Hooray for Ares! What about Enyo?

edited July 2013 in Enyo 2
There is a veritable crapton of extant Enyo issues that the team not only has not fixed, but in many cases hasn't even gotten around to screening even after 6+ months in pre-screen purgatory. The vast majority of recent JIRA activity is around discussing issues and features in Ares, while nearly 500 reported Enyo bugs are unresolved. Should I expect to continue having to implement hackish workarounds for Enyo bugs because of the team's focus on shiny rather than fundamentals?

Comments

  • Pull requests gladly accepted.
  • The reason you're not seeing much activity on JIRA is that a lot of that has moved to an internal JIRA instance for the work we're doing on the TV platform. However, I agree that we should be paying more attention to the public JIRA. Are there specific issues that you want us to focus on more?
  • @Sugardave I'm not sure why I would spend time submitting fixes to any of the 313 issues which no one has looked at yet, some of which are mine, to a project that hasn't released anything in 6 months and has no roadmap.

    @unwiredben I guess a lot has changed since you said this: "the Enyo team will still be working on it as an open source project at LG. Our mission hasn't changed."

    It's not very encouraging when the team said you could rely on them are actually now spending a lot of their time working on a secret fork with a secret JIRA. I've built quite a few applications with Enyo over the past year (we are in the developer directory), but now that Sencha Touch 2 no longer requires Webkit, I don't know why I would use Enyo for my next project.
  • edited August 2013
  • @TravisWebbUSA, we have been doing an awful lot in our open source repos. While Onyx hasn't been getting much attention, there are a ton of changes in enyo and a small number of changes in layout. The master branch of enyo is the one that we've been using internally for projects; we're not holding back anything there. The main hold up on an official beta of 2.3 is API stability. The main new features of 2.3 are

    - data binding
    - observers
    - computed properties
    - kind mixins
    - data models & collections
    - controllers
    - application objects
    - data-bound repeaters and lists
    - enyo.super-decorated methods for speedy super calls

    There are implementations of these all in enyo now, but they are still changing. Some people are making productive use of them, but we haven't been ready to stamp they with our "have at it approval" stamp because we've been seeing problems and making adjustments continuously since February. Just this week, Cole and I were simplifying a lot of the implementation and making it run faster.

    I think we're really close. I've been posting occasional updates on architectural changes to the enyo-development list.

    We're going 90 miles an hour internally right now, but we don't have the programmer capacity internally at the moment at LG to do much support of Onyx. I really don't like that, and we're trying to hire additional people.

    Here's our roadmap: get 2.3 ready for beta in the very near future and release it, while working endlessly to improve framework performance.
  • edited August 2013
    I don't know where to get bug requests legitimized. I see you guys working in github but I'm not too comfortable with pulling from master, especially considering the current level of flux. I'm not sure what you guys are actually building, and so even if I did want the bleeding edge, there's no documentation. Being involved in the process is one thing, which I would like, but I currently feel as if I'd need to be a enyo core developer in order to make effective use of what's happened in the last 3-6 months. Perhaps you can point me to some documentation or even design proposals or very well-commented code so I would be able to use it... whatever "it" is.

    Most projects solve the long release problem by maintaining more than one current release; develop new features on 2.3-SNAPSHOT, keep releasing related bugfixes on 2.2.1, 2.2.2, etc. Also, the website mentions nothing anywhere about 2.3, what features to expect, and when it might be done. If you don't know, then say so: "Sometime in 2013" is more informative than "". In the world of community FOSS, silence precedes death. Most people, when choosing what framework to use, don't scour forums to figure out what features it has and what direction its heading in. They look at the main website for this sort of guidance.

    If I've filed bugs that you guys don't have the bandwidth to fix, feel free to assign them to me in JIRA. That everything there is lingering in nebulous purgatory is the main problem that needs to be solved; many hundreds of issues simply not even being screened is a huge problem. I'm sure you guys are doing great work, but liberating your output from obscurity is just as important as the output itself.
  • That's good feedback. We do know that the website is really stale, and I think we'll ask our devrel team to see what they can do. We've honestly felt like we were about to release 2.3 beta for months now, so it has always seemed like we could wait for that event to do updates. They have done some outreach and added things like the application spotlight site, but they've also been busy building up http://lgsvl.com and working on internal developer support.

    It does seem like a good idea to do a 2.2.1 release. For what its worth, onyx + layout + canvas could be taken from master and work. It's enyo that's had the grand bulk of changes, and backporting small changes from it to a branch based on 2.2.0 would be tough.

    The timing of things earlier this year is a bit to blame... we just released 2.2.0 at the beginning of February, just before my son was born. I was out for a couple of weeks, and came back in work just in time to see the LG acquisition. We were in limbo for a little then as we had computers and servers swapped out, and since late March, we've been on this major development project that's had us all occupied.

    I personally think enyo core is in good shape if you don't use the new features. It's much more testing over the last few months than in all the time before, although most of that has either been desktop Chrome or on new embedded hardware.

    As for docs -- we're behind schedule on that. I was originally tasked with doing the docs on new features, but it's taken too long for me to come up with examples, mostly because I've had a hard time understanding some of the designs. Now that I've used the features, I've got a better feel for them, so I hope I'll be able to get to that work as we release the beta. We've got inline docs for a lot (see http://latest.enyojs.com/api/), but the conceptual docs and tutorials are still very much in flux.
  • hang in there guys! love palm and enyo
  • Huge enyo fan that can't wait for 2.3. @unwiredben, can't your kid code yet? The team needs to do some short term cloning to get more hands on deck. Does enyo have a cloneSelf function yet?
Sign In or Register to comment.