Our Manifold Journey

This resource was developed to document our journey as we published our first textbook on Manifold. But before we get to our notes and discoveries, here is a little background about our project.

Background

Our team of four members from the University of Hawai‘i Community College System’s OER group was selected to participate in the OEN Manifold pilot. In participating in the pilot, our team had three main goals:

Publish content that was developed and housed in our closed Sakai LMS system as an open textbook on a platform that would allow broad access.
Evaluate Manifold to see if it would meet the needs of our faculty. (Faculty point of view)
Evaluate the level of support/training Manifold would require. (Support point of view)

Considerations

Source content

The content for our textbook was published within our LMS, which required copying out into a format that could be uploaded into Manifold.
The content was distributed over six modules.
Our content included text, lists, tables, images, and embedded videos.

Authoring

The content was a collaborative project with two authors and several contributors.
The content may be reviewed and updated annually.

Publishing

Texts can be uploaded into Manifold in in one of the following formats::

Reflowable EPUBs
HTML
Markdown
Microsoft Word Documents of the DOCX variety
Google Docs
Manifest Ingest (a single Text from a collection of independent files and assets, and a YAML manifest file)

With very limited support staffing, it is likely authors will be responsible for the creation, editing, and possibly the uploading of content into Manifold.

Process

After discussion, the team decided to try out all of the upload options listed above except for Markdown files. We felt most of our faculty members are not familiar with markdown files, which could make the creation of OER less appealing, while requiring more support and resources than we are able to provide.

For each of our trials, we checked our source files to ensure our content (formatting, tables, images, etc.) met accessibility criteria, and replaced embedded videos with hyperlinks since embedded videos were not supported in Manifold. Additionally, in most of our trials, only a portion of our content was uploaded (usually just the introduction and/or chapter 1).

Trial #1: Google Doc Upload

Believing Google Documents would be the easiest for our faculty to collaborate and co-author content, we started out by copying the content out of our LMS and into Google Docs, creating one document (chapter) for each of the six modules. We also created an introduction document.

Pros:

Customized header formatting maintained.
Image placements and sizes maintained.
Customized hyperlink formatting maintained.
Table formatting maintained.

Cons:

Unwanted line spacing added.
List formatting altered and hanging indents not applied.

Takeaway:

This method has potential, but the line spacing and list formatting issues were visually displeasing and could be a “deal breaker” with faculty.

Trial #2: MS Word Upload (converted Google Doc)

Wanting to allow faculty to continue to author in Google Doc, our next attempt was to take the Google Docs from Trial #1 and download them as MS Word (.docx) files which were then uploaded into Manifold.

Pros:

Line spacing was generally acceptable.

Cons:

Customized header formatting was altered. Customized colors were not applied; headers appeared in black..
Image placements seemed to default to the center-aligned placement of the first image, which altered the intended layout of the images.
Images that were sized to fit the layout in the Word Docs were not maintained. It appears the original image sizes and not the edited sizes were on display in Manifold.
List formatting such as hanging indents applied, but the lists were aligned outside of the margin and not indented.
Table formatting was not maintained.
Customized hyperlink formatting was not maintained. Customized colors were not maintained and links appeared in Manifold’s default green.

Takeaway:

There were too many layout issues with this method to be considered.

Trial #3: MS Word Upload (converted Google Doc, formatting stripped & reapplied)

Theorizing that some of the issues could have been caused when the formatting of the Google Doc was converted to MS Word, we tried stripping the formatting from the Word file from Trial #1, and then reapplying the formatting manually in Word before uploading the file into Manifold.

Pros:

Line spacing was generally acceptable.
Customized hyperlink formatting maintained.

Cons:

Customized header formatting such as colors were not maintained and headers.
Image placements seemed to default to the center-aligned placement of the first image, which altered the intended layout of the images.
Images that were sized to fit the layout in the Word Docs were not maintained. It appears the original image sizes and not the edited sizes were on display in Manifold.
Table formatting not maintained.
List formatting such as (hanging indents applied, but the lists were aligned outside of the margin and not indented).

Takeaway:

The results of this trial generally mirrored Trial #2. There were too many issues with this method to be considered.

Trial #4: MS Word Upload (created from scratch)

We made one more attempt uploading a MS Word document. This time we started with a blank Word document and recreated the content from our LMS using a “paste as plain text” process.

Pros:

Customized header formatting maintained.
Line spacing was generally acceptable.
List formatting (indented, hanging indents applied).

Cons:

Images placements defaulted to left-alignment, contrary to original layout.).
Images that were sized to fit the layout in the Word Docs were not maintained. It appears the original image sizes and not the edited sizes were on display in Manifold.
Table formatting not maintained.
Customized header formatting (such as customized colors were not maintained applied and headers appeared in black).

