Re: [mnemosyne-proj-users] One and Five hour Pimsleur delay

[Peter Bienstman]

Hi,

Mnemosyne uses the SM2 algorithm which does not use sub day intervals, although it shouldn't be too hard for someone to write a plugin for a Pimsleurs scheduler.

Cheers,

Peter

thro...@yahoo.com wrote:

The Pimsleur scheduled learning would have a learning session after one and then 5 hours and THEN 24 hours. Mnemosyne appears to job straight from status "1" (memorized) to "2" (24 hours).

Or maybe I'm not Mnemosyne correctly?

Thanks,
Mark

Paul Pimsleur's intervals for language learning are:

    5 seconds
    25 seconds
    2 minutes
    10 minutes
    1 hour
    5 hours
    1 day
    5 days
    25 days
    4 months
    2 years

[Vit]

   Hi, Mark.
To do  1hr and 5hr   sessions, I set a Timer .
-------
 I decided to investigate what MY personal  5 first intervals are.
Obviously, they are NOT 5sec, 25sec, 2min, 10min, 60min.
    The Browser  is a convenient vehicle to do this study.
5 sec is what it roughly takes  to recall word 1 and 2 - at top of the List;   
5" - to repeat words 1, 2;
5" - words 3, 4;  
5" - repeat 3,4.   That amounts to  about 30 sec. from the start.
    Repeat the cycle for words 5-8;    and for words 9-12.
Now you are at 2 min from start; 
review words 1-12.
Set your timer at 10 min.
 Review more new words as described.
 As soon as timer sounds, go to Main screen and do the 10min interval there.
PS. If you have more then 15 new words, tag them in the Browser ,  ahead of Review.  Then, review cards tagged "New" as ONE chunk.
PPS. I am glad that these first 5 intervals are not in algorithm in that the numbers should be customizable - which Peter, most likely, would not implement.

Good day.

[thro...@yahoo.com]

Hi Vit,

I'm not sure I understand what the advantage of the browser is.

I'm thinking of using "cram" mode for the first reviews until a better solution comes along.

I wouldn't expect any software to handle the first reviews at 5, 25, and 120 seconds -- Sometimes it takes longer to work through the material than the time limit! Its just that there is no built-in transition from 10 minutes to one day. Those reviews at 1 and 5 hours can be important. Of course, the 24 cycle probably fits most people's schedule better.

Thanks!
Mark

[George Wade]

That series of intervals for reinforcing learning seems to be related to, though not identical to samurai schooling: which may have been borrowed, or stolen, from Korea or China.

It is short to medium term in emphasis except that no mention is made of using a multiplicity of senses to learn and recall. It would be nice to have software to assemble individualised cards with all the elements to build associations: everybody could have their own style ready to pop into Mnemosyne decks. I'd have videos with editable sub-titles and titles for preference. People doing things; models working; experiments happening. Still photos and other graphics as fall back. While it seems like a lot of work the learning is that much deeper.

Presentation software seems very nearly custom made. The resulting videos can be compressed a lot.

George

[Vit]

  Hi, Mark.
 I have read a lot on memory. The verdict for cramming (if u have along term goal) is:    bad  bad  bad. 
1. Waste of time
2. Does nothing to develop medium / long term memory.
3. One needs an accelerating spacing and ONE rep per interval.
 ------------------
It's almost imposible to understand someone's process without trying it.   If u have questions - I am ready.

  For the 10 days since my post, I have a reliable picture of
what my Intervals are NOT.  I am talking about Vocabulary cards only.    Memorizing the salient points of text is another story.
   I got 100% recall for 5-10 sec; and 30 sec;  I skip them from now on { the process is for Fresh cards - not for Failed once }.
   Other change I made - I schedule and track results with pencil and  paper -  like a  Spreasheet.  So far, looks like i will drop 2-3min and use 5 min  20  60 5hr  1d 3 5 6 10 day (like in Supermemo).
  After that, cards go into Mnemosyne.
Another process is implemented in Browser {mature cards}.

Good day.

On Monday, February 11, 2013 12:26:47 PM UTC-5, thro...@yahoo.com wrote:

Hi Vit,

I'm not sure I understand what the advantage of the browser is.

I'm thinking of using "cram" mode for the first reviews until a better solution comes along.

..........

