Finally, my hope for an ADN messaging client has been realized. ThreadOne, created by developer Aaron Vegh, has hit the App Store and it is exactly what I have wanted in an ADN chat client. While Kiwi was functional1, it struggled with presenting private messaging because of its relatively slow refresh times and mechanics more suited to a stream-based message list. I basically wanted a version of Whisper on my Mac and ThreadOne has so far delivered on that wish.
It is a focused effort for a first release but Vegh has plans for further development to add features down the road. 2 I’ve been using it and have noticed a few things that ThreadOne does better than any of the ADN messaging alternatives I’ve used on the Mac so far:
So far, ThreadOne does exactly what I needed it for. More features would be nice (font size and choices would be my highest priority) but it is doing the job I bought it to do and shows what a good developer can do with the well-designed, fast and capable API ADN has put together.
There have been a lot of critical things said about the latest version of Dark Sky since its last release. Some of it may be warranted (and I’ll get to that at the end of the piece) but some have been focused so much on some elements of the app’s design that they put the actual daily usefulness of the app aside to make their points.
From the Dark Sky website:
Dark Sky uses state-of-the-art technology to predict when it will rain or snow — down to the minute — at your exact location, and presents it to you alongside the most beautiful weather visualizations you’ve ever seen.
I have long sung the praises of Dark Sky’s ability to predict bad weather with uncanny accuracy. It’s not always correct, but I understand that weather is an unexact science. Sometimes it predicts rain, like right now, when I’m actually getting snow so heavy I can’t see across the street. The inflection point of a temperature change along a front is understandably a hard thing to predict with 100% accuracy. I think we all understand that as reasonable people.
So that being said, as reasonable people, we would know that looking at a chart, no matter how many labeled axes and tickmarks, no matter how many data points which show a temperature trend inching upwards or downwards twelve hours from now, it is really just an approximation of what the actual weather might be. We understand, as reasonable people, that the weather might be something quite different when that time arrives 12 hours from now. It turns out that it wasn’t 14F, but 18F or even -4F (no thank you).
Yet I have read many words over the last week criticising the lack of exactingly-labeled graph axes. The line on the charts have tickmarks labeled with a time and a temperature when there is an inflection point. It’s that simple. Sometimes it doesn’t change at that exact datapoint because the line is a smooth curve, not a jagged chart. Given that, the savvy designer will know that the line can’t change quickly when there are two temperature inflection points close to one another but yet that is the criticism I’m hearing.
The other knocks are that there are two stacked representations of data on the “Next 24 Hours” screen. One is labeled “The Sky” and shows what his happening in the sky over the next 24 hours and has an x-axis for the next 24 hours starting from now. That makes intuitive sense.
The chart below is labeled “Temperature” and has an approximation of temperature across the same 24 hour period. Looking up at the chart you can interpolate temperature to time but you can intuit that 12 hours from now is halfway.
“How can I see what the temperature is fourteen and a half hours from now, smart guy?”, they bellow.
To them I say, “You are missing the point. This is a weather app and even if there was an exact 24 hour chart with a 100x zoom to the minute, the temperature will still be approximate 14 hours from now. That is the point of these charts and that is fairly well conveyed to me when I look at them.”
I’d go so far as to say that if it does bother you, this probably isn’t the weather app for you.
There are also some pretty smart people criticising the lack of labels on each page to denote what they represent. I’m with Dr. Drang on this one when he dissects Jared Sinclair’s complaints about the app.
I can only assume he thinks “casual users” are idiots. It’s true that having a heading on only one of the screens is inconsistent, but I don’t see how anyone could mistake what these three screens are for.
If you could make a color on a chart representing rain, what color would it be? A watery blue, right? That makes sense. How about snow? Maybe an icier blue? Check. Clouds should be gray and fog should probably be a darker gray. Check and check. Those all make sense intuitively and those are the exact colors used on charts when those weather events are taking place.
The complaints about colors (or lack thereof) appear to me to be uncompromisingly harsh given what the app is describing. You could have criticised the last version for being too dark.
There is nothing but a pulsing circle when you open the app.1 The pulsing continues until it gets enough information to display and then the screen updates with up-to-the-minute information.
