About my talk

Last year, I gave my talk at PyCon Hyderabad which was held from October 5th-9th. The talk was about optimizations that one can do while using Django to build a website. I used Django during my Google Summer of Code and Outreachy projects. In my talk, I majorly covered three topics — 1. Why and when should one use Django?; 2. Caching and 3. Database Optimizations.

These are the related links: Proposal, Video.

How did the experience help me?

Back then I was fresh out of college and relatively new in the open source world. Submitting a proposal for PyCon itself was overwhelming, let alone speaking there. I had applied thinking that there is no chance that they’d select a newbie like me to speak at a coveted conference like PyCon. But they did! And now when I look back I’m glad they did. Experience of attending and speaking at PyCon has acted as a big push for me in the past year.

It reinstated my believe in myself. After I got selected, I always kept on asking this question — do I really deserve this? Is my talk going to be worthwhile audience’s time? Can’t they simply learn all this from Django documentation and Stack Overflow answers? All these questions made me doubt myself. But giving up wasn’t an