incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Julian Hyde <jul...@hydromatic.net>
Subject Issues raised during Optiq 0.9.0 release vote
Date Mon, 25 Aug 2014 18:17:41 GMT
A few issues were raised during the vote for Apache Optiq 0.9.0 (incubating) [0]. I started
a separate thread to avoid muddying the vote. I address them below.

On Aug 23, 2014, at 10:36 AM, jan i <jani@apache.org> wrote:

> I am a bit concerned about the xml files. suppressions.xml references:
> <!DOCTYPE suppressions PUBLIC
>        "-//Puppy Crawl//DTD Suppressions 1.1//EN"
>        "http://www.puppycrawl.com/dtds/suppressions_1_1.dtd">
> 
> and checker.xml does not have any dtd reference. These are just examples,
> it continues. Its not a problem, I would just have expected the project to
> supply their own .dtd files and use them for all xml files.

This DTD is used as a URI that represents the particular version of the XML grammar and is
not referenced at run time. (As evidenced by the fact that the project can be built using
“mvn -o”, i.e. offline.) If we copied the DTD it would lose its effect of specifying the
particular version of the XML grammar. Other projects (e.g. Hive) use the same DOCTYPE header
for their checkstyle suppressions file.

> Sources look ok from a license perspective (it is a bit confusing though
> that src is nearly empty).

Not much in root src because it is the parent module (i.e. serves to combine several child
modules); most of the code is in child modules, principally core.

$ for i in $(find . -name src); do wc $(find $i -type f) | awk -v i=$i '$4 == "total" {print
NR, $1, $2, $3, i}'; done
Files   Lines   Words     Chars Directory
=====  ====== ======= ========= ================
   34   6,546  21,753   195,415 ./avatica/src
  964 263,367 927,737 8,710,388 ./core/src
   20   3,077  12,147   127,812 ./mongodb/src
   15   1,979   7,719    70,028 ./plus/src
   15   1,480   5,061    51,791 ./spark/src
   20   2,514   8,548    78,994 ./splunk/src
    5     733   2,770    35,749 ./src
    3     258     948     7,639 ./ubenchmark/src


On Aug 23, 2014, at 3:14 PM, John D. Ament <john.d.ament@gmail.com> wrote:

> I'm not sure that I'm comfortable with the fact that all of your source
> code is under net.hydromatic.optiq rather than org.apache.optiq.

There is an issue for this: https://issues.apache.org/jira/browse/OPTIQ-296. Code re-org is
disruptive, both for this project and for dependent projects, and we chose to get a release
out first.

> Your
> apache headers also look a bit weird.  Every other apache licensed that
> I've seen looks like [1], rather than your mix of single line and multi
> line headers.  But I think that's just a preference thing.

I think we should comply with the norm. I’ve logged https://issues.apache.org/jira/browse/OPTIQ-385.

On Aug 23, 2014, at 6:48 PM, Justin Mclean <justin@classsoftware.com> wrote:

> You may want to consider changing for the next release:
> - change name format of the released files [1]

We actually tried this release, but backed off when it seemed that we were fighting maven.

I’ve logged https://issues.apache.org/jira/browse/OPTIQ-384.

> - place files in correct place [2]

I plan to do this as I roll out the release to mirrors. I didn’t know if it was even possible
to populate “dist” until you have an approved release.

> - given there' nothing extra in LICENCE then there's probably no need for "Please see
LICENSE for additional copyright and licensing information." in the NOTICE
> - NOTICE is missing "The Initial Developer of the Original Code, ..." which I assume
is the case as the package names are not apache ones


I’ve logged https://issues.apache.org/jira/browse/OPTIQ-386.

Julian

0. http://mail-archives.apache.org/mod_mbox/incubator-general/201408.mbox/%3CCD77DBD8-DBA6-435D-858F-688EF72C8418%40hydromatic.net%3E
1. http://incubator.apache.org/guides/releasemanagement.html#best-practice-unique-names
2. https://dist.apache.org/repos/dist/dev/incubator/twill/
Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message