creativecommons / termination-of-transfer Goto Github PK

Changed a fail to a flag on the basis of legal review (pre-78 creation should no longer fail)

Additionally, added flag to "don't know" answer to post-78.

See proposed decision tree for details.

This might be a good project for clinical assistance. I'll edit this post, adding needed entries as they arise.

1. Copyright Notice
2. Copyright Registration
3. Special order or commission (currently links to work for hire entry)

And logo.

"Next" button is clickable, but clicking does nothing (doesn't advance the tool).

If the user does not specify that the work was registered, don't ask for the copyright registration number.

This should be in the rule logic for the previous question.

The description issued at the end of the questionnaire should also reappear in the PDF results ("Based on the information provided . . .")

Can't click "next" at that point.

This was only added as a comment to an earlier issue—sorry about that.

Right now, we always ask whether a work was registered (question 1.c). Technically, we only need this information for pre-78 works. It might be desirable to do more robust public domain testing on 89-and-earlier works. But it won't be relevant to works from '90 onward.

Only need to ask for 1989 and earlier.

James Grimmelmann had two very similar cases resulting in failure at this point:

The first:

Also loops back to the start. I haven’t been able to get to the details of the work on any set of inputs; it either freezes or restarts no matter what I enter.

When was the work created? 1950
Has the work been published? yes
When was the work first published? 1950
Works from 1989 and earlier usually display a copyright notice. Did the work have a copyright notice? yes
Has the work been registered with the United State Copyright Office? yes
When was the work registered with the United States Copyright Office? (Leave blank if unknown) 1950
What is the date of the agreement or transfer?
If the agreement predates the work’s creation, please enter the creation year.

The second:

When was the work created? 1950
Has the work been published? yes
When was the work first published? 1951
Works from 1989 and earlier usually display a copyright notice. Did the work have a copyright notice? yes
Has the work been registered with the United State Copyright Office? yes
When was the work registered with the United States Copyright Office? (Leave blank if unknown) 1951

I'm playing around trying to isolate what's wrong, because similar entries work fine for different years (e.g., clustered around 1928 or 1960)

Present output (both blurb and PDF) doesn't go through the availability logic.

That is, a termination right —

• may have existed and expired
• may presently exist
• may not yet exist, but be expected in the future

This is the " [[is presently/was/will be]]" language in conclusion key A. Should use the tense appropriate given the relationship between the present time and the termination window.

Problem example pictured below (in the case of a future right)
Related issue to be created regarding output of window information.

For the input pictured here. Google Chrome on Mac OS X El Capitan v 10.11.2.

By both CC (peterspdx) and Authors Alliance (mike)

Due May 17

As per Grimmelman's UX suggestion. See suggested change in proposed decision tree.

On advice of counsel, let's not force an answer that makes the result overly specific to a given case.

(decision tree step 1.b.i.)

Should instead for a year on or after creation.

I think the current tool decision tree treats "no" and "don't know" the same, but we don't want people who don't know to be turned back.

We have the main pages, but need to come up with a complete inventory of the pages that are linked to from the tool that provide additional info about legal concepts and similar in preparation for our legal review and transfer of pages and content to Authors Alliance.

With 2002 as date input, tool won't go forward from 1.f.i.

At present, it reveals a page a code, not a document with the responses and preliminary determination of termination rights as expected.

• Ability to step backwards after the tool completes is a must (currently, users must start over)
• Do not frontload metadata entry (it's only relevant if tools completes as a positive)
• Frontload date questions — the present approach is duplicative (it asks both regarding work for hire and regarding general eligibility)
• Don't automatically redirect to the PDF on completion—generate a link to it instead.

The screen froze with this search:

When was the work created? 1992
Has the work been published? yes
When was the work first published? 1993
Has the work been registered with the United State Copyright Office? no
What is the date of the agreement or transfer?
If the agreement predates the work’s creation, please enter the creation year. 1995
Did the agreement or transfer include the right of publication? yes

After providing a summary of the answers, at this point: “Both notice window and copyright status look good, let’s get some more details!”, both the “back” and “Next” buttons did not seem to function at all. (Note that the searcher purposefully had the transfer date be after the publication date — just to see what the tool would do with that type of situation).
.

This is the session where we'll meet virtually to review code and decision tree changes in depth: please block half of the day starting at noon pacific. May 17th.

See Decision Tree document.

To be conducted by Authors Alliance by end of January 2016; secondary review by CC US.

The flag looks looks good here, and the windows look good, but it says up top that notification is possible—contradicting the flag that says we're outside the notice window.

We could just scrap the big bold conclusion, but that text is duplicated on the questionnaire page so it would be best to have it reflect the circumstances if possible.

1. Under Termination of Transfer Information Sheet at top, should read:
   Creative Commons and Authors Alliance
2. Just under that, in bold and italicized (can be smaller text), move up from the first paragraph under the division line:
   This is a beta information sheet produced for testing purposes only.
3. Under the top line, revise first sentence to read as follows: This is an information sheet generated by the Termination of Transfer Tool currently being developed by Creative Commons and Authors Alliance.
4. Change the link to the FAQ at the top from the old beta to the new beta that is hosted at: http://labs.creativecommons.org/demos/termination-of-transfer/

• Include a discrete list of all possible outcomes

Checking code base against annotated decision tree 0.1

Due May 17

Let's be sure to include at least the:
• notice window (beginning and end)
• termination window (beginning and end)
(these are the most important things)

But I think it would be also good to have (if applicable):
• the publication date
• the registration date
• the creation date

Particularly relevant for $ 304(d).

Unfortunately, the don't know breaks the result, by causing possible false negatives. This could be fixed in the long run, but best answer for the time being is to scrap the possibility of answering "don't know."

Just need to add a short paragraph above the tool itself on the questionnaire page. Rob, if I can get you some text, could that be quickly added?

We have two scenarios where there might be two possible termination windows:

Under § 304 we can have the normal termination or the $d_term.

Under § 203, if we don't know whether publication rights were transferred, the term can be a regular term or a $p_term.

Presently, neither $d_terms (and their notice periods) nor $p_terms (and their notices periods) seem to be generating int he PDF output.

If the user selects the "Back" button, changes the information they have input, then goes forward they may change the flags that should be applied. But we do not remove flags as we go back. This may lead to incorrect flags in the input.

Change the back button behaviour to strip specified flags, or simply remove the back button if this would be too complex in the available time (if we would have to reference count flags for example).

The analysis in 1.f of the decision tree should only apply to transfers from '78 or later (begins with this question about the right of publication).

"next" button is available, but clicking does nothing.

A range of dates entered into "where was the work created" results in no progression (Chrome/Firefox/Safari on Mac OS X)

Information gathering (decision tree section 3) is unnecessary where failure is known.

I think this might be handy for users, but is also invaluable for our own testing purposes.

The inputs above should also trigger flags:
B.1 (triggered at stage 1.a.i.1.b of the decision tree) (no copyright notice)
C.1 (triggered at stage 2.e.i of the decision tree) (possible renegotiations)

(these flags also do not appear when the engine is run solely to trigger them individually)

In the proposed decision tree. Two changes:

First: step 1.e (§ 304 analysis) — added test / flag for possible expiration (If $cright_year < 1950 : flag B.iv (Possible expired copyright—possibly public domain);)

Next: copyright term expiration (post-78 works + § 303 works only — i.e., "life of the author" terms) — added test as new 2.b.

The back button logic is broken, it doesn't work for the second question.