Emacs ringing system: bug reports

Reported Fixed Description Comments
1997-06-16 New Emacs session. To recreate this bug you need to use desktop.el. Save a desktop with a msiril place notation file loaded (s8 in my case) in your home directory so that it will automatically get loaded when you start a new emacs session. Here is my message window. This bug does not seem to happen when s8 is loaded as normal.
Loading mSiRiL-mode...
Loading cl-macs...
Loading cl-macs...done
Loading mSiRiL-mode...done
File mode specification error: (wrong-type-argument number-or-marker-p nil) [2 times]
1997-06-16 `ringing:set-variables' function gave default prompt for the following three directory variables. It got the _correct_ default from the value already defined. Having accepted the default value, your code then inserted "/user/me/lisp/bells/source/" into the first and third variables below and saved my .ringing.el file. Now all the source code files I load from "/user/me/lisp/bells/source/" are in msiril-mode!!
method-library:directory "/user/me/lisp/bells/mslibs/"
place-notation:default-methods-directory "~/lisp/bells/mslibs/"
method-groups:groups-directory "/user/me/lisp/bells/groups/"
I've tried to fix this, but it's still not right yet. Will look at it further.
1997-06-16 Postscript printing looks like it will do the job, but then I save it to method.ps and print it to get a blank page. I tried several printers. Could you not pipe the buffer to a unix command like "lpr". Even better, allow the user to configure the print command in a variable. Not completed yet!
1997-06-16 "*Ringing report*" and "*Clash*" buffers not tidied up by `ringing:tidy-up' function.
1997-06-16 `placenotation-explore' of "-34-56-7-56-34" on 8 bells actually wrote out x34x56x78x56x34x - OOPS! last cross should be omitted.
1997-07-04 1.12
Antelope Place Doubles:
3.4-1-1-1, 4
please write place notation as "+......" or "&.....,+le"
place-notation:recreate-string-from-changes needs to know about symmetric methods
1997-07-8 1.11 Handstrokes and backstrokes aren't distinguished Add another set of faces -- perhaps italic for backstrokes?
1997-07-11 1.11 Standing a touch before making any mistakes gets it recorded in your touch-book as rung perfectly
1997-07-11 1.11 mSiRiL-mode does not munge method names into sensible form -- that's used only in the completion code.
1997-07-19 Cannot now give zero to reset observation bell in mSiRiL-mode prefix arguments, as that is now taken to mean 10. Make negative numbers reset it?
1997-07-27 Some confusion about loading method files when a group calls for a method that isn't present at that stage.
1997-07-28 Methods produced by combination fail to get methods with the same work listsed
1997-08-05 Compilation function failed. `load-file-name' gate a path/filename that did not exist!
1997-08-05 C-u 0 M-x byte-recompile-directory seems to work with the following warnings:-
Compiling file /users/me/lisp/bells/source/leads.el at Mon Aug  4 15:39:15 1997
  ** reference to free variable ringing:lead-count
  ** assignment to free variable ringing:lead-count

Compiling file /users/me/lisp/bells/source/mSiRiL-mode.el at Mon Aug  4 15:39:22 1997
While compiling mSiRiL-mode:print:
  ** reference to free variable ringing:latest-ps-title-style
  ** assignment to free variable ringing:latest-ps-title-style
  ** reference to free variable ringing:latest-ps-style
  ** assignment to free variable ringing:latest-ps-style
While compiling the end of the data:
  ** The following functions are not known to be defined:
    ringing:ps-title-styles, ringing:ps-styles
Compiling file /users/me/lisp/bells/source/method-groups.el at Mon Aug  4 15:39:39 1997
While compiling method-group:print-group:
  ** reference to free variable ringing:latest-ps-title-style
  ** assignment to free variable ringing:latest-ps-title-style
  ** reference to free variable ringing:latest-ps-style
  ** assignment to free variable ringing:latest-ps-style
  ** reference to free variable ringing:latest-ps-group-style
  ** assignment to free variable ringing:latest-ps-group-style
While compiling the end of the data:
  ** The following functions are not known to be defined:
    ringing:ps-title-styles, ringing:ps-styles,
    ringing:ps-group-styles, ringing:postscript-for-group
1997-08-05 ringing a method. Oh krickey whats going on? If I do not press a key for a while you press the "right arrow" key rather than the "correct arrow" key. You are supposed to put the answer in for me (up, down, place) and record a clash whatever. Hence, if left alone, the computer would ring all the method and record a clash for every row. At the moment you press right arrow and then check to see if that was the right answer, if not record a clash. That is incorrect. That was what I thought you'd asked for -- i.e. ring too late if you can't think of where to go in time! Will fix, what you suggest now is much more sensible.
1997-08-05 Ringing random leads from group. Buffer named "One lead" - good. What method am I ringing? The minibuffer flashed it past too fast. Please print the method name (and place bell?) in the buffer above the blue line.
1997-08-05 Postscript printing when awol. Probably our system though. Eventually after turning the printer off and then back on again it came out.

