Copy of email to Vik Rajagopal

MJ. Smith

[quote]

I am very pleased with the progress being made on bug removal in the software, with the direction of the software implied by the betas in the web app, and the hints that the direction has changed in development from "fit is quickly into the current structure" to "make it fit the users' needs". These changes were overdue and are much appreciated. However, what is not visible is similar efforts on the resource/data side of the picture. Here, a number of design and implementation "bugs" make the software unnecessarily difficult for an average rural user to use. Yes, the use of Logos for religious education in sprawling rural parishes where travel into church is a planned hour-or-more trip is my measure of success.

Issues:

1. When a single book is divided into multiple parts to accommodate Bible text and languages, it is nearly impossible to find the parts to pull together in order to read the book as was originally intended. Logos needs to provide a view of the book that represents the original, hiding the multi-part Logos implementation.

2. Logos also needs to package a book, leader's guide, participant's manual, reading plan, and workflow into a visible unit. For example, John Climacus' The Ladder of Divine Ascent should be associated with its traditional Lenten reading plan. Bryan Chapell's Christ-Centered Preparation should be tied to his book teaching it. Tying the various elements together introduces the user to the Logos features that give value beyond an e-reader and may serve as a minor marketing tool for those elements which are separate services. I would suggest that Logos have a mechanism by which users may provide study guides, reading plans, workflows ... for consideration to become Logos products.

3. Logos needs to help the user find the older books with British spellings - just a few basics like or/our (flavor/flavour) and er/re (theater/theatre) being included as "match all forms" often exposes several additional resources. Expecting the user to build a search for "favor OR favour" is unrealistic.

4. Logos needs to make the data consistent. Currently in order to access all the sermons on the seven last words of Christ, I must (1) use the sermon label to identify one group. (2) search resource type:sermon to identify the next group, and (3) use a separate search to find sermons without labels in books outside type:sermon e.g. type:"collected works". In short, incomplete tagging make sermons, lectures, personal letters, journals ... very difficult to access. For sermons, the user can build their missing labels to alleviate the situation. This should be true for all labels and Logos should consider ways to harvest these labels, review them, and make them official. As an aside, a label for essays within a book that parallels articles within a journal is needed.

5. Book types are inconsistent partially because of an incomplete effort to move books out of the monograph type. This makes it difficult to build targeted searches resulting in results that are overly broad and time-consuming to scan. The effort needs to be completed and definitions of the types provided to the user.

6. The data for many of the labels is best understood if given in a table/facet format so that one can see patterns in the whole rather than just individual values. That is why I value the interactive tables. Making the table/facet function generic for all the "all the altars/battles/meal etc. in the Bible" and the interactive would give the sense of familiarity that users have from Bible handbooks and make the data more accessible than just the label search or detail information. I would also like to see a way for teacher/professors to provide Logos with additional tables to incorporate in Logos e.g. all the canticles of the Bible, all the type-scenes of the Bible, . . .

7. Similarly, much of the tagging that spans more than one word is underutilized and not understood in part because the span of the tag is not clear. For example, how am I to learn what the Longacre genre means if I can't easily see the span of text it covers? And while the unit of text studied is pericope-paragraph-sentence-clause-phrase-word, the clause data is so isolated from the rest of the information that it gets little use even when wants to use it. Instead, users pretend morphology=grammatical role=semantic role ... i.e. the tool leads to misuse of the data. 

I could easily double this list, but this covers many of the big issues with an emphasis on those that can be done in the current environment. I will also post this in the forums to see what feedback others have.

Sincerely,

MJ Smith (MVP)