-
1. Re: Warp replacing JSFUnit? yes? no? make the message more visible
aslak Aug 31, 2012 11:34 AM (in response to anthonyhib)More or less.. But Warp is only Alpha1. As it matures the message will become clearer.
-
2. Re: Warp replacing JSFUnit? yes? no? make the message more visible
kito99 Aug 31, 2012 5:16 PM (in response to aslak)Why create a new project instead of working with the existing JSFUnit code base?
-
3. Re: Warp replacing JSFUnit? yes? no? make the message more visible
aslak Aug 31, 2012 6:04 PM (in response to kito99)Basically because they work completely different. Not much to reuse at all, and starting from scratch you can focus on the new instead of being bugged down by the old.
-
4. Re: Warp replacing JSFUnit? yes? no? make the message more visible
lfryc Sep 4, 2012 6:10 AM (in response to aslak)Hey Anthony,
you are totally right.
There are currently no reference documentation, just a blog and functional tests showing the usage.
But as we are moving project forward, the documentation starts to be critical piece.
If it's something you would like to help with. Let me know,
@Kito: Warp is so important because it enables things that were not possible with JSFUnit and its redesign trying to shape it to work more closely to Arquillian was basically at least as complex as writing Warp. More at https://community.jboss.org/thread/198339?start=0&tstart=0
-
5. Re: Warp replacing JSFUnit? yes? no? make the message more visible
kito99 Sep 4, 2012 8:15 AM (in response to lfryc)Thanks for the info. The thread is really helpful. It'll also be nice to have the project maintaiend by someone who has the time to work on it.
-
6. Re: Warp replacing JSFUnit? yes? no? make the message more visible
ssilvert Sep 4, 2012 8:54 AM (in response to kito99)Kito,
Let me chime in and say that I totally support what Aslak is saying. JSFUnit was based on Cactus and HtmlUnit. Its internal design and much of the API reflects that fact.
If I were writing JSFUnit today it would look a lot different.
I gladly passed the reins of JSFUnit over to the Arquillian team, and not just because they have the time to work on it. There is only one major thing missing from the current JSFUnit. That is speed of execution, which is what you will get with Arquillian. It's going to be really great.
Stan
-
7. Re: Warp replacing JSFUnit? yes? no? make the message more visible
anthonyhib Sep 4, 2012 9:52 AM (in response to ssilvert)Hey Stan,
what about adding an announcement there http://www.jboss.org/jsfunit/
Simply telling users to *forget* JSFUnit and keep an eye on Arquillian ?
That would avoid a lot of confusion.
@Lukas
Having been a very active Hibernate member during more than 4 years (community forum/support, documentation, ...) I know how expensive it is to maintain a good documentation.
Unfortunatly, my workload is BIG here (as all of us at Red Hat), I can't promise anything.
In a general manner arquillian code is moving ways faster than the documentation, that is an issue, I agree.
Cheers,
Anthony
-
8. Re: Warp replacing JSFUnit? yes? no? make the message more visible
lfryc Sep 4, 2012 11:08 AM (in response to anthonyhib)Anthony Patricio wrote:
Having been a very active Hibernate member during more than 4 years (community forum/support, documentation, ...) I know how expensive it is to maintain a good documentation.
Aslak's concept of self-generated Reference might play nice for Warp.
-
9. Re: Warp replacing JSFUnit? yes? no? make the message more visible
anthonyhib Sep 4, 2012 12:05 PM (in response to lfryc)certainly.
Something that could also help and I'm sure it already exists (but not easy to find) is a way to grab "everything".
When it's time to invest time in trying a new framework, honestly we don't care about modularity, that problem comes later, we want to test a bunch of features.
It seems arquillian is extremely fine grained in terms of modularity and that is a pain when you spend hours trying to fix your pom.
-
10. Re: Warp replacing JSFUnit? yes? no? make the message more visible
kito99 Sep 4, 2012 1:40 PM (in response to anthonyhib)Amen! It's also a royal pain if you want to use it without Maven.
-
11. Re: Warp replacing JSFUnit? yes? no? make the message more visible
anthonyhib Sep 4, 2012 1:45 PM (in response to kito99)I'm afraid it would be even more complicated without maven (and I'm not a fan of maven at all but ...).
I'm talking about about a mega dependency that would bring everything arquillian, drone, warp, graphene, persistence, shrinkwrap (+all its extension) ...
-
12. Re: Warp replacing JSFUnit? yes? no? make the message more visible
alrubinger Sep 4, 2012 1:53 PM (in response to anthonyhib)While the modularity is necessary to keep concerns correctly separated, I don't think a "dist" POM (with accompanying build to generate a ZIP/TAR) to combine all dependencies together is a terrible idea.
For instance ShrinkWrap has: https://github.com/shrinkwrap/shrinkwrap/blob/master/dist/pom.xml
-
13. Re: Warp replacing JSFUnit? yes? no? make the message more visible
ssilvert Sep 4, 2012 1:58 PM (in response to anthonyhib)Anthony Patricio wrote:
Hey Stan,
what about adding an announcement there http://www.jboss.org/jsfunit/
Simply telling users to *forget* JSFUnit and keep an eye on Arquillian ?
That would avoid a lot of confusion.
Cheers,
Anthony
Yes, it's probably time to do that.
@Aslak, is there a particular page I should point people to? It would be great if we had landing page for former JSFUnit users.
Stan
-
14. Re: Warp replacing JSFUnit? yes? no? make the message more visible
aslak Sep 4, 2012 2:08 PM (in response to ssilvert)@stan Let's hold of until we have a Warp Guide + Alpha2 out. Without a guide users are just frustrated. And Alpha1 has some serious sharp edges which should be fixed for Alpha2.
My 2 cents.. @Lukas ?