Heyo! I thought I'd update the blog a bit. I haven't been focusing on my development for a bit because I'm gearing up to go back to school. I've been admitted to the Rochester Institute of Technology for their MS in Software Engineering. It's very exciting.
I do hope to continue some of my personal software exploits while I am there and also to continue updating this blog when I can. I've been messing with a servo controller that might make for some interesting activity eventually for instance. Android development seems to occasionally hit a creative block from time to time.
Random observation: have you noticed that the YouTube client behaves much much MUCH better when it plays an ad versus when it plays content? This shouldn't surprise anyone, but it came home to me again when, for the first time I can recall, the ad I was watching took a moment to buffer in the middle. These commercials almost NEVER do that. I'm not complaining, believe me. It's a wonderful service; it's just interesting to see how clearly YouTube knows who butters its bread.
Monday, November 21, 2011
Wednesday, November 9, 2011
Grad School (again?) and Other Things
Well, my application to RIT's SE master's program is officially in. I await their decision. I would really like to get back to RIT to formalize my educational credentials so I could get a job. Sounds like not an unreasonable hope right?
On a totally different note, someone contacted me earlier today asking me to make a product like Uzu for the iPad. I took a look and found youtube videos that really impressed me. It's a very apt comparison to Finger Fireworks too.
The thing that most caught my eye though was the fact that the hardware on the iPad is so good. I have not had the opportunity to mess around with an Android tablet, so I don't have any idea how they would perform, but it's pretty clear that the iPad has a pretty heavy duty GPU.
Additionally, the ten-point multitouch is something I really want to experiment with. I just looked up the Xoom's specs and apparently some (all?) also support ten-point multitouch. My Android phone is a two-point multitouch device. This is a perfectly understandable design decision, as screen real estate is so limited, but it's also BORING.
I need more money to buy gadgets...
On a totally different note, someone contacted me earlier today asking me to make a product like Uzu for the iPad. I took a look and found youtube videos that really impressed me. It's a very apt comparison to Finger Fireworks too.
The thing that most caught my eye though was the fact that the hardware on the iPad is so good. I have not had the opportunity to mess around with an Android tablet, so I don't have any idea how they would perform, but it's pretty clear that the iPad has a pretty heavy duty GPU.
Additionally, the ten-point multitouch is something I really want to experiment with. I just looked up the Xoom's specs and apparently some (all?) also support ten-point multitouch. My Android phone is a two-point multitouch device. This is a perfectly understandable design decision, as screen real estate is so limited, but it's also BORING.
I need more money to buy gadgets...
Monday, October 31, 2011
Dialogs
Android dialogs are grey, boring, and chunky by default. I've been trying to figure out how to create dialogs that go with the visual scheme of my game. I've seen it done; it's just a matter of getting the "style" framework to work.
The documentation says it's like CSS, in that it allows you to format innumerable documents simply by setting a few key properties. I have yet to see a complete reference for the properties that can be manipulated in this manner for dialogs (or for anything for that matter). The only things I can find are incidental examples. This kind of thing frustrates me. I like to know about the full panoply of options.
I'm pretty sure I've said that before.
The documentation says it's like CSS, in that it allows you to format innumerable documents simply by setting a few key properties. I have yet to see a complete reference for the properties that can be manipulated in this manner for dialogs (or for anything for that matter). The only things I can find are incidental examples. This kind of thing frustrates me. I like to know about the full panoply of options.
I'm pretty sure I've said that before.
Tuesday, October 25, 2011
Eclipse Weirdness
So I discovered what looks like a brand-spanking-new Facebook Android SDK tutorial yesterday. It makes using the SDK actually doable for me, as I couldn't make any sense of it before, although I suppose I've learned a lot since I last tried.
I booted up the ol' compy today and started Eclipse, only to find that the references to the SDK were not registering in my test program any more. Eclipse just had no idea what I was talking about any more when I said "import com.facebook.android.Facebook". The project reference was still there. I did all kinds of stuff to see if I could get it to register. No effing reason (can you tell I'm not happy with this?) why it shouldn't work.
