Maqetta: Open Source HTML5 Editor From IBM 121
PybusJ writes "IBM has released an online HTML5 editing tool called Maqetta, hosted by the Dojo Foundation. eWeek calls it an open source answer to Flash and Silverlight. That remains to be seen, but it does look interesting."
Re:Pronunciation (Score:5, Informative)
from the FAQ;
'"Maqetta" is a spelling variation of the Spanish word for mock-up ("maqueta"). The team members pronounce the name as if the "q" were a "k".'
So, yes. :)
Re:Holy fuck. It makes Eclipse and VS feel fast. (Score:5, Informative)
Well the entire thing is coded in dhtml, not html5 but pure dhtml which scales down to firefox 3.5.
And actually it is not that slow, it is ok and what you can expect from a pure html + javascript solution.
The neat thing about this is, it puts out pure dojo widget code, which is heavens sent if you want to generate
dhtml forms rather quickly.
The downside is it pushes out the more easily readable pseudo tag code
which is slower on browsers which do not have document.querySelectorAll than the programmatic initialisation (nevertheless pseudo attributes are a no go before html5 as well where the syntax for those things was finally finalized)
Either way thanks ibm and this is a neat thing. The dojo library undeservedly is rather unknown it is one of the most extensive javascript libaries out there and one of the oldest as well, and one of the cleanest designed ones as well.
Jquery is utter junk compared to it, but it does not reach entirely the code quality of the YUI 3 lib.
Let's be fair... (Score:2, Informative)
Could be that it's Slashdotted, not that it's inherently slow.
Re:Holy fuck. It makes Eclipse and VS feel fast. (Score:3, Informative)
Maybe you should try it when the site's not been slashdotted...
Been using Maqetta for past week (read news via eWeek) and it's not that the site is slashdotted, it simply darn slow.. The things is slow even when running locally (the distribution is bundled with jetty 7).
Re:Holy fuck. It makes Eclipse and VS feel fast. (Score:4, Informative)
Yeah, I would agree that the documentation is extensive on dojocampus.org if it wasn't so frequently vastly and utterly WRONG. I've been developing with Dojo since somewhere around .08 (circa 2005 I think) and even back then it was attrocious - but rightly so - it was a new project with an overwhelming number of people who jumped in the pool. And for versions 1.3 through 1.5 the documentation was often never updated. The documentation would say in big bold headers - This documentent is outdated please click here for the new documentation - which every single one of those links pointed to dojocampus.org - THE HOME PAGE, not the actual updated documentation for said specific feature - because more often than not that page DIDN'T EVEN EXIST. I'm not lazy and am willing to then research for the new answers except that if they existed they were so frequently wrong that I personally deem the dojo project and utter failure. Sure I was able to bend the framework to my needs when forced to, and figure out why the hell line 75,278 threw a JS error, but I would never start a project with that POS again.
And to even suggest books as an alternative indicates to me that you haven't worked with DOJO for a LONG time (by tech standards anyway) because the last books published for this pile of garbage were in 2008 - and they were some of the last books I will ever purchase on computing due to their innaccuracies and outdated mode of technical coverage. Even those books admittedly report that the content in the books is deprecated but should be applicable until version 2 - which was over 3 years ago - and they are still only at 1.6 and they are still just as innaccurate as they were in 2008 because those Russell is about as naive to think the interfaces would be backwards compatible for that long as equally naive he is on the practicallity of the framework to begin with.
And even had they been more accurrate (a feat I don't blame Russell for because no one could be expected to properly document about it in it's persistent state) they still let you down - Mastering Dojo should be a lot more than just how to use their widgets out of the box. I would have much rather seen more than 20 pages of documentation on how to use their abstract databinding interfaces for remote data in all of its variant forms with regard to a few key controls than over 100 pages of What is AJAX and Why use Object Oriented Programming in Javascript (WHICH IS A FUCNTIONAL LANGUAGE!!!!) Instead you get a grade school level of understanding when you expect a Black Belt level of intimacy as the pretty cover and double entendre would suggest.
When you have a javascript client side framework that is over 21 MB ZIPPED!!! you have a fundamental architectural and maintenance problem that is there to stay. It is a perfect concrete representation of the conceptual frustrations I have the the JAVA community as a whole and why I'm ashamed that I couldn't release my HTML based HTML composer that produces clean developer quality html markup before IBM released Maqetta because it's the type of solution that if done correctly can really spurr development versus wasting cycles on building a page from scratch every time - but when that development comes with the cost of DOJO that solution could just turn out to be a scar that nobody wants to revisit but for the wrong reasons.
Even the DOJO build systems are terrible - they do not work as designated and past that have platform incompatibilities that produce entirely different results from when you build on Windows versus Linux. The architecture promotes the direct importation of over 50 separate libraries without the build, which from a "designed for the masses" perspective is a FAIL because you end up with people who hack in the results and produce a lag time on page load that is worse than YUI in it's early days.
The widget aspect is nice, but the dependencies and points of frequent failure between them is unbearable. I've worked with some difficult and poorly documented sy