Download: Episode 75.
This episode we talk about Amazon’s AI recruiting fail, WordPress accessibility issues, Google+ and more.
Community Event Planning pre-order. Still time to get in on the book previews!
Survey for event organizers
Issue 10 – Science! It’s shipping. Back order sale use code READER18 for buy 2, get 3rd 1/2 off!
Amazon scraps secret AI recruiting tool that showed bias against women | Reuters
I have resigned as the WordPress accessibility team lead. Here is why. – Rian Rietveld
A Plan for 5.0 – Make WordPress Core
Google is shutting down Google+ following massive data exposure
Google faces mounting pressure from Congress over Google+ privacy flaw – The Verge
Tampered Chinese Ethernet port used to hack ‘major US telecom,’ says Bloomberg report
Zotero Blog » Blog Archive » Improved PDF retrieval with Unpaywall integration
Now Broadcasting LIVE most Fridays
We broadcast our episode recordings LIVE on most Fridays at 12pm PT. Mark your calendars and visit recompilermag.live to tune-in.
We love hearing from you! Feedback, comments, questions…
We’d love hearing from you, so get in touch!
CHRISTIE: Hello and welcome to The Recompiler, a feminist hacker podcast where we talk about technology in a fun and playful way. I’m your host, Christie Koehler.
AUDREY: Hey, Christie.
CHRISTIE: How’s it going?
AUDREY: It’s Saturday.
CHRISTIE: It is Saturday. It’s 13th of October, 2018. We’re doing our live recording of Recompiler podcast Episode 75. This week, we’re talking about Amazon’s AI recruiting fail, some WordPress accessibility issues, Google+, and more. I think we got some announcements first.
AUDREY: Yes. Mostly, I think we just want to talk about the Community Event Planning book, our 2nd Edition. The pre-order shop is still open. So if you’ve been thinking about what kind of events you’re going to run in 2019, then maybe this is something that would interest you. We’re doing a survey for event organizers as part of this where we’re asking people to tell us a little bit more about what they’re working on and what’s gone well, what they’re kind of learning or trying to figure out now. And then, I’m just going to go straight through all the announcements.
CHRISTIE: I love it.
AUDREY: We also have Issue 10 – Science in the shop. It just started shipping the other week. So, that’s a new Recompiler content to check out.
CHRISTIE: Awesome. And is the backlist sale still going on?
AUDREY: Yep. We have a deal where if you’ve been thinking about picking up a couple issues, you can buy two and get the third one half off.
CHRISTIE: Is it something like READER18 for that?
CHRISTIE: Cool. All right. Awesome. So, AI (Artificial Intelligence) Machine Learning, it’s a big thing. A lot of companies are doing it. That sounded a really weird way to put it.
AUDREY: It’s very popular conceptually right now.
AUDREY: Whether or not in practice it’s doing the things that maybe people thought of.
CHRISTIE: A lot of people hiring data scientists. Amazon, for a little while, was trying to use it for recruiting or, I don’t know. I guess it’s still part of the recruiting pipeline when you’ve got resumes.
AUDREY: Resume filtering and maybe locating resumes.
CHRISTIE: How did that work out for them, Audrey?
AUDREY: Well, they’re not doing it officially anymore. They never got a good result. Not only did they never get a very useful result, but it amplified bias in the system, gender bias in the system.
CHRISTIE: So basically, they trained the models using sort of their historical recruiting data. And so they ended up creating a model with that bias encoded and amplified in it. And something I thought was pretty funny was…I mean, you know what I mean by “funny” in this context. Where is this bit in the Reuters article? Oh, it says, “Gender bias was not the only issue. Problems that the data underpinned the model’s judgments meant that unqualified candidates were often recommended for all manner of jobs, the people said. With the technology returning results almost at random, Amazon shut down the project.”
AUDREY: I suspect that that’s really what [inaudible]. They might have thought, “Oh well, the bias we can work on,” but it actually never produced a useful model.
CHRISTIE: They encoded the Peter Principle model.
AUDREY: We’d like a random dice roll in there like, “How high do you want to go?”
CHRISTIE: Yeah. I know it’s not exactly what it did but that was like my first thought and then it made me chuckle.
