* maximization (also partially fixes strange Firefox behaviour)
* sane initial window placement
Firefox is stranger than most; it appears to maximize to double
X Screen width even on single head displays. Fixed (on multihead).
The function print_progress uses the pointer infile.
If fclose is called before print_progress then the build process
can fail to build the files:
doc/C/cde.dti/CDEDOC/dtsearch/CDEDOC.d99
doc/C/cde.dti/CDEDOC/dtsearch/dtsearch.ocf
doc/C/cde.dti/cde.oli
To fix the issue move the call fclose(infile) after print_progress()
Formatted text currently gets corrupted by DtTerm when copied into its
history buffer.
As soon as a line of text is copied into the history, the text of each
segment of formatted text is altered so that it contains the text from
the start of the line. For example:
echo -e '\e[1mbold\e[m \e[4munderlined\e[m'
When the text has scrolled off the screen, scrolling back to it reveals:
boldbbold under
This is fixed by adding the calls to _DtTermPrimBufferGetCharacterPointer
that were missing which would update the buffer pointer when inserting
text into the history buffer.
When using a non C locale such as fr_FR.ISO8859-1, the CUPS messages
returned by lpq are translated. However, the parser in ParseJobs.C can only
handle messages in English. As a result, dtprintinfo exits immediately with
an error message. To fix that problem we force a LANG=C environment when calling
lpq so that messages will be in English and will be processed correctly by dtprintinfo
parser.
* insserv requires a shebang line
* the rc script can hang if a daemon is started in the foreground
* "pgrep dtlogin" apparently can match the init script, if it is started
with "service dtlogin start" or similar.
* reduce code duplication
