Apple Hardware Mozilla

Third Party iPhone Apps

Apple is said to be wrestling with the idea of allowing third party apps. In reality they are deciding if they should aid developers. I think it’s pretty clear there will be hacks to get third party software on there. The question is if Apple will bless the efforts or not.

Come on Steve… release the SDK and let us have some fun and make your product even more useful. The availability of good software is what has kept the Palm Treo and every Windows Mobile device alive despite Blackberry fever.

It would be an interesting platform for a Mobile XULRunner via Mozilla’s mobile efforts.

Don’t forget. Part of the Mac mini’s success is that it is small enough to embed in all sorts of places. Apple has even taken advantage of that press by putting it on their Mac mini page (lower right “Big Ideas”). Hackable products sell. Just ask Linksys about the WRT54G.

Mozilla Open Source

Open Source Joost

You all know about Joost (formerly “The Venice Project”) by now right? You know it’s based on XULRunner right? Well it’s using a ton of open source stuff, and it displays it rather proudly. You can see it all by visiting their open source site.

They use a lot of open source.

Mozilla Software

The Venice Project

I got an invite to the much-talked-about Venice Project a few weeks ago, and managed to take it for a spin a few times. GigaOm among others like Engadget and BusinessWeek have also mentioned it. A few noteworthy things:

  • It’s running on Mozilla’s XULRunner! Confirmed by Allan Beaufour.
  • Video quality is extremely good, and buffering time is minimal even when getting only about 4Mbps broadband (during busy holiday times as I typically do 6-8Mbps). This means we should expect Linux/Mac versions to follow.
  • Content is extremely sparse. This is really the Achilles’ heel of the venture. Without content it’s worthless, despite being technologically cool.
  • It doesn’t quit when you close the app by default. It minimizes so it can still share data. This isn’t a very cool thing if you care about bandwidth consumption (you may, or you may not). I understand why they do this though.
  • UI is very fluent and almost TiVo-ish, though not always as intuitive as one would like. It is still a beta, so I’d expect things to improve. One thing I did email about was getting tooltips to show up sooner, which would help remedy this.
  • Doesn’t seem to pierce through firewalls as easy as Skype does (though still does a pretty good job). Not sure if this is on the todo list, of if it’s something that can’t be fixed. Perhaps Skype just set a really high standard.

I don’t want to go into a very serious review because it’s still in development, and it doesn’t seem I can include screenshots other than those released just yet. I plan to revisit this with a more thorough review at a later date. Perhaps when it’s permissible to include screenshots, or when a beta picks up a feature warranting a little more info. As someone who loves seeing new innovative software, and businesses, this has been a real treat to check out. As someone who appreciates the tech behind the scenes (go XULRunner!) it’s really a treat.

Edit: Now known as Joost.


XULrunner vs. Dashboard

Daniel Glazman says:

Forget Konfabulator, forget Dashboard, forget Tcl/Tk. XULrunner will kill them all. As usual, it’s not planned to kill them. It’ll be just a side effect of the existence of XULrunner.

I agree, but with an exception. Konfabulator, and Dashboard have much better management of widgets than XULrunner appears to have (I’ve yet to see any documentation planning something better). XULrunner manages widgets more like Java. It simply doesn’t. It relies on the OS to do so. It’s purely a runtime.

If XULrunner could be run in a Dashboard type mode (run multiple widgets, background, be rather graceful and beautiful), and good installation methods to add widgets and manage them. It would be a killer app. Especially since it’s so portable and powerful.

But as a runtime (which is essentially what XULrunner is at this time), it’s not quite a killer. It’s apple’s and oranges.

I’d personally love to see the ability to run XULrunner in a “Dashboard mode” so that it managed XUL based widgets similar to how one of those other products manage widgets. That really would be an impressive feat. Perhaps that would be a great XULrunner app for someone to write. A portable, fast, lean XUL based Dashboard app?


XUL Runner First Impressions

Downloaded XUL Runner (mentioned a while back) today to just check it out, and have a few notes. I’m not bashing anyone or the project. They are just things that I feel are incorrect, could be better, places for improvement. It’s not even at it’s first production release. Takes most products several releases to get really good. So it’s already ahead.

  • If I launch xulrunner.exe (or os equivalent) I should be prompted to open something, be brought to a shell prompt, told to open a file directly, or something. Nothing is incorrect.
  • 5.8MB download isn’t bad. But if I’m making a “Hello World” program, that’s a big download. I think the documentation needs to say the bare minimum needed for a distribution and which files needed for feature X, and feature Y. So that download size can be kept down. I may not need SSL for my app. I don’t want to bundle it.
  • Performance is pretty decent. Can’t complain.
  • Should recommend developers create a batch file on windows rather than open command prompt and enter commands. It’s just friendlier and shows how it can be used to create an application.
  • App #1 should be a Xcode for XUL like App. Think about it. A simple way to manage files, a decent editor, and a simple “build” which bundles things appropriately… it will be a pretty solid platform. Perhaps a good mozdev project.
  • Overall totally cool and I’m looking forward to some time to play with it and have some fun.

XUL Runner

I wanted to comment on this the day I got back from vacation, but have been busy.

Way to cool! I can’t wait to get a chance to play with this.

I’m hoping a few things happen when this step is done:

  • Simple IDE – And I really do mean simple. Start with Composer/NVU as a base (since they rock). What both Composer and this IDE would need is color syntax editing (already in the works), and autocomplete. Additionally a XUL WYSIWYG editor would be great, but that could come later on. IIRC people are talking about a WYSIWYG editor already. The most important thing is good JS, and XUL editor. This could be repurposed for composer anyway, since many authors would like to edit code with ease, XUL is XML, which would be good for XHTML editing as well. JS is used on both web development and mozilla development.
  • Ability to ‘compile’ – By this I mean we should have a tool that essentially contains a stripped down FireFox. When you ‘compile’ you then create the equivilant of the browser.jar file. And put everything together Leaving you with a simple release ready build, with an executable, and necessary libraries. Ideally this tool would come with binaries for a few OS’s (Windows, Linux, Mac), so a simple checkbox on the compile dialog would allow you to build your product for all operating systems (take that .net). This tool ideally would be part of the larger IDE.

With the above, the Mozilla platform would really be ready for developers. Part 1 would minimize the learning curve, by helping you code. Part 2 would help you put together really nice packages (for multiple platforms). The great thing about the ‘compile’ feature would be that you can develop for more than 1 platform with ease on your platform of choice.

Just my $0.02