[ANN] Bugzilla upgraded to 4.4.10

[Robinson Tryon]

> On Sat, Sep 19, 2015 at 09:30:55AM -0700, julien2412 wrote:
>> I'm not sure and, perhaps it's not related, but it seems docx attachments
>> have mimetype "application/zip" instead of
>> "application/vnd.openxmlformats-officedocument.wordprocessingml.document"
>> (eg: https://bugs.documentfoundation.org/show_bug.cgi?id=94359). Even if
>> confirmed, not a blocker of course ! :-)

Because OOXML and ODF both use ZIP as a container for a bunch of XML,
I wonder if the type-detecting mechanisms aren't sufficiently nuanced
to detect and differentiate between ZIP and ZIP+(DOCX guts).

Here's a quick test using the 'file' command (v5.14)

qubit at loopbackoffice:~/libreoffice/bugs$ find -type f -print0 | xargs -0 file
./90736/TableCellBorderLineStyle.pptx:                 Microsoft
PowerPoint 2007+
./93865/fubar.odt:                                     OpenDocument Text
./93853/checkbox bug_pdf-test-export.pdf:              PDF document, version 1.4
./93853/.~lock.checkbox bug.odt#:                      ASCII text,
with no line terminators
./93853/checkbox bug.odt:                              OpenDocument Text
./93356/testsort.ods:                                  OpenDocument Spreadsheet
./92703/Character.Controls.rtf:                        Rich Text
Format data, version 1, ANSI
./91293/impress-file-saved-as-pptx.pptx:               Zip archive
data, at least v2.0 to extract
./91293/.~lock.impress-file-saved-as-pptx_try-2.pptx#: ASCII text,
with no line terminators
./91293/impress-file-saved-as-pptx_try-2.pptx:         Zip archive
data, at least v2.0 to extract
./93868/Lecture 1 African Historical Background.pptx:  Microsoft
PowerPoint 2007+
./93851/Untitled 1.ods:                                OpenDocument Spreadsheet

Looks like ODF files are recognized quite well on this system, but
it's struggling with OOXML. Is there a correlation between authoring
software and recognition?

On Sat, Sep 19, 2015 at 1:27 PM, David Tardon <dtardon at redhat.com> wrote:
> That's nothing new. The attachment format auto-detection has never really
> worked.

We did have much worse file-type detection when we were hosted at FDO,
but I believe that things have gotten significantly better.

David: Is there something we could do to help upstream utilities more
easily recognize OOXML files?  (both files that LibreOffice creates,
as well as ones from other sources)


Thanks,
--R

-- 
Robinson Tryon
QA Engineer - The Document Foundation
LibreOffice Community Outreach Herald
qubit at libreoffice.org
802-379-9482 | IRC: colonelqubit on Freenode