Zamansky: GitHub Classroom 2

Posted on 2019-02-18 by jcs

Mike Zamansky has another video out on how he uses GitHub Classroom for teaching. Like the first, it’s not really about Emacs—although he does use Emacs in the process, of course—but it’s an interesting study on a tools-based approach to otherwise boring or rote tasks.

The use of GitHub Classroom makes things a bit easier but the same ideas would work with any setup that has a server with something like Git installed on it. Here, for example, is the method John Kitchin uses to accomplish pretty much the same thing. Still, as Zamansky says, GitHub Classroom abstracts away a lot of the administrative tasks and you don’t have to worry about setting up your own server.

If you’re a teacher, you should definitely take a look at the video but even if you’re not, it’s an interesting use case of leveraging everyday tools—shell, sed, make, echo, and so on—to get a job done as easily as possible. You may not be a teacher but if you’re reading Irreal you probably often have occasion to process a collection of related files and that’s what Zamansky’s video is really about.

As a final note, I’ve been using Sed for decades but didn’t know that s/** Name: // was legal. It turns out the special regex operators that don’t occur after other text are not treated specially. So I learned something new and that’s yet another reason to spend the 16 minutes 43 seconds on the video.

Posted in General | Tagged , | Leave a comment

Copying the Whole Buffer

Posted on 2019-02-17 by jcs

Tim Bray tweeted a nice bit of elisp to copy the current buffer to the kill ring:

The normal method—or at least my normal method—is to mark the buffer with Ctrl+x h and then copy it to the kill ring with Meta+w. That’s pretty simple and it’s burned into my muscle memory but I do that operation a lot so it probably makes sense to get it bound to a single key sequence. In any event, if you often find yourself wanting to capture the buffer so you can paste it somewhere else, Bray’s code is just what you need.

My only quibble is binding it to Meta+a. By default, that’s bound to either backward-sentence or org-backward-sentence, which I use all the time. If you use or think you might use the go to beginning and end of sentence functions, you should probably bind Bray’s code to something else.

Posted in General | Tagged | Leave a comment

Selling 911 Location Data

Posted on 2019-02-16 by jcs

Speaking of John Cook’s Data Privacy twitter account, as I did yesterday, here’s an interesting item:

It turns out that the carriers are selling 911 location data. This isn’t just the cell tower data that I wrote about before, but actual GPS data that is meant to be used only in support of the 911 emergency system. In the previous post I naively and reluctantly suggested that the government should get involved and make that behavior illegal. The thing is, using the 911 location data for anything other than 911 (or other things required by law) is already illegal. Ars Technica has a detailed article on the matter that notes the selling of this data is against federal law.

The FCC can and should levy fines on every carrier found guilty of doing this but the fines need to discourage the behavior. Something like 10 times the income from the practice would be appropriate. Otherwise, the fines are just written off as a cost of doing business. If I make \$10 from some illegal activity and you fine me \$1 for doing it, guess what’s going to happen. If the fines don’t do the trick, taking senior management on a perp walk might. It probably wouldn’t take more than one.

Posted in General | Tagged | Leave a comment

New Data Privacy Tip Twitter Account from John D Cook

Posted on 2019-02-15 by jcs

I’ve long been a fan of John Cook’s TeX Tips and often learn something new from them. Cook has a number of such Twitter accounts covering various aspects of Mathematics and other technical areas. I was, therefore, happy to see that he’s added a new account for data privacy tips:

It’s already seeded with several tips and worth taking a look at.

Since digital privacy is one of my hobby horses—as most of you know to your dismay—I was delighted to see this addition. Anything that helps raise awareness of the issues has my welcome. Drop by and take a look and perhaps follow the account if you’re on Twitter or bookmark the site if you aren’t.

UPDATE [2019-02-15 Fri 14:51]: and → as.

Posted in General | Tagged | Leave a comment

Securing Your iPhone

Posted on 2019-02-14 by jcs

This is an Apple-centric post but I know a lot of Irreal readers have iPhones or iPads so it’s worth passing on. Over at Computerworld, Jonny Evans has an excellent article on How to stay as private as possible on Apple’s iPad and iPhone.

Regular readers know I’m very paranoid conscious about security matters so much of Evans’ advice was already familiar but I still learned a few new things so the article is definitely worth reading if you have an iPhone or iPad. It turns out that Apple provides a lot of help in locking things down. It was mostly this information that I didn’t know.

If nothing else, going through your settings as you read the article will ensure that you haven’t missed something. It’s astounding how many apps want to be able to access your location, microphone, or camera without any real reason to do so and it’s worth reviewing the apps that have that access and turning it off if the app has no reason to have it.

I’m sure Android has many of the same features but, sadly, this article won’t help Android users; it specific to Apple devices. On the other hand, if you are an Apple user you should, as I said, definitely give it a read.