I will get in trouble with sysadmin if I print anymore methods as they get stuck in the print queue and don't come out. I will have to test this another way. Works OK on SunOS4 but not SunOS5 (Solaris 2)

This is Level 2 PostScript, and so will not work on very old printers. Also, remember it is actually running the method on the processor inside the printer, which might be quite a slow one (although ours is pretty quick at them).
1997-08-05 Textual output to a buffer when ringing/showing a method has changed. I don't like alternate rows in italics so I turn that off. Now I get the second and subsequent rows in bold -- yuck! No bold thank you. Don't know how it does that! There's nothing in the code I can think of that would do that. Will investigate. Could you try loading test.el from the emace-ringing directory, run M-x ringing:test-faces and telling me what the output it produces is like?
1997-08-05 Cambridge S. Minor looks silly in printout. Can you make the line print a fixed size. If the page is not big enough for all the selected methods, then start a new page.

The appropiate size for methods (I think):-

  • For minor, the size of line achieved when fitting 6 minor methods to landscape A4.
  • For Major, Royal and Maximus, the size achieved when fitting 3 methods to portrait A4.
The method size if fixed whatever the size of the paper. The methods are still the same size even when there is only one being printed on the piece of paper. Of course, whitespace between and in the margins should be calculated to be equal to make it look pretty. The only other calculation you need to perform is how many methods you can fit on a page. Eg. does the standard size of minor method fit 5 or 6 time in landscape? Well the 6th, 7th and 8th methods will have to go on a new page then....
Which line style are you using? Not quite fixed; for styles with the numbers shown, there is a default size, and if that doesn't fit the paper, it is reduced according to the most restricting dimension. Line-only styles always take the full drawing area given to them.
1997-08-05 Optionally left or center justify the methods on each page. For center justify, the white space between and in each margin should be equal left to right. And in a similar but separate calculation for top and bottom justification. For left justify, center all methods vertically on the page, but print them layed out as if the page was taking the maximum possible number of methods per page. The reason for this is, when printing a page and a halfs worth, the second page of methods will be aligned with the first page.
1997-08-05 Scale factor for printing. Decide on well proportioned standard sizes and then multiply both dimensions by the user specified scaling factor. (Default = 1.0)
1997-08-05 Marking ephemeral groups requires the buffer to be writable. I want to keep by place notation files read only hence by default the buffer is read only. Can you allow selection of methods in a read only buffer (or just make it writable if it isn't). Will do (to do this kind of thing, you temporarily make the buffer writable while putting the colouring in -- mainstream highlighting code has to do this, too, as far as I know).
1997-08-05 Title too large. I think it should be no wider than the width of the method and should be able to wrap on to a new line if necessary.
1997-08-05 Some or all operations on a single method result in the whole pn file being parsed which causes a delay that I dislike. Ah! I've altered `ringing:always-scan-all-libraries'. May be that will work now.
1997-08-05 I stood my worst lead twice and it told me I had passed?!?!
1997-08-05 Path names still get messed up. I have to hand edit those.
1997-08-05 The lines printed over the numbers are a fraction too high. I dislike the numbers in italics on alternative rows, normal face for all will do. I'll put in a switch at the Lisp level to control what the PostScript does with this. For now, you can get round it by editing the files in ps-title-styles/* where they say things like
  /handstrokefont /Courier findfont fontsize scalefont def
  /backstrokefont /Courier-Oblique findfont fontsize scalefont def
Just the obvious changes!
1997-08-07 Method name highlighting is getting into touch-book
1997-08-07 Problems with red and green lines getting mixed up when blue line starts at bell 2.
1997-08-07 Stand (to quit) doesn't always work first time
1997-08-07 Blobs in wrong places, at least on compact-line
1998-04-23 Automatic ringing on timeout doesn't work in random leads from mSiRiL buffers This goes through the teaching code, which calls teach-ringing:get-user-stroke, which has no timeout facility, rather than through ringing:get-user-stroke where the timeout facility is.
1998-04-27 The following printing styles seem to hang:
  • course-bell (fixed -- missing pop in displayfirstrow)
  • circular (fixed -- missing pop in displayfirstrow)
  • one-column (fixed -- missing pop in displayfirstrow)
  • stereo-grid (fixed)
1998-05-11 Full-grid editing does strange things if you click in first place.
1998-09-18 Things get on the wrong row in non-leftmost columns when column labelling is off.
Diopside l +-56-4-56-36-34-5-34-1-4-5.4-4.36.5-5.4-5-
from s8 doesn't look right to me
1999-03-19 method-library:report-on-loaded-methods doesn't work

[John's ringing page] [John's home] Back to emacs-ringing index.
John C. G. Sturdy
Last modified: Fri Mar 19 15:33:30 1999