AUDREY: This is the hazard of thinking that AI or machine learning is going to fix everything. You can put a lot of effort into a lot of data into it and get something that, again, [inaudible] your biases and produces no useful results. Even accepting the bias, it doesn’t sound like it was producing good filtering, useful filtering.
CHRISTIE: Amazon’s sort of trying to deny they weren’t using it but also said they didn’t dispute. I can’t remember the exact wording about this. They were like, “Oh, we’re not using it to make hiring decisions.” But they also didn’t dispute the fact that they were using it to analyze certain resumes or something. And then the Reuters article goes on to sort of talk about some other examples of different machine learning applied to hiring pipeline. I found this a little disturbing. There’s this company, HireVue. It says the firm analyzes candidates’ speech and facial expressions in video interviews to reduce reliance on resumes.
AUDREY: I feel like we talked about that at some point.
CHRISTIE: Did we? Okay.
AUDREY: Go back and have a look. Either we talked about it in one of the episodes or we had put it into our kind of queue of stuff to look at. But yeah, there’s a couple of these systems that are using video interviews, like automated video interviews with some kind of a processing backend to weed out or filter through just large candidate pools.
CHRISTIE: So does that mean you submit a video interview instead of a resume?
AUDREY: Yup. There’s like an app that you need to sit down with and follow the prompts.
CHRISTIE: Do you think they’re sort of using that to replace the initial phone screen?
AUDREY: I think yeah, but I think also beyond that, the stuff is getting used in industries that would normally just have those big walk-up dates.
CHRISTIE: What’s a walk-up date?
AUDREY: I don’t know what the formal term is. I mean like there are certain kinds of jobs where they have a hiring day. And you get the application, everybody fills it out, you all wait in line, you put in your application, you talk to somebody for a couple of minutes, whatever. And so, they just have these big hiring events. I think that some of the stuff is coming out of a desire to not spend the resources on that.
CHRISTIE: Okay. It reminds me a little bit of the companies that employ that sort of personality testing or some kind of automated testing. I’ve never actually passed one of those. I don’t know what it is.
AUDREY: I’m not sure I’ve ever had to do one for an interview. About Amazon specifically, one of the things that really struck me about this is that this could be a major lawsuit. There’s a reason that they need to be very careful in saying…I mean, I hope it’s true that they didn’t actually use this in any real hiring decision. But they need to be very careful about saying that because if their application, their AI thingy was weeding out women aggressively, that’s a big hiring problem. I mean, like a legal liability problem.
CHRISTIE: It’s definitely not legal here in the States.
AUDREY: Even if hiring managers could see this information, then it could have biased their opinions. I think that that could be potentially a discrimination suit.
The Reuters piece also talks about…it says LinkedIn offers algorithmic rankings of candidates based on their fit for job postings. I’m really curious about that. So, the ACLU is talking about algorithmic transparency fairness. The article ends by saying another person said a new team in Edinburgh has been formed to give automated employment screening another try, this time with a focus on diversity.
AUDREY: I’m very curious how they’re going to do that. I think that the whole concept is flawed. Resumes can’t be standardized in a way that makes them unambiguously interpreted.
CHRISTIE: Not unless there is a standard that is defined and people start adhering to that.
AUDREY: You have to give us a series of checkboxes.
CHRISTIE: It has to be very structured.
AUDREY: As long as people are writing their own resumes, you can’t have a system that really is going to do this in equal and fair way. People are going to use different language, different terminology even talking about the same tasks. And it’s not a straightforward problem to figure out how to rank those kinds of things.
CHRISTIE: Right. And that was one of the things at the piece that said that certain language ended up being favored over others, like executed and captured.
AUDREY: Or even if you just go based on job titles. Not all companies use the same job titles. They don’t have the same scale, the same progression. People could have job titles that don’t match their actual abilities or what they’re doing. I don’t know. I kind of hate every part of the hiring process.
CHRISTIE: You mean in general?
AUDREY: Yeah. I’ve encountered some blind screening too. Blind in the sense of like the people don’t see the names on them. I think even that’s problematic because it’s sort of expecting that resumes that are generic in a way that I’m not sure that it can be, if you’re actually just telling a story about what you know how to do.
