Friday, April 20, 2018

My First Three Days Driving For Uber




A little over a week ago I applied online to become an Uber driver. The good news was that the application process was rather easy. The bad news is that the application process was rather easy. I will explain in a moment.

I picked an interesting time to start driving for Uber, as they have just released a new and improved, revamped driver app. This rendered the large amount of YouTube how-to videos obsolete. The videos did however let me see some of the things to expect. They also contained some good tips for new drivers.

One tip involved a method, albeit not a great one, to see where other Uber drivers are located given that Uber does not provide any way for their drivers to see where other drivers are positioned. The way to do it is to use the Uber Ride app which shows the nearest six Uber vehicles. If you are online in the Uber Driver app, one of the vehicles displayed in the Uber Ride app will be your own so you get to see where the five nearest Uber vehicles are positioned relative to your current position.

I have been using the Uber Rider app as suggested from day one. What I have determined is that in the suburbs of Chicago there are way too many active Uber drivers at any time. Today for example, after an early morning fare to O'Hare and a subsequent fare to the Loop, I ended up at Woodfield Mall. As I sat waiting for the Uber Drive app to ping me I switched over to the Uber Ride app and noticed that there were five vehicles very close to where I was. Subsequent checks of the User Ride app as I ended up in different positions near Woodfield always produce five other Uber vehicles close to me. After a while I realized that there must have been a hundred Uber drivers in the general area of Woodfield Mall. Although I did get a fare here and there it was a long wait between them. The same has applied for the other various locations I have positioned myself in during the past three days.

If you think about it from Uber's point of view, it is a good thing to over saturate areas with drivers. When someone requests a ride the response time is great since there is most likely a driver nearby. A vehicle showing up within minutes of a request is very good for business.

Over saturation is not a good thing for the suburban drivers. This is where the ease at which anyone with a newer vehicle and a clean background check can become a Uber driver is great for Uber and its passengers but bad news for the drivers.


Monday, January 1, 2018

The NFL app on my Android phone is great but buggy



My brother Jack came up with the clever saying, “The smarter the smartphone, the dumber the apps”. I really like the official free NFL app on my Android phone and use it frequently during the NFL season, but the “Drive Chart” feature does have a couple of glaring bugs.

I have noticed one in particular all season and have a hard time believing it has not been fixed. If you look closely at the above partial screen shot from my phone, Arizona had the ball first and 10 at their own 34 when Kerwynn Williams ran for 9 yards. It is now second and 1 at the Arizona 43 yard line however the down and distance line displayed under the score still shows “1st & 10 ARI 34” as if a play had not taken place. For most plays during a game the down and distance display is one play behind the current state of the game leaving me to do the math to compute the real down and distance.

The second bug, which I noticed yesterday for the first time, is that the play description for kickoff returns shows the incorrect final position of the ball. For example, Arizona kicked off to New Orleans with the play description stating that the kick went 54 yards and was returned 15 yards to the New Orleans 11 yard line. This is incorrect. The 54 yard kick was fielded at the New Orleans 11 and after the 15 yard return was actually positioned at the New Orleans 29 yard line.

A third problem is an annoying 'feature'. If you click on a play description in the drive chart, it is expanded to to show the full detail of the play but then automatically reverts back to the abbreviated description after two seconds. In many cases two seconds is not nearly enough time to read and comprehend everything that is there. This feature really should be click to expand and click when done.



Sunday, December 24, 2017

Owning A Windows PC Guarantees Stoopidity



As Yogi Berra said, “It's like deja vu all over again”. The feeling I get each time our Windows PC does something randomly stoopid out of the blue. Every Windows PC owner knows the accompanying frustration well, and that lost feeling of trying to solve the problem as the unwanted “time suck” occurs, sometimes stretching across hours and days. I recently had two such experiences. Perhaps they sound familiar to you even if you don't use the exact same software involved in my problems.