Many designers have said that it is bad design to display just a pulsing circle. I am not exactly sure what they feel should be displayed while data is being fetched. If it was cached data, as some suggest, it clearly wouldn’t be up-to-the-minute data. The app needs to connect to a server to get its information. That’s just the way it is and if it takes a while, that’s a function of a lot of things that aren’t entirely in the app’s control.
I fail to see where a spinning globe or a dancing pickle would alleviate your pain. Just wait a second and you’ll get your information. If not, check back in a few minutes. Also, it’s the actual weather so maybe poke your head outside and take a look.
After spending some time with the new app, there are some things that could use some help. The radar view always is zoomed out to view the entire eastern seaboard for me. I am not quite sure why but they should fix that. It should be about a 35 mile radius around my current location.
The feed from alerts.weather.gov depicting winter warning news is showing an ugly non-mobile-formatted webpage right now. It’s a minor thing but it would be nice if it was easier to read. Knowing that I’m about to be hosed by awful weather is easier to take when I don’t have to zoom in on a webpage that looks like something best viewed in Netscape Navigator.
I get that people hate change. Having a healthy critique of a beloved app that underwent a pretty major makeover is healthy however much of the comments I have been reading over the last few days has seemed like design buffs trying to one-up their peers on which pixel ended up in the wrong place. It’s the “gotcha” school of collective design.
Sorry if this seems a bit critical. I was ignoring this whole echospheric event until I caught up on the ADN comments this morning and felt like I needed to clear the air a bit. I’m also cranky and miserable that the clouds above (as seen as a pretty rainbow of reds and oranges in Dark Sky) are dumping another 8-10” on snow on my car outside right now.
That has been a big criticism I’ve seen as well. I like that its a circle but again I’m in the minority. Those of you who hate the circle and don’t know why it is a circle, well, I don’t know what to tell you. Would a square be better? A smiling raindrop? A cute puppy? ↩
Over a year ago, I started a journey to eliminate the seemingly-endless filing of paper. Bills, bank statements, car service documents, insurance documentation, licenses – the list goes on and on. The implementation of my paperless system was thought through for months and when I finally pulled the trigger on the methodology and the gear, I was confident in it working pretty well.
Now that more than a year has gone by, I thought it would be a good idea to do a reassessment and determine if I had hit the mark with my system or not.
The Scansnap has worked well.
The only problem I had with it was the scanning software. About three months ago or so, the makers released a software update that changed how the scanning software worked which made it considerably more annoying to use. In the older release of the software, you could Scan to Folder and the files would be whisked away to their destination (in my case the Action folder) without any intervention. For a system like mine, this was perfect.
The change made it so that, post-scan, a dialog would pop up asking for a file name to replace the default. The cursor was “conveniently” placed on the default file name expecting you to change it. This sounds good in thoery. In practice, when you scan files that’s really all it is convenient to do because, post-scanning, the software pops up a window demanding your attention. If you happen to be typing anything else on your keyboard you erase the file name highlighted in the textbox leaving you no choice but to type a new one. Maybe it was the software maker’s attempt to make me less focused on multitasking or maybe it was their funny plan to create lot of unidentifiable files named “fuck.pdf” and “fuck1.pdf” in my Action folder. In reality, it just made me annoyed whenever I used their software. An option would have been nice.
I’m happy to say that I’ve found a way to use it to my advantage but more on that later.
I will shamefully admit that sometimes I get too busy to scan on a regular basis. Papers pile up next to the dormant scanner and, when I start scanning, I can sometimes end up with dozens of documents in my Action folder awaiting filing. When my OCR-based Hazel rules fail me, it makes for a lot of work on my end to figure out what each file is and how to consistently rename it. This involved “touching” each file several times.
All of a sudden that annoying pop-up form didn’t seem so annoying anymore. I needed a better solution.
For months after creating the system, each document I scanned would generate OCR text that could be used to determine the file’s contents. I had Hazel dig through that content for keywords and then use that data to generate new file names and move them to their target locations. In practice, this often didn’t work perfectly and I’d be left with a lot of files requiring inspection, renaming and moving by hand.