CHRISTIE: We’ve talked about this before. I don’t know that we’ve talked about it in the podcast. Whenever a conference or some processes like, “We do planned screenings as sort of under the diversity column…” I try to figure out like do we actually have any studies about that? And the one I’ve been able to find so far is related to symphony performance.
AUDREY: As far as I know, that’s where the practice comes from.
CHRISTIE: Yeah, and I think that is really hard to equate with other types of…because there, you’re able to isolate the performance from the person in a way that is much harder to do with technology work unless you’re doing a performance coding test or something which has its own issues.
AUDREY: In a symphony, the type of audition you do is actually a good representation of how you’re going to perform in the symphony.
CHRISTIE: You’re saying it’s more like what it’s actually like to be in a symphony than a white burning test or arbitrary coding test?
AUDREY: Yes, you’re going to play music that demonstrates your skills. You’re going to do in a way where people can hear your music ability. I think the only thing that you can’t do from behind the screen is show how you interact with the other musicians. But there’s a level of technical competence that you need before it’s even worth looking at that. And those screened interviews work perfectly.
CHRISTIE: Can you imagine them trying to do something like that for a different role in the symphony, like a conductor?
AUDREY: It wouldn’t happen. And I think when we’re talking about programming jobs or technical jobs, it’s a combination of our technical knowledge and expertise and our ability to work the problem and work with each other that makes us good in those roles. So there’s only a fairly small part of it that you can actually extract in this way.
CHRISTIE: Have you ever done one of those symphony type auditions?
AUDREY: Yeah, actually.
CHRISTIE: What was your experience of it?
AUDREY: It was a little frustrating but I did think that it was pretty reasonable.
CHRISTIE: The only thing…I’m trying to…
AUDREY: And I always like to have feedback in the middle, like how am I doing, how is this going. And the lack of feedback was the part that I found frustrating. But I did think that it equalized things.
CHRISTIE: So it’s pretty [backblocks]. You let go and you sit down, you play your piece.
AUDREY: Or you submit a tape.
CHRISTIE: Okay. The only thing I have that’s similar is like a concert band and I don’t remember what the term was…auditioning or playing for the chair order. I remember doing that once. But that’s not the same thing. Cool. Thanks for sharing your experience there.
I saw this post going around about WordPress which I try to keep general tabs on what’s going on in the WordPress world because I use WordPress a lot. And it’s actually possibly one of the open source projects I’ve used the longest, certainly up there with PHP and Apache web server and Firefox or something. But this one caught my attention because it’s related to a big release of WordPress that’s coming up. And so it says, “I have resigned as the WordPress accessibility team lead. Here is why.” And I knew that WordPress has been working on a new editor interface for quite some time. And I sort of had a sense that it’s looming, but haven’t totally kept up-to-date with this.
So I did a little bit of research and I think you did too, Audrey. My first question is, so Gutenberg’s the new editor. Is that the thing that’s currently on WordPress.com or no?
AUDREY: That was actually one of the things that I didn’t get around to figuring out.
CHRISTIE: Okay. We’ll put that on the still-to-research list. Because I looked at some of the screenshots and I couldn’t quite tell.
AUDREY: And I’ve been using both the older version and the newer one that’s on WordPress.com.
CHRISTIE: And in fact when I’m doing The Recompiler stuff, I start with the old version because there’s some things I like about it or there’s some things I just don’t know where to find in the other version. But the one thing I like in the newer version is it does a much better job of updating the URL especially if you’re scheduling…
AUDREY: If you change the title, you mean?
CHRISTIE: You change the title or schedule it, it will update that URL and give you…oh, it’s not just that. You can actually get the whole URL with one click.
AUDREY: Yeah, it definitely has some nice details that way. I go back to the old version of WordPress admin all the time though because there are things that I have not found the same functionality in the newer interface.
AUDREY: There are just things that aren’t there.
CHRISTIE: Right. And even on my self-hosted sites, if you go to click on ‘try the new improved editor’, it takes you to WordPress.com. And part of me is like, “I don’t want this to be a dependency I’m forced into all the time,” but I just haven’t had a chance to figure it out. Well, those are all the things we don’t know.
