Patch : editing "Posted" time of transactions.

Geert Janssens geert.gnucash at kobaltwit.be
Fri Jan 29 06:47:08 EST 2016


On Friday 29 January 2016 11:44:53 gLETTERyYuMEANSj LETTERyOt wrote:
> Dear John,
> 
> Thank you very much for the help your comments provide to me !
> 
> I accept without additionnal question all of your comments that are
> not quoted below, and will work on them.
> 
> 2016-01-28 19:54 UTC+01:00, John Ralls <jralls at ceridwen.us>:
> > Don't remove function description comments; do make them Doxygen
> > markup.
> I did that because some function description comments were removed
> between gnucash-2.6.5 and end of 2015. Should I also revert those
> comments?

Can you provide some links to commits that remove comments ? There may have been other 
reasons this was done.

> >> * if TZ is not defined, gnucash crashes immediately.
> > 
> > The crashes alone make this patch incomplete and unacceptable
> > as-written.
> Can someone help me, by showing the right way to recover the time
> difference due to the timezone settings at a given UTC date?
> 
> > But we still haven't decided whether to drop times altogether from
> > date_posted, and this change, even if otherwise perfect, can't merge
> > until we do.
> 
> Should I add changes to
> gnucash-2.4.12/src/gnome-utils/glade/preferences.glade to let the
> user decide, at runtime, if he wants to see times or no ?

Well definitely not to gnucash-2.4.12. That's a version of gnucash that's no longer supported. 
Are you working from an old tarball for your development ?

Secondly, the glade files have been converted to gtkbuilder format since 2.6.x, so the file would 
now reside in src/gnome-utils/gtkbuilder

And lastly, if it's decided to drop time altogether from gnucash, it won't be stored in the data 
file anymore either, so a preference setting wouldn't make sense anymore. If I remember 
correctly, John opened two contradicting feature requests on uservoice to gather user opinions 
on which direction to take on this. You can raise your voice there and motivate why you believe 
gnucash should keep time stamps [1]. Personally, I don't see much use for it as I use gnucash, 
but I'm not the only user :)

> > BTW your
> > https://github.com/gijut/gnucash/commit/885314788855f74d920e0637ea41
> > b3289f5b55b4 incorrectly copies a GCC bug:
> > https://sourceware.org/bugzilla/show_bug.cgi?id=15366 *and*
> > overrides a #define that belongs to the libc/libc++ headers.
> > (The bug was the result of the GCC folks following an effort by the
> > ISO C committee to tell the ISO C++ committee what to do. The ISO
> > C++ committee ignored them, but the GCC folks didn't. The ISO C
> > committee fixed their overstep in C11. See
> > http://en.cppreference.com/w/Talk:cpp/types/integer.
> > 
> > Please remove that commit.
> 
> Is there a better way than reverting the source, then typing `git
> commit -a` ? I am new to git, I mainly use cvs.

There is.

You can use interactive rebasing to prune out the bad commit
git rebase -i <base branch>

Where <base branch> can be origin/master for example. This works best if your branch is a 
fast-forward from the base branch (that is ALL commits from the base branch are also included 
in your branch).

The man page for git-rebase has more details.

> I wonder if configure
> should or not bail out in front of gcc 4.8.4 (Gentoo 4.8.4 p1.0,
> pie-0.6.1).
> 

I'm afraid I don't know that.

> Best Regards, Daniel
> 
> NB: Thanks Geert for clarifying gnucash --help and helping me in this
> thread.

You're welcome :)

Geert

[1] Refer to these links:
http://gnucash.uservoice.com/forums/101223-feature-request/suggestions/6194125-fix-the-moving-entry-date-problem-by-getting-rid-o
http://gnucash.uservoice.com/forums/101223-feature-request/suggestions/6194177-add-an-optional-entry-time-field-to-the


More information about the gnucash-devel mailing list