Windup Meeting Minutes - 2014-09-24
by Lincoln Baxter, III
Minutes:
http://transcripts.jboss.org/meeting/irc.freenode.org/windup/2014/windup....
Minutes (text):
http://transcripts.jboss.org/meeting/irc.freenode.org/windup/2014/windup....
Log:
http://transcripts.jboss.org/meeting/irc.freenode.org/windup/2014/windup....
Meeting summary
---------------
* Agenda (lincolnthree, 14:07:09)
* Status Reports (lincolnthree, 14:08:33)
* This past week I worked on merging several larger PRs, including the
configuration / graphImpl changes (lincolnthree, 14:09:41)
* I've also worked on merging the XML PR, and a few others
(lincolnthree, 14:10:00)
* This week I am preparing for JavaOne so not working on windup as
much, but still trying to review things (lincolnthree, 14:10:24)
* I have the weekly BU meeting today so we'll see what they want :)
(lincolnthree, 14:10:35)
* Next week I will be mostly offline because I'll be at JavaOne
(lincolnthree, 14:10:48)
* I also worked on creating the Roaster command line tool
https://github.com/forge/roaster#cli so that folks can work in their
own code format and still submit code that matches the project
(lincolnthree, 14:11:42)
* I worked on finishing up the port of visitors from the earlier
brad/grapher branch (jsightler, 14:13:02)
* I completed that, and then worked on finishing up some tweaks to the
reporting, as well as fixing a few problems with the charts that
lincolnthree found (jsightler, 14:13:24)
* Now, I am working on getting rules written in an xml syntax to work
within windup (jsightler, 14:13:44)
* I also worked on tweaking the groovy syntax a little to make it less
verbose, and to provide an example of basic Java hinting in Groovy
(jsightler, 14:14:25)
* I sent PRs for: (ozizka, 14:15:25)
* WINDUP-301 Skip certain packages when scanning for *Model classes.
(ozizka, 14:15:25)
* WINDUP-299 Refactor FurnaceClasspathScanner, *Filter, et al.
(ozizka, 14:15:25)
* WINDUP-235 Decompile single class (ozizka, 14:15:25)
* WINDUP-272 Refactor GraphContextFactoryImpl and GraphContextImpl
(ozizka, 14:15:25)
* (ozizka, 14:15:25)
* Windup Status (lincolnthree, 14:17:20)
* We're actually looking really good! we've basically captured all of
the functionality from windup 1 at this point and met our promise to
deliver a more poiwerful rule syntax (we'll actually deliver this in
3 ways afaik) (lincolnthree, 14:19:47)
* as far as internal tech preview goes, I think we're pretty much
there, now that effort is being reported in the reports
(lincolnthree, 14:20:52)
* Next steps (lincolnthree, 14:23:19)
* ozizka's plans: (ozizka, 14:23:26)
* Will update the tutorial, and example ruleset. (Sande asked for it)
(ozizka, 14:23:26)
* Then according to priority. Perhaps the decompilation pruning.
(ozizka, 14:23:26)
* LINK: https://github.com/forge/core/blob/master/dist/pom.xml#L71
(lincolnthree, 14:32:35)
* LINK: https://github.com/forge/core/blob/master/dist/pom.xml#L33
(lincolnthree, 14:33:06)
* LINK: https://issues.jboss.org/browse/WINDUP-88 (ozizka, 14:33:43)
Meeting ended at 14:45:47 UTC.
--
Lincoln Baxter, III
http://ocpsoft.org
"Simpler is better."
9 years, 7 months
Windup terminology question
by Sande Gilda
Ondra and I have created a glossary on the Windup Wiki to document
terminology that is used by Windup. The goal is to finalize the correct
terms we want to use in Windup and avoid confusion down the road.
One of the open issues is whether to describe a "ruleset" (formerly
referred to as "rulebase") as a "Forge add-on" or as a "Windup plug-in".
Does anyone have a preference?
The glossary page is located here:
https://github.com/windup/windup/wiki/Glossary. As you come across other
Windup terms, feel free to update the page directly or email/ping me so
that we include them on that page.
Robb, I included some terms from your presentation and took an educated
guess at the meanings. Please let me know if they are incorrect.
Thanks,
Sande
9 years, 7 months
Double building?
by Ondrej Zizka
Hi (Lincoln),
could you please describe the issue with forge resolver more closely?
I think I was hit by that few times - the app is behaving differently
when I build only one module, then run test with other, and in whole run.
What's the cause and what exactly does it cause - what should I count with?
On unrelated note - I changed few things, rebuilt few times, and now
instead of
HintsClassificationsTest.testIterationVariableResolving:162
expected:<3> but was:<0>
I have
HintsClassificationsTest.testIterationVariableResolving:162
expected:<3> but was:<4>
Which is good - the whole run sticks with the same graph :)
But a bit weirder - having more nodes than the test expects...
More tomorrow.
Ondra
9 years, 7 months
Windup Meeting Minutes - 2014-09-17
by Lincoln Baxter, III
Minutes:
http://transcripts.jboss.org/meeting/irc.freenode.org/windup/2014/windup....
Minutes (text):
http://transcripts.jboss.org/meeting/irc.freenode.org/windup/2014/windup....
Log:
http://transcripts.jboss.org/meeting/irc.freenode.org/windup/2014/windup....