AUDREY: The general trend that we’re seeing is toward these very separate frontend applications. Like what we see in the new editor is that there’s a lot of frames replaced within the application and it’s not like the older PHP application would be a series of web pages basically that they go through. You close one, go to the next one, close one, go to the next one. I think we are trying to figure out what the name for this general kind of application is where contents replaced in place in the frame.
CHRISTIE: Is it just single page app or single…why am I messing up the acronym? I think it’s like single page app. Yes, single page application, SPA.
CHRISTIE: And it’s interesting because if you go to WordPress.org/Gutenberg, the preview of it, it shows it in the WordPress admin frame, the visual that I am used to seeing which is not how the editor I’m used to in WordPress.com looks like. So anyway, back to the original issue. Basically, this person who was working on the accessibility team and I couldn’t tell if this was on a paid role or in a volunteer role. It kind of sounded like volunteer but high level or like core contributor sort of level. Basically she says Gutenberg still has a long way to go in terms of accessibility. And if it’s sort of released as it is now, it’s a significant step back in terms of anyone who needs to use assistive, like a screen reader or voice interface assistive technology.
AUDREY: Keyboard-only input.
CHRISTIE: Keyboard-only input, yeah. So it’s a significant step back. And they kind of talk about what went wrong. They talk about all the things they tried to do. And one of the things they point out is that the people really knowledgeable about accessibility were not knowledgeable about the front-end framework that they chose to write Gutenberg in, which is React. And so, it made them hard to contribute towards improving accessibility.
AUDREY: And presumably, the folks that knew React really well were not spending a lot of time paying attention to the accessibility issue to begin with.
CHRISTIE: They also mentioned that it was sort of like they did a lot of work and then the accessibility folks looked at it. And then along the way too, certain things were fixed and then got reported as the editor was continued to be developed.
AUDREY: And to me, that sounds like a really hard way to do accessibility. I mean, there are kinds of decisions that you need to make up front. And if that’s not well integrated into the process, it’s just going to be like putting stickers on top, like what if we just label it a little.
CHRISTIE: It’s like security and we’ve talked about this before, how security needs to be part of the conversation from the beginning. So it says, “In hindsight, what I would have done differently is communicate more often and louder we needed a skilled Gutenberg/React developer way earlier in the process, convince the developers that keyboard testing is a must, support Andrea more when he asked for a second voice and for testing, convince the AT (I think that’s assistive technology) testers to give it a go again.” It sounds like they got frustrated with the process.
AUDREY: Yeah. I mean, I don’t know what. What are they getting out of it if they’re not seeing a good response?
CHRISTIE: Right. And if a lot of them are volunteers too, unpaid volunteers.
CHRISTIE: And then lastly, invest more effort in recruiting accessibility experts and companies to help. So yeah, why is there a timeline for this? This post is from October 3rd.
AUDREY: It sounded like they were planning to release this pretty soon.
CHRISTIE: That’s what I thought.
AUDREY: Maybe it wasn’t going to become anybody’s default.
AUDREY: It’s one way of saying we have introduced this thing to you and…
AUDREY: Templating, probably.
CHRISTIE: I’ve seen this where a lot of people…I mean, we’ve had sort of static site generators for a while now like Jekyll, I think it’s been around for a while. Jekyll is possibly what GitHub uses to power GitHub pages basically.
AUDREY: Yeah, that’s where it came from.
AUDREY: I mean, you always [inaudible] to the extent that you do in a much more old school web site.
CHRISTIE: Okay. You’re right.
AUDREY: I don’t know that it would do anything specific with the generator, but it seems like you’d be able to integrate all those same kinds of things.
AUDREY: And the browser is being designed for it.
AUDREY: I tried to do a little bit of reading on this. And this is me reading it from the outside but I had this feeling that people are having the same conversation over and over again when it comes to advocacy for accessibility and something like React because I saw fairly recent blog posts that were covering things that I’m sure came up five years ago.
CHRISTIE: And React is also not that old too. So, I think there’s something about the speed at which these things are adopted.
AUDREY: Well, five years out of 20 for web applications or something.
CHRISTIE: It’s just half the life…. [crosstalk]. Yeah.
AUDREY: I mean, for web pages in general, there is an accessibility standard to follow. I don’t think that, from what I’ve seen anyhow and I’m like super not an expert here, but I didn’t see a lot that seemed to point to a similar standard for these kinds of frontend applications.
