meeting_2007-03-13

Meeting with Polderland 13.3.2007

Participants:

  • Peter Beinema
  • Sjur Moshagen

Agenda

  • Since last time
  • Possible issues
  • Next meeting
  • Todo items for the next meeting

Since last time

Polderland:

  • test software sent, does not recognise UTF8 -> investigating
    • upgraded test software, now handles utf8 (requires option -u8). Sorry for forgetting to mention the option immediately.
  • lexicon issue: "WI" words ignored under certain circumstances. This is on purpose: used to be regarded as somewhat "lesser" word class. investigating impact of changing this.
    • Divvun is presently pregenerating the closed POSes including clitics, which means we do not need the clitics for the W entries
    • but PLD is interested in finding out what led to this way of handling data
  • windows installer: check if executable winzip version will run setup.exe automatically
  • send all files + instructions to create mac .pkg/.pax installer

Divvun:

  • new speller test tools working fine with the UTF-8 bug fixed
    • found a new, smaller issue with the test tool skipping lines beginning with non-letters (ie numbers, hyphen). It should test those as well.
  • derivations added to the PLX dataset, but the conversion to PLX is now extremely slow - working on fixing it
  • automated testing partially built, but will need some more work
  • also created an AppleScript to run text documents through the speller using MS Word - it will allow us to test spellers for other languages, and compare their behaviour with ours, as well as check consistency between the command-line tool and the MS Word behaviour: -) (it is very slow, though) (it is in gt/script/*.applescript, usage example towards the end of gt/Makefile)

Possible issues

Updating installer packages

We (ie Divvun) can now easily compile our own speller lexicons - thanks for the tools: -)

But we also need to place that speller lexicon file over the old one within the installation packages. For that I assume we need the original installation package script (or whatever it is called), and for Win we also - I assume - need the software used to create the installer? (the corresponding software is included on the Mac, as part of the developer tools).

The tools used:

  • Windows: InstallShield + small executable asking for the language to install
  • Mac: Apple Installer tool

Next meeting

Next Tuesday (20.3.) at the usual time.

TODO:

  • check handling of lines by test software: lines starting with non-alpha skipped? cf. types-err-sme.txt: first 7 lines skipped
  • check: does the Adobe hyphenator differentiate between "word boundary" (strong) hyphenation points and "word internal" (weak) ones?
  • send all files for mac installer + instructions
  • setup-text for windows installer: does language chosen refer to installer interface, or to language to be installed?
  • installer for windows: should be multi-lingual for Norwegian, Swedish, Finnish, English, N.Sami and L.Sami. Check which installshield version we need
  • discuss: how to include new lexicons in installer (PLD)
  • check: installer on Mac should be able to:
    • move existing proofing tools out of the way (into a zip archive?)
    • restore "original" proofing tools on deinstallation
      • yes, but some residues remain after de-installation.
      • solution probably requires admin rights - will document
  • check if Sami-codes are included in Office 2008 / Mac (PLD) (when beta arrives with codes) (Polderland)
  • try to find proper compiler version for Adobe Indesign (old version will probably do) (Polderland)
  • send speller lexicon (or CVS tag) for the beta release to Polderland ( Sjur)