Wanted: improvements to Interlisp bibliography

[masinter]

The bibliography we have is of mixed quality:

• There are bad links, bad metadata, and few annotations
• There are many documents in our docs and MEDLEY-PDFs and ENVOS-PDFs that haven't been cataloged
• etc

I was hoping that the web site bibliography could be clustered by (half?) decade, to separate (late 60s) (Early 70s) (Late 70s) ... in order to correlate with Teitelman's History of Interlisp and the Interlisp/history repository.

The long term goal is for our bibliography to be on par with: the Interlisp Family bibliography maintained by the Software Preservation Group.

---

https://scholar.archive.org/search?q=interlisp&filter_type=everything has a lot of material that mentions Interlisp and Interlisp-D, and the History and Timeline documents have some bibliographic references. From looking at this, there are various categories; from papers and articles ABOUT Interlisp, to citations that merely mention Interlisp in passing (the latter not so interesting.

I'd like to collect metadata and online copies (if available) of Interlisp Reference Manuals (many of them I donated to CHM several years ago.

[masinter]

@AbeJellinek wrote https://github.com/Interlisp/Interlisp.github.io/pull/3 I don’t think it’ll run from a file:// URL because it makes AJAX requests, but you can test it out by running:

$ cd /path/to/Interlisp.github.io/
$ python -m SimpleHTTPServer 8080  

or (python 3)

$ python -m http.server 8080

Then visit localhost:8080/bibliography/ in your browser.

Part of the way down the page, there’s one item with an annotation added so you can see what that looks like. We obviously need to clean up the collection a bit (some funny OCR artifacts in there), but it’s looking good.

[masinter]

was hoping for something where what was visible was very simple. For example, “year” in bold, “title” as a hyperlink, “author” (if important to say), “genre” (e.g. “book”, “article”, “publication”, with a link/icon to show (in a hover/popup).

Going to Zotero every time….. results should be cached?

[nbriggs]

How do we fix the citations? If the current style is supposed to be IEEE, can we choose something else that gives better formatting?

[AbeJellinek]

Yes! Pretty easily. There's an editor to create the XML style definitions that the citation generator uses. I can create one following that spec.

[nbriggs]

I mean the source of the citations -- in Zotero? A number of entries have garbage (e.g., bogus authors, OCR errors, ...)

[masinter]

The citations were picked up from a scan of different sources for the word “Interlisp” – scholar.archive.org (which says it is beta with poor quality metadata, but more full-text PDFs), scholar.google.com (more records), dl.acm.org (I have a lifetime membership), patents.google.com, plus, for good measure, a few of the documents we’ve collected. I was just getting an idea of what we might do with Zotero’s current capabilities. To get an idea of what is there, you should get Zotero desktop.

There are a bunch of people on Fiverr who offer to gather, clean up Zotero references. The going rate seems to be about $0.50 / reference.
https://www.fiverr.com/search/gigs?query=zotero so maybe $100? Or sticking to sites that QA their citations ? This is all exploratory.

[masinter]

I found someone on Fiverr who has a software background to edit the Zotero bibliography. $15 for 45 entries. Marked with the tag “sz” or “s2z”. (sana_zafar).

Zotero doesn’t seem to have a built in facility to convert from subcollections to tags and vice versa. I made subcollections “About Lisp” “Common Lisp” “Interlisp” “Interlisp-D” “Applications written in Interlisp” “Mention Interlisp in passing” “Irrelevant”

And

“Unsorted” – things that belong in one of the previous categories.

I was imagining the web site would show some but not all of these categories, sorted by Decade, within a Decade by Item Type (or some categories of item types) and dates

Anyway, please review https://www.zotero.org/groups/2914042/interlisp/library At some point soon we should take this as a baseline, close this issue, and open any additional work as separate issues marked "Documentation".

[masinter]

@pmcjones note

[masinter]

Status of Interlisp bibliography: @pmcjones put together a wonderful annotated bibliography which tells a story. http://www.softwarepreservation.org/projects/LISP/interlisp_family#Interlisp-D_ and quotes my recollection with some facts I got wrong. I had made a couple of attempts at putting together a more detailed "History" and "Timeline" before shifting toward a strategy of linking citations and notes, and using that to generate the https://interlisp.org/bibliography . I see now this is a research project. But I think it's still worth trying.