CHRISTIE: And do you think the…because there’s a working group, web content accessibility guidelines, there’s three working groups around this. I’m curious like does there need to be specific best practices for these frontend frameworks? Is that kind of what you mean?
AUDREY: Yeah, and tooling, like they were talking about with keyboards support. Isn’t that something that the framework potentially could enable by default? With a standard set of conventions even.
CHRISTIE: And I don’t know enough nearly about React to know if it does any of that. They have an accessibility page on their docs. We should find someone who knows a lot about this and get them on the show.
AUDREY: Yeah. I do want to hear a lot more and I have no doubt that there are people working on the stuff that care a lot about it. But if it’s not [inaudible] from the start like you were saying, there’s just so much that won’t happen. It’s really easy if you’re not one of the people affected by these kinds of usability things to just not notice. You’re doing everything that you wanted to perfectly fine. It’s working fast. I have these feelings about high bandwidth demand in web apps too. It’s working for you, so you don’t look at the other thing. You don’t test at some different way that doesn’t match your working style.
CHRISTIE: Yeah. I’ve heard similar complaints about accessibility in Electron apps like Slack. I think Slack is still an Electron app.
AUDREY: And that’s GitHub’s platform.
CHRISTIE: Well, everyone uses it for basically packaging web apps as desktop apps.
AUDREY: But it came out of the GitHub Atom editor. Is that the right thing?
AUDREY: Hold on. My audio input dropped.
CHRISTIE: It sounds like it’s back.
AUDREY: Yeah. You remember how we said there was a thing with the bathroom door?
CHRISTIE: That is so weird.
AUDREY: I still don’t know what’s actually happening but it did in fact make the audio drop.
CHRISTIE: We need to get an oscilloscope and I don’t know what other diagnostic equipment, and come over and do some experiments and figure out like what…do we need to Faraday cage your office so that doesn’t happen?
AUDREY: I don’t know. It’s so weird. It makes the speakers pop. I don’t even understand why would the microphone drop. Yeah, we need some kind of testing equipment here.
AUDREY: That toaster incident isn’t the only way that my house is haunted.
CHRISTIE: No. Old houses are so fun.
When we decided we were going to talk about this, I thought, “I’m really curious about the evolution of sort of how people are designing, developing, deploying web apps.” And I’m curious to see like what are the patterns and stuff. I haven’t found a complete analysis or compendium or graphic or anything but one of the things I did remind myself of is sort of if you go back, I don’t know, five, a few years, I think some of the momentum around switching to loading stuff to the frontend and having micro-services was actually driven by the demands for mobile development and people not wanting to produce a whole app for the desktop web browser and then a whole app for the mobile browser or even the mobile phone itself.
AUDREY: At first, you couldn’t.
AUDREY: I mean, at first this is the direction that Apple was pushing everybody to go in with the iPhone.
CHRISTIE: They wanted everything to be a web app, right?
CHRISTIE: There was no App Store to begin with.
CHRISTIE: Why didn’t that work? Just because the browsers couldn’t do enough yet?
AUDREY: I think some of it’s that the frameworks and there were a lot of things that weren’t…there’s more that’s accessible on the phone from the browser than there was at the start. But there were things that you just couldn’t do like upload photos. There were things that there was no API for.
CHRISTIE: I was sort of remembering that a lot of that push towards having, “Okay, this server just gives you a big chunk of JSON,” or something. And then the app or the web browser renders that data, and does the logic and stuff. So that kind of filled in a big gap for me.
AUDREY: You mentioned just briefly the deployment aspect but that was something else that I was looking at as part of kind of a different thing that the transition from just plain HTML and CSS to the LAMP Stack required that there was a way to deploy it. So you got these Apache modules that could be set up in a variety of hosting [inaudible]. And then as we transitioned out of that, there was another round of this where for example, Rails deployment wasn’t very straightforward. People had to set up their own hosting pretty much from scratch.
CHRISTIE: It was much more like the way you might have to deploy a Java app where you have like intermediaries.
AUDREY: Yeah. So, we’ve been kind of through rounds of that too and I think that this drive toward micro services is requiring other kinds of systems. I think that the changes in the browser obviously have become their own cycle to support it. So, it’s kind of interesting seeing what the pieces on the outside of it are too.
