pretext
pretext copied to clipboard
PreTeXtBook
WeBWorK: make valid inside PROJECT-LIKE
This makes it so you can have (introduction?|webwork|conclusion?) in a PROJECT-LIKE. PROJECT-LIKE is similar enough to inline exercise that this did not take much work.
Motivated by @mitchkeller 's request to turn the Active Calculus preview activities into WeBWorK exercises.
Examples in the sample chapter?
The final line of current commit looks like our recent time sinks? List each PROJECT-LIKE in the match?
Examples in the sample chapter?
Will do.
The final line of current commit looks like our recent time sinks? List each PROJECT-LIKE in the match?
What do you think about a PROJECT-ANCESTOR-FILTER among the entities? Listing them all out makes me wonder if the list changes, will it be updated in all the right places.
There was another place I saw that same construction already in use while I was searching for instances of PROJECT-FILTER. I could remove that at the same time? Or should I keep it separate?
Examples in the sample chapter?
Done on a second commit.
The final line of current commit looks like our recent time sinks?
For posterity: we chatted offline and this should not be so much of a time sink. In -html, it is within a variable selection. In -latex, within a match but for a modal template.
@Alex-Jordan notes that the (serial) number of a PROJECT-LIKE and the number of am "inline exercise" can be identical, which is a bit of a train wreck in the problem set formulation. Projects can be "on their own counter" or not, so sometimes a problem, sometimes not. At any rate, this PR is paused.
Thanks, Alex.
- Made an issue for the serial number limitation of WW:
https://github.com/PreTeXtBook/pretext/issues/1859
-
Should we announce on
pretext-announce? [NEW]? Briefly mention numbers in problem sets? (You or me, your call.) -
Using
@copyin the tests is great. But to have our only example that way makes me nervous. Such as, the change in the problem sets is negligible. (Doing a cut/paste copy of some other problem would be better.) -
The
match="*[...]construction we discussed is not so bad. (A) runtime seems unaffected. (B) it is a modal template so only gets fed a limited number of elements anyway. It is wildcard, filter, all elements that is dangerous. I missed that yesterday. But I silently changed it anyway, and added aPROJECT-LIKEnearby. -
String parameter
webwork.project.static, and its cousins, will be deprecated very soon. Weeks, not days, not months. Your call if you want to update documentation now or wait.
String parameter webwork.project.static, and its cousins, will be deprecated very soon. Weeks, not days, not months.
Not true! Doing it now.
Midterm exams and Thanksgiving-related stuff had my attention when this was in the final stages.
My guess is that with this merged, it's OK to have a webwork in a PROJECT-LIKE, but the archiving process ignores such webwork. So for now, no standalone .pg file, and no set definition file for all the PROJECT-LIKE. Does that sound right? If so I might try to get archiving working, then do the announcement.
Is this a guess, or did you try? My careful reading of the code just now is that they won't be missed. Can you do a check? Will want such a thing in the sample chapter in any event, if it is not there already.
But if you want three set definitions (in exercises, inline exercise, project-like) then some adjustments are needed.
It's a guess, I haven't tried anything yet. But in pretext-ww-problem-sets.pl, line 81, it has <xsl:apply-templates select="$document-root//exercise/webwork-reps" />. That makes me thing that webwork inside a project-like will be missed for the .pg file creation.
Meanwhile it looks to me that there will be two kinds of set definition files built. One for each exercises and one with all the rest, including those in a project-like. (Is that what you were seeing?) But it's no good if the .pg file doesn't exist. And besides, I think there should be one set def for exercises, one for each worksheet, one for all of the section's inline exercises, and one for all of the section's project-like. And I guess one for all of the section's reading-exercises. So it's also missing that aspect and lumping all the latter into one def file.
On my radar. Just 1.5 more weeks of teaching to get through...
Well, yes, I glossed over that one. Since we place the webwork-reps we know just where they are, and thus, the exercise step can just be scrubbed. Rather than a complicated version with all the PROJECT-LIKE.
There's just a bit more though to this. Currently, if there is no title to a webwork, the pg file comes out like this: 4_2_4.pg. That's inline exercise number 4 from section 4.2. So if section 4.2 also has project-like number 4, there would be a clash. So something should be done. I'm not sure if we want to make files like Project-4_2_4.pg or put things in a folder like Project/4_2_4.pg or something else entirely.
There is a PreTeXt-wide notion of a "default title". That could be impelmented for WW with no author-provided title.
There is a PreTeXt-wide notion of a "default title".
On second, thought, no. It'd bleed through into too many places. Let's keep anything like this in the dedicated stylesheet. A utility template could go into pretext-common.xsl since we just need text for a filename.
Do we think Active Calculus thinks they can put a WW in a task in a project?
I think we could live without that. WW has enough structuring options that just using the WW as a child of project should be enough without wanting a WW inside project/task. (I think the only part of Active Calculus currently using task is the vector calculus chapter, but long term, I do want to migrate toward task and away from lists.)
Using @copy in the tests is great. But to have our only example that way makes me nervous. Such as, the change in the problem sets is negligible. (Doing a cut/paste copy of some other problem would be better.)
Now using the sample chapter to test completing this addition. We need another example that does not use @copy, and as such it might as well be something other than project.