[ 
https://issues.apache.org/jira/browse/NETBEANS-54?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Jan Lahoda updated NETBEANS-54:
-------------------------------
    Attachment: categorized-1.zip
                convert-headers-1.zip

I was trying to look at a tool that would help us categorize licenses in the 
main sources. The tool is attached in "convert-headers-1.zip".

The (current) output is in "categorized-1.zip". It contains 4 kinds of sources:
-do-not-have-cddl0: list of files that do not contain string "CDDL"
-have-cddl-not-recognized-filetype0: list of files that contain string "CDDL", 
but the tool cannot (yet) detect the header
-lic*: a detected variant of the licence header (normalized - leading comment 
marks, leading and trailing whitespaces removed, paragraphs joined on one line, 
copyright years replaced with "<YEARS>"), with a list of files that contain 
this specific license header
-par*: the header above is split by paragraphs, and each unique paragraph gets 
a "par" file. Seems it may be easier to review the paragraphs rather than the 
full license headers (which may contain many duplicated text, with just small 
differences among variants).

Currently, the do-not-have-cddl0 list has 14627 files (so we to do some more 
work to categorize that), the have-cddl-not-recognized-filetype0 has 667 files, 
there are 155 license header variants, and 132 variants of the license header 
paragraphs.

Any feedback/opinions on this? Does this look like a viable direction?


> Clarify the status of donated files from ApacheNetBeansDonation1.zip
> --------------------------------------------------------------------
>
>                 Key: NETBEANS-54
>                 URL: https://issues.apache.org/jira/browse/NETBEANS-54
>             Project: NetBeans
>          Issue Type: Bug
>            Reporter: Bertrand Delacretaz
>         Attachments: categorized-1.zip, convert-headers-1.zip, 
> JavaApplication9.java
>
>
> Reviewing the software grant [1] provided with the first NetBeans code 
> donation, my understanding is that not all files found in that zip are 
> donated.
> The grant indicates that the contents of that file [2] ("Code File") are 
> donated "excluding any third-party and separately licensed material contained 
> within the Code File". 
> I don't think we can consider that code ready to release as is, due to the 
> uncertainty that this creates w.r.t to which files are donated and which are 
> not.
> The zip archive contains about 45'000 files.
> I see the following options:
> a) Review all files and include only those which are not "third-party and 
> separately licensed material". How such files are identified needs to be 
> defined and documented, either in scripts under version control or as a set 
> of rules. The review process must be documented and traceable, so that we can 
> go back to it if problems are found in the future.
> b) Get an updated grant which does not mention such exclusions.
> If someone sees other options please discuss here or on the NetBeans dev 
> list, referencing this ticket.
> [1] in svn under private/documents/grants/oracle-america-netbeans.pdf , ASF 
> members have access
> [2] under http://bits.netbeans.org/download/apache-donation/  - 
> SHA256(ApacheNetBeansDonation1.zip)=
> 7f2ca0f61953a190613c9a0fbcc1b034084b04a4d55d23c02cefffc354e7c24a



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

Reply via email to