Which OpenLaszlo version should I use for new projects and when migrating applications to the latest version in 2012?

Question

As a former committer on the OpenLaszlo project, I've been getting a lot of questions regarding the best OpenLaszlo version to be used when migrating applications using either a 3.x version or 4.x version lower than 4.2, or when starting a new project with OpenLaszlo.

The difficulty in making this decision is related to the fact that there has not been a minor or major release of OpenLaszlo since October 2010, and that there has been close to zero information by Laszlo on planned releases, any kind of marketing or participation in developer conferences by Laszlo employees in the past 2 years (check the News and Event section on the Laszlo System homepage, or the OpenLaszlo announcement mailing list).

Still, OpenLaszlo is a very powerful platform for building either HTML5 or Flash/SWF based applications.

I will answer my own question here, hoping for others to join and report their results with using either the 4.9 or unreleased 5.0 version of OpenLaszlo in projects.

Aneroid

Answer 1

Last official release 4.9 or OpenLaszlo 5.0 (trunk)
The last official release of a minor OpenLaszlo version was the release of OpenLaszlo 4.9 in Oct 2010. The last official release viewed by the OpenLaszlo team as a major release was version 4.2 in Dec 2008. Since the 4.2 release added ActionScript 3 and SWF9 support to OpenLaszlo - therefore a new runtime - it definitely should have been a new major version.

After the 4.9 release, a lot of bugs had been fixed until April/May 2011 - when development slowed down (based on the number of messages sent to the laszlo-reviews mailing list). In 2012, Laszlo did a small number of improvements on the DHTML runtime, and I have personally added SWF11 runtime support to OpenLaszlo, which is checked into the flex4.6 brnach.

Many of the bugs fixed in trunk and new features were not ported back to 4.9, and developers using OpenLaszlo to target the DHTML/HTML5 runtime and the SWF11 are well advised to use the unreleased version of OpenLaszlo 5.0, or the unreleased flex4.6 branch (the first OpenLaszlo version with the SWF11 runtime support I added to the platform).

DHTML/HTML5 browser support in 5.0 trunk
Browser support for more modern browsers in 5.0 trunk seems relatively stable for IE7, IE8, the group of Webkit based browsers and Firefox. Opera might work, but will not be fully tested by QA, as far as I know. IE9 is not supported at the moment, although that might change if Laszlo manages to release 5.0 at some point in the future. OpenLaszlo runs on the iPad - and I think that most optimizations for a mobile DHTML runtime were done with regard to the iPad, back in 2009 and 2010.

Flash Player 11 API support and the SWF11 runtime
In March 2012 I started work on adding SWF11 runtime support to OpenLaszlo, and checked the code into the new OpenLaszlo flex4.6 branch in July. The flex4.6 branch is a fork of OpenLaszlo trunk, and contains some DHTML runtime improvements added by Laszlo. If you want to utilize Flash Player 11 APIs with OpenLaszlo, the flex4.6 branch is currently the only option. Since I've stopped working as a committer in August (cannot accept the new Laszlo policy, that critical mails sent to the laszlo-dev and laszlo-user mailing list are blocked), I don't know if Laszlo will manage to integrate the SWF11 runtime support into trunk in the near future.

OpenLaszlo 5.0 release date - when is Laszlo Systems going to release 5.0?
The BIG question which has been popping up in the mailing lists over and over. I heard dates ranging between end of 2009 and end of 2011 - but there is still no release. It's hard to tell what is going on at Laszlo, since the engineering management refused to make any announcements on release dates for the past 3 years. After the acquisition of Laszlo by Critical Path in early 2012, the only public announcement regarding OpenLaszlo is a sentence in an image header on the OpenLaszlo.org website:

"CP (Critical Path) is committed to continued sponsorship of the OpenLaszlo Open Source project and its global community."

Who is using OpenLaszlo 5.0 trunk in production?
I don't know many products using OpenLaszlo trunk, I initially thought that Gliffy's HTML5 version of the diagram editor was built using OpenLaszlo, but it looks like they switched to jQuery and HTML5 canvas with some JS libraries instead (the Flash version of Gliffy was built using OpenLaszlo). Laszlo seems to use 5.0 (trunk) for customer projects with Webtop, although I never got an official confirmation for that.
OpenMeetings - an open source web conferencing tool and Apache Incubator project - uses OpenLaszlo, and they have started using 5.0 trunk with SWF11 support.
I've been using 5.0 trunk for a project in the past 15 months, and haven't run into problems which could not be resolved. The application I'm working on uses both the SWFx and DHTML/HTML5 runtime.

Status of documentation in 4.9 and 5.0
Many of the newer features are not well documented, since there isn't anyone responsible for upgrading the documentation at the moment. Laszlo used to have a resource working full-time on the documentation until the end of 2007, but since then it seems that the documentation has not been well maintained. Most developers tried to document new features when committing new code, but that is not equal to having a technical writer working on the documentation.

Support and the community
The visible OpenLaszlo community is really small by now, I've seen activity in the forums and mailing lists by 30-40 people at most in the past 2 years. I'm trying to move the discussion away from the OpenLaszlo forums (which are pretty dead) to Stackoverflow.com, since you'll at least get reputation points here when answering questions, and it's not guaranteed that someone will pull not the plug on the OpenLaszlo.org website in the future. It looks like most questions on Stackoverflow tagged openlaszlo will be answered, but the number of questions tagged with openlaszlo is still below 100.

