Documentation and analysis of failures to predict in projects ranks in importance with building an accurate predictive model.

I enjoyed reading about the innovative and uncompromising Father Busa.

This link is not working any more. I could not find the code on http://www.matthewjockers.net/

Only noting a typo in case Dr. Graham is looking for these. The content is clear.

Great open notebook sample for a DH project.

Broken links for me. (Not unexpected, content keeps changing) As an aside, this is a good example of rot in digital sources and perhaps a need for digital historians to take into account version controls in the raw information we process and also the idea that having a more reliable digital source of history from a point in time means backing it up along with information that is linked to it.

I would like to see the data, at least samples of it to get the structure of what went into the visualization tool. (Maybe it is linked from somewhere here.) This goes back to the positive of open notebook DH.

I recommend the short blog at this link, it led to http://app.rawgraphs.io/ It looks like a neat tool for visualization. It`s used in the diagrams above.

I was wondering if library(rJava) would be library('rJava'). They both work in R studio though.

When I did this a second time (git pull origin master), it said Everything up-to-date.

For me (Windows 10) git pull -u origin master gave me error error: unknown switch `u'. git pull origin master worked.

This is the transcriber's computer name? - Follow up: No, I think this is the article's author Ryan Cordell.

Caution in how I will cite digitized sources is noted. As items are digitized, are corrections made? New notes added?

The game for genealogists is to learn more about their family. I think the game for many local historians is to learn more about a place.

What if bringing the crowd together is the first step? The historical resources of a locality that are available digitally are probably the tip of the iceberg compared to what exists on paper: photos, school records, deeds, wills, contracts, cemetery diagrams, diaries or family bibles to name some examples. Also oral history. Knowing where things happened. This is probably more true in an area that only recently received high speed internet. The crowd when brought together also brings their knowledge of paper or oral resources that likely few other people have and perhaps no one on the Internet yet.

This could also serve to bring the creators of the content together. Crowd sourcing brings together a crowd. Just like most Wikipedia articles are improved when there is more than one contributor, crowdsourced local history could move ahead by bringing knowledgeable people together for synergy.

This link no longer seems to work.

We still use a mainframe where I work for a healthy amount of the transaction processing and data storage we do. As governments and companies "move to the cloud" workplaces are moving back to accessing more centralized computers, often in huge data centres run by large companies such as Microsoft. Sort of more mainframe than ever.

Monopoly or creator's rights? If I was talented enough to produce music or write a book I would want to get royalties from the work, not have copies of my work downloaded for free (unless I chose to do that). I like music and books, to get new music and books musicians and authors need to make a living to produce new works. Royalties are part of that living and worth protecting.

I ran this, and then saw I was getting a lot of files so I cancelled the run. I did not want to pollute DHBox (my output.txt was 340 mb) So I refined my search using Shawville as the city. I kept getting huge downloads, regardless of the time period. I also used a couple other towns. "Why am I getting all of this for New Germany, Nova Scotia?" I thought. I got better results when I deleted the working files before running $ ./ canadiana.txt I added this to the top of the program: # clean up from the previous session rm results.txt rm cleanlist.txt rm urlstograb.txt

I think the double quotation marks "" is a typo, should be one ".

I'll use the page number from the viewer (example: page 19) instead of the page number from the document (page 32 or 33) in this example.

In GitHub, when I viewed subfolder "tei-hist3907" I had to click on repository module3-wranglingdata to download it all as a zip. (Not sure if there is a slicker way to just get only the subfolder of the repository)

Even if content does exist, does it matter if it can not be found?

This is a very honest way to write and would take courage too, especially as a professional with a reputation at stake.

I tried exporting as pdf from Dillinger.io, but it displayed an error. When I exported html it worked. Did you have this issue? Follow up: When I did this (twice) I was at a coffee shop using wifi. I just tried this at home (still using wifi) and it works quickly. I can't explain the original error. Dillinger.io seemed to time out.

If you are looking for a collaborator I am too.<br> (Please see replies, I misread the instructions, no live collaborator needed.)

I started 4.6 the next day after 4.5 and logged back on to DHBox. When I did a cd .. I was put into the /home directory. So I did a cd jeffblackadar to get back to my working directory (I don't know the real name for this) Suggest doing a pwd first to check where you are, then if you are in the repository directory do a cd .. to get to the directory above it. If you are working straight through, you can ignore this.

