Getting LibreOffice Writer 5.0.6 bibisected regression free against LibreOffice 4.4.0 (was: LibreOffice ESC call, ...)

[Bjoern Michaelsen]

Hi,

On Wed, Mar 23, 2016 at 10:19:58PM +0000, Michael Meeks wrote:
> * QA stats
> * Bibisected bugs open: whiteboard 'bibisected'
>   + 399/1305 404/1297 406/1292 412/1284 408/1277 403/1260 398/1248
>     + http://tdf.io/bibibugs
>  
> * all bugs tagged with 'regression'
> 	+ 773(-12) bugs open of 4798(+10) total 28(-2) high prio.
>  
>         * ~Component   count net * high severity regressions
>                   Calc - 11 (-2)
>                 Writer -  9 (+0)
>            LibreOffice -  4 (-1)
>                   Base -  2 (+0)
>             Extensions -  1 (+1)
>                Impress -  1 (+0)
>                 + http://bit.ly/1HWHb3E
>  
>         * ~Component   count net * all regressions
>                 Writer - 242 (-1)
>                   Calc - 166 (-7)
>            LibreOffice -  56 (-3)
>                Crashes -  51 (-3)
>                Impress -  46 (+0)
>                     UI -  41 (-1)
>         graphics stack -  36 (+1)
>                   Base -  34 (+0)
>                Borders -  32 (+0)
>                   Draw -  26 (-1)
>    filters and storage -  22 (+0)
>                  Chart -  14 (+0)
>                  BASIC -  14 (+1)
> Printing and PDF export-  10 (-1)
>         Formula Editor -   7 (+0)
>             Linguistic -   4 (+0)
>             Extensions -   3 (+0)
>           Installation -   2 (+0)
>                    sdk -   1 (+0)
>              framework -   1 (+0)
>                 + http://bit.ly/1BUdI8i

So, we have a few more week until we close the 5.0 series with the 5.0.6
release. I would think it would be awesome for us to aim for closing all
triaged, confirmed and bibisected regressions that were introduced betweem
LibreOffice 4.4 and 5.0 by then. As writer has the highest absolute regression
counts, I would suggest to start with that goal there for closing the 5.0
series. Looking at:

 https://bugs.documentfoundation.org/report.cgi?x_axis_field=component&y_axis_field=bug_severity&z_axis_field=&no_redirect=1&query_format=report-table&short_desc_type=allwordssubstr&short_desc=&product=LibreOffice&bug_status=NEW&resolution=---&longdesc_type=allwordssubstr&longdesc=&bug_file_loc_type=allwordssubstr&bug_file_loc=&status_whiteboard_type=allwordssubstr&status_whiteboard=&keywords_type=allwords&keywords=bibisected%2C+&bug_id=&bug_id_type=anyexact&emailtype1=substring&email1=&emailtype2=substring&email2=&emailtype3=substring&email3=&chfieldvalue=&chfieldfrom=&chfieldto=Now&j_top=AND&f1=version&o1=regexp&v1=^5.0&format=table&action=wrap

LibreOffice Writer has only 8 unresolved, fully confirmed, triaged and
bibisected regressions today. So getting LibreOffice Writer "bibisected
regression clean" for 5.0 should be doable. By priority IMHO the remaining
unresolved issues are:

- tdf#94063: critial, all platforms
- tdf#94570, tdf#94449, tdf#97103: non-critical, all platforms
- tdf#95836, tdf#94597, tdf#95194: all caused by same commit, Windows only
- tdf#96716: Windows only

I love TDF to be able to claim that LibreOffice Writer 5.0.6 has no bibisected
regression against LibreOffice Writer 4.4.0. Its a good story for users, it
shows that we got better by a long way from our earlier releases and -- its
also motivating for those doing tireless work on QA and bibisecting.

Best,

Bjoern