[ESS] ESS 15.09-beta: please try it before release!
v|ncent@gou|et @end|ng |rom me@com
Fri Sep 18 18:14:00 CEST 2015
I downloaded the zip file again using a cleaner approach and everything compiled normally.
*Sigh* Sorry for the noise!
> Le 2015-09-18 à 12:05, Lionel Henry <lionel.hry using gmail.com> a écrit :
> Hello Vincent,
> Does it work if you compile outside of your scripts?
> I cannot reproduce the problem with `make` (same version
> as yours, also on last OSX).
> What Emacs version are you using? I have 24.5.1.
>> On 18 sept. 2015, at 17:42, Vincent Goulet <vincent.goulet using me.com> wrote:
>> Building 15.09-beta fails for me with my usual scripts. It appears to be at the byte compiling of the 'lisp' directory stage. The last few lines of output from 'make' are
>> Source file `/Users/vincent/Emacs-modified/macos/ess-15.09-beta/lisp/ess-toolbar.el' newer than byte-compiled file
>> [ess-site:] before (ess-r-s-versions-creation) ...
>> [ess-site:] before creating ess-versions-* ...
>> [ess-site:] (let ... before (ess-r-versions-create) ...
>> [ess-site:] after ess-versions-created ...
>> [ess-site:] after ess-check-R-prog... ...
>> [ess-site:] *very* end ...
>> Source file `/Users/vincent/Emacs-modified/macos/ess-15.09-beta/lisp/ess-custom.el' newer than byte-compiled file
>> Fatal error (6)make: *** [ess.elc] Abort trap: 6
>> make: *** [install] Error 2
>> make: *** [ess] Error 2
>> On a side note, I noticed the following warnings while scrolling the messages. They're probably harmless and may even have been there in previous versions:
>> In toplevel form:
>> ess-custom.el:248:1:Warning: defcustom for `ess-can-eval-in-background' fails
>> to specify containing group
>> ess-custom.el:391:1:Warning: defcustom for `ess-tab-complete-in-script' fails
>> to specify containing group
>> ess-custom.el:1620:1:Warning: defcustom for `ess-getwd-command' fails to
>> specify containing group
>> ess-custom.el:1623:1:Warning: defcustom for `ess-setwd-command' fails to
>> specify containing group
>> This is all on OS X Yosemite with GNU make 3.81 from XCode 7.0.
>> Hope this helps.
>>> Le 2015-09-18 à 07:11, Martin Maechler <maechler using stat.math.ethz.ch> a écrit :
>>> Dear ESS users,
>>> The ESS core team plans to release ESS 15.09 in a few days (probably Sept.23).
>>> Those (daring ones) of you who work from the github or MELPA
>>> development version of ESS are indeed beta-testing ESS 15.09
>>> Others are kindly asked to try it, possibly most easily by getting one of the packaged beta versions (showing 'https' below; 'http' also works) :
>>> and giving feedback either here or by opening "github issue" at
>>> The biggest changes are in indentation behavior which has become
>>> much more consistent and still more flexible,
>>> notably thanks to our new team member
>>> Lionel Henry (thank you, Lionel!)
>>> Many other improvements / bug fixes have been incorporated since
>>> the spring release 15.03. See below for the current documented
>>> list of new features for 15.09.
>>> For the ESS Core Team:
>>> Martin Maechler, ETH Zurich
>>> Changes and New Features in 15.09:
>>> * ESS[R]: The indentation logic has been refactored. It should be
>>> faster, more consistent and more flexible. A test framework has
>>> been set up in order to prevent any regression on future code
>>> change. Indentation offsets have been renamed as well and now all
>>> start with the prefix 'ess-offset-' and other control variables
>>> start with 'ess-indent-'. See 'ess-style-alist' for detailed
>>> description of the new indentation system.
>>> The most visible user change in the indentation setup is that
>>> customisation is now centered around 3 settings:
>>> 'ess-offset-arguments', 'ess-offset-arguments-newline' and
>>> 'ess-offset-block'. The first two settings control how arguments
>>> are indented inside function calls and index brackets, while the
>>> latter controls the indentation of statement blocks. They can be
>>> set to 'prev-call' (indentation relative to innermost function
>>> call), 'prev-line' (indentation relative to last line indentation
>>> as in RStudio) or 'open-delim' (indentation relative to opening
>>> delimiter). See the documentation strings of those variables for
>>> more information. Indentation of delimiters refer to these three
>>> values depending on the context in order to offer a consistent
>>> indentation. 'ess-arg-function-offset' has been replaced by
>>> 'ess-indent-prev-call-lhs' and has been generalised to
>>> assignements. This setting works with both statement blocks and
>>> expressions. It now only takes effect for offsets set to
>>> 'prev-call' in order to produce a consistent indentation.
>>> * ESS[R]: 'ess-brace-imaginary-offset', 'ess-expression-offset' and
>>> all delimiter offsets are deprecated. The settings for indentation
>>> of continued statements have been replaced by
>>> 'ess-offset-continuations'. It can be set to either 'cascade' or
>>> 'straight' (the default).
>>> * ESS[R]: A new RStudio style is provided to mimic as closely as
>>> possible R files indented via RStudio. To reproduce the setup of
>>> some of the RStudio users, you will also need to set
>>> 'ess-offset-arguments' to 'prev-line'.
>>> * ESS[R]: Roxygen fields will now be indented on paragraph refilling
>>> in order to make the documentation more readable. You can also
>>> refill commented lines in the 'examples' field without squashing
>>> the surrounding code in the comments.
>>> * ESS[R]: ESS can now format your code. This is controlled through
>>> the settings 'ess-fill-calls' and 'ess-fill-continuations'. When
>>> activated, '(fill-paragraph)' will format your calls and your
>>> formulas/continuations and make sure they don't go past
>>> 'fill-column'. Refilling with a prefix produces an alternative
>>> formatting with one element per line. Repeated refills cycle
>>> through different styles (see the docstrings for more details). By
>>> default, the refilled region blinks. Set 'ess-blink-filling' to
>>> nil to prevent this.
>>> * ESS[R]: Fix occasional missing error location fontification in
>>> inferior buffers.
>>> ESS-help using r-project.org mailing list
>> ESS-help using r-project.org mailing list
More information about the ESS-help