CoCoA-5: Issueshttps://cocoa.dima.unige.it/redmine/https://cocoa.dima.unige.it/redmine/redmine/favicon.ico?15088260392015-06-11T06:26:36ZCoCoA Project
Redmine Feature #733 (Closed): Primary Decomposition in CoCoA-5.1.2: zero dimensional, finite characteristichttps://cocoa.dima.unige.it/redmine/issues/7332015-06-11T06:26:36ZAnna Maria Bigatti
<p>First prototype</p> Feature #732 (Closed): Primary Decomposition in CoCoA-5: port Luis Garcia's package(cocoa-4)https://cocoa.dima.unige.it/redmine/issues/7322015-06-11T06:25:41ZAnna Maria Bigatti
<p>Luis Garcia wrote a package for CoCoA-4. Port it to CoCoA-5.</p> Bug #726 (Closed): TopLevel cannot "import" a package variablehttps://cocoa.dima.unige.it/redmine/issues/7262015-06-06T20:34:29ZJohn Abbott
<p>I would like to have a user controllable "verbosity level" in a package. How should this be done?</p>
<p>I had thought of having a package "global" variable <code>$.verbosity</code> but then how can this be tested inside package functions?</p>
<p><code>TopLevel verbosity;</code> does not work (perhaps because the variable is not exported?)</p>
<p><code>TopLevel $.verbosity;</code> gives a syntax error at <code>$</code> saying it was expecting an identifier.</p>
<p><code>TopLevel $PkgName.verbosity;</code> gives the same syntax error.</p> Feature #719 (Closed): Readline: consider creating a textual interpreter with readlinehttps://cocoa.dima.unige.it/redmine/issues/7192015-05-21T15:47:53ZJohn Abbott
<p>This was discussed in March 2009 via email. I will copy the essence of the messages here.</p>
<p>Christof would greatly appreciate a readline textual version (since he's an emacs heathen).</p> Bug #699 (Rejected): Emacs UI: set back "cocoa5-executable" to "CoCoAInterpreter"?https://cocoa.dima.unige.it/redmine/issues/6992015-05-12T14:47:14ZAnna Maria Bigatti
<p>In earlier versions <code>cocoa5-executable</code> was <code>CoCoAInterpreter</code>.<br />Now it is the script <code>cocoa5</code> (and this caused problems on Windows because scripts do not work there).</p>
<p>The <code>cocoa5</code> script works very well for calling the command-line version from any directory.<br />But this is not necessary for emacs (because emacs needs to be configured anyway).<br />So, we cannot remember why we did this change: probably to allow spaces in path.</p>
<p>Proposal: change back and test with spaces in path. If it works we can unify Mac/Linux and Windows EmacsUI settings.</p>
<p><strong>Rejected</strong>: the script <code>cocoa5</code> is used on linux to choose the executable 32/64 bit (May also be used similarly on Mac).<br />Related issue <a class="issue tracker-1 status-5 priority-4 priority-default closed" title="Bug: Emacs UI: problems with packages on windows (Closed)" href="https://cocoa.dima.unige.it/redmine/issues/698">#698</a> solved using <strong>system-type</strong></p> Bug #684 (Closed): Emacs UI: wrong name after auto-inserting enddefinehttps://cocoa.dima.unige.it/redmine/issues/6842015-04-14T13:37:51ZJohn Abbott
<p>The following in complete code adds the wrong commented out name after doing <code>C-c C-e</code> (calls <code>cocoa5-close-block</code>):<br /><pre>
define fn(X)
N := 1;
// Now define the blah blah...
</pre></p>
<p>The commented out name which is inserted is <code>the</code>, presumably taken from the comment.</p> Feature #652 (Closed): Flag to enable/disable printing of list of loaded packageshttps://cocoa.dima.unige.it/redmine/issues/6522014-11-28T10:06:23ZJohn Abbott
<p>Add a command-line flag to enable printing of package names as they are loaded. Currently the names are always printed out (one per line) which wastes lots of vertical space (and is probably of no interest to normal CoCoA-5 users).</p> Design #634 (Rejected): Symbol in the coeff ringhttps://cocoa.dima.unige.it/redmine/issues/6342014-10-22T14:08:11ZJohn Abbott
<p>Consider the following mini-session<br /><pre>
QQab ::= QQ[a,b];
K := FractionField(QQab);
Use P ::= K[x,y];
a; // in which ring???
