Add a "recommended version" to the homepage?

[Carl Erik Kopseng]

I was very enthusiastic about the BusterJS project for a while, but ended up scrapping it for daily work for a much more manual and harder to use workflow with mocha-phantomjs. Simply because Buster seemed to break too often ... Now, I realize this is due to Buster being in very active development, and that you are working very hard to make new features and the 1.0 release, something I am looking very much forward too. But ... waiting isn't doing much for adoption.

So what I was wondering was if there is a recommended version to use, or possible a somewhat more stable, patched, older version? Or if there could be one ... ? Like the Wine project having 1.2, 1.4, etc as "stable" versions, and 1.1, 1.3, etc as development versions, with stability patches being backported.

I can see there being problems with this, mainly that this project is run by a handful of developers in their spare time, and using time on backporting fixes can be time consuming and affect progress. Which again can affect motivation. Thought I might raise the issue anyhow.

[Christian Johansen]

Hi,

Since the release of 0.7 buster is pretty stable. We still have things
to do for 1.0, but there will be no more user-facing breakages.

Welcome back! :)

Christian

[August Lilleaas]

Since we have not yet reached 1.0, there isn't a stable version yet. If 1.0 existed, that's what I'd recommend :) When it is out, we'll be doing semver with dev branches, so 1.1 will also be stable.

So for now, I think the promise of an upcoming 1.0 is the best we can do.

And your e-mail motivates me to keep up the work, it's good to hear that some developers prefers the workflows in Buster to existing solutioks, despite the fact that numerous alternatives has been released since buster was architected a couple of years ago.

Sent from my smartphone.

Carl Erik Kopseng <carl...@gmail.com> wrote:

I was very enthusiastic about the BusterJS project for a while, but ended up scrapping it for daily work for a much more manual and harder to use workflow with mocha-phantomjs. Simply because Buster seemed to break too often ... Now, I realize this is due to Buster being in very active development, and that you are working very hard to make new features and the 1.0 release, something I am looking very much forward too. But ... waiting isn't doing much for adoption.

So what I was wondering was if there is a recommended version to use, or possible a somewhat more stable, patched, older version? Or if there could be one ... ? Like the Wine project having 1.2, 1.4, etc as "stable" versions, and 1.1, 1.3, etc as development versions, with stability patches being backported.

I can see there being problems with this, mainly that this project is run by a handful of developers in their spare time, and using time on backporting fixes can be time consuming and affect progress. Which again can affect motivation. Thought I might raise the issue anyhow.

--
You received this message because you are subscribed to the Google Groups "Buster.JS" group.
To unsubscribe from this group and stop receiving emails from it, send an email to busterjs+u...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[Andrew Gibson]

I'd like to second this request. We tried using buster for a year at our company because the unified server-side/browser-side framework makes a lot of sense. However, we've since given up and moved over to Mocha - similar to Carl. Issues that finally made us move away were:

1. Even though we have bandwidth, there doesn't seem to be an active issue list that people can volunteer to fix and merge back. There didn't seem to be a way to avoid waiting for you guys to have capacity to do it yourselves. We also found that the large number of packages made it hard to track down bugs.

2. Documentation proved hard to navigate - it seems like all the information's there somewhere, but it's less than easy to find what you need if you encounter an obscure issue. Docs don't seem to be open source / collaborative.

3. Occasional unexplained issues with the browser-side becoming unresponsive. There are similar issues with other testing products (e.g. Selenium et al.), but with phantom/html-verifying assertions, we were able to avoid the issue entirely.

I still think buster is potentially one of the best - but not as it currently stands.

[August Lilleaas]

Andrew,

Would just like to say that I completely agree with you. I would only recommend buster if you can live with some rough edges. If you said that after 1.0 was released though, I would curl up in a ball and  go cry in a corner. But as it currently stands, we're (painfully) aware of the problems and will try to fix them all.

We would certainly like to have more "core" contributors (i.e. people making architectural decisions etc) for the core modules , and I think that will happen after 1.0 when Buster is stable and a very good choice for more people. Until now, buster has been in a state of pending rewrites that only existed in my and Chris' heads. We already have lots of great contributors for extensions that provides additional functionality, though, that's awesome!

Documentation is a never ending story.. I personally prefer more structural docs, and less prose. We have a lot of prose now. Some modules have good docs that lists all their functionality, but we lack some "meta" docs for everything you can do with the combination of modules that makes the whole that is Buster.

Do you experience occasional browser unresponsiveness with 0.7 btw? If so I'd certainly like to hear more about it.

[Andrew Gibson]

As I recall, the problems with browser not responding were after an exception had occurred during a test run, causing subsequent test runs to be ignored. but would have to dig to find more details.

On 25 October 2013 16:08, August Lilleaas <aug...@augustl.com> wrote:

browser