W3C home > Mailing lists > Public > www-amaya@w3.org > April to June 2005

Re: lots of bugs in Amaya

From: Vincent Quint <Vincent.Quint@inrialpes.fr>
Date: Mon, 30 May 2005 16:38:55 +0200
To: Wolfgang Jeltsch <wolfgang@jeltsch.net>
Cc: Vincent.Quint@inrialpes.fr, www-amaya@w3.org
Message-Id: <20050530163855.043d09d0.Vincent.Quint@inrialpes.fr>

Thanks for the detailed bug report. This is very helpful.
Several bugs you reported have been fixed in the CVS base.
A snapshop is scheduled for tomorrow: it will include these fixes.
However, some bugs you reported will take more time. They are
expected to be fixed in the next release.

1. DISTURBANCE OF STRUCTURE IN CONNECTION WITH MATHML

This one is deferred. Although its cause is identified, it would require
some more testing to make sure it can be fixed without unwanted side effects.

2. MISSING ELEMENT TYPES IN ELEMENT TYPE LIST

Fixed.

3. AUTOMATIC INCLUSION OF UNWANTED LINEBREAK IN SOURCECODE

Deferred

4. PROCESS OF CHANGING ENTITY REFERENCES TO CHARACTER REFERENCES IS MADE 
VISIBLE

Fixed. All entities are always displayed in the same way in the structure view.
Invisible characters, such as &ApplyFunction; or &InvisibleTimes; are displayed
as entity references, while visible characters such as &alpha; or &empty; are
displayed as a single character.

5. SLOWNESS AND NASTY GUI THINGS

What is the clock frequency of your processor? the size of core memory?
We have not noticed what you describe on our computers.

6. EVENTS REACH WRONG WINDOW

Deferred

7. FORMULAS NOT LOOKING NICE

Could you be more specific on this? A few samples exhibiting the problems
could help.

8. REMOVAL OF FENCES

Fixed.

9. ESCAPE KEY NOT HANDLED BY CHARACTER ENTITY WINDOW

Fixed.

Vincent.

