Riding the escalator at OSCON, someone asked me if Mason was dead. I was a bit taken aback, but I figure if one person was wondering, then maybe others are too. Certainly, development has slowed down a fair bit in the last year or two, but that's not necessarily a bad thing.
But here's a summary of what's up with Mason development, and I'll try to do this on a semi-regular basis:
- Mason 1.30 is very close to release. I'll go out on a limb and say it will be out by mid-August, if not sooner. This is the next big release of Mason, and brings the following changes ...
-- Support for mod_perl 2 out of the box. You should be able to configure Mason exactly like you did with mod_perl 1:
-- Various speed enhancements from Jon Swartz and Amazon. This mostly affects "static_source" mode, though performance should be a bit better in all cases.
-- A new plugin system that lets you hook various parts of Mason's execution. Right now there's just a few plugin points but more can always be added later.
Those are just a few of the highlights, as there's lots of other changes to (see the Changes files once 1.30 is released).
Now, this release did take a while to get out, but I think the main reason for this is that Mason 1.2x works so darn well! I cannot remember the last time we had any serious bug reports, and the docs seems to be good enough that we get many fewer beginner-type questions than when I first started using Mason.
In the past couple years, there's also been more ways to extend and use Mason via the MasonX modules. This is something I've been particularly interested in. Using MasonX::WebApp I can add lots of additional behavior to my web applications without touching the Mason core code. This is a good thing, as it means Mason provides enough of a complete API to be useful in many contexts, and I think this is a sign of a very mature piece of code.
So what's next? I'm sure that after the release of 1.30, there will be a number of bug reports related to new features, so that will occupy our time. I hope that people start playing with the plugin system and that we see even more interesting extensions uploaded to CPAN. If I get enough round tuits, I may find time to update the online copy of the book for 1.30. I should note that the book's POD source is available, and patches are welcome.
I use it in all of my web applications, and I know it has many, many corporate users. Take a look at the jobs listing on the MasonHQ site, courtesy of jobs.perl.org. There have been seven job listings in the first seven days of August alone that mention Mason. As one of the jobs site admins, I see most of the listings, and I'm pretty sure Mason is mentioned in postings more than any other Perl templating system. So Mason is most definitely not dead.
Re:to be considered alive, Mason needs...
autarch on 2005-08-08T14:23:46
Heh. I do AJAX (actually AJAHB where HB = HTML blob) with Mason. My AJAX code makes a request for/ajax/... and that executes a Mason component which returns a JS data structure. The structure is a hash where the keys are element ids and the values are chunks of HTML. Then the JS code just finds the element and sets innerHTML. Piece of cake, no special AJAX support needed.
The idea that I'm going to export server-side methods/components, return some sort of data, parse it, and then mess with the DOM is just madness. The server side is good at generating HTML, so I'd like to do as much HTML generation there are possible.
Hmmm... It looks like they are looking for people to eventually port their stuff to java servlets.We are looking for PERL gurus with major strength in an OO language – C++ or Java. Linux, Apache, mod_perl, mason, Remedy, cgi, Oracle – these are the tools we want today; but C++/Java, jsp, servlets and much more are the direction we want for the future.
Re:Speaking of Mason and Amazon
autarch on 2005-08-27T17:01:49
Heh, from what I've heard there's a _lot_ of different sub-groups inside Amazon trying lots of stuff. I wouldn't take any one job posting as gospel. If you search for software dev jobs mentioning Mason on their site it gets 20 hits, and only 2 have that wording AFAICT.