Meeting summary
---------------
* Agenda (lincolnthree, 14:32:05)
* Status Reports (lincolnthree, 14:36:00)
* I've been handling JBDS release stuff, as well as meeting with
Pete/BU to discuss planning and trying to get good requirements out
of them (lincolnthree, 14:36:57)
* I've also worked on refactoring and merging for WINDUP-253 and
WINDUP-222 (lincolnthree, 14:37:13)
* I'll be continuing to work on WINDUP-222 as well as more meetings
(yayyyy!) (lincolnthree, 14:37:32)
* Status: I finished the refactoring and now working on single class
decompilation. It's almost working. It could be ready tomorrow.
(ozizka, 14:37:54)
* I will update the tutorial and example rule (quickstart) (ozizka,
14:38:50)
* It was shown that there was really one commit that got lost from the
xmls. With Licoln we got it together and PR is there (mbriskar,
14:40:41)
* jsightler WINDUP-271 - BUG - addTypeToElement does not check for
duplicate values - This was hurting performance and causing some
items to be duplicated in reports (lincolnthree, 14:41:35)
* jsightler WINDUP-245 - Catalog properties file contents, and report
on all properties files in the overview report (lincolnthree,
14:41:35)
* jsightler WINDUP-243 - Include a hash of all archives in the report
(lincolnthree, 14:41:35)
* jsightler WINDUP-242 - Import some hibernate metadata (from mapping
files and cfg files) into the graph (lincolnthree, 14:41:35)
* jsightler The coming week I will focus on missing gaps for the
upcoming preview/beta. This includes making sure that
classifications/hints are displaying properly in the reports, and
making sure that estimated effort levels appear in the reports as
well. (lincolnthree, 14:41:45)
* Release Status (lincolnthree, 14:41:53)
* WINDUP-280: we have 3 open sub-tasks, one of which is a business
task and we are helping them with it. (lincolnthree, 14:43:14)
* I think we are looking good/on schedule to release as planned.
(lincolnthree, 14:43:28)
* We will be doing some planning with BU to do better requirements for
future releases. (lincolnthree, 14:44:10)
Meeting ended at 15:01:43 UTC.
--
Lincoln Baxter, III
http://ocpsoft.org
"Simpler is better."
9 years, 7 months
Performance - Forge scanning through Weld
by Ondrej Zizka
It seems that Forge spends considerable time (around 10 seconds on my
laptop) scanning the classes for annotated types;
Somehow related to
org.jboss.forge.furnace.container.cdi.impl.ContainerServiceExtension.processRemoteServiceTypes()
This repeats for each test in which Forge is involved.
Could this scanning be sped up using jandex? If we have 40 tests, then
it would buy us over 5 minutes.
Ondra
9 years, 7 months
Asciidoc - lists
by Ondrej Zizka
Hi Sande,
is there any way to prevent asciidoc put <p> to each <li>? Makes the
list tall and breaks the whole page layout.
Thanks,
Ondra
9 years, 7 months
Fwd: Re: GitHub problem rendering AsciiDoc lists
by Sande Gilda
This is interesting. GitHub uses Asciidoctor to render the asciidoc files!
Ondra, where did you report this issue? If you didn't report it to
Asciidoctor, I will.
-------- Original Message --------
Subject: Re: GitHub problem rendering AsciiDoc lists
Date: Thu, 11 Sep 2014 10:08:00 -0700
From: Jess Hosman (GitHub Staff) <support(a)github.com>
To: Sande Gilda <sgilda(a)redhat.com>
Hi Sande,
Thanks for your patience. After some digging, it looks like we use the third party tool Asciidoctor [1] for the conversion, which is where the extra `<p>` tags are coming from. I'd recommend opening an issue in the Asciidoctor repo about this. Once changes are made there, we should be able to pull them into GitHub.
Cheers,
Jess
[1]: https://github.com/asciidoctor/asciidoctor
9 years, 7 months
Doubled matching
by Ondrej Zizka
(23:56:23) ozizka-FN: Regarding java matching:
2014-09-10 23:58:11 INFO
org.jboss.windup.rules.java.TestJavaClassTestRuleProvider$2 perform
First rule matched:
rules/app/java/src/test/java/org/jboss/windup/rules/java/HintsClassificationsTest.java
2014-09-10 23:58:11 INFO
org.jboss.windup.rules.java.TestJavaClassTestRuleProvider$2 perform
First rule matched:
rules/app/java/src/test/java/org/jboss/windup/rules/java/JavaClassTest.java
2014-09-10 23:58:11 INFO
org.jboss.windup.rules.java.TestJavaClassTestRuleProvider$2 perform
First rule matched:
rules/app/java/src/test/java/org/jboss/windup/rules/java/HintsClassificationsTest.java
2014-09-10 23:58:11 INFO
org.jboss.windup.rules.java.TestJavaClassTestRuleProvider$2 perform
First rule matched:
rules/app/java/src/test/java/org/jboss/windup/rules/java/JavaClassTest.java
(23:56:41) ozizka-FN: This is a result of:
(23:56:50) ozizka-FN:
JavaClass.references("org.jboss.forge.furnace.*").inFile(".*").at(TypeReferenceLocation.IMPORT)
(23:59:22) ozizka-FN: Any idea why that is doubled?
Ondra
9 years, 7 months