On Fri, 27 May 2005 12:51:32 +0200 Wolfgang Jeltsch <wolfgang@jeltsch.net> wrote:
> 
> 
> Hello,
> 
> I use Amaya for editing DocBook XML documents with embedded MathML.  So I use 
> the generic XML editing facilities and the MathML editing facilities.  In my 
> opinion, Amaya might have the potential to become the leading free software 
> editor for text-oriented XML.
> 
> But it annoys me that Amaya is really full of bugs.  And what is even more 
> annoying is that the developers seem to not see these bugs or to not care 
> about them since otherwise the list on Amaya_open_bugs.html would be twice as 
> large.  So there seems to be the danger that the next Amaya release will be 
> as buggy as the current release.
> 
> That's why I'm writing this e-mail with certain bug reports.  Please 
> appreciate that I've spent a notable amount of time analysing bugs, working 
> out examples etc.  I would be happy if the developers of Amaya would take more 
> effort in ensuring quality so that users don't have to deal with that many 
> bugs and don't have to do so much of this time-consuming work of writing good 
> bug reports.
> 
> Okay, now lets stop the preface and start with the reports.
> 
> 
> Metadata applying to all reports:
> 
>     Release number: 9.1
> 
>     Platform: Debian GNU/Linux sarge on an i686 with Kernel 2.6.8
> 
> 
> 1. DISTURBANCE OF STRUCTURE IN CONNECTION WITH MATHML
> 
> Open the following XML file in Amaya:
> 
>     <?xml version="1.0"?>
>     <para xmlns="http://docbook.org/ns/docbook">Hello!</para>
> 
> Set the cursor at the beginning of the text and insert a mathematical formula.  
> Save the file, exit Amaya, restart Amaya and open the file.  Now you won't be 
> able to insert another formula in the text.  If you open the structure view, 
> you see that the text "Hello!" is errornously placed inside the math element.  
> I also experienced segementation faults in similar situations.
> 
> This problem seems to appear only at the end of the file.  If I have elements 
> after the last para element containing formulas, it seems to work correctly.
> 
> 
> 2. MISSING ELEMENT TYPES IN ELEMENT TYPE LIST
> 
> Create a file docbook.css with the following content:
> 
>     para { display: block }
>     itemizedlist { display: block }
> 
> Open a file with the following content with Amaya:
> 
>     <?xml version="1.0"?>
>     <?xml-stylesheet href="docbook.css" type="text/css"?>
>     <para xmlns="http://docbook.org/ns/docbook">Hello!</para>
> 
> Press the reload button under "XML element type" in the Tools subwindow.  You 
> will now see that para is missing on the list.
> 
> 
> 3. AUTOMATIC INCLUSION OF UNWANTED LINEBREAK IN SOURCECODE
> 
> Take the stylesheet file of 2. and open the XML file of 2. with Amaya.  Place 
> the cursor inside the text.  Press the reload button under "XML element type" 
> and double-click on "itemizedlist".  Now open the sourcecode view.  You will 
> see that there is a linebreak before the <itemizedlist> tag.  Actually, this 
> is wrong since it introduces significant whitespace.
> 
> 
> 4. PROCESS OF CHANGING ENTITY REFERENCES TO CHARACTER REFERENCES IS MADE 
> VISIBLE
> 
> When editing documents with formulas, Amaya sometimes changes entity 
> references like &alpha; to character references.  Until now, I don't know 
> what rule Amaya follows in deciding when to do these changes.  A big problem 
> with them is that every elementary change is reflected by the structure view.  
> So if you have the structure view open, you can see Amaya removing an element 
> reference, adding a character reference, removing the next element reference 
> and so on.  This looks ugly.  In my opinion, it belongs to the basics of GUI 
> programming that one first makes all the changes and updates the GUI once 
> after being done with them. A big problem with these reference changes is 
> that they take a lot of time.  It can happen that I have to wait for over 
> five seconds because Amaya is busy with changing references!
> 
> 
> 5. SLOWNESS AND NASTY GUI THINGS
> 
> In general, Amaya is terribly slow.  In addition, sometimes during editing, 
> the user interface blocks for over five seconds.  It can happen that during 
> editing of MathML via the formatted view, visual appearance of the formula is 
> changed just to be changed again.  The thing said under 4. applies here as 
> well: First do all the calculations, then change the GUI once.
> 
> 
> 6. EVENTS REACH WRONG WINDOW
> 
> If I try to open the character entity window via C-m C-e, it regularily 
> happens that I have to wait for a while because Amaya is so terribly slow.  
> If I enter some characters before the character entity window has popped up, 
> these characters are inserted into the view I work in, not into the entity 
> window's editing field.  This is wrong.  The sequence C-m C-e a l p h a, for 
> example, should always result in the string "alpha" appearing in the 
> character entity window, regardless of how much time it takes for the window 
> to pop up.  After pressing C-m C-e, events must not be handled until the 
> action corresponding to C-m C-e, i.e., opening the entity window, is 
> completed.
> 
> 
> 7. FORMULAS NOT LOOKING NICE
> 
> The formulas are typeset badly.  &bigcup; is way to big, spacing is often not 
> sensible, fences are way too large if the fenced part appears inside an mrow 
> etc.
> 
> 
> 8. REMOVAL OF FENCES
> 
> Open the file from 1. with Amaya, place the cursor inside the text and insert 
> a formula containing an mrow with fences.  Now save the document and look at 
> the sourcecode of the saved file.  You will discover that the mo elements 
> representing the fences have empty content.  This problem seems to not appear 
> when editing a separate MathML file.
> 
> 
> 9. ESCAPE KEY NOT HANDLED BY CHARACTER ENTITY WINDOW
> 
> The character entity window which is invoked via C-m C-e should cancel when 
> the ESC key is pressed.
> 
> 
> Okay, let this be enough for now although there are other things to report 
> yet.  Maybe, I will write further reports later.
> 
> Best wishes,
> Wolfgang
> 
> 


----------
Vincent Quint                       INRIA Rhône-Alpes
INRIA                               ZIRST
e-mail: Vincent.Quint@inria.fr      655 avenue de l'Europe
Tel.: +33 4 76 61 53 62             Montbonnot
Fax:  +33 4 76 61 52 07             38334 Saint Ismier Cedex
                                    France
Received on Monday, 30 May 2005 14:48:15 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 14:53:33 UTC