[hardy] G-C-U FTBFS and Possible Merge

Bug #172717 reported by Kow
2
Affects Status Importance Assigned to Milestone
gnome-chemistry-utils (Ubuntu)
Fix Released
Undecided
Unassigned

Bug Description

Here is the scenario/problem:

MOM wants us to merge with debian 0.8.4. The problem is G-C-U 0.8.4 will FTBFS against libgoffice-0.5.1 or newer (our only version in hardy.) G-C-U devs have released 0.9 (development) but this will still not build against 0.5.1 or newer. Why?

G-C-U changelog portion:
----------------------
2007-09-28 Jean Brefort <email address hidden>

 * configure.in: added goffice/math directory.
 * goffice/Makefile.am: ditto.
 * goffice/data/go-data-simple.c: moved maths to goffice/math.
 * goffice/data/go-data.c: ditto.
 * goffice/goffice.c: ditto.
 * goffice/graph/gog-axis-line.c: ditto.
 * goffice/graph/gog-axis.c: ditto.
 * goffice/graph/gog-chart-map.c: ditto.
 * goffice/graph/gog-chart.c: ditto.
 * goffice/graph/gog-data-set.c: ditto.
 * goffice/graph/gog-error-bar.c: ditto.
 * goffice/graph/gog-grid-line.c: ditto.
 * goffice/graph/gog-grid.c: ditto.
 * goffice/graph/gog-legend.c: ditto.
 * goffice/graph/gog-plot.c: ditto.
 * goffice/graph/gog-reg-curve.c: ditto.
 * goffice/graph/gog-renderer.c: ditto.
 * goffice/graph/gog-style.c: ditto.
 * goffice/gtk/go-graph-widget.c: ditto.
 * goffice/gtk/go-rotation-sel.c: ditto.
 * goffice/math/Makefile.am: ditto.
 * goffice/math/go-complex.c: copied code from gnumeric.
 * goffice/math/go-complex.h: ditto.
 * goffice/math/go-cspline.c: moved maths to goffice/math.
 * goffice/math/go-cspline.h: ditto.
 * goffice/math/go-fft.c: copied code from gnumeric.
 * goffice/math/go-fft.h: ditto.
 * goffice/math/go-math.c: moved maths to goffice/math.
 * goffice/math/go-math.h: ditto.
 * goffice/math/go-rangefunc.c: ditto.
 * goffice/math/go-rangefunc.h: ditto.
 * goffice/math/go-regression.c: ditto.
 * goffice/math/go-regression.h: ditto.
 * goffice/utils/Makefile.am:
 * goffice/utils/go-cspline.c: ditto.
 * goffice/utils/go-cspline.h: ditto.
 * goffice/utils/go-format.c: ditto.
 * goffice/utils/go-geometry.c: ditto.
 * goffice/utils/go-line.c: ditto.
 * goffice/utils/go-marker.c: ditto.
 * goffice/utils/go-math.c: ditto.
 * goffice/utils/go-math.h: ditto.
 * goffice/utils/go-path.c: ditto.
 * goffice/utils/go-rangefunc.c: ditto.
 * goffice/utils/go-rangefunc.h: ditto.
 * goffice/utils/go-regression.c: ditto.
 * goffice/utils/go-regression.h: ditto.
 * plugins/plot_barcol/gog-1.5d.c: ditto.
 * plugins/plot_barcol/gog-barcol.c: ditto.
 * plugins/plot_barcol/gog-dropbar.c: ditto.
 * plugins/plot_barcol/gog-line.c: ditto.
 * plugins/plot_boxes/gog-boxplot.c: ditto.
 * plugins/plot_boxes/gog-histogram.c: ditto.
 * plugins/plot_pie/gog-pie.c: ditto.
 * plugins/plot_radar/gog-radar.c: ditto.
 * plugins/plot_surface/gog-surface.c: ditto.
 * plugins/plot_surface/xl-surface.c: ditto.
 * plugins/plot_xy/gog-xy.c: ditto.
 * plugins/reg_linear/gog-exp-reg.c: ditto.
 * plugins/reg_linear/gog-lin-reg.c: ditto.
 * plugins/reg_linear/gog-lin-reg.h: ditto.
 * plugins/reg_linear/gog-log-reg.c: ditto.
 * plugins/reg_linear/gog-polynom-reg.c: ditto.
 * plugins/reg_linear/gog-power-reg.c: ditto.
 * plugins/reg_logfit/gog-logfit.c: ditto.
 * plugins/reg_logfit/gog-logfit.h: ditto.
 * plugins/smoothing/gog-exp-smooth.c: ditto.
 * plugins/smoothing/gog-moving-avg.c: ditto.
----------------------

So at this point G-C-U will FTBFS using the current version (0.6.3) or the upstream debian version. Debian has been bypassing this issue so far by sticking with libgoffice-0.4. I have not verified but I expect that the current binary will not run properly in hardy (if at all). I found this problem while attempting to create a merge candidate (0.8.4-3ubuntu1). Here is what MUST be done:
(1) Fix G-C-U so it will build from source and run in hardy.

How?

1) I recommend someone notifies G-C-U devs to this problem and have them fix their code to build against libgoffice >= 0.5.1.
2) Pending the results of (1) if G-C-U devs release a new version we can directly import this into ubuntu because debian will likely stick with libgoffice 0.4 for a while. Otherwise we will need to heavily patch the 0.8.4 source from debian to build against our current (and only) libgoffice which is 0.5.1.

Or we can remove G-C-U from the hardy repositories.

Revision history for this message
Kow (kow) wrote :
Revision history for this message
LaserJock (laserjock) wrote :

G-C-U should be built against libgoffice 0.4 . Debian has it and we should sync it for Hardy. Then we can merge G-C-U 0.8.4

G-C-U 0.9.x are unstable development releases and we should *not* put them into hardy, especially as it is an LTS release.

Revision history for this message
Jean Bréfort (jean-brefort-normalesup) wrote :

I confirm Jordan's comment. You need both libgoffice-0.4 and libgoffice-0.5 (will be 0.6 soon) for Hardy. The first is needed by gcu and abiword, and the second by gnumeric. The removal of gnome-print support in goffice-0.5 makes it impossible to use this latest version in both gcu and abiword.

Revision history for this message
Jean Bréfort (jean-brefort-normalesup) wrote :

One more point, gcu needs goffice with gnome support, goffice-gtk is not enough (I'll fix the requirement in 0.8.5).

Revision history for this message
Kow (kow) wrote :

Okay update this bug report when 0.4 makes it into the hardy repos and I'll test this merge candidate for g-c-u that I have.

Revision history for this message
LaserJock (laserjock) wrote :

Sorry, I forgot about this bug. I merged g-c-u last night now that goffice0.4 is in Hardy.

Changed in gnome-chemistry-utils:
status: New → Fix Released
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.