XULRunner 1.9 without ‘system-level integration’?

I’ve noticed a recurring theme here: those who say we can’t put out a XULRunner 1.9 are saying it’s because we need system-level integration: one XULRunner instance supporting several applications on the same machine at the same time, I think. This is the JRE runtime model I’ve seen a couple people throw around.

I want to throw out a radical idea: what if we simply dropped that requirement for 1.9?

Simply put, I think it’s a nice-to-have convenience, but I as an application developer do not need it. If that’s the primary point of contention preventing a XULRunner build being blessed as 1.9, maybe we can have our cake and eat it too.

Perhaps we simply need to ask a better question: What does the community need, and what can they live without?

8 thoughts on “XULRunner 1.9 without ‘system-level integration’?”

  1. If you throw out the shared, system-level requirement, then which remaining part of the announcement is causing you grief?
    (From Alex: The “We’re not going to bless a XULRunner build” part. The system-level requirement was never a concern for me. I’m simply suggesting if we threw out the requirement, we might be able to get a release.)

  2. I think this is a capital idea. I’d much rather see standalone packages with XULRunner versions that are known to work than a common install with different add-ons.

  3. Earlier, I was completely convinced early on that a “singleton runtime” was the way to go. These days, I’m not so sure.
    I think it would be beneficial, for 1.9 at least, to focus on tools for ease of deployment, rather than trying to cobble together a single unified runtime to “rule them all”.
    A bit like the “Extension developer’s extension” for Firefox, only for XULRunner. An easy to tweak build system, some support for easy creation of installers. Essentially, build and deployment tools.
    Application _users_ don’t care one bit if it’s a single runtime juggling all applications or one runtime per application (download size with good compression is negligible).
    Generally, I think you’d probably see _more_ end-user problems with the “singleton runtime”, due to versioning problems and such.
    When deploying a separate XULRunner per application, the developers know exactly what they’re dealing with. It makes testing easier. And the released software always comes with a runtime that “works”.
    Focus could probably also fall on making various separate XULRunners co-exist gracefully. Making sure they don’t step on eachothers’ toes. Possibly, that’s done through deployment tools, not necessarily requiring XULRunner code changes at all.

  4. What is meant by one instance?
    If it’s meant that several apps run in one process, that’s a terrible idea. I prefer firefox+thunderbird+… to the mozilla suite any day. In fact, I’d often like to start a separate process for them, to improve reliability and/or performance (using one thunderbird for mail and another for news would be nice).
    If it means potential profile sharing (sharing cache / prefs), that’s something that should be considered (and is needed for running multiple instances of the same app)

  5. I agree with you. If xpcom was designed at the application level, for me seem ok that xulrunner will be an application framework. I mean, its power and utility worth to include it with your application.
    I am not saying not to allow a JVM behavior in the future, just that now it is not vital.
    A good way of solve this is, suppose that mozilla reduce its efforts an 50%, the community can do that part, helping mozilla in focus on Fx and not ‘leaving’ xulrunner at all.

  6. So, what you are asking for is kind of a continuation of what is being done for XULRunner in 1.8? Soft of stealth released milestone builds that you can link to from your product’s web page?
    (From Alex: I’d prefer it wasn’t stealth, and that we had better docs, but essentially yeah: repeating what was done for 1.8 with 1.9 would be fine.)

  7. Whilst having a single runtime with smaller downloads for the applications would be nice one day, it’s far from essential. Bundling XULRunner with each app isn’t a huge concern for me in these days of DSL and CD-ROMs. Disk and memory are cheap enough that a few independent XULRunner instances aren’t likely to be a big problem.
    If lack of a shared runtime is the main reason for Mozilla to not want to produce official 1.9 builds then I think they need to think again. A non-shared runtime is better for a developer than no runtime at all. Remember that many developers, and potential developers, are coming to XUL from the web world. The use of JS and an XML language appeals to them. Having to crack out a compiler and build our own runtime does not. It’s enough to immediately turn off a lot of casual developers who just want an exe that they can code like a web page.

Comments are closed.