Recommendation This is my personal opinion, don't blame me if you run into problems following my recommendations. As a former committer I know the source code well enough to make changes to the OpenLaszlo server, if you don't have that kind of knowledge, things might be different for you:

1. Always use 5.0 (trunk), when migrating to a newer OpenLaszlo version.
2. For new projects, consider the risk of Laszlo or Critical Path (the new owner of Laszlo and OpenLaszlo) discontinuing support of the platform. If they continue to sponsor the project as little as they have done in the past year, the end of OpenLaszlo is very near - unless we as the community fork the project.
3. New projects using OpenLaszlo for HTML5 apps: If you only target the DHTML/HTML5 runtime, other options or frameworks might be better. OpenLaszlo makes it difficult to modify the DOM or integrate with existing JavaScript applications. With the uncertainty around the future of the project it's a relatively big risk to use this technology.
4. New projects using OpenLaszlo for SWF11 based apps: You should be fine using OpenLaszlo, since the product is very stable, you can use any ActionScript 3 API, and even compile to an iOS or Android native application (using the Adobe AIR SDK and a few build scripts). I've successfully tested OpenLaszlo 5.0 with the new Apache Flex SDK (the first release of Flex as an Apache Incubator project), and it's easy to upgrade the Flex SDK to newer versions, using these documents I've created.

Aneroid

It's discouraging if Gliffy has dumped OpenLaszlo for the HTML5 version in favor of JavaScript. They must have such a large code base written in OpenLaszlo, and still basically re-created the application. – Forfeiture

I'm assuming that Gliffy is still an OpenLaszlo application compiled to the HTML5 (formerly DHTML) run-time. Is this the case or did they drop OpenLaszlo and create a new HTML5 application without using OpenLaszlo? – Objectivity

I thought it's an OpenLaszlo app, but when I looked at the client source code (HTMl and JavaScript), I didn't see anything OpenLaszlo specific in it. But then, the Gliffy blog states: "That’s why Gliffy has spent 2012 rebuilding the software from the ground up in HTML5 — and now you can take a look!" That's surprising... – Aneroid

Ah, I see in your comments now that Gliffy did indeed drop OpenLaszlo, I was skimming while reading before so I missed it. Honestly, if we didn't already have a large application written in OpenLaszlo already, I'd probably drop it too. It is a great platform, but we plan to move to HTML5 and mobile apps soon and I don't think OpenLaszlo in its current management state is the right tool for that. – Objectivity

Answer 2

I would recommend using the Nightly Build 5.0.x because there are many bug fixes, new features, and much improvement to the HTML5 run-time (formerly called DHTML run-time) that are not available in 4.9.0.

OpenLaszlo 5.0.x seems to be very stable. I have a fairly complicated application (30,000+ lines) written in OpenLaszlo 4.9.0 and I tested compiling it under version 5.0.x and all the features of my application seem to function correctly under both the SWF10 and HTML5 run-time modes.

Unless you have an existing application in a version of OpenLaszlo less than 4.2.x already there is no good reason to continue to use that platform, you should just start with 5.0.x. If your application is between versions 4.2.x and 4.9.0 you should try to compile it to 5.0.x and if you get it working, then use 5.0.x from then on.

The nightly builds of 5.0.x are available here:

http://download.openlaszlo.org/nightly/trunk/

Also, there is another branch with SWF11 (Flash 11) support (written by https://stackoverflow.com/users/410652/r-bitter) that has not been integrated with the main 5.0.x nightly builds yet, if you need SWF11 support the nightly builds of that branch can be downloaded here:

http://download.openlaszlo.org/nightly/flex4.6/

Objectivity

Answer 3

Taking a quick look at the old openlaszlo site (which I built for Laszlo), they haven't done any significant work on it since I left. The forums were last updated in 2009. (heck, the openlaszlo.org site still shows a 2008 copyright. they didn't even change that).

This isn't a project that's being actively supported, if they can't even fix the copyright in the footer...

Melioration

It is true that the former owners and the new owners of openlaszlo.org have not really done anything to actively support OpenLaszlo for a long time (except continue to host the openlaszlo.org site). But the forums do still work well, if it ain't broke why fix it? The main problem is the dwindling community since OpenLaszlo appears dead since no official version has been released since Oct 22, 2010. – Objectivity

And the non-existent marketing. Compiling into JavaScript has become so popular in the past years, but Laszlo never managed to position the technology as a true alternative for building powerful HTML5 applications. I haven't seen anyone from Laszlo presenting the technology at a conference or meetup. Which is a shame, since the company is directly in Silicon Valley, and there lots of meetups with an interested crowd. – Aneroid

Yeah, it's a shame. OpenLaszlo 4.2+ can already do what Adobe wants to be able to do in the future with FalconJS, namely compiling an XML+ActionScript language into pure HTML5+JavaScript. But since there is no marketing barely anyone hears about OpenLaszlo these days. OpenLaszlo had a great opportunity to become a serious option for HTML5 and mobile apps but they kind of missed the boat by sitting idle and doing nothing for nearly the past two years. – Objectivity

Adobe is not going to work on FalconJS any more. They are going to open source the proof-of-concept, which was created by a single developer. It's up to the Apache Flex Community to create FalconJS as a product, but I doubt that it's going to happen. – Aneroid

if you don't invest in the community, it stagnates and dies. The lack of any updates to the site is something a community picks up on and a lack of interest by the leadership convinces everyone else to go elsewhere.. – Melioration