The bibliography we have in Zotero has a lot of entries somewhat at random: I did searches for "Interlisp" in https://scholar.google.com https://scholar.archive.org https://dl.acm.org. I also paid someone on Fiverr.com to "fix" the references we have by completing the metadata including abstracts, URLs and other fields. They spent more than a week on the project.

@hjellinek -- it's a complicated "digital library" story.

[hjellinek]

We need a precise definition of what it means to resolve (complete, close) this issue. Given that no bibliography will ever be finished, I propose to engage a bibliographer/Zotero expert, probably on Fiverr, with carefully-worded instructions to take the current 'sz' Zotero entries, compare them with Paul McJones's annotated bibliography, and (1) flesh out the items tagged with 'sz' with the information from Paul's bibliography they do not include; (2) add the items from Paul's bibliography that are not already present in Zotero.

We'll name the result something that clearly identifies the source.

[masinter]

There were other sources for the entries that sz worked on. I spent several hours looking for "Interlisp" in scholar.google.com and scholar.archive.org. As well as patents.google.com and the links in the two Google docs linked from the History section of Interlisp.org. It's fine to start small, but don't start with the 'sz' tagged items. Start with Paul's document; look for citations we have; tag them with a tag that indicates which sub-collection they were in, and move them to a new "checked" sub-collection. We can publish that on our web site.

For citations from Paul's that we don't already have (I don't expect any) I'd say add them to a "New" subcolletion.

[pmcjones]

Herb,

I spent some time going over one of the subcategories of Larry’s Zotero bibliography, and filled in a few missing dates and so on. I must say Larry’s bibliography is probably much more comprehensive than mine: I was focused mostly on papers and documents by the Interlisp group about the core system, so I missed a lot of the application stuff, which is very important.

Larry,

In your message of August 23, you said:

@pmcjones https://github.com/pmcjones put together a wonderful annotated bibliography which tells a story. http://www.softwarepreservation.org/projects/LISP/interlisp_family#Interlisp-D http://www.softwarepreservation.org/projects/LISP/interlisp_family#Interlisp-D_ and quotes my recollection with some facts I got wrong.

If there are incorrect facts in http://www.softwarepreservation.org/projects/LISP/interlisp_family/, I’d like to correct them.

Paul

On Sep 13, 2021, at 4:55 PM, Herb Jellinek @.***> wrote:

We need a precise definition of what it means to resolve (complete, close) this issue. Given that no bibliography will ever be finished, I propose to engage a bibliographer/Zotero expert, probably on Fiverr, with carefully-worded instructions to take the current 'sz' Zotero entries, compare them with Paul McJones's annotated bibliography, and (1) flesh out the items tagged with 'sz' with the information from Paul's bibliography they do not include; (2) add the items from Paul's bibliography that are not already present in Zotero.

We'll name the result something that clearly identifies the source.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/Interlisp/Interlisp.github.io/issues/23, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAHARYQLQGDZQJDJN6OZE33UB2FO7ANCNFSM42FG6YOA. Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

[hjellinek]

Let me rephrase what I wrote above:

I will start (the Fiverr person) with Paul's document; look for citations we have, paying particular attention to the sz ones, which may need to be updated or enlarged; and move the result to a new collection/library.

I'm not clear on what you mean by tagging them by "what sub-collection they were in." Also, libraries/collections can be moved under other collections/libraries at any time, so we're free to design a hierarchy we like at any time. Or so I'm told....

[masinter]

@pmcjones re "facts I got wrong" let's just say that in the note to you that you quoted, there are things that I said with more certainty than they deserved. It was so common a failing at PARC that I can hear Jerry Elkind ask "Do you know, or are you just guessing?" (or something like that, I don't know for sure).

[masinter]

@hjellinek

I'm not clear on what you mean by tagging them by "what sub-collection they were in

you can see the rough categorization "Common Lisp" "See Also for" etc, what I was trying to do with sub-collections would probably be done better with tags.