Ultimately I just copied the whole package from the Facebook SDK into my project. It seems to work. There are a couple of drawable resources that I copied over to make it work too, but it shouldn't be that bad. As it is, I'm pretty sure the SDK is just a very thin wrapper over some web service calls. Admittedly, I don't know how to manipulate these calls myself, but that's why I suspect just coppying the code like this should work without a problem.
Fingers crossed.
Edit: upon closer inspection and fiddling, it seems these problems revolve around the version of Android associated with the project, as apparently the SDK uses features not available in certain versions. It hasn't resolved all the issues yet, but it's a step forward. One reason why this was a hard problem (for me at least) is because these problems specifically prevented Eclipse from compiling the R file, which unleashes calamity when it happens.
I booted up the ol' compy today and started Eclipse, only to find that the references to the SDK were not registering in my test program any more. Eclipse just had no idea what I was talking about any more when I said "import com.facebook.android.Facebook". The project reference was still there. I did all kinds of stuff to see if I could get it to register. No effing reason (can you tell I'm not happy with this?) why it shouldn't work.
Ultimately I just copied the whole package from the Facebook SDK into my project. It seems to work. There are a couple of drawable resources that I copied over to make it work too, but it shouldn't be that bad. As it is, I'm pretty sure the SDK is just a very thin wrapper over some web service calls. Admittedly, I don't know how to manipulate these calls myself, but that's why I suspect just coppying the code like this should work without a problem.
Fingers crossed.
Edit: upon closer inspection and fiddling, it seems these problems revolve around the version of Android associated with the project, as apparently the SDK uses features not available in certain versions. It hasn't resolved all the issues yet, but it's a step forward. One reason why this was a hard problem (for me at least) is because these problems specifically prevented Eclipse from compiling the R file, which unleashes calamity when it happens.
Monday, October 24, 2011
Help
I Google a few dozen times a day to get help on the stuff I'm trying to do. Multifarious sources are pretty much the name of the game here. I find it funny though when I read through someone's guide, then at the very end they mention, just by the way, that it doesn't work. Huh? What the crap did I read this thing for?
Lulz. Oh well, better (read: cheaper) than buying a reference for everything I'm interested in I guess.
Lulz. Oh well, better (read: cheaper) than buying a reference for everything I'm interested in I guess.
Sunday, October 23, 2011
Progress
I found a database with a RESTful web service that I think I can use with my new project. As with anything, there can be hidden drawbacks, but I'm pretty enthusiastic about it:
http://platform.fatsecret.com/
I pictured the ability to scan the SKU of various items from the grocery store and authomatically put the data into the user's food diary, but I will still have to get the sku data from somewhere. More issues to mull over I guess.
EDIT: So, the people who made the database have made a (pretty sweet) app. Of course. Not sure why I thought any different was possible. I'm still going to mess around with it.
http://platform.fatsecret.com/
I pictured the ability to scan the SKU of various items from the grocery store and authomatically put the data into the user's food diary, but I will still have to get the sku data from somewhere. More issues to mull over I guess.
EDIT: So, the people who made the database have made a (pretty sweet) app. Of course. Not sure why I thought any different was possible. I'm still going to mess around with it.
Friday, October 21, 2011
New stuff is exciting, but
I've just spent the better part of a week and a half trying to get something up that would run a Facebook app on a nice fluffy cloud. I'm a newb and a half when it comes to these things, so despite having learned a TON, I don't have much in terms of a working service to show for it.
I've decided to return to what I know for now: the Google App Engine and Android.
That said, I am excited to be on a new project in which I will attempt to make it easier for people to track the nutritional information in the food they consume. I say easiER because there's still a lot of effort involved. People will still need to be inclined to measure their food and drink and occasionally input unknown items into the database for food and or drink. I still think it is potentially a great tool. I've been doing this for myself with an excel spreadsheet for a while now, and am always surprised by how effective it is at helping me make the right decisions.
This project has the potential to dwarf all my other ones combined if I want to do it right, but that too is a good proposition for me at this point. Here's hoping things work out...
I've decided to return to what I know for now: the Google App Engine and Android.