Thanks!
Mark

 

[pharmtech]

On Feb 8, 7:59 pm, throa...@yahoo.com wrote:
> The Pimsleur scheduled learning ...

Thanks for posting that. I always feel like something is missing with
SRS systems in this regard (faster reviews when initially acquiring
information). I hadn't thought of Pimsleur, but that's a good example.
A marriage of both Pimsleur and SRS would be nice, IMO. Pimsleur for
initial acquisition, SRS for intervals of 1 day or more.

Have you looked at Anki 2? It added a "learning mode" which
accommodates this. You can specify the "steps" (intervals of minutes)
a card must go through to "graduate" to normal, day-sized, SRS (SM2)
reviews. Also "steps" if you forget a card, and it enters
"relearning." I wish it was more of a seamless transition. It feels
abrupt as cards are only reviewable once per day.

[Vit]

 pharmtech wrote:

The Pimsleur scheduled learning ...

Thanks for posting that. I always feel like something is missing with
SRS systems in this regard (faster reviews when initially acquiring
information). I hadn't thought of Pimsleur, but that's a good example.
A marriage of both Pimsleur and SRS would be nice, IMO. Pimsleur for
initial acquisition, SRS for intervals of 1 day or more.

------ I do learning for 25-20 days.  Mnemo takes over after that.

---------------------------

Have you looked at Anki 2? It added a "learning mode" which
accommodates this. You can specify the "steps" (intervals of minutes)
a card must go through to "graduate" to normal, day-sized, SRS (SM2)
reviews. Also "steps" if you forget a card, and it enters
"relearning." I wish it was more of a seamless transition. It feels
abrupt as cards are only reviewable once per day.

----------I downloaded  Anki2.  Intend using it along with Mnemo

till I learn all daunting sounding options -- I read Anki forum; seems
very bad UI { if u venture beyond basic }.

  I am stuck at step-1:  cant change the Font or background. Can u help me with that?   Where do I get grey from?  WHy font doesnt change when I increase the number ?    

Thanks in advance.
Vit

[pharmtech]

On Feb 26, 11:50 am, Vit <vrem2003-tempor...@yahoo.com> wrote:
> till I learn all daunting sounding options -- I read Anki forum; seems
> very bad UI { if u venture beyond basic }.

It depends on your experience with computers, SRS systems, etc. Those
with heavy experience like it. But, new users can be bewildered.

>   I am stuck at step-1:  cant change the Font or background. Can u help me
> with that?   Where do I get grey from?  WHy font doesnt change when I
> increase the number ?

This is probably not the forum for this. You could ask on Anki's forum.
[1]

If I wanted to change the background color of a card, I would open a
deck, click "Add" at the top of the window (to add a card to the
deck), and then click "Cards" to get into the "Card Types" window
where the "template" for the front/back of the card is defined. This
is where you specify how you want the fields to appear.

(Side topic: Your card might have more than a "front" and "back"
field. You might break up a complex front/back into smaller units of
information that you could layout in the "front" and "back template".
To create those fields, you'd click "fields" instead of "cards" at the
beginning of the prior step.).

There is a "Styling" window on that screen with the default Anki
styling:

.card {
font-family: arial;
font-size: 20px;
text-align: center;
color: black;
background-color: white;
}

You can change the font-size of all the fields by changing that CSS
(stylesheet) property. Also the background color.

If you want your fields to be sized and colored differently, you'd do
this:

<span id="field1">{field1}</span>, <span id="field2">{field2}</
span>

span#field1 { font-size: 16px; background-color: blue; }
span#field2 { font-size: 28px; background-color: red; }

This is a good example of how Anki has some powerful features, giving
you a great deal of control by separating data from format and style.
But, it's also a good example of how a new user might think "I have
one field and just want it to be green. Why do I have to learn all
this HTML stuff?"

It's hard to find a good balance between abstraction/flexibility and
concreteness/simplicity. I like Mnemosyne's simplicity, but I'm not
yet sold on how plugins (all the way down to the level of defining
card types, multiple fields, etc.) will work. If I distribute a deck
based upon complex fields, layout, etc., it seems like anyone who uses
that deck will have a tedious learning curve identifying, downloading
and installing all the plugins required by that deck. If the system
treated decks like a package-manager, resolving "dependencies" for the
user, it might make sense.