CHRISTIE: Outside meaning like the deployment.
AUDREY: Yeah, the structures that allow the application to happen.
CHRISTIE: Right. And the maintenance. The thing that I’ve been sort of puzzling over is like I’m very familiar with the workflow and the complexity involved in like a LAMP Stack or something similar where you’re running a web server and you have a database and there’s a caching layer. And I have a certain comfort level with that. The sort of other methodology of we have our content and we’re going to run through it, it’s dynamic content that we’re changing a lot so we usually have some kind of…you might be working with Markdown or you might have a database but then you spit out some sort of static content and then push to a CDN or application delivery network. There’s all kinds of complications that come up with that too. And to me, it’s sort of like, “Well, we haven’t actually reduced the overall…there’s no way to reduce the overall amount of complication and complexity.” You’re just kind of like pushing around where it happens in trying to find the best fit for your particular organization or use case.
AUDREY: And then I start wondering about the labor angle of this too. Do we have a lot of developers who can do certain things and so we’re prioritizing putting the application development effort into certain places because of that? There’s another aspect to what is or isn’t going to work here.
CHRISTIE: I do think that this sort of explosion of front end development has had a positive feedback cycle with the amount of front end developers, like if you had told me I discontinued to be surprised just like how much everything feels like it’s written in Node these days and just how much that developer demographic has just become so so large. And so, I think those are feeding into each other.
CHRISTIE: But you are now.
CHRISTIE: Yes. I was just thinking that because I was thinking the flip side of that is that you have things like Go and Rust that are being used to write really optimized fast backend services. That’s sort of like the through the looking glass component of that.
AUDREY: Yeah. It’s interesting to get another look at it since that isn’t the kind of work that I’ve been doing. I was working further and further into the backend, the more stuff was happening in the frontend. But it means that I had less and less interaction with the stuff as it was changing.
CHRISTIE: I mean, we say you can make an argument that five years is actually a pretty long time when it comes to the web but also it’s not a long time in terms of you could take a position with a company and be there for five years doing one particular thing and a lot of this stuff evolves very quickly.
AUDREY: Yeah. And people find that quite difficult when they get back on the job market.
CHRISTIE: Another thing that can be difficult is when you have a crappy social media network that no one uses.
AUDREY: Somebody was.
CHRISTIE: I couldn’t find the exact amount of affected users for this Google+ thing. But I remember when I first saw the headline, I actually inflated like I was one position off and the number was 40-something. And I think I originally thought it was 40 million and then it turned out to be like, I don’t know, it was some level smaller because that’s just how little use Google+ has been getting. I guess back in March, Google discovered that one of their Google+ APIs had been allowing third party developers to have access to private information that was not intentional. It was not the design of the API. Do you know what prompted the disclosure now? Was someone going to leak it?
AUDREY: I don’t know if it was that or the more general congressional inquiry to privacy and data handling.
CHRISTIE: Part of the problem is that the Wall Street Journal reported on this and we have yet to figure out how to deal with the Wall Street Journal paywall. If anyone has any tips, let us know.
AUDREY: We tried the library. I thought that made sense but I don’t know.
CHRISTIE: I’m really tempted. Next time I’m near there, I might pop in and ask because for some of the stuff if you can access it from the library computer, that might be worth spending an hour or so there. Anyway, I distracted myself. So, they announced this breach. Did you read the actual…what was the name of the…strobe light or something. Project Strobe, yeah. You can tell Google hires very good PR people because part of their announcement of the breach was totally mixed in with this announcement. Project Strobe: Protecting your data and improving our third party APIs, and sunsetting consumer Google+.
AUDREY: So we’re protecting you and we’re getting rid of that thing that we didn’t protect.
AUDREY: I have a lot of confidence here and how this is being handled.
CHRISTIE: Finding 1: There are significant challenges in creating and maintaining a successful Google+ product that meet consumers’ expectations. Action 1: We are shutting down Google+ for consumers.
We’re kind of being tongue in cheek about it but I mean, I do think it’s a good business decision that if something is underperforming, don’t let it just be poorly secured and whatnot. Maybe they’re overdoing it shutting it down.
