Medias Sucias: Tinta Digital

** More... Trello Clone **

Today I made some git branches... not that cool, but something I'd like to get into the habit of doing. The downside is that I had an unsaved file when I committed and I'm not sure what happens now.. ah well, I'll figure it out tomorrow morning.

I now can sort lists, that's cool. Tomorrow I'll finish up adding interaction to cards and making sure to persist order data... I'll have to hack around this and will promptly look for more elegant solutions.

** Trello Clone **

I'm beginning to think that nested forms may be required...

I figured out that I needed to provide all of my data up front, that is at the launching of my backbone app to avoid having to add conditionals to my custom parsing methods on models. I am not encountering a different issue when saving a list; apparently saving a list causes a model's parse method to be called at which point it is parsing some data, presumably a json representation of a list or model, which does not include a list attribute. There may be several solutions... one being nested forms and another being the modification of the response given by the database (I'm sure there are more).

** Trello Clone **

Today I spent some time working with modals. I ended up having to combine both css and jquery bootstrap classes / data tags / methods to come up with a working solution; in the end, I'm satisfied with what I have.

I now am able to sign-in, sign-out -- both with validations and error handling --, create a board, navigate to a board show page... and that's about it. Tomorrow will be a big day as I'll work on getting my board view (composed of sub views) working with draggable and sortable.

** Trello Clone **

Heroku:

It turns out that when deploying to heroku with Rails 3.x it is important to not initialize an instance of the application when heroku tries to copy your application. Secondly, if you have installed Backbone, you must put placeholder files in all of the folders you are going to use as git will ignore the empty folders.

Git:

Luckily, I only lost a few hours of work today... I somehow managed to replace an entire folder (git and all) with a trashed version of that folder. This occurred after realized I had a Rails 4.0 application and needed to recreate the app. Unfortunately I lost the work I had completed in the morning... fortunately it was mainly user creation and session information.

Bootstrap:

This is great. Quite simple to use and allows you to create somewhat decent looking applications with relative ease. I'll have to make sure to balance functionality with UI... I do think I'll work in slices. I'll build feature by feature.

** Gist Clone, Day 2. **

Another solo day down. Next week we begin out final project; I'm intrigued by Trello. I haven't used it much, but I like the client and server side challenge it presents.

** Gist Clone **

Today was a slow day. I do think that I'm understanding backbone's flow now, but still not in a way that's quick enough to not cause moments of confusion... as in, "what was I trying to do?" Often times when I get confused, I'm not exactly sure what is causing the confusion... sometimes I think that I'm over-thinking what I need to get done and thus obfuscating the problem I'm trying to solve. For example, as the day went along today and I had to add a favorite model... I was trying to find solutions using the existing architecture, when I really should have made some adjustments to make my life easier.

** News Reader **

Just when I thought (by the end of yesterday) that I was getting a hang of the backbone flow today happened. It's as if at the beginning of the day I had forgotten what had happened yesterday, and once I needed to nest json and use rabl I became utterly confused. Needless to say, I really enjoy backbone. It's pretty amazing what you can do on the client-side of things and today we got to work a bit more on how the client and server interact. I do think that a lot of the confusion is not being able to visualize the data flow that occurs between the server / client and the client's pieces; not knowing what needs to exactly happen at each step makes it easy to get quickly lost in the sea of steps involved -- this sea is really quite small, but feels rather large at the moment.

** Journal App - Backbone.js **

Today we ventured into the world of backbone. Backbone is quite nice, but I have to admit that there are many more things to keep track of now and because of this it's easier to get sidetracked. Some of the issues we encountered today were around trying to manipulate the wrong object-types, comparing with the wrong data-types, and forgetting to prevent default for actions. By the end of the day, however, I do have a much better feel for the backbone relationships between the router, views, models, and templates.

Client-side JavaScript MVC

Ajax requests are pretty cool, but if you want to have a mode dynamically generated single-page application it make sense to package up your client side code with a MVC framework. It makes sense, but it was difficult to create.

Although the client-side MVC is very similar to its server-side counterpart, it was hard to put into practice through code. Part of this may have been from not knowing exactly the purpose of creating a client-side MVC and part may have been with still feeling like I need to master modular JS. Yes, today we found ourselves in callback hell and were only saved by our friend .bind().

Ajax

We are now putting together JS (jQuery, AJAX) and our rails applications. I have to say that it was pretty cool to be able to make a background request. Especially entertaining when we slept a controller action and would temporarily change a status to show a user that a request had been made.

Making AJAX requests is pretty straightforward in that you're essentially moving the form's method and action from the form into the AJAX request, plus some additional information (what to do with the response upon success, etc.).

jQuery

Dynamically generated web pages?! A layer of fun? jQuery is awesome. I have to say that the beginning of the day was a bit slow... I honestly didn't really know where to begin, but as the day went along I felt more comfortable manipulating DOM elements. Today we built a UI for Tic Tac Toe and Towers of Hanoi. This probably shouldn't have taken too long, but, as I mentioned, I had never really seen a css file or generated a page by dynamically adding elements to the DOM.