That said, I am excited to be on a new project in which I will attempt to make it easier for people to track the nutritional information in the food they consume. I say easiER because there's still a lot of effort involved. People will still need to be inclined to measure their food and drink and occasionally input unknown items into the database for food and or drink. I still think it is potentially a great tool. I've been doing this for myself with an excel spreadsheet for a while now, and am always surprised by how effective it is at helping me make the right decisions.
This project has the potential to dwarf all my other ones combined if I want to do it right, but that too is a good proposition for me at this point. Here's hoping things work out...
Thursday, October 20, 2011
Fun fun fun
I've received multiple messages now from Google Checkout about my TIN being incorrect. They need it so that they can report my income to the IRS. This is standard business practice; far from unusual.
The annoying thing comes when Google thinks (wrongly) that my information is incorrect. I have received multiple messages now asserting that my information is incorrect. I suspect that there is a string compare on my name or something that is not recording a character-for-character match with the name on file with the government for my SSN (which I used and can use in place of a TIN that an incorporated entity might use). I got out my social security card and copied character for character. But who knows if this was the problem.
This is why you need real people to interact with your partners/customers. I'm sure at a level higher than my business there are many ways to get this kind of problem solved, but I guess I'm not important enough. You need to have a person to find out and explain what the problem is, and maybe on occasion to gloss over a frivolous problem.
Instead, I found a typical hands-off, no-human-contact-allowed form of support. The email informing me of the taxpayer ID issue was your classic no-reply deal. The support forums are full of people with the same problem and no help. The alleged customer help form presents an array of radio buttons that reveal more radios when clicked on, never allowing for textual input. When I received a robotic response from that last bit, I responded for shits and giggles with what my problem was. Today I received a message that seems to reflect someone having read my email (miraculously), but failed to respond in a useful way. They basically told me where to input the TIN, which is not helpful since I clearly already have used that form.
It really is amazing how quickly Google has gone from being a helpful business partner to being a giant, faceless corporation uninterested in plebes like me. Like I said in a forum post in their support area, I still have considerable esteem in Google, but if they cancel my Checkout account because they can't get their act together enough to use the CORRECT information I have given them or work with me to make it right, then I will be spending my time and energy elsewhere than Android development.
The annoying thing comes when Google thinks (wrongly) that my information is incorrect. I have received multiple messages now asserting that my information is incorrect. I suspect that there is a string compare on my name or something that is not recording a character-for-character match with the name on file with the government for my SSN (which I used and can use in place of a TIN that an incorporated entity might use). I got out my social security card and copied character for character. But who knows if this was the problem.
This is why you need real people to interact with your partners/customers. I'm sure at a level higher than my business there are many ways to get this kind of problem solved, but I guess I'm not important enough. You need to have a person to find out and explain what the problem is, and maybe on occasion to gloss over a frivolous problem.
Instead, I found a typical hands-off, no-human-contact-allowed form of support. The email informing me of the taxpayer ID issue was your classic no-reply deal. The support forums are full of people with the same problem and no help. The alleged customer help form presents an array of radio buttons that reveal more radios when clicked on, never allowing for textual input. When I received a robotic response from that last bit, I responded for shits and giggles with what my problem was. Today I received a message that seems to reflect someone having read my email (miraculously), but failed to respond in a useful way. They basically told me where to input the TIN, which is not helpful since I clearly already have used that form.
It really is amazing how quickly Google has gone from being a helpful business partner to being a giant, faceless corporation uninterested in plebes like me. Like I said in a forum post in their support area, I still have considerable esteem in Google, but if they cancel my Checkout account because they can't get their act together enough to use the CORRECT information I have given them or work with me to make it right, then I will be spending my time and energy elsewhere than Android development.
Tuesday, October 18, 2011
Postgre
I've been trying for a while now to learn how to establish a connection with the Postgre SQL database that Heroku uses. It's a bit like besieging a city; I'm getting there, just ever so slowly, mostly because I'm also carrying all the weight of unfamiliarity with the language.
The fact that Heroku doesn't document its stuff very well doesn't help either. There do not seem to be nearly as many examples available as there are for say, the Google App Engine or for the Android platform. I'm guessing it's a smaller community.