[pmcjones]

Re:

I'd like to collect metadata and online copies (if available) of Interlisp Reference Manuals (many of them I donated to CHM several years ago.

see:

Larry Masinter Lisp materials donated to Computer History Museum, Lot X6058.2011 Cataloged by Paul McJones February 6, 2011 (updated April 8, 2011) https://mcjones.org/masinter/

On Aug 30, 2022, at 9:03 PM, Larry Masinter @.***> wrote:

https://scholar.archive.org/search?q=interlisp&filter_type=everything https://scholar.archive.org/search?q=interlisp&filter_type=everything has a lot of material that mentions Interlisp and Interlisp-D, and the History and Timeline documents have some bibliographic references. From looking at this, there are various categories; from papers and articles ABOUT Interlisp, to citations that merely mention Interlisp in passing (the latter not so interesting.

I'd like to collect metadata and online copies (if available) of Interlisp Reference Manuals (many of them I donated to CHM several years ago.

— Reply to this email directly, view it on GitHub https://github.com/Interlisp/interlisp.github.io/issues/17, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAHARYR5ENBDCVWQZGK7Z2LV33KRTANCNFSM6AAAAAAQA7Z2ZY. You are receiving this because you were mentioned.

[stumbo]

@masinter @hjellinek Are there any specific tasks left in this issue? My quick read says the goal of this issue ended up being to update the Zotero bibliography.

Shall we close this issue? If there are specific tasks that still need to be done, open a new issue with them explicitly stated?

[masinter]

I think this is an issue that I talked about with @ekaltman as a possible project for students of computer history. The bibliography we have is of mixed quality -- there are bad links, bad metadata, and hardly any annotations. I'd hired someone on Fiverr to work on it, and the result was better than they'd started with, but far from ideal.

There are many documents in our docs and MEDLEY-PDFs and ENVOS-PDFs that haven't been cataloged. And so on... I don't know what we need to do to reach the standards of, say, https://www.softwarepreservation.org/projects/LISP/interlisp_family#Interlisp-D_ . Getting to that level would be a lot of work, but not infinite.

I was hoping that the web site bibliography could be clustered by (half?) decade, to separate (late 60s) (Early 70s) (Late 70s) ... in order to correlate with Teitelman's History of Interlisp and the Interlisp/history repository, and perhaps link to demos of features organized by year of introduction.

There are some other technical gaps -- can we serve our own documents on interlisp.org rather than from zotero?

I'm not sure when it's optimal to close the Main iddurand leave residuals, but I suspect we're still at "forest" level with this one and not quite ready for "trees".

[masinter]

In a bibliography, the terms "tags" and "keywords" are often used interchangeably, but there is a subtle difference in how they are used in various bibliographic management tools and platforms.

Keywords:

Purpose: Primarily used for searching and filtering references within a bibliographic database. Format: Typically a list of words or phrases separated by commas or semicolons. Source: Often extracted from the reference's metadata (e.g., title, abstract) or added manually by the user. Visibility: Usually not displayed directly in the bibliography entry but used behind the scenes for organization.

Tags:

Purpose: Similar to keywords, used for organization and retrieval, but often more flexible and user-defined. Format: Can be single words, phrases, or even hierarchical structures (e.g., "machine learning/neural networks"). Source: Primarily added manually by the user to categorize and label references according to their own needs. Visibility: May be displayed alongside the bibliography entry in some tools, providing additional context.

Examples:

Keywords: artificial intelligence, machine learning, deep learning Tags: AI research, ML algorithms, neural networks, project X

Key Differences:

Flexibility: Tags are generally more flexible than keywords, allowing for personalized categorization and hierarchical structures. Visibility: Tags are sometimes displayed with the bibliography entry, while keywords are usually hidden. Source: Keywords are often extracted from the reference data, while tags are primarily user-defined.

In Practice:

Bibliographic Management Software (e.g., Zotero, Mendeley): Often allows using both keywords (sometimes called "subject headings") and tags, offering the benefits of both. Reference Managers (e.g., EndNote): May primarily use keywords for organization and search. Online Platforms (e.g., CiteDrive): Might use the term "tags" even though they function more like traditional keywords.

Ultimately, the choice of whether to use tags, keywords, or both depends on the specific tool you are using and your personal preferences for organizing and managing your bibliography.