Episode #1
The first occurred a little over a week ago when I clicked on the tray icon for Abobe Lightroom 6 in order to import and edit a single photograph. Instead of the familiar main dialog of Lightroom I was greeted with a tiny dialog titled “Server Busy”. It had three buttons; “Switch Too” which did nothing, “Retry” which did nothing, and “Cancel” which was disabled. Figuring that perhaps Lightroom had not closed out properly last time I used it I decided to restart Windows thus beginning one of the most common “time sucks” in the Windows world, waiting for the system to start and become usable. Tick, tick, tick. After the restart I clicked on the Lightroom tray icon and voila … “Server Busy”.

A Google search produced a nice little article from Adobe itself about the server busy problem with a list of four possible solutions. The first suggestion was to upgrade Lightroom to a newer version. This was the last thing I wanted to do, especially since the last upgrade from version 3 to 6 was a battle in and of itself. The second suggestion was to disable the Windows Superfetch service. I have a slight advantage over the average PC user when it comes to enabling and disabling services since a huge chunk of the product I work on for my job runs as Windows services. So I stopped the Superfetch service and disabled it, after which Lightroom properly started.

It would have been nice if that was the end of this particular saga but it isn't. Although Lightroom was now working properly I noticed several tasks, like starting Chrome and browsing common web sites were very sluggish. What I have run into is an all too frequent catch 22 in Windows. There are multiple background tasks such as Superfetch and Indexing that help a PC run faster yet have bugs and sometimes use an excessive amount of CPU or cause disk thrashing. You can find a lot of posts where people turn off these services to solve slowness problems but you can also find a lot of posts where people turn on these services to solve slowness problems. In my case I have now turned Superfetch back on and made sure that Lightroom would start properly (it did). I have the feeling that I will see the “Server Busy” dialog sometime in the future and have to revisit this issue. Tick, tick, tick.


Episode #2
I have been using Carbonite to automatically back up files on our Windows PC for the last five years. About three years ago, when reports of ransomware attacks started to surface, I realize that the Carbonite backup could be made worthless if it automatically backed up maliciously encrypted files. Doing periodic backups to an external hard drive, something I had done for many years before buying our current PC, seemed like good insurance. So I did a little research and decided to use highly rated ShadowProtect by StorageCraft.

A few days ago I started a full backup before going to bed in much the same way I have many times over the past three years. In the morning I discovered that the backup had failed with read error -31, unexpected end of file. Once again turning to Google it looked like the first thing to try was running Windows Chkdsk. So I went through the process of requesting Chkdsk run after a restart and away it went. Slowly, very slowly. Tick, tick, tick. So slowly that I left it running and went to work. When I got home from work I fired up ShadowProtect. Periodically checking on the backup status, the progress bar was advancing as it always had. Until I came in to check on it and Windows was at the logic screen. The backup had cause a blue screen of death. Yes, this was becoming a major “time suck”.

Next suggested solution was to upgrade to the latest version of Shadow Protect. Which I did. Tick, tick, tick. Once the new version was installed I fired up another backup, immediately noticing that it was progressing along slower than the prior version (new compression algorithm using more CPU?). Tick, tick, tick. Over three hours later, blue screen. I opened a case on the StorageCraft web site. It's the Friday before Christmas. I have the feeling I won't hear back from them for days. I also have the feeling that when they do get back to me the next set of suggestions will venture into the “Stoopid Support Punt Zone” (remove all 3rd party software?).

So I go back to an old friend and download Acronis TrueImage 2018. Once installed I initiate a full backup. Tick, tick, tick. It worked!

The Wait Begins
So for now life on little Windows PC is back to normal. Looking forward to hearing from StorageCraft so I can ask for a refund of the $19.99 yearly license fee I just payed less than a month ago. Also looking forward to the next episode of Windows PC stoopidity, which is inevitable. Tick, tick, tick.







Saturday, December 2, 2017

Windows Update Has Been Stoopid Forever



Toshiba Netbook NB505 and Acer Switch Alpha 12

Everyone who thinks Windows Update is great please raise your hand! Oops. I don't see any hands. That's because Windows Update is stoopid and has caused grief forever. Yesterday I had some spare time and decided to update two devices. As you guessed, if both updates had gone smoothly I would not be writing this.

