Lessons from a deployment deadline

This is a really overdue post, but it's the first time in a while i'm properly awake in the day ^^ which brings me to the first lesson i've learnt:

• Sleep is an essential part of life. All repayment of owed sleep often comes with exponential interest.

For the last week, my group of 4 people, Laurence, Carolyn, Eldwin and I, have been doing night shifts of varying lengths. I, for one, can't really do coding marathons over several nights, but I.. managed somewhat. I do believe, though, that there has been at least one night where I failed to respond despite being online... having fell asleep on my keyboard. The full aftermath, however, came in the days after submission. I can't recall sleeping through most of the day for a few days on end in recent years ^^ Now that the first submission is over, I realise it could have been partially averted by observing the next lesson i've learnt:

• Read the manual. Especially the fine print.

Our application was, up till a really short while before submission, hosted on a virtual server within our school. I was not the one who submitted the request, but was told that the space would definitely be sufficient for a collection of text files. That claim turned out... partially wrong.

On the night of our submission, I uploaded one rather essential web script onto the server. After prompting for confirmation to overwrite, it deleted the file, and attempted the write the new one.

It failed. Horribly. And left us staring at a "Disk quota exceeded" message.

Suffice to say, we had a bit of a crisis. However, we couldn't, for the life of us, imagine *how* the total file size could have exceeded what i now knew as 1GB of space! But after a while of frantic digging, we found a rather large file on the server, named error_log. It was... 1.3GB. Apparently, we had forgotten to lower PHP error reporting, and a whole clump of warnings to the tune of "Variable not declared". Exasperated, we proceeded to delete the file, but the quota busted error still remained. We guessed this status was probably refreshed on a periodic basis, and couldn't really do much at that point. Luckily, we managed to find a saviour nearby, and with him came my next lesson~

• Always have a backup plan. Especially so during the last moments

When we realised we couldn't save our server that night, we intended to shift to AWS, which we weren't on initially as we wanted to save the credits for the Final Assignment. While trying to figure out how to connect to it's SQL service (also part of the reason we weren't on AWS), Frank offered to host it on his server. So now the application is running on a third-party host, and we actually had something to show for the submission.

So, thanks to all the help we received in the last hours, we actually managed to submit the alpha version of our as-yet-unnamed application (officially, anyway). We now have a less buggy system, a really good design, and a humongous backlog of features yet to be implemented. So I'm off to sleep a bit more for now, and resume coding very soon~

On Choices and Names

Wednesday, 13th January. Our group has decided to go all out with our Facebook app, or die trying~ In other words, Penguin Award, here we come!

It might seem strange that we're striving for the award given to the most spectacular failure... but it's not just that which we're aiming towards. Among us 4, we've bounced our dream FB apps off each other, grabbed a few ideas off different points made by different people, and have come up with a dream app that all of us, individually and as a team, aim and want to see through. And i think therein lies two very important points. Firstly, a team works much much better when everyone wants it for their own personal and professional pride, and not just because it's their job or they were coaxed by someone else onto the team. Secondly, dream big. It doesn't matter if you fail initially and / or repeatedly.

Despite starting early, our assignment Aspirations weren't complete till just before 11 pm yesterday, the day of the mid-assignment submission for the Facebook assignment. Not that we were lazy, but simply because that wasn't our priority. This was, and still is, the first Flash app Eldwin and I have created from scratch, and we cracked our brains for quite some time over the format / syntax of Flash. Concurrently, Laurence and Carolyn were busy designing the logic and graphical interface of our app, both having to learn on the go as well. Consequently, the Aspirations were put on some figurative shelf for quite some time.

Even the name for the App wasn't decided till really late. In a world where all you see of a product before you use it is its name and a short description, these are the first attention grabbers for new users. We wanted something with a catchy ring to it, while referring to our content in some way. Our placeholder name, Codename Gentoo, is our internal reference to it, on account of Gentoo being a family of penguins ^^. An interesting name I found while page hopping was Booooooom, which I initially thought was weird since it was hard to remember its exact length, till I saw its logo.



Okay, that's quite cool~ ^^ It'd be quite hard for someone with their logo to forget them. It even reminds users of the services they provide, as the company is actually a design company.

Names and logos are, especially in Facebook, what many users will see first. They need to be related to your app for their first impressions to be accurate. They need to be catchy enough for them to want to know more. And above all, they need to be simple enough to be remembered.

I wonder what our app will be called in a few weeks time~