On Mon, 19 Sep 2005 17:28:59 -0400
Chris Lawrence <[EMAIL PROTECTED]> wrote:
> querybts is probably querying xpdf (binary), while reportbug queries
> xpdf (source) by default. That is almost certainly the discrepancy
> here.
So you're saying there's two bug lists for most packages, one for
binary, and one for source. The one I'm used to has a BTS page like
this:
http://bugs.debian.org/cgi-bin/pkgreport.cgi?pkg=xpdf
...but users can link from there to here:
http://bugs.debian.org/cgi-bin/pkgreport.cgi?src=xpdf
...which, as you say, has those "missing" bugs. Which makes sense in
cases where users are compiling from source, if they find bugs in the
code.
OTOH here's a section of 'reportbug' (source bugs) output:
Querying Debian BTS for reports on xpdf (source)...
108 bug reports found:
Outstanding bugs -- Normal; Patch Available (1 bug)
1) #320631: manpage xpdf.1 broken
Outstanding bugs -- Normal; Unclassified (11 bugs)
2) #221521: update-xpdfrc includes files from packages removed but
not purged
3) #227913: xpdf: Non-latin1 characters in find dialog
4) #243132: xpdf: one pdf document can't print
5) #280767: xpdf: Rasterization of dark gray text looks bad
6) #284734: xpdf-reader: pdftoppm requires integer resolution
7) #298757: /usr/bin/pdftotext: some big5 chars dropped
8) #320509: PDF 1.6 not supported, xpdf warns and continues
9) #322318: xpdf: interface looks reversed (colors)
10) #326888: xpdf-reader: sets document bounding size arbitrarily
11) #327170: xpdf-reader: Always fails to allocate fonts on first try
12) #329112: xpdf resize problem
Outstanding bugs -- Minor; Patch Available (1 bug)
13) #280460: xpdf-reader: zxpdf doesn't remove its tmp file at exit
And I finally notice where it says "reports on xpdf (source)", which I
hadn't appreciated until you mentioned it.
Except the answer opens up some questions...
Do most of those bugs look like source bugs? No. For example:
#243132 xpdf: one pdf document can't print
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=243132
...there's no code in that bug, it describes the symptoms, the
program's output. Are most source bugs misfiled? Is 'reportbug'
itself misfiling them by using source as a default?
And not all packages have so many noticeable differences as 'xpdf' in
their 'source' and 'binary' BTS pages. It's why I never
noticed 'em before.
Example: a 'querybts -w reportbug' and 'reportbug --query-only
reportbug' are the same... proof:
# send reportbug output to a temp file
% reportbug --query-only reportbug 2>&1 | tee /tmp/rb.txt #hit <enter>
a lot for this...
# separate the bug numbers, sort, send to another temp file.
% grep '#' /tmp/rb.txt | sed -e 's/.*\#//' -e 's/:.*//' | sort -g |
uniq > /tmp/rbnum.txt
# do the same for 'querybts'
% yes N | { querybts reportbug ; echo ; } 2>&1 | grep '#' | sed -e
's/.*\#//' -e 's/:.*//' | sort -g | uniq > /tmp/qbnum.txt
% cmp /tmp/[rq]bnum.txt ; echo $?
0 # they're the same
Why should 'querybts' and 'reportbug' have different defaults
anyway? The 'whatis' descriptions don't mention that:
% whatis querybts reportbug
querybts (1) - view outstanding bug reports on a debbugs server
reportbug (1) - reports a bug to a debbugs server
Nor do their respective man page "DESCRIPTION" sections say that
both programs display different (sometimes overlapping) lists. I have
used these utils for years, and didn't know this.
Conclusion: thanks very much for for answering the question; though
from my naive view it looks like the answer implies some new bugs. Hope
somebody can clue me in if I've got it all wrong.
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
