You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@uima.apache.org by "Marshall Schor (JIRA)" <ui...@incubator.apache.org> on 2008/09/02 19:31:44 UTC
[jira] Commented: (UIMA-1163) set svn:eol-style to native for
source files in UIMA project
[ https://issues.apache.org/jira/browse/UIMA-1163?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12627758#action_12627758 ]
Marshall Schor commented on UIMA-1163:
--------------------------------------
Among the kinds of files in our svn repository are scripts. Some of these are intended for windows (they end in .cmd or .bat). Others are intended to run on linux or other unix (they end in .sh).
How should the svn:eol-style be set for these?
Here are some choices:
1) no setting at all
2) "native"
3) CRLF for the Windows files, LF for the *nix files
Do we have a goal of having the line endings for the scripts correspond to the Operating System category (windows or *nix)?
It may be true that most (all?) target OSs don't care about this. It might be "safer" though to have these match.
If we want them to match, then 2) above fails - because if we "build" the distribution on a Windows system, it would have CRLF for all files, and if we build on a *nix system, it would have LF for all files.
1) and 3) would work for these, provided if choosing 1) we insured the line endings were the right "kind" for the file. This seems easy to break, though - I think most of our "mixed line endings" situations were caused by copy /paste actions in the past (perhaps copying in a license header, etc.).
So, *if* we have a goal of having the line endings for scripts correspond to the OS category, I think 3) is the best.
Other opinions?
If no dissent, I will try and achieve this as I go thru fixing up line endings (but it's not hi priority work)
> set svn:eol-style to native for source files in UIMA project
> ------------------------------------------------------------
>
> Key: UIMA-1163
> URL: https://issues.apache.org/jira/browse/UIMA-1163
> Project: UIMA
> Issue Type: Improvement
> Components: Core Java Framework
> Affects Versions: 2.2.2
> Reporter: Marshall Schor
> Assignee: Marshall Schor
> Priority: Minor
>
> set the svn:eol-style to native for java source files in the UIMA project. (Other sandbox projects may also want to do similar things, and the uimaCPP project, and the site project).
> Don't set this for scripts - the ones that are intended to run on windows probably can have either line ending style, but I think the ones that run on the various linuxes need the unix style. This is so that extracts for builds done on a windows machine can properly build a linux distribution.
> Don't set this for test data - some UTF-16 or UTF-8 test data may have things intentionally present that might look like line endings, but are not, and should not be disturbed by checkout.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
Re: [jira] Commented: (UIMA-1163) set svn:eol-style to native for source files in UIMA project
Posted by Burn Lewis <bu...@gmail.com>.
+1 for #3 since it means we only have to set the appropriate eol-style when
adding new scripts while #1 means we in addition have to check on every
update that the line-ends are correct.
On Tue, Sep 2, 2008 at 1:31 PM, Marshall Schor (JIRA) <
uima-dev@incubator.apache.org> wrote:
>
> [
> https://issues.apache.org/jira/browse/UIMA-1163?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12627758#action_12627758]
>
> Marshall Schor commented on UIMA-1163:
> --------------------------------------
>
> Among the kinds of files in our svn repository are scripts. Some of these
> are intended for windows (they end in .cmd or .bat). Others are intended to
> run on linux or other unix (they end in .sh).
>
> How should the svn:eol-style be set for these?
>
> Here are some choices:
>
> 1) no setting at all
> 2) "native"
> 3) CRLF for the Windows files, LF for the *nix files
>
> Do we have a goal of having the line endings for the scripts correspond to
> the Operating System category (windows or *nix)?
>
> It may be true that most (all?) target OSs don't care about this. It might
> be "safer" though to have these match.
>
> If we want them to match, then 2) above fails - because if we "build" the
> distribution on a Windows system, it would have CRLF for all files, and if
> we build on a *nix system, it would have LF for all files.
>
> 1) and 3) would work for these, provided if choosing 1) we insured the line
> endings were the right "kind" for the file. This seems easy to break,
> though - I think most of our "mixed line endings" situations were caused by
> copy /paste actions in the past (perhaps copying in a license header, etc.).
>
> So, *if* we have a goal of having the line endings for scripts correspond
> to the OS category, I think 3) is the best.
>
> Other opinions?
>
> If no dissent, I will try and achieve this as I go thru fixing up line
> endings (but it's not hi priority work)
>
>
>
>
>
> > set svn:eol-style to native for source files in UIMA project
> > ------------------------------------------------------------
> >
> > Key: UIMA-1163
> > URL: https://issues.apache.org/jira/browse/UIMA-1163
> > Project: UIMA
> > Issue Type: Improvement
> > Components: Core Java Framework
> > Affects Versions: 2.2.2
> > Reporter: Marshall Schor
> > Assignee: Marshall Schor
> > Priority: Minor
> >
> > set the svn:eol-style to native for java source files in the UIMA
> project. (Other sandbox projects may also want to do similar things, and
> the uimaCPP project, and the site project).
> > Don't set this for scripts - the ones that are intended to run on windows
> probably can have either line ending style, but I think the ones that run on
> the various linuxes need the unix style. This is so that extracts for
> builds done on a windows machine can properly build a linux distribution.
> > Don't set this for test data - some UTF-16 or UTF-8 test data may have
> things intentionally present that might look like line endings, but are not,
> and should not be disturbed by checkout.
>
> --
> This message is automatically generated by JIRA.
> -
> You can reply to this email to add a comment to the issue online.
>
>