A couple of days ago, Sanjiv Jivan announced the names of the new members of the Gwt-Ext project. This comes after his decision to step down as a project lead, which was triggered by the Ext JS license change (and the discussion that followed). I don’t really want to summarize the argument between Sanjiv and Jack Slockum (Ext JS main contributor). Below are the links to the relevant blog posts:
- Gwt-Ext user catplusplus writes in a forum topic about a fork of Ext JS called OpenEXT Jack Slockum responds to that post in that topic.
- Sanjiv Jivan responds to Slockum on his blog
- Slockum posts his thoughts on the license change, and answers to (what he calls) personal attacks. This post is located here, but if you want to read the original (unedited) version, you can download it here. The original is significantly longer and addresses Sanjiv Jivan’s arguments.
- A few days later Sanjiv posts an update on the future direction of Gwt-Ext, in which he announces his decision to step down as a project lead.
As of now, the Gwt-Ext project supports only the LGPL(ish) Ext JS 2.0.2, and not the GPL 2.1 version. I strongly support that decision, but cannot help but to wonder if the Gwt-Ext project still has a future. Gwt-Ext is a GWT wrapper for Ext JS, and as such should do it’s best to support the current version of the wrapped library. On the other hand, a huge selling point for it was that it was free (as in free beer), and going in the other direction, would mean loosing that advantage. How will the new contributors deal with adding new features to Gwt-Ext? As long as they were wrapping Ext JS, they only had to wait for the Ext team to add some feature and wrap it, now they will have to start thinking about adding new stuff by themselves. Won’t that be too much of a burden? I really admire Sanjiv Jivan’s work, so it’s not easy for me to say that I don’t really see a future ahead of his project. It’s by no means his fault, and I really hope I’m wrong. There is just no right direction in which the Gwt-Ext project can turn, and any decision made by the team will probably be a wrong one. This pretty much signals that Gwt-Ext will soon face extinction.
This brings me to the main point, which I meant to discuss in this post. How much this Ext JS license changing business affects the GWT library landscape? You may remember Gwt Site having a contest for the most popular GWT library, here you can find the results of that contest, and below you can read the Top 5 GWT 3rd part libraries as of January 14th 2008:
- MyGWT
- GWT-Ext
- GWT-SL
- Gwittir
- Hibernate4gwt
As you can see, the two most popular of those libraries are Ext Js based. MyGWT, now an Ext product rebranded to Ext GWT (for the sake of clarity I’ll refrain here from using that name in favor of MyGWT), is a library that directly competes with Gwt-Ext, but written completely in Java. The earlier versions of MyGwt used only Ext JS images, version 1.0 uses also it’s CSS stylesheets. With Darrell Meyer joining the Ext team, the license/business model change also affected MyGWT (despite Darrell announcing that MyGWT will stay LGPL in 1.0 earlier on MyGWT forums).
Today of the two most popular GWT libraries, one changed it’s licence, and is no longer freeware, and the other one’s fate is still uncertain. I don’t have a problem with paying for MyGwt, mind you. Actually, if Darrell announced the change a couple of months back, I probably wouldn’t complain at all. It’s pretty common to pay license fees for widget libraries after all. The problem for me is that the licensing terms seem no longer to be in Darrell’s hands, but Jack Slockum’s and I just plainly don’t trust that guy at all.s For all I know tomorrow MyGwt might be AGPL/Commercial and God knows how the commercial licence terms will change in future. Charges per deployment / user/ CPU - everything seems possible if Jack decides that he’s still making not enough money out of his library.
I have mentioned before the OpenEXT project, which is going to be a set of patches that you apply onto the Ext 2.0.2, this way conforming to it’s license and being effectively a fork of Ext JS. I think it would be great to see someone doing something simmilar with MyGwt. MyGwt was licensed under pure LGPL license, with the exception of the so called “assets” (css and image files). Maybe it would be possible to create a real fork of this project given that it’s only dependency on Ext JS lies in the assets? I don’t feel up to the task, but I still hope that someone will eventually step up:)
For me, both MyGWT and Gwt-Ext are now practically useless. It’s a real shame since these are really wonderful tools for creating slick-looking back-office applications. However I cannot invest my time (and therefore money) in creating software using a tool, which might stagnate or become no longer supported (Gwt-Ext), and I don’t want to give money to an unethical and greedy company owned by Jack Slockum. So what can be used instead of those two libraries? I’ll propably check out IT Mill Toolkit which is licensed under Apache 2.0 licence, or maybe just stick with plain old GWT widgets - the new showcase application looks pretty nice, and there will be three styles to choose from. Anyway I think that with recent developments, a place has opened in the market for a new Gwt-Ext/MyGwt-like library. Hopefully it will emerge soon.
When Ext license change topic was still hot, one of the bloggers wrote a nice post titled “Ext Discovers Step 2 of the Slashdot Business Model?“. To thank you for reading this pretty long post, I’ve decided to add an episode clip from South Park, which I thought about immediately after reading his post. Unluckily for us, Ext JS team has discovered what the step 2 is…
Cheers!
Psyho