At the end of the day, Tic Tac Toe and Towers of Hanoi worked quite well, but I know for sure that our code was not the most elegant. I definitely appreciate being able to look over solutions to see ways in which we could improve our code. In our case we had all of our game happening in one call back... I'd argue that for Tic Tac Toe this isn't too egregious, but it would definitely be more flexible if it were to be broken up into smaller pieces.

Snake:

We worked on Snake solo today... I didn't get too far, but I'm looking forward to playing around with this when I get a chance. If anything this will be good jQuery / callback practice.

Curried Functions, Inheritance, and Asteroids

Functions can take any number of arguments?! This is nice, in a way... no need for a splat operator, but can also lead to quiet fails. If, for example, a function is called with zero arguments and it needs one, the function will still run... where is the error?

What is interesting to note, however, is that every time you call a function it has an arguments property that is set to an array-like object containing all of the arguments passed to that function. This feature allows the currying of functions -- breaking down a function that takes multiple arguments into one that takes only one returning itself until the larger function has enough arguments to run. Cool.

Inheritance: it can be tricky.

Don't do this: Dog.prototype = Animal.prototype (Dog's prototype points to Animal's prototype which points to its proto property...) ==> THIS WILL NOT WORK, because when you add methods to Dog's prototype, all Animals will get these methods.

The only way to set a class' proto property is through the new keyword; we could say Dog.prototype = new Animal(); -- this would have unintended consequences. To get around these consequences we use a surrogate function who's prototype points to Animal's prototype...

JavaScript

Today we redid some of the first exercises of the course in JS... I have to say that the problems were quite easy, but were a great way to become familiar with the language. Some of the more interesting findings:

[1] += [2] => '12' [1, 2, 3].push(4) => returns the length of the array. null + "word" => "nullword"

Not recommended, but:

myFunction(); function myFunction() { ... }

... works because of JS Hoisting... essentially hoisting will "hoist" named functions and variables, but not variable assignments.

You pass functions that can be called later (callback).

So much to learn. I'm looking forward to continuing this tomorrow; I do have to say that learning JS the second time around makes much more sense than my first foray through CodeAcademy.

** Assessment Retake / Reddit Clone **

Today was the first time in 5 weeks that I have not felt nervous before an assessment. I felt quite confident that I could finish the assessment without a problem, and that was the case today. I have to say that having several solo days to implement the information we had been learning was quite nice and helpful in resolving any gaps that existed.

Reddit Clone: recursive partials! I think I have my comments (with nested comments) rendering properly... I'll have to test it a bit more to see if it is truly doing what I want it to be doing.

Javascript: this weekend I'll get to revisit code academy's JS curriculum; I'm glad that this time around I'll be using JS to build something rather than for the sake of learning a coding language.

Reddit Clone / Nested Forms

I think I may not be getting enough sleep... or the slight lack of sleep on a daily basis has been building up, because today I made many silly mistakes and it was hard to keep track of the moving pieces.

Yes, nested forms can be a bit difficult, but I do think after a bit of repetition it shouldn't be so hard. I have to say though, that I'm quite glad that we're steering clear of using accepts_nested_attributes_for and #collection_name_ids= ... there was too much mystery involved. There is more code involved in manually creating objects for other tables, but then again you have more control. You can catch errors, and alert users overall leading to a better experience. I also have to say that it is easier to understand what is going on since you have to build it yourself.

A handy trick (not really a trick) that I learned today was that there is this folder that holds initializers where I can add custom methods, inflections, etc. if I need to (for example adding an #all_blank? method to ActiveRecord::Base).

Assessment tomorrow: I feel prepared. I redid the assessment on Monday and, with the additional practice this week, I was able to finish it in about 1.5 hrs. This time I tried not to forget my usual testing methods (server logs and the browser) along with using the provided assessment rspec.

Social Thingamajig

Today we built a social thingamajig... The app would allow users to log-in, sign-up, and log-out; wahoo. The great thing was that creating this logging in, logging out and signing up process felt way more natural than it had before... still required a bit of thought, but mostly around validations, rather than the concept. Other things this app would be able to do: a user could create a circle of friends to share posts with (friends had to be other users); posts would have a title, a body, and links => when a user created a post we needed to create the associated objects with ids(posts and friend circles) and the associated objects with attributes (links). To do this we used nested forms, mass assignment, and some handy methods that rails creates for us ( accepts_nested_attributes_for and #{attribute_name}_ids= ).

Rails Lite / Rails Lite II

Today I finished Rails Lite: the trickiest part was understanding how RegEx .match worked to extract route params. It's quite satisfying to finish a project, although my solution isn't as clean as Ned's... some day.

I sometimes wonder: what is the most effective way to review solution code? Typically there isn't a lot of time to digest the code, but would it be helpful to type the code out? I guess this varies by person, but perhaps there is a recommended study method.

Assessment 04

I can't tell whether or not the assessment was particularly hard. I can say that I hadn't encountered rspec capybara documents before, which made it more difficult to troubleshoot errors... I also realized that once I had specs, I stopped doing what I knew (checking logs and running rails server) leading to more confusion.

I did learn a few things: to capture params (without explicitly passing them with a redirect) you can render as this does not require a separate request -- this is useful when, for example, you want to pre-fill a form with information from a failed save; and, if you forget about non-custom validations you can always write a custom one.