Posted in General | Tagged , | Leave a comment

Sharing Data Among Org-mode Blocks

Posted on 2019-02-13 by jcs

John Kitchin has another excellent post on leveraging Emacs and Org-mode. This time, he considers sharing data between Org-mode code blocks. Most Org-mode users know about naming code blocks and using the results of the computation in another code block but it turns out that the facilities for data sharing are much richer than that.

You can, for example, arrange that the results of a code block are not recalculated no matter how many times it’s called. Or if the block generates a huge amount of data or binary data, you can save it to a file and access the file from another block.

Kitchin covers these and other techniques. As he says, which technique you should use depends on the situation and what your data looks like. He lays out ways of sharing the data among blocks and leaves it to the users to pick the one most appropriate for their use case.

As is often the case with posts involving Org files, I found it easier to follow Kitchin’s post by reading actual Org source file. Happily, Kitchin always provides a link to those files so if you also like to see the original file, he makes it easy.

This is a great post and I recommend everyone using Org code blocks read it.

Posted in General | Tagged , | Leave a comment

Unix and Linux History from Maddog

Posted on 2019-02-12 by jcs

Older readers may know of Jon “Maddog” Hall. Although we hear less from him these days, he’s been a force in the open source movement for many years. I just came across a video in which he talks about Unix and Linux history. As Hall says, 2019 is a year with a lot of anniversaries for our community including 50 years of Unix and 25 years of Linux.

Much of the talk represents his personal memories of events and he tells some stories I hadn’t heard before. For example, it was Hall’s personal intervention and efforts that provided Linus with an Alpha machine for the Linux Alpha port. In those days, Alphas cost about $30,000 so this was a significant achievement.

There are other interesting stories too. The video is just short of 56 minutes so you’ll need to plan ahead. The history he recounts is part of our shared heritage and hearing those stories is worth your time.

Posted in General | Tagged , | Leave a comment

NeoTree and Perspectives

Posted on 2019-02-11 by jcs

A lot of people like to organize their workflow around projects, mostly via Bozhidar Batsov’s excellent projectile package. Some of those people go further and use a package like perspective to provide each project with its own workspace. I don’t organize my workflow that way but it’s an attractive option that I occasionally consider adopting.

Another group of people swear by packages like neotree that provide a file tree off to the side of their Emacs windows. I definitely don’t want to sacrifice screen real estate for that but there’s a large population of users who disagree.

Calum MacRae is in both camps and has neotree set up to track his current perspective. That works well most of the time but when he switches between two existing perspectives, neotree does not update to the new perspective. Of course, it’s Emacs so MacRae wrote a bit of Elisp to fix the problem. If you like the perspective idea and want a file tree rooted at the top directory of each project, you should definitely take a look at MacRae’s post and grab his code.

Posted in General | Tagged | Leave a comment

Org Capture and Refiling

Posted on 2019-02-10 by jcs

Like me, Matus Goljer (fuco1) is a heavy user of Org-capture. I’ve got 11 templates that cover all sorts of events that I want to capture. Goljer has a similar number of templates but his taxonomy is much richer than mine. I’m generally content to file the note at the end of one of my agenda files and the templates take care of that automatically. Goljer has more target files and headings within those files that are possible targets for the notes. As a hypothetical example, consider a TODO note. It may be associated with one of his projects and within that project it may be concerned with some task and within that task, some subtask and so on. That makes putting the note in the proper place difficult and, as Goljer says, it’s not always obvious at capture time what the proper place is.

To solve that problem, Goljer’s templates dump many of his captures into a holding file that he can come back to later to refile the notes. The problem is that having so many possible targets—Goljer says the number can grow into the tens of thousands—makes refiling very slow.

As I’ve written before, Goljer is an expert at identifying and eliminating friction in his workflow. He eliminates this particular friction by temporarily restricting the number of targets. He’s got a clever scheme for doing that and ties everything together with a hydra that makes refiling entries in the holding file easy. Take a look at his post for the details. If you have a complicated taxonomy like Goljer’s, you may want to adopt or adapt his system.

Posted in General | Tagged , | Leave a comment

Using Keyboard Macros to Ease Drudgework

Posted on 2019-02-09 by jcs

I usually use keyboard macros to repeat some editing operation that’s too complicated for iedit, multiple cursors, or search and replace. It’s astounding how useful it can be and every Emacser should learn to use them.

Recently, Arne Babenhauserheide tweeted a link to an old post of his on how to use macros to ease drudgework. The boring chore he wants to streamline is (basically) to update a LaTeX table with a column having an identifier for each entry by searching for the all instances of the identifier in a config file and using the information found to update the table. The basic strategy is to move down the table row-by-row, use that row’s identifier as the target for an occur in the config file, and copy the necessary information from the occur results into the table.

Babenhauserheide automates a lot of the process with a keyboard macro. What’s significant about the macro is that it’s being used to help with a search rather than to edit some text and that the macro “visits” multiple windows. That last aspect is reminiscent of abo-abo’s use of a similar strategy in one of his refactoring processes.

Babenhauserheide’s post serves as a reminder that keyboard macros have more power than we usually take advantage of.

Posted in General | Tagged | Leave a comment