Takeaway:

This method has potential, but image issues were visually displeasing and could be a “deal breaker” with faculty. It could also be unappealing to faculty to collaborate using MS Word (as compared to working in Google Docs).

Trial #5: EPUB Upload (converted Google Doc)

We next looked at uploading an EPUB file. We took the Google Docs from Trial #1 and downloaded it as an EPUB file which was then uploaded into Manifold.

Pros:

Customized header formatting maintained.
Customized hyperlink formatting maintained.
Image sizes and placements maintained.

Cons:

Unwanted line spacing added (extra spacing added).
List formatting indent and hanging indents applied, but bullets/numbering not applied.
Table widths not maintained.

Takeaway:

The lack of bullets/numbers in lists potentially creates accessibility issues, making this method not a consideration.

Trial #6: HTML Upload (converted Google Doc)

We next looked at uploading an HTMLfile. We took the Google Docs from Trial #1 and downloaded it as an HTML file which was then uploaded into Manifold.

Pros:

Customized header formatting maintained.
List formatting was maintained with indents and hanging indents.
Image sizes and placements maintained.
Customized hyperlink formatting maintained.

Cons:

Text formatting was inconsistent, such as font size suddenly changing mid-sentence or mid-paragraph.
Unwanted line spacing added (extra spacing added).
Table widths not maintained.

Takeaway:

This method has potential, but the text formatting, table widths, and line spacing issues were visually displeasing and could be “deal breakers” with faculty.

Trial #7: HTML Upload (reformatted Google Doc)

The results of Trial #6 looked promising, so we decided to try it again, but this time fixed some of the formatting in the Google Doc before downloading it as an HTML file. We found instances where the text size was inconsistent, deleted it and manually reentered it. We also removed any customized line spacing and only allowed spacing defined by heading styles. Once the Google document was reformatted, it was downloaded again as an HTML file which was then uploaded into Manifold.

Pros:

Customized header formatting maintained.
Line spacing is generally acceptable.
List formatting (indented, hanging indents applied).
Image sizes and placements maintained.
Table formatting improved.
Customized hyperlink formatting maintained.

Cons:

Unexplainable text formatting inconsistencies improved over Trial #6 but some occurrences remain (font size suddenly changes mid-sentence or mid-paragraph).

Takeaway:

While this method has potential, faculty might find the process of reformatting the Google Doc confusing or frustrating as some of the actions seem counterintuitive. For example, to increase the width of the table, the table in the Google Doc needed to be stretched beyond the width of the page. Also, by removing the additional line spacing after paragraphs, the resulting content became visually unappealing.

Trial #8: Manifest of HTML files

Next, we looked at trying a Manifest Ingestion. Using this method creates a product that presents a single, cohesive text that is partitioned into smaller, more digestible sections. Rather than requiring viewers to go in and out of six separate texts, they could click on the “Next” button at the end of one section to begin the following section from the top of the page. That makes for faster load times and allows readers to better understand where they are in the context of a larger whole. In this first of two manifest trials, we used the HTML files from Trial #7.

Pros:

Customized header formatting maintained.
Line spacing is generally acceptable.
List formatting (indented, hanging indents applied).
Image sizes and placements maintained.
Table formatting improved.
Customized hyperlink formatting maintained.

Cons:

Unexplainable text formatting inconsistencies improved over Trial #6 but some occurrences remain (font size suddenly changes mid-sentence or mid-paragraph).

Takeaway:

This method successfully created the desired “Next” button and improved navigation but required the same reformatting and had the same issues as Trial #7. However, and more importantly, this process required considerable knowledge and skills to work with the HTML files. Image files needed to be renamed and their references in the HTML edited in order to avoid errors caused by “bundling” the multiple HTML files. For this reason, this method is not a consideration.

Trial #9: Manifest of Reformatted Google Doc files

Lastly, we tried a Manifest Ingestion using the reformatted Google Doc created in Trial #7 (before downloading them as HTML files).

Pros:

Customized header formatting maintained.
Line spacing is generally acceptable.
Image sizes and placements maintained.
Table formatting improved.
Customized hyperlink formatting maintained.

Cons:

List formatting was indented, but hanging indents not applied.

Takeaway:

This method has potential. It successfully created the desired “Next” button. Using the Google Docs required the reformatting done in Trial #7 which faculty could find confusing or frustrating as some of the actions seem counterintuitive. However, it did not require working the HTML files as in Trial #8.

Conclusion

In the end, we landed on the Manifest Ingest using the reformatted Google Docs (Trial #9). We felt this provided the best balance of what we wanted in terms of navigation and aesthetics, ease of use, and support needs. We anticipate the need to create a few templates and instructional guides, but we are hopeful that this method will be appealing to faculty and sustainable from the support standpoint.

Introduction and Overview