</pre><br />In which ring should the value of <code>a</code> lie?<br />Currently, it lies in the ring <code>P</code> so that expressions such as <code>x+a</code> are accepted (<em>i.e.</em> no problem of <em>mixed rings</em>).</p>
<p>If we allow automatic mapping from a subring to a superring then we could put the value of <code>a</code> in the smallest subring which contains it (in this case <code>K</code>), then in the expression <code>x+a</code> the value of <code>a</code> is automatically mapped into <code>P</code> before being added to <code>x</code>.</p> Feature #622 (Closed): New function: RandomSubsethttps://cocoa.dima.unige.it/redmine/issues/6222014-09-12T08:38:30ZAnna Maria Bigatti
<p>Sometimes "all subsets" are far too many (compute how many! ;-) to fit into memory.<br />So checking a property on a suitable number of random subsets can be interesting.</p> Feature #590 (Closed): package for subalgebrashttps://cocoa.dima.unige.it/redmine/issues/5902014-07-16T10:27:31ZAnna Maria Bigatti
<p>In cocoa-4 there was a package to compute with subalgebras.<br />Restore it.</p> Design #490 (Closed): Duplicate fns: valuation and FactorMultiplicityhttps://cocoa.dima.unige.it/redmine/issues/4902014-03-21T13:35:18ZJohn Abbott
<p>CoCoA-5 has two virtually identical fns <strong><code>valuation</code></strong> and <strong><code>FactorMultiplicity</code></strong>.</p>
<p>We need only one of them!<br />Decide the precise semantics, and fn name.</p> Feature #485 (Closed): Initialization for CoCoA-5: file init.cocoa5https://cocoa.dima.unige.it/redmine/issues/4852014-03-20T09:10:02ZAnna Maria Bigatti
<p>In CoCoA-4 there was a file called <code>init.coc</code> with some "initial" settings.<br />Now most initialization are done by CoCoALib and by the Interpreter.<br />The only missing feature is the missing "prealoded" ring QQ[x,y,z].</p>
<p>CoCoA-5 starts with such a ring predefined and called <code>R</code> but is not loaded.<br />G.Lagorio said it was not easy/clean to load it at interpreter level, but should be done by automatically reading an "init.cocoa5" file.</p> Feature #359 (Closed): RingOf for listhttps://cocoa.dima.unige.it/redmine/issues/3592013-05-27T07:40:16ZAnna Maria Bigatti
<p>It is often useful to know whether all elements in a list belong to the same ring.<br />A function of this kind was <strong>RingEnvSet</strong> in CoCoA-4.<br />It has been adjusted to CoCoA-5 as <strong>RingSet</strong> (see manual) but I'm not happy with the name: maybe it should be called <strong>RingsOf</strong>?</p>
<p>Moreover it is quite flexible (also deals with lists of lists, and with INT and RAT): too flexible?</p> Design #332 (Closed): Approx pts preprocessing fnshttps://cocoa.dima.unige.it/redmine/issues/3322013-03-27T18:30:31ZJohn Abbott
<p>I have just added the approx pts preprocessing fns to CoCoA-5.<br />This issue is about deciding the most appropriate interface; right now it is just what was easiest for me to impl.</p>
<p>Points and tolerances are specified as matrices (automatically guarantees that all coords are in the same ring, and that the point all have the same number of coords -- do need to check that points and tolerances are compatible). Result is as a list of lists currently; shoudl this be a matrix too?</p>
<p>Each point is represented as a row in the matrix.</p>
<p>Opinions?</p> Bug #164 (Closed): CoCoA-5 emacs interface bug (minor): Source & SourceRegion with troublesome fi...https://cocoa.dima.unige.it/redmine/issues/1642012-05-20T09:25:09ZJohn Abbott
<p>By accident I created a filename containing a newline.<br />Sending the file to CoCoA-5 in emacs (by C-c C-f) produced the following error:<br /><pre>
ERROR: Newline characters are not allowed inside string literals (you can insert them using escape sequences, for instance, \n or use multi-line string literals)
</pre></p>
<p>I suspect something bad may occur if the filename contains double quote characters too. We should add some safety checks.</p>