For clarity I think this "branch" matches "branchname" in the previous command. It's clear in the example a couple lines below, "experiment" is the new branch name. Just mentioning per a double-take I had.

After I type git log, the cursor returns for me. For whatever reason I don't need to type q to get out of the git log. Maybe because my log is short right now?

I created a repository in Gitbub. Then I was asked to do a "Quick setup — if you’ve done this kind of thing before." Apparently I need to run some commands to set up necessary files like a readme.md. This is different than when I made my first repository. I am wondering if I am seeing this because I already made a repository and the first one gets set up with all of the necessary files (like readme.md) out of the box. I am not sure what to do, I expect I will learn about the commands Github expects me to run in exercise 4. I mention this because I can`t really proceed to the next part of this exercise until I solve this.

markdown is straight forward to compose in, but I admit I am struggling with why it is an advantageous way to compose documents. It's easier for me to open a simple word processor and write, with the draft version of the document looking more pleasing to the eye than text with symbols. The .md format is nicely human readable and more future proof than Mictosoft's .docx format (or other proprietary formats), but a user could choose to save files in .md. A disadvantage of the format is the loss of meta-data that may be useful to future historians and present users and computers. A lot of information can be organized in tags and used adding richness to the document beyond the text.

This is a very useful paper. Minor comment about .csv and .tsv. Commas and tabs can be problematic as delimiters for data meant to be structured on columns when a field may contain a comma or a stray tab. Choose a delimiter that is not as risk of being in the data already.

This paragraph is an excellent summary of the benefits and obligations of sharing data: more interdisciplinary use (contribution to the body of knowledge), reduction of redundancy in the collection of data (re-use it, don't re-collect it), transparency (how were conclusions formed, repeat-ability of results, what work did funds pay for?)

I suggest to watch the report form URL. If it is https://via.hypothes.is/https://goo.gl/forms/V201a2LUgiADg7GD3. I think that the "https://via.hypothes.is/" part of the URL confuses ReCAPTCHA, because ReCAPTCHA would not work unless I removed this part of the URL and reloaded the page. If you came to this page directly you may not be affected by this.

Test annotation in Klein and Gold's Digital Humanities: The Expanded Field

I`m wondering if the "technical" aspect of DH gets linked with the type of hostile exclusivity that happens to women in some university engineering departments (per anecdotes related to me.) Seeing my daughter study programming in high school has shown her education in that course to be male dominated, I am not sure why this problem persists. I have worked with a lot of programmers who are women over the last 20 years or so (and hired or promoted some of them), but I have not yet been on an IT team that had gender parity, which is where we should be.

Computers can be expensive and that expense represents a barrier. I`m going to try to access DHBox from my low cost Raspberry Pi computer and see how that goes. Some cloud computing providers like Microsoft Azure and IBM Bluemix offer a free tier of service (with limits), Maybe others. Having access to something like DHBox lowers barriers too.

It's a danger only if the computer is over relied on. We do the same thing when we consult paper archives - we research certain questions the archives can help with, but don't expect to find out the kinds of things in an archive that oral history can tell us.

I pasted the code into the page, but it was just displayed as text. In Wordpress I clicked the Text tab of the page editor and then highlighted the stuff I added and tagged it as "code" with the code button. It was marked up with <code> so I believe it's correct now. Follow up: That worked. Follow up 2: I was getting the "code" format of fixed width courier type font text leaking from my list of annotations over to the menu on the right side. An ugly hack of adding a second close </code> code tag at the end "fixed" the formatting. See example (<code>[hypothesis user = 'jeffblackadar']</code></code>) I don't like it but I thought I'd mention it.

I agree, this is a neat tool.

Will this change how we research and write history? I seem to see historians working as individual specialists with perhaps a small group of researchers for support. Will history be written by larger inter-disciplinary teams who can deal with these huge amounts of available information?

I agree with Baker suppressing these excellent maps from his book. History is about making an argument based on evidence. If the maps are not directly relevant to telling the history they do not really belong in the book, good editing would leave them out. The maps are important, and as @sarahmcole points out they are useful in a blog post like this, or a supplement to a book for those historians who want to delve into the methods the author used in research. Future researchers can also build on the author's work, repeat the experiments and perhaps get more conclusive outcomes. So, in fact it's important that work like this be published somewhere.

What biases does DH introduce to history? I don't have a good answer to this yet, other than to see DH as an important method of inquiry, but not to over-rely on it to draw a conclusion.