Published 2008-05-07 00:00:00

I think that basically sums up my reaction to Jack's GPLv3'ing a library. For those who are not aware, As I was not, when I visited the extjs web site a few days ago to grab a copy of the old 1.1 version and found all references to download it had disappeared. Digging a bit deeper on the site, I started spotting a few comments about the new licensing.

While I can say that Jack as the author (of I presume most of extjs?) has the right to change the license to whatever he likes, I think he has probably just destroyed the project. I could not commit to writing new code with a "Library" that is GPL, unless I was working on a GPL project (which is unlikely at present - got bills to pay). And for commercial or spec projects, that are not turning revenues yet, I can't really justify my time in committing to develop stuff that may, or may-not be able to fund today's and whatever Jack feels like charging in the future for the non-commercial license.

I've seen too many people burned by this closed source dependencies that they build their businesses around, only to have the effective rental for their office yanked through the roof, and no other option than a huge effort moving to another library or software causing chaos.

All that said, I'm not sure if v2.0 and v2.1 are really worth bothering with anyway, I've tried them on 2 projects so far, and the general sense I get, is that compared to v1.1 they are a little finicky, and tend to produce slightly unpredictable results. Which is tempting me to stick with 1.1..

The problem I see though is that the community that has built up around extjs has been supported by quite a few handy tools, the wiki, the doc's and the rather nasty forum (which is a good way to waste time finding answers to issues). So I guess someone setting up a openext? (I saw something on the net about it) should probably sort out those issues first, then start solving the technical issues about how to replace all the images and css in extjs which where not previouslylicensed under LGPL.

I would hate to have to go looking at the alternatives again, So sticking with v1.1+ hacks may be the best long term plan for me anyway.

Mentioned By:
www.dzone.com : Ouch, what a mess, - Extjs goes GPL3 (232 referals)
www.planet-php.net : Planet PHP (53 referals)
google.com : december (40 referals)
google.com : extjs (40 referals)
www.google.es : Versin traducida de http://www.akbkhome.com/blog.php/View/163/Ouch_what_a_mess__Extjs_goes_GPL3.html (27 referals)
google.com : extjs fork (21 referals)
computer-internet.marc8.com : Ouch, what a mess, - Extjs goes GPL3 - Alan Knowles | Computer & Internet (15 referals)
google.com : extjs timer (14 referals)
google.com : extjs vs dojo (13 referals)
google.com : extjs sound (12 referals)
google.com : extjs gpl3 (11 referals)
google.com : extjs 1.1 download (10 referals)
www.dzone.com : DZone - New links (9 referals)
google.com : extjs alternatives (9 referals)
google.com : extjs lock column (9 referals)
google.com : what a mess (9 referals)
google.com : extjs locked column (8 referals)
planet.dsource.org : Planet D at Dsource (7 referals)
google.com : extjs accessibility (6 referals)
google.com : extjs grid lock column (6 referals)

Comments

Dojo baby
I don't know how, but I had avoided the YUI / extjs library until stonecobra sent me over there a few weeks ago. I'll admit that it's pretty shiny and nifty in a few places. But I guess my choice of Dojo many moons ago might turn out to be a good one. In addition to their accessibility work (ARIA), their grid just kicks ass. It allows you to lock multiple rows/columns from scrolling, and is still fast, unlike ActiveWidgets. Further, the Tango license was inspired by the Dojo Foundation's dual license... Good stuff by Alex Russell and gang over there.

Hopefully you can join us in Dojo fun?

BA
#0 - boorad ( Link) on 2008-05-07 04:17:11 Delete Comment
Dojo doesnt cut it
Unfortunatly, Dojo (from experience and even a quick look at the updated demo's) doesnt appear to cut it. (although I could be wrong)

- The email demo is jiberish / a mess of code.
- The HTML extra tag's stuff is good for adding small effects to web pages but not feasible for larger application interface writing.
- The docs' are missing any event information, and are a world away from ExtJs.



#1 - Alan Knowles ( Link) on 2008-05-07 08:02:38 Delete Comment
Stick a fork in it.
I used to be a big evangelist for Ext, well before the project became an all-in-wonder library and was just a collection of widgets on Jack's blog. Honestly, his team has managed to get Javascript/HTML to do things that even some desktop environments don't do right. The lib is easy to use, performs decently, and is *very* well documented.

But to slam the door on the ecosystem that LGPL afforded them is a colossal blunder. Once upon a time, I was considering writing 508 compliant peers for that lib*, and now I can see that would have cost me dearly.

That said, ExtJS is going to fork so fast, it'll break the sound barrier; which is a shame since the corporate customers there were pushing some of the API improvements through. It looks like I'll be joining Brad on the Dojo front soon enough.

(*Thanks to Jack telling me, in the time-wasting forums, that they didn't deem accessability features to be at all important. Nevermind that only the entire Federal Government of the USA requires it.)
#2 - Pragma ( Link) on 2008-05-07 12:00:28 Delete Comment

Add Your Comment