AUDREY: Probably. I think I’ve heard before that they would pull people off of these teams because they weren’t [inaudible]. Obviously, what happens then is that you have a lot of stuff that doesn’t get fixed and a lot of things that just don’t get addressed. If you’re at that point where you’re not putting resources into maintaining it, you probably should shut it down. It’s not really great or a safe thing to be running a social network that doesn’t get any support.
CHRISTIE: Something that did catch my eye was…well, two things. One is that they’re keeping it going for their enterprise clients or their non-consumer clients, which again kind of reminded me who they’re willing to put resources towards.
AUDREY: I think there’s internal use or at least there used to be.
CHRISTIE: That came up and all this stuff with all the lawsuits and stuff. There’s been in the news with Google that a lot of…it’s clear that they have internal communication. I had that thought too, “By enterprise clients, do they mean themselves?”
AUDREY: It counts, I guess.
CHRISTIE: It sure does. But that would just be kind of funny.
AUDREY: Yeah. What was interesting to me about this aside from like last week, we talked about how long is too long to wait to disclose a security vulnerability and I think six months and you had to tell Congress and you’re shutting down the product is definitely too long. Part of what I thought was interesting is that when Google+ started, their selling point was that they were going to have a better privacy model. The selling point was that they were going to do a better job of modeling the way they you share information with people. And that didn’t stay very long as a priority. But it is quite a path from there to here.
CHRISTIE: I don’t even remember that as being…I don’t think I was ever excited about Google+. I was mostly annoyed that they pushed so hard. And I read something and I don’t know if it was a Twitter thread or something else. I haven’t been able to find it again that talked about, I think it must have been a Twitter thread by someone who previously worked at Google. So maybe you’ve seen this, Audrey. But it talked about how people were pushed really hard. Product teams are pushed really hard to integrate Google+ and then it caused a lot of divisiveness.
AUDREY: We saw that from the outside, like how many different Google things we’re using where they wanted to integrate your profile.
AUDREY: If you use Google groups still, there’s weird stuff that happens when you try to manage your subscriptions because of all of these different waves of kinds of Google profiles that have existed.
CHRISTIE: Right. I don’t even know where that leaves…I haven’t really kept track of the state of that and where it leaves. If they are deactivating Google+, where does that leave that integrated profile stuff?
AUDREY: I think the YouTube suffered. They got peeled back off of it. That was where I noticed that push the strongest.
CHRISTIE: It was Google+? Sorry, for YouTube?
AUDREY: For YouTube, yeah, because there was kind of a terrifying thing that they did where at some point I logged in and there might have been a couple different accounts I was looking at. But it wanted you to make a decision right now about how you were integrating your profile. And I did not want to do that at all.
AUDREY: I remember like looking at this thing and going, “How do I make it stop? What do I do?”
CHRISTIE: Yeah, I recently experience that with the Microsoft accounts. I think I might have accidentally merged my Microsoft work account with my Xbox profile which is so the inverse or the opposite of what I would ever want to do.
AUDREY: I am not surprised with Skype. They had us all upgrade our accounts to a unified one.
AUDREY: And I typed in my last name and suddenly I’m seeing family members that I never talk to and I’m like, “There’s no way that they’re using Skype.” That doesn’t strike me as something they’d be interested in. But I thought, “Oh, but it’s like the Xbox and maybe a Microsoft office or Hotmail thing.” Suddenly this is all in the same context search field.
CHRISTIE: It was another one of those, like I was trying to get something done and I was having trouble getting things logged in and like a thing for work was using my LinkedIn which seemed more reasonable. But then things just got all scrambled and I still need to figure that out but I’m a little afraid to. I guess, get your data out of Google+ if you had some data in there you care about.
AUDREY: Photos maybe.
AUDREY: I’m not really sure what the current stage of Google’s photo sharing universe is either, but that would be the kind of thing to be concerned about.
CHRISTIE: They definitely have a photo sync because it’s integrated with Android and whatnot. We’ll let you all figure that out.
AUDREY: Good luck.
CHRISTIE: Did I put this in here? Did we want to follow up about the China spying chips?
AUDREY: I just grabbed a quick link about that because Bloomberg had another thing that they reported that sounded a little bit more possible to substantiate. And they are still holding firm on their initial stuff even though pretty much everybody who could trade or refute that Apple knew that there was a problem has done so, including various intelligence agencies.
