*** GMX Spamverdacht *** Re: Parley: words remain in inital-stage

[Anna Vernerová]

Hi P.,

Parley is currently set to work with as many initial stages (called
"preGrades" in the code) as there are normal levels. Their number is set in
the variable KV_MAX_GRADE which is not part of the parley repository, but
comes with the libkdeedu library - on my system, it is defined in the file

/ </usr/local/include/libkdeedu/keduvoctext.h>
usr/local/include/libkdeedu/keduvoctext.h

What I find particularly useful about the initial stages is that they
increase the total number of levels to 14 (7 initial and 7 normal), AND
allow me to skip directly from level 0 to level 8 (from New word to normal
level 1 or 2, I am not quite sure how exactly it works) by marking my
answer to a "New word" question as correct. I can skip the initial frequent
revisions for words that I currently know but want to make sure that I do
not forget them in the future. I use blocking times from 3,5 minutes to one
week for the initial stages and from 2 weeks to a full year for the normal
level - see below.

Anyway, if you would still prefer to go completely without the initial
stages (for example because it is annoying that there is no visualisation
for your progress across the initial stages), I think you really have to go
back to some previous version of parley. I am not sure about compiling the
old 0.9.4, but I guess it should work. The last commit before the initial
stages were introduced is a33aa6ae8c654d54a74f987741b613c6262da900 , so you
can get the code by issuing the following in the command line:

git clone git://anongit.kde.org/parley
git checkout a33aa6ae8c654d54a74f987741b613c6262da900



Thank you for filing the wish report.

Anša


p.s. Here are my settings for the blocking times:

initial stages (in the file parley/src/collection/entryfilter.cpp):
    0,
    3 * 60 + 30,                // 1: 3.5 minutes
    7 * 60,                     // 2: 7 minutes
    15 * 60,                    // 3: 15 minutes
    16 * 3600,                  // 4: 16 h (next day)
    42 * 3600,                  // 5: almost 2 days (but allow a 6 h
difference in the practice time)
    90 * 3600,                  // 6: almost 4 days (but allow a 6 h
difference in the practice time)
    7 * 24 * 3600,              // 7: a week
normal levels (blocking times set in parley settings):
   2 weeks       level 1
   4 weeks       level 2
   2 months     level 3
   4 months     level 4
   6 months     level 5
   10 months   level 6
   12 months   level 7

This means it takes about three years for a word to get all the way to
level 7, so it provides some extra long term revision.


2015-06-14 15:53 GMT+02:00 h3x32000-sf <h3x32000-sf at yahoo.de>:

> Hi Anša,
>
> thank you very much for you explanation.
> But if i got it right, it's not possible to switch off the initial-stages
> or to reduce their number
> by tuning the blocking times. That's a pity.
>
> I have now made a wish report as you suggested and i will try to downdate
> to 0.9.4. I had already
> done this in the last weeks by installing the old wheezy-packages in the
> jessie-environment. It
> worked for me for some weeks and then probably after updating something
> else it began to crash every
> time.
>
> Perhaps its possible to to compile the old 0.9.4 sources against the new
> environment?
>
> Greets, P
>
>
> On 13-06-2015 23:15, Anna Vernerová wrote:
> > Hi P.,
> >
> > there are seven hard-coded initial stages, after which the words go
> through
> > the usual "pots" (levels) as before. The time intervals for the initial
> > stages are all between a few minutes and 8 hours. The idea is to provide
> > extra intensive training of new words (and to replace the old "require
> > three consecutive right answers" option). Currently, the only way to
> change
> > the behaviour is to compile parley from sources (and change the intervals
> > for these initial stages before compiling). In all other aspects, the
> > initial stages behave just as usual levels, that is, if you answer a word
> > correctly, it goes onto the next initial stage (or, from the last initial
> > stage, to the usual level 2), and if you answer incorrectly, it drops all
> > the way to the beginning.
> >
> > You can find the sources here:
> > https://projects.kde.org/projects/kde/kdeedu/parley/repository
> > I am not very experienced with compiling, but here is what I do on a
> > Kubuntu system. The following procedure should work on Kubuntu 14.10 or
> > 15.04 - it does not work on 14.04 or older, in which case you would have
> to
> > clone some older version of the code from the repo (just write to me and
> I
> > will tell you which version works for me on Kubuntu 14.04).
> >
> > - uninstall parley
> > - open the konsole
> > - the following command will create a directory called "parley" in the
> > current directory, and download the source files into it:
> > git clone git://anongit.kde.org/parley
> > - alter the intervals right at the top of the file
> > src/collection/entryfilter.cpp   (the format is self-explanatory)
> > - then start Kdevelop and create a project containing the "parley"
> directory
> > - press the "build" button
> > - the file your_current_directory/parley/build/src/parley is the file
> that
> > you should use for opening parley
> >
> >
> > It would be great if you could file a "bug/wish report" at
> > https://bugs.kde.org/buglist.cgi
> >
> > Anša
> >
> > 2015-06-12 10:12 GMT+02:00 h3x32000-sf <h3x32000-sf at yahoo.de>:
> >
> >> Hello,
> >>
> >> i have a question about the parley stages:
> >> Since updating to Parley 1.1.1 under debian, once wrong answered words
> >> remain in an inital stage and
> >> do not re-enter in the vocabulary-pots anymore. Also the
> >> blocking-configuration is not working as
> >> expected for these words. It seems, there is a second level-structure
> >> called "confidence" that is
> >> counting up. So, a right answered word doesnt jump to the secound pot
> but
> >> is remaining out of the
> >> pots and the confidence level is counting up. In the top right corner in
> >> the practice window is then
> >> written something like "initial, confidence 3"
> >>
> >> How do i get rid of this?
> >>
> >> Thak you for some help, i already asked google, but without sucess...
> >>
> >> Greets, P.
> >> _______________________________________________
> >> kde-edu mailing list
> >> kde-edu at mail.kde.org
> >> https://mail.kde.org/mailman/listinfo/kde-edu
> >>
> >
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/kde-edu/attachments/20150615/f5534a2c/attachment.html>