If you have questions, feel free to email me or post to the Anki group
(maybe with a subject: Pharmtech). I don't think we should go much
further on Mnemosyne's forum.

[1] http://groups.google.com/group/ankisrs/topics?hl=en

[Peter Bienstman]

On 02/27/2013 05:55 AM, pharmtech wrote:
> It's hard to find a good balance between abstraction/flexibility and
> concreteness/simplicity. I like Mnemosyne's simplicity, but I'm not
> yet sold on how plugins (all the way down to the level of defining
> card types, multiple fields, etc.) will work. If I distribute a deck
> based upon complex fields, layout, etc., it seems like anyone who uses
> that deck will have a tedious learning curve identifying, downloading
> and installing all the plugins required by that deck.

As I told you a few times before in private conversation, please don't
make any assumptions like that without knowing how libmnemosyne2 is
designed. User-defined card types are stored in the database without the
need for plugins, and interoperate well with the sync and export
functionalities. Also, there is no issue whatsoever with the card browser.

If you want 'proof' of this, for cloned card types you also don't need
to distribute any plugins. User-defined N-sided card types can make use
of exactly the same machinery, it's just that the UI side of things has
not been written yet...

Peter

[George Wade]

The Chinese have a way of thinking about this: thinking is not doing — so go out and do it for a year, or five or twenty five. Thinking each evening while planning the next day's work and study. If we do well enough we may write about it and become a respected master.

I don't know if you have noticed the Chinese and Indians taking over the world's banking, industry and health. We are lucky to have Peter as a master.

George

[Peter Bienstman]

Thanks, but I think you are giving me too much credit, though :-)

Cheers,

Peter

On 02/27/2013 09:23 AM, George Wade wrote:
> The Chinese have a way of thinking about this: thinking is not doing � so go out and do it for a year, or five or twenty five. Thinking each evening while planning the next day's work and study. If we do well enough we may write about it and become a respected master.

[pharmtech]

On Feb 27, 12:29 am, Peter Bienstman <Peter.Bienst...@UGent.be> wrote:
> User-defined card types are stored in the database without the
> need for plugins, and interoperate well with the sync and export
> functionalities. Also, there is no issue whatsoever with the card browser.

Where can I find a card type plugin, to change field names, add
fields, etc?

[Peter Bienstman]

As I mentioned in the part of my mail you deleted, the UI functionality
for that still needs to be written.

You can vote for the feature here

https://mnemosyne.uservoice.com/forums/164265-general/suggestions/2901330-ui-support-for-n-sided-card-types-30-work-units-

But I'm sure you've already done so :-)

Peter

[pharmtech]

On Thursday, February 28, 2013 12:38:34 AM UTC-7, Peter Bienstman wrote:

As I mentioned in the part of my mail you deleted, the UI functionality
for that still needs to be written.

Somewhere I got the impression card types (the ability to define field names) were to be added via a plugin. I think I came to that understanding from the manual:

Mnemosyne comes with a few plugins for different card types, like Cloze deletion (to selectively hide one or more words in a sentence), Map (specifically tailored to study blank and marked maps) or Sentence (a cross between Vocabulary and Cloze to help you study languages using sentences). Have a look at the plugin description for more info on how to use them.

I'm not deliberately ignoring what you've said. I'm just having trouble putting it together between what exists now or is a planned feature. When you said the UI wasn't ready for something, I thought you were talking about automatically distributing a card-type plugin. I thought, between that comment and the manual above, it was still possible to implement a card type via a plugin.

[Peter Bienstman]

On 02/28/2013 06:24 PM, pharmtech wrote:
> I'm not deliberately ignoring what you've said. I'm just having trouble
> putting it together between what exists now or is a planned feature.
> When you said the UI wasn't ready for something, I thought you were
> talking about automatically distributing a card-type plugin. I thought,
> between that comment and the manual above, it was still possible to
> implement a card type via a plugin.

Ah, I see the confusion. Cloze, Map, etc are built-in system card types
distributed with every install of Mnemosyne. They are in a plugin
because many new users don't care about them and activating them by
default would clutter the UI needlessly.

User-defined card types like cloned types now and N-sided card types in
the future, are stored in the database, and not in a separate plugin
python script that needs to be distributed. The GUI component to create
and edit these N-sided card types will be in plugin, however, as this is
advanced functionality.

Hope this clears things up.

Peter