[Gzz-commits] navidoc doc/pegboard/pegboard_format--tjl/peg.r...

[Tuomas J. Lukka]

CVSROOT:        /cvsroot/navidoc
Module name:    navidoc
Branch:         
Changes by:     Tuomas J. Lukka <address@hidden>        03/06/06 16:04:24

Modified files:
        doc/pegboard/pegboard_format--tjl: peg.rst 
        navidoc/mp     : uml.mp 
Added files:
        navidoc/mp     : roughEdgeTester.mp 

Log message:
        Rough edge tester

CVSWeb URLs:
http://savannah.gnu.org/cgi-bin/viewcvs/navidoc/navidoc/doc/pegboard/pegboard_format--tjl/peg.rst.diff?tr1=1.2&tr2=1.3&r1=text&r2=text
http://savannah.gnu.org/cgi-bin/viewcvs/navidoc/navidoc/navidoc/mp/roughEdgeTester.mp?rev=1.1
http://savannah.gnu.org/cgi-bin/viewcvs/navidoc/navidoc/navidoc/mp/uml.mp.diff?tr1=1.10&tr2=1.11&r1=text&r2=text

Patches:
Index: navidoc/doc/pegboard/pegboard_format--tjl/peg.rst
diff -u navidoc/doc/pegboard/pegboard_format--tjl/peg.rst:1.2 
navidoc/doc/pegboard/pegboard_format--tjl/peg.rst:1.3
--- navidoc/doc/pegboard/pegboard_format--tjl/peg.rst:1.2       Fri Apr 25 
03:52:06 2003
+++ navidoc/doc/pegboard/pegboard_format--tjl/peg.rst   Fri Jun  6 16:04:24 2003
@@ -4,25 +4,26 @@
 
 :Authors:  Tuomas Lukka
 :Stakeholders: Asko Soukka
-:Last-Modified: $Date: 2003/04/25 07:52:06 $
-:Revision: $Revision: 1.2 $
-:Status:   Incomplete
+:Last-Modified: $Date: 2003/06/06 20:04:24 $
+:Revision: $Revision: 1.3 $
+:Status:   Current
 
 This META-PEG deals with formatting PEGs.
 
 Issues
 ------
 
-    - Our own directives?
+    - Our own RST directives?
 
-        RESOLVED: Not yet.
+        RESOLVED: Not yet. Of course, there's UML but that's
+       used in other docs. Nothing pegboard-specific.
 
     - Should we use urn-5 instead of serial numbers to avoid
       collisions? This could be a nice application.
 
         RESOLVED: Not. Not human-readable enough. Collisions
-       are resolved by using the names described below instead
-       of serial numbers.
+       are resolved by using the individual user names 
+       described below instead of serial numbers.
 
     - What should we say about the issues section?
 
@@ -57,8 +58,51 @@
 
        is much better, since it at least gives more information.
 
+       The point is that issues once raised will continue to be raised,
+       and having the reply *with the reasons* at hand helps avoid
+       getting stuck in loops.
+
     - Do we need the Scope and Type fields?
 
+       RESOLVED: They don't hurt.
+
+What is a PEG?
+--------------
+
+"PEG" used to be short for "Proposal for Enhancing Gzz". At the moment,
+it's just PEG, with no real acronym meaning.
+
+At the moment, PEGs are used in all the Gzz-split projects, i.e. Fenfire,
+Libvob, CallGL, Alph, Storm, ...
+
+A PEG is a short document detailing, for example,
+
+- An architectural plan for an extension
+
+- Changes to a core interface
+
+- Plans for experiments
+
+- File format specification that is to be used somewhere
+
+- Code style specifications
+
+- User interface specifications
+
+- Roadmaps
+
+I.e. *any* kind of plan or specification that needs to be discussed among the 
+group members.  The PEG process is about writing out some plan, submitting
+it for peer review among the group, resolving issues raised by the group and
+clarifying the relevant parts and finally voting (or dictatorial acceptance by
+the project leader) -- i.e. the PEG process is about achieving **consensus**.
+
+PEGs are especially important for the Jyu Hyperstructure Research group because
+we're geographically outspread and open, so PEGs are vital for everyone
+knowing what the others are thinking about.
+
+PEGs are also an educational instrument for learning to write English text 
comfortably.
+
 Process
 -------
 
@@ -72,20 +116,26 @@
 
 As long as the PEG is in Incomplete or Current state, the author
 may change its status to Current, Incomplete (for revision),
-or Irrelevant (e.g., other PEGs superceded the current one).
+Irrelevant (e.g., other PEGs superceded the current one) or
+Postponed (i.e. while making the PEG it became clear that the time is not yet 
ripe
+for it, something else needs to be done first.). Postponed PEGs may be moved 
to the other
+non-accepted states freely by the author.
 
 **NOTE**: PEGs should not be made Current as long as there are unresolved
 issues. If the responses from the list raise issues that cannot be immediately
 resolved, the PEG should be put into Incomplete state again.
 
-At some point after this, the architect (tjl) will make a decision
+At some point after this, the project leader (varies for the different 
subprojects) 
+will make a decision
 to either accept, force revision, declare rejected, or declare
 irrelevant. No-one else may change the status of the PEGs 
-to Accepted or Rejected.
+to Accepted or Rejected. While it is allowed, the project leader should avoid 
accepting PEGs
+that have not been posted to the mailing list for review in their latest form.
 
 If the PEG was accepted, rejected or declared irrelevant, it may not
 be edited any more, except for the architect or author
-declaring it implemented or partly implemented.
+declaring it implemented or partly implemented from the accepted state.  
+For rejected or irrelevant PEGs, new PEGs should be started.
 
 A PEG int the Revising state is editable by the authors, similarly to
 an incomplete PEG.
@@ -101,7 +151,8 @@
     move_vobs_2--benja
     meta_new_fields--tjl
 
-These names are used for the directory names.
+These names are used for the directory names. The last part is the handle of 
the author, 
+to ensure uniqueness.
 
 Inside the directory, there should be one reStructuredText file,
 called ``peg.rst`` and any number of other files (images, example code ...).
Index: navidoc/navidoc/mp/uml.mp
diff -u navidoc/navidoc/mp/uml.mp:1.10 navidoc/navidoc/mp/uml.mp:1.11
--- navidoc/navidoc/mp/uml.mp:1.10      Mon May 26 12:20:20 2003
+++ navidoc/navidoc/mp/uml.mp   Fri Jun  6 16:04:24 2003
@@ -19,7 +19,7 @@
 % MA  02111-1307  USA
 % 
 
-% $Id: uml.mp,v 1.10 2003/05/26 16:20:20 humppake Exp $
+% $Id: uml.mp,v 1.11 2003/06/06 20:04:24 tjl Exp $
 
 %
 % Written by Tuomas Lukka
@@ -475,7 +475,9 @@
     np := jitterpath(np, round(40+uniformdeviate(30)), .025 / scale * j);
 
     draw np withpen pencircle scaled (4pt/scale) xscaled .4 rotated -37;
-    clip currentpicture to q;;
+    if known(q):
+       clip currentpicture to q;;
+    fi;
     
 enddef;