There seems to be some support on Stack Overflow, but often questions are posed by people who have a much bigger clue what they are doing than I do.
I have had some success: I am running a copy of Postgre on my computer and did successfully connect to it through the Ruby test console. The connection will be established in a different way though when it is up on Heroku. In my local server, I specified the location, port, user, password, and a few other things, whereas it seems the Heroku connection must be established with a single string with all that information compacted inside. There is, as far as I can tell, a version of the method call I made that takes a single string, but I'm suspicious as to whether it will work.
Anyhow, back to besieging the city.
The fact that Heroku doesn't document its stuff very well doesn't help either. There do not seem to be nearly as many examples available as there are for say, the Google App Engine or for the Android platform. I'm guessing it's a smaller community.
There seems to be some support on Stack Overflow, but often questions are posed by people who have a much bigger clue what they are doing than I do.
I have had some success: I am running a copy of Postgre on my computer and did successfully connect to it through the Ruby test console. The connection will be established in a different way though when it is up on Heroku. In my local server, I specified the location, port, user, password, and a few other things, whereas it seems the Heroku connection must be established with a single string with all that information compacted inside. There is, as far as I can tell, a version of the method call I made that takes a single string, but I'm suspicious as to whether it will work.
Anyhow, back to besieging the city.
Saturday, October 15, 2011
Facebook Apps
I'm attempting to learn Ruby so that I can use a service called Heroku to implement a Facebook app. I was originally going to use PHP, but Heroku seems not to have that much support for PHP. Specifically, I couldn't figure out how to establish a connection with the database. This will undboutedly be easier with time, as more examples will crop up as their PHP support continues. As it stands now, I think their support for PHP is about a month old and they specifically limit it to Facebook apps. That could mean they really only are interested in social graph functionality for the time being.
That would be ridiculous, but people are ridiculous some times.
In any case, Ruby should work. It's just going slowly. I'm hoping ultimately to be able to tie a Facebook app together with an Android app so I can market the Android app better. I've already had some success with marketing stuff through Facebook, so I think it could be a real step forward.
That would be ridiculous, but people are ridiculous some times.
In any case, Ruby should work. It's just going slowly. I'm hoping ultimately to be able to tie a Facebook app together with an Android app so I can market the Android app better. I've already had some success with marketing stuff through Facebook, so I think it could be a real step forward.
Friday, October 14, 2011
What is Wernerware
Hi, my name is Daniel Werner. I am a law school grad that got back into programming after discovering what a bleak, awful place legal practice really is (and after not really even getting a chance to gain employment in it, really). Yeah, it's that bad.
In any case, I created Wernerware, my own little sole proprietership (i.e. there's no legal entity yet, it's just me doing stuff), initially to create and sell apps on the Android market. That activity remains my primary source of income, but I am branching out to bigger and better things every day.
My most recent efforts have been focused on web technology and cloud services. I have integrated the Google App Engine into my Android apps in the past. At the moment I am attempting to learn Ruby and Ruby on Rails to create a Facebook application running on Heroku. It's kind of slow going, but very educational.
I started this blog because I found myself posting a lot to my Facebook account when I learned something or wanted to say something about my efforts in programming. My friends, though, are mostly law school grads like myself, and therefore most likely honestly don't give a crap about that stuff, so I decided to create this blog. You can look forward to my random musings here instead then.
In any case, I created Wernerware, my own little sole proprietership (i.e. there's no legal entity yet, it's just me doing stuff), initially to create and sell apps on the Android market. That activity remains my primary source of income, but I am branching out to bigger and better things every day.
My most recent efforts have been focused on web technology and cloud services. I have integrated the Google App Engine into my Android apps in the past. At the moment I am attempting to learn Ruby and Ruby on Rails to create a Facebook application running on Heroku. It's kind of slow going, but very educational.
I started this blog because I found myself posting a lot to my Facebook account when I learned something or wanted to say something about my efforts in programming. My friends, though, are mostly law school grads like myself, and therefore most likely honestly don't give a crap about that stuff, so I decided to create this blog. You can look forward to my random musings here instead then.
Subscribe to:
Comments (Atom)