2737

Webnet
17 Feb 2007, 7:52 PM
You're on the right track jack :)

SonicBog
18 Feb 2007, 5:32 PM
Hi Jack,

Could I emplore you to not go down the deployment license route.
We are a small company and money is tight. Many of our clients have verly limited funding and so adding the cost of a delpoyment license would negate our ability to use Ext. We have had to drop the use of many products (exceptional ones like Ext) due to our clients funding limitations.

I am waiting eagerly to see what is in Ext 1 as I have been using YUI-Ext for some time now.
Again, keep the developer costs reasonable as many good utilities go commercial and get left behind due to under/over pricing.

It's a difficult balancing act trying to bring in enough money to live on yet keeping the product within the reach of many.

Keep up the great work :) [/code]

KimH
19 Feb 2007, 12:18 AM
SonicBog,

did you see Jack's reply a couple of response up in this thread?


This is a bump for those subscribed:

http://www.yui-ext.com/forum/viewtopic.php?t=2737

You should really be checking out this URL ;-)