Jaxer is a dead technology. Some necromancer hacker may resurrect him, but he does not look likely.
Jaxer 1.0 was released in January 2008 and after that received only minor updates, mostly fixing typical 1.0 errors and limitations. The latest release 1.0.3 is about three years old as I write this.
Three years is a long wait for an open source product update, especially considering that Jaxer died almost the same way the entire Javascript speed race began. Jaxer is based on Firefox 3.0, so it does not have TraceMonkey and JägerMonkey has been promoting Mozilla since its release.
Jaxer code has since been put into the repository, but no significant commits have been made since the initial check.
the old support forums are closed , and the new Jaxer mailing list contains some content . This mainly consists of unanswered questions.
Aptana, the company that created Jaxer, was bought by a desktop-oriented manufacturing company. Appcelerator Acquisition Frequently Asked Questions Aptana says they intend to continue developing server-side products for Python, PHP, and Rails. Jaxer is not counted, described as “in maintenance mode” with ads that appear at the end of the first quarter of 2011. Since then crickets.
Prior to the buyback of the Apperator, Apatana released Jaxer from Studio download. You could still download it separately, but this little comfort, given that the support for it, it seems, "if it breaks, you get to save both parts."
Even if Jaxer was a healthy, prosperous project, it still has some serious flaws:
The Javascript browser, the Jaxer platform is built, does not know how to run the code in reliable separate sandboxes, so you need a separate interpreter for each site hosted on the server. It costs memory, one of the largest drivers of higher hosting fees.
It would not be so bad if the main technology was an economical RAM user, but it is not: this is Firefox, a puzzle program. While debugging a server crash due to out of memory, I found that the main problem was that starting an external process started using the 35 MB virtual machine. This was a one-time hit only for the first run, not a memory leak, but multiply this by the number of Jaxer processes you serve connections with (3, by default, with higher efficiency), and you have already finished 64 or 128 MB for some cheap hosting plans. The absolute bare minimum for hosting a functional Jaxer site is 256 MB, and 512 MB is much safer.
You can get around some of Jaxer's voice nature by working on virtual machine technology that allows you to have swap space, so things like the GUI libraries that Jaxer (!) Uses do not remain in RAM, but many do not. Often you have to use more expensive hosting to get something like Xen, and not a more efficient VM system like OpenVZ, which does not allow you to use the swap space.
Since Jaxer is no longer patched, it still has errors that can lead to server crash or freeze. I was faced with a situation where some random bot on the "network" accessed random URLs on my server, causing Jaxer to start for each of them, as a result of which the system was out of resources. I had to rebuild the way my site was built, so bot hits were served only by Apache, limiting Jaxer to one subtree of the site. This is just a workaround, because all you would need to do to start it up again is access to lots of random URLs in the new subtree. Every month or so I have to restart the VPS because it consumes all the RAM, probably because someone has stumbled upon this design weakness.
The bottom line, Jaxer is completely unsuitable for cheap, shared hosting.
So that you do not think that I hate the hater, let me protect myself by indicating that I was an early adopter of Jaxer (before 1.0), was one of the most active posters on the original forum and had two sites running on Jaxer. Yes, I'm upset, but mainly because one of these sites was transferred from one dead technology to another, which died during the year of this move. (The other was created from scratch after I successfully transferred the first.) Now I am stuck wondering if I need to move again or continue shipping on a platform that looks like it will no longer move forward. You do not want to be in the same boat as much as I like the company.