Bugs

171107:
Status: pending
Seriousness: benign.
Description:
As Perl has changed since the development of Csgrouper, some new warnings are now polluting the output in spite of “no warnings”.
Solution:
“use experimental” should get rid of some warnings related to smart match.  Other ones should be avoided by cleaning the code, one day…

171107:
Status: pending
Seriousness: annoying.
Description:
Deselecting a row in the main table doesn’t suffice to process the Y functions. This is due to the fact that Yfuncs don’t check wether a row is selected or not but rely on the content of the targeted sets. If a row pertains to a set Rythmic canon or Ensemble will fail when it is deselected.
Solution:
In order to recompute Rythmic canon or Ensemble without the deselected row, one has to change its set number too, for example, if row 10 is member of set 4 and this set is targeted by a Rythmic canon, then at present, the only way to obtain a valid computation after deselecting it, is to change its set attribution in the main table, to a set number that is not existing, e.g. 40 instead of 4. This bug should be corrected soon.

 

131020:
Status: pending
Seriousness: annoying.
Description:
The decision mechanism is not influenced by parameter “Comp. type” (self vs random) as it should. This will be fixed in the next release. It should be noted that this bug is touching a very fine part of Csgrouper’s process, things like wether to put a silence here or not, where to start and stop ensemble or r-canon subsections. Till now these decisions were taken according to static internal comparisons, in the future, they will be sensitive to the chosen comparison type. Users should not expect too obvious changes through this means since it is always combined with other internal static decision procedures, however using a random comparison type makes the score somewhat different each time the project is reloaded.

130830:

Status: corrected.
Seriousness: annoying.
Description:
Tk::Table->clear didn’t destroy checkboxes.
Solution:
Upgrade Tk.

130708:

Status:  corrected.
Seriousness: disabling.
Description:
Glissandi weren’t usable because of a bad comparison operator.
Solution:

replace in Music.pm, line ~671:

goto ENDXGLI unless ($ref->ocmp2 =~ /^($ref->ocmp3)$/); ## indio == inoto

by:

goto ENDXGLI unless ($ref->ocmp2 eq $ref->ocmp3); ## indio == inoto

130705:

Status:  corrected.
Seriousness: disabling.
Description:
Rythmic Canon and Ensemble weren’t usable in case of non selected rows within
their target set because of a loose condition in yryc() and yens().
Solution:

replace in Music.pm, lines ~865 and ~1031: 

 push @ordseqs,$seq unless ( $seq ~~ @ordseqs );

by: 

 push @ordseqs,$seq unless ( $seq ~~ @ordseqs || not($seq ~~ @{${$$Csgrouper::SectionsH{$set}}{'sequences'}}) );

130327:
Status: corrected.
Seriousness: disabling.
Description:
Xfuns weren’t usable anymore – probably since rev. 0.3 – because of a typo in $XYfuns.

130225:
Status: corrected.
Seriousness: benign.
Description:
seq_del() didn’t delete correctly. ins_del() didn’t delete in the %Project as well as in %$CsgObj->instruments.

120626:
Status: corrected.
Seriousness: annoying.
Description:
The reload() command was inverting sequence selection.
Solution:
-validatecommand option in Tk text entries is run at creation time, thus our rows are deselected. We had met this problem already and needed only to reuse the same $STARTEXEC fake init time trick.

120625 :
Status: corrected.
Seriousness: disabling.
Description:
As soon as a parameter was missing from the instrument text ins_parnum was not numbering parameters correctly anymore, in consequence parameters had to be verified and even listed with unuseful variables to ensure their taking into account. Unfortunately this situation arises normally for the duration param p3 and sometimes also for params prepared but not exploited yet.
Solution:
ins_parnum has been written better and now accepts (and even warns for) missing
params, no need to list them specially in the beggining of instrument text anymore, no error in consequent parameter numbering. The only param that can be ommited without warning is the last one for any instrument, in that case Csound will complain about an incorrect param count.

111018 :

Status: corrected.
Seriousness: disabling.
Description:
While deleting multiple sequences but after objects have been created and struct_ctl has been run: lots of sequences all the possible sequences are created.
Solution:
I had the following line in the begining of the struct_ctl() loop:
$self->sequences->{$tpref.”_id”}//= “”;
so the objects were created.

111018:
Status: not corrected.
Seriousness: annoying.
Description: *** Perl/Tk-related bug ***
In spite of my effort, instrument fields continue to record bad multibyte chars (e.g., from Emacs-like control-char reflexes). If a .csg project refuses to load, it may be because an unwanted multibyte character has accidentally been inserted; this will cause XML:Simple to die with an error message mentionning the concerned line.
Solution:
Currently, the- Solution: is to find and delete the bad characters using nano or vi (they will appear with a leading ^).
120323 :
Status: corrected.
Seriousness: annoying.
Description: *** Perl/Tk-related bug ***
The Podtext parameter disturbs proper Tk display here. See : my $Man_top_tw = $Tabs{‘man’}->Scrolled. Perlmonks 120323.
Solution:
use Tk:option (actually Tk:Cmdline) to override default color.

120323 :
Status: not corrected.
Seriousness: benign.
Description:
Tk:Pod:Text overwrites the main window title with the head1 entry.

120323 :
Status: not corrected.
Seriousness: annoying.
Description:
Tk:Table : Long strings are not displayed correctly, horizontal scrolling is corrupted.

120323 :
Status: not corrected.
Seriousness: benign.
Description: *** Perl/Tk-related bug ***
Tk:TextEntry -validatecommand is called on widget creation (is it a bug?).

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s