CHRISTIE: Right. And we haven’t had any more corroborating accounts come forward, right?
AUDREY: No. When I was looking to see, aside from us who looked at the original story and said, “This doesn’t really make sense,” even one of the people that’s quoted went on the record afterward to say that looking at the article, he was no longer comfortable with his comments the way that they were included.
CHRISTIE: That’s a bad sign.
AUDREY: Yeah, because what had been the sort of theoretical ‘oh, if this happened, what this would mean’, to see that next to things that were not being demonstrated with any real proof and no named sources that could actually show anyone else what they were talking about. Even the security folks that were quoted didn’t seem to be very comfortable. But it was interesting that they had kind of a follow up thing that actually referred to a type of hardware hack that has been seen before and that had all of the telecoms going, “Not me, not me, not me.”
CHRISTIE: So, some more information but really it’s still kind of a WTF.
CHRISTIE: Okay. Are we to things we love on the internet this week?
AUDREY: I think so.
CHRISTIE: Cool. I got something. I’m on a ball this week with this. Zotero, which is, I’ve been calling it citations manager. They call it your personal research assistant. It’s a free, easy to use tool to help you collect, organize site and share research. So it’ll basically keep a database of all kinds of research materials, articles and it keeps the information about it. You can create questions, share them, create bibliography entries. There’s extensions for browsers, so you can you look in an article in your browser and save it to Zotero. It’s really handy.
So, they added a thing. It says, “We’re excited to make it easier than ever to find PDFs for the items in your library.” So I’ve been able to download PDFs automatically. They’re often behind publisher paywalls. So, they’ve added integration with Unpaywall which has a database of legal, full-text articles hosted by publishers and repositories around the world. So if you save an item in Zotero now and it can’t find or access a PDF, it will search Unpaywall for it. I just think that’s really cool. We’ll link to the article because it talks about Impactstory which is the group behind Unpaywall. And if you’re someone who does research and you’re trying to figure out how to manage all that and haven’t checked out Zotero, check that out. It is cross-platform. It’s free. I like it a lot. What have you got, Audrey?
AUDREY: Since we often talk about zines, I had just a couple of tech zine things to reference. If you haven’t looked at it before, Julia Evans does a bunch of different hand-drawn zines about various technology things. I think some of her recent stuff also talks about working with your manager, other workplace stuff that’s really useful. But she started selling some of them for download and she wrote up a blog post that talks a little bit about that. So it’s worth checking out the zines as well as some of the background about how they happen. And then there is a zine about pair programming that I backed on Kickstarter and I was hoping that maybe they would have put up like an additional Shoplink or something by now, but I didn’t see it when I looked. So I’m just going to mention you should keep your eyes out for it if you didn’t back the Kickstarter because it’s kind of a cute introduction to pair programming.
CHRISTIE: Awesome. Marlena Compton put this together. I worked with Marlena at Mozilla. So, it’s cool to see her doing this.
Okey-dokey. I think that’s our show. Thanks, Audrey, for joining me for a special weekend edition of The Recompiler.
AUDREY: That makes me think we need a little outro music too.
CHRISTIE: You mean a special outro music?
AUDREY: A special weekend edition.
CHRISTIE: Yeah. All right, I’ll get on that. All right, bye everyone.
CHRISTIE: And that’s a wrap. You’ve been listening to The Recompiler Podcast. You can find this and all previous episodes at recompilermag.com/podcast. There you’ll find links to individual episodes as well as the show notes. You’ll also find links to subscribe to The Recompiler Podcast using iTunes or your favorite podcatcher. If you’re already subscribed via iTunes, please take a moment to leave us a review. It really helps us out. Speaking of which, we love your feedback. What do you like? What do you not like? What do you want to hear more of? Let us know. You can send email feedback to firstname.lastname@example.org or send feedback via Twitter to @RecompilerMag or directly to me, @Christi3k. You can also leave us an audio comment by calling 503 489 9083 and leave in a message.
The Recompiler podcast is a project of Recompiler Media, founded and led by Audrey Eschright and is hosted and produced by yours truly, Christie Koehler. Thanks for listening.