no nightly builds of the spec

[scabug]

Now that I see spec changes I guess I should open a ticket for the absence of fresh specifications.

http://www.scala-lang.org/node/212/pdfs

Has not had new scala references for a while.

[scabug]

Imported From: https://issues.scala-lang.org/browse/SI-4208?orig=1 Reporter: @paulp Attachments:

• i.tgz (created on Feb 4, 2011 5:33:21 PM UTC, 8346 bytes)
• LuxiMono.zip (created on Feb 4, 2011 5:34:07 PM UTC, 199660 bytes)

[scabug]

@lindydonna said: Toni, are you the right person to handle this? If not, please re-assign or tell me to re-assign.

[scabug]

@cunei said: This seems to be a genuine bug in Ghostscript: the problem disappears removing the lines between begin and end in the following section (lines 8394 to 8399 as of r24170):

\begin{lstlisting}
  Byte    <-> java.lang.Byte
  Short   <-> java.lang.Short
  Char    <-> java.lang.Character
  Int     <-> java.lang.Integer
  Long    <-> java.lang.Long
  Float   <-> java.lang.Float
  Double  <-> java.lang.Double
  Boolean <-> java.lang.Boolean
\end{lstlisting}

The issue reappears again if at least two non-empty lines of text (any text) are inserted between the begin and end. I will ask Fabien to try Ghostscript 9.00, we'll see whether there's any improvement.

[scabug]

@cunei said: Correction, lines 8394 to 9399 are:

\begin{lstlisting}
  Byte  -> Short
  Short -> Int
  Char  -> Int
  Int   -> Long
  Long  -> Float
  Float -> Double
\end{lstlisting}

The issue is as described above.

[scabug]

@cunei said: reduced version of problematic tex source file

[scabug]

@cunei said: Luxi Mono font

[scabug]

@cunei said: (In r24193) Ugly workaround to address a LaTeX issue introduced in r23753.

Failure in the sense that TeX gets confused for reasons unknown. See #4208 for further information.

[scabug]

@cunei said: After some investigation, it turns out this is a LaTeX failure; a TeX wizard may be required to fully understand what is going on. I am enclosing a reduced version of the TeX source that exhibit the problem (see ticket attachments); removing some random line from this small file seems sufficient to make the example typeset again.

The anomaly in the spec source was introduced in the spec file in r23753. I added an ugly workaround in r24193, but it would clearly be preferable to understand the root cause of the matter. Any suggestion is appreciated.

[scabug]

@cunei said: (In r24195) Slightly better looking ugly workaround, see #4208.

[scabug]

@lexspoon said: It looks like a bug that occurs when a listing crosses a page break. There's nothing special about the listing itself, and if I put a \clearpage before it, the thing builds fine. Give that, the least bad thing to do right now might be to put a \clearpage before any listing that has the problem. Additionally, for longer listings, it might be better to put them inside a float so that they won't possibly cross a page boundary.

To prevent people from following a dead end, I can also add that surrounding the listing with \begin{samepage}...\end{samepage} didn't fix the problem.

The ideal fix would be in the Listings package itself. I don't have time right now to look into that, but perhaps someone else does. Also, I'll email Brooks Moses, the most recent maintainer I can find for the Listings package, and include a link to this bug report and minified sample file.