When trying to name files by hand, however, I had the same problem I often have when tagging – I forget my exact taxonomy and end up creating slight differentiations in the tags making the system breakdown. In order to make things a bit more consistent, I reduced the number of keywords I was using and also reduced the number of Hazel file renaming rules (to zero).
Currently I am using a method that requires touching each file only once. It employs TextExpander, some modified Hazel rules and takes advantage of that damn pop-up screen.
First, I hit the glowing blue button on the Scansnap. When the file is done scanning, the software runs OCR 1 and then the previously-annoying pop-up is displayed.
20%fill:year%-%fill:month% - %fill:whosefile% %fill:subject%%|%key:enter%
The keywords are simple things like “bank”, “house”, “car”, “lifeinsurance” and “carinsurance”. They are simple, repeatable and easy to remember. Combined with a name like “honda”, “jeff”, etc. it is pretty easy to create a Hazel rule to act on each file properly.
Hazel takes over at this point by executing rules based on file name keywords. It is less sexy than scanning OCR’ed text but it is reliable and consistent.
The downside of this method is that I need to be concentrating solely on the task of scanning and it can be a test of patience when waiting for a large file to be processed by the scanning software. That said, since I am skipping far more time-consuming steps (of previewing, memorizing, renaming and moving) down the road, the job of processing my papers much takes less time overall.
Hiccups aside, implementing the paperless system was one of the best time and money investments I’ve made in the last few years. I have used the ability to seach my documents from anywhere in the world dozens of times over the last sixteen months. When we were getting our documentation together to buy a house recently, what would have taken hours of searching seventeen months ago took almost no time at all. Several times during the process I received a panicked phone call from the loan company requesting a specific document they had forgot to mention. Rather than drive home from wherever I happened to be to find the document, I logged into my home machine from my iPad and send them the file within minutes.
This type of system takes some time and effort to get working but the dividends are pretty incredible when you get it set up.
The OCR’ed text is still useful for Spotlight searches. ↩
The trigger is “;pf” which stands for “Paperless Fill-in”. ↩
I had to put a few tweaks like the “%key:enter%” in there to speed things up. I also use Shift-Tab > Space to quickly select the “Save” button on the scanning software dialog pop-up. ↩
The other day, Gabe over at Macdrifter started a series of articles that will get to the root of how he manages his tasks. There are few topics I have thought more about than task management since I started this website. It is integral to how I do my job and, once GTD demonstrated its ability to solve my difficult work problems, I naturally started applying what I learned to every aspect of my life.
While Gabe’s take may seem, upon first read, to be overly ambitious taking a from-the-rivets-up approach like his is really the only way to approach it. I have undertaken a few attempts to replace OmniFocus in the last couple of years, partially sharing Gabe’s frustrations but also just to peer deeper at what makes my process tick. What I’ve found during those forays into madness is that it turns over every stone you have based your organizational life on. It sounds extreme but the whole point of a GTD system is that it become ingrained, automatic and habitual – it has to be those things if it is to work. Changing the thing that is the foundational layer of your system is disruptive and time-consuming.
I’ve tried other systems but they have never felt as satisfying as OmniFocus. It was the first task management system that I felt “got” me. It did things the way I would expect them to happen, so it made it very easy for me to groove good habits. The success fed on itself, making the tool more and more effective as months wore on. The other tools that were in the running never quite had the features I felt I needed when trying to find an OmniFocus replacement.
I think my problem was that I was trying to replace OmniFocus rather than replacing what I do to manage my projects and tasks. That’s why the cellular-level exploration happening over at Macdrifter is so compelling. That said, the end result might be something that works great for Gabe and, despite being plain text and A-OK by me, still won’t be able to do what I need it to do on a daily basis to manage my life.
As Gabe himself says:
There will be no satisfying conclusion for anyone other than me. A task management system is not as generic as the self-help app market would like us to believe.
I could not agree more. Despite his caveats, I will be watching Gabe’s valiant attempts to reengineer his process very closely. At the very least, I am going to let it challenge every assumption I currently make regarding my current task management solutions and, if pieces of it look like they might work better than what I have, I look forward to integrating them into a personalized system that makes me a better, more organized person.