The first device was a Toshba NB505 netbook which came with Windows 7 Starter. Purchased 6 years ago, it doesn't get used nowadays. For a couple of years I kept it alive because it was the only portable device in the house that had Windows and an Ethernet port, allowing me to it direct connect to our wireless router when there was a problem with the router or the router needed firmware updates. When we moved into our new house a little over a year ago there was an existing Ethernet run to the bedroom where our main desktop PC is located. This provided a direct connection to the router and put the little netbook out of business.

Every few months I get the netbook out and do Windows Update out of habit and for no other reason. Yesterday was such a day and 23 updates were installed without a hitch. It was a slow laborious process taking a couple of hours because the Toshiba NB505 netbook is a glacially slow device and everything it does is a test of patience.

The second device was an Acer Switch Alpha 12 2-in-1 I purchased back in September. I opted for the i5 processor (middle ground between the i3 and i7). It came with Windows 10 Home. It is a nice little device and can run circles around the Toshiba. Windows 10 will automatically perform Windows Update in the background but in this case I manually went in and checked for updates. This began the process of downloading and installing the Windows 10 Creator Update (version 1079), a major update to the operating system with new features I didn't particularly know about or care about. It took over seven hours to download and install the update. Seven hours!

Needless to say I was off doing other things while the update crawled along. Checking on it periodically there were several times during the “preparing for installation” and “installing” phases that I would have sworn that Windows Update had hung. Fortunately for me I had a recent experience with a PC at work where Windows Update took way longer than usual (over an hour in that case, not seven hours) and I had learned to simply walk away and give Windows Update as much time as it wants to get the job done.

Seven hours for an update is ludicrous but that's not my main gripe. After all of these years you would think that Microsoft would have added a better progress indicator to the Windows Update mechanism. If “Preparing For Updates” or “Configuring Updates” takes a long period of time, having a percentage indicator “freeze” is unacceptable. During that long period of time when the percentage does not advance, many shorter duration operations are being performed. It would not be all that difficult to display something like the name of the file being processed or the key of the registry entry being updated as the software is installed. This common practice serves as a decent pacifier during long updates and reassures the user that the process is indeed not hung. Will Microsoft ever make this sensible improvement? Doubt it.


Sunday, November 26, 2017

Waze is brilliant (when it is not being stoopid)

Back when navigation systems first became popular we had a lunch table discussion about their virtues at work. I remember making the point that someone with local knowledge of traffic patterns would make better decisions than a system which solely relied on a map and posted speed limits. I cited cabbies and fire truck drivers in Manhattan as examples.

Of course, technology marched on and it was not long before two major changes to navigation systems came along; First, real time traffic conditions were added to the decision making process, improving the initial selection of the route and facilitating detours around backups during travel. Second, the storing of historical traffic patterns on a server allowed apps such as Waze and Google Maps to estimate travel times based on time of day and day of the week thus leading to the selection of a better route.

Yet with all of the improvements, my original contention that local knowledge trumps navigation technology still holds true. I use Waze to navigate and have found that it makes both good decisions and stoopid decisions.

The most common stoopid decision on my way to work occurs as I approach the intersection of Algonquin Road and Old Sutton Road heading east. The straight forward way for me to get to work at that point is to continue east to Barrington Road and turn south towards work. Waze however often suggests that in order to save one minute I turn right on Old Sutton Road and take a zig zag route additionally using Penny Road, Route 59, and Route 72. The problem with Waze's suggestion is that the Old Sutton Road route includes a two way stop sign crossing of a high volume 55 MPH two lane highway plus a subsequent grade crossing of the Canadian National railroad tracks. Getting blocked by a slow moving freight train is not fun. My preferred alternate, which is to continue straight on Algonquin Road to Barrington Road, has no stop signs and a bridge over the railroad tracks. When I ignore Waze and it recalculates there is at most a one minute time difference and often no time difference between the two routes.

My ride home features Waze suggesting a similar reversal of the route. The left turn from Old Sutton Road onto Algonquin at a stop sign is pretty much a killer during the evening rush hour with continuous streams of cars on Algonquin Road from both the left and right. Nobody in their right mind would take Waze's route and attempt that left turn however Waze suggests it every commute home.

Waze's stoopid decisions are not limited to my commute and perhaps you have noticed some of the same blunders:

Waze frequently takes me to a tough left turn at a stop sign when I could have been on a nearby parallel road allowing me to make the left turn at a traffic signal.

Waze has routed me on roads that are closed to through traffic due to construction.

Waze has routed me on roads which are marked by signage as having no though traffic allowed.

On a commute home last winter when the roads were icy and traffic was slowed Waze came up with a clever route that was less traveled. It was less traveled because it included a road with a dangerous steep hill that nobody in their right mind would drive on when icy. Some times faster is not better.

Despite all of this I do like Waze and use it frequently go get from here to there. I'll take some stoopid decisions over the good old days of navigating by maps.


Friday, November 17, 2017

Arrivo's Denver Hyperloop, is it hyper-poop?

Earlier this week there were several online news articles about Hyperloop firm Arrivo creating above ground tubes along side an existing highway that would allow cars to zip along at 200 MPH between downtown Denver and the Denver International Airport. Cars would travel on trays, with the trays being magnetically levitated. The system would be powered by electricity. Unlike competitor Hyperloop One's concept of a hyperloop which hopes to achieve higher speeds, Arrivo will not attempt to lower the air pressure in the tubes.

These articles really got me thinking. Thinking not about what a wonderful idea it is, but instead thinking about how every article I read about hyperloops is lacking in details and fails to address any of the possible problems they may encounter in building and maintaining said hyperloop. For example, Arrivo co-founder Brogan BramBrogan stated that the tubes could transport 20,000 vehicles an hour. Sounds wonderful, but I have trouble figuring out how this could possibly be implemented. Let's take a look at what it takes to get cars into and out of the tube. Remember that the cars are going to travel on trays. Therefore every car that enters the tube has to drive onto a tray which is then accelerated up to 200 MPH. At the end of the journey the tray has to be decelerated and when it has come to a complete stop the car can drive off of the tray and go on its merry way.

Let's investigate what it will take to achieve the stated goal of 20,000 cars an hour. Dividing 20,000 by 3,600 (the number of seconds in an hour) we need to feed a smidgen over 5.5 cars a second into the tube. Assume the the following steps are required to get cars into the tube using a single loading lane:

A car drives onto a tray

The tray heads down the mag lev track and merges onto the 200 MPH main tube track.

Another tray is moved into position so that the next car can drive onto it.

If the above process takes 15 seconds, which all things considered is probably a low estimate, there would have to be over 80 such lanes to feed 5.5 cars a second into the tube. 80 plus lanes! This would require a football size plot of land on each end of the tube just to handle the incoming cars. As if that is not enough of a a formidable challenge, how exactly do you get 22,000 cars an hour into the loading area without creating gridlock in all of the surrounding roads? Remember, the whole idea of spending a zillion dollars on the hyperloop is to reduce traffic on the highway between downtown Denver and the airport. Wouldn't this just create a bigger mess?

As if figuring out how to load cars into the tube isn't bad enough, the problems at the other end of the tube are worse where you have to get 5.5 cars a second off of their trays and onto the roads leading away from the tube. Needless to say, traffic on the roads leading away cannot back up or the unloading process will grind to a halt requiring all trays in the tube to slow down or stop.

There is one last challenge at the end of the tube and I view it as the toughest one to deal with. All trays have to decelerate from 200 MPH to a stop so that the cars can drive off of their trays. The trays however cannot decelerate on the main track in the tube or there will be a ripple effect back down the track and every tray behind it will have to slow down. Therefore trays will have to be “switched” onto multiple deceleration lanes which will then feed the unloading lanes. This switching will have to be done at 200 MPH! Good luck with that.

Considering all of the above I highly doubt that a rate of 20,000 cars an hour can be achieved. I'm not even so sure that 10,000 cars an hour is doable. Get much lower than 10,000 cars and hour and it would be a heck of a lot cheaper to simply add a lane in each direction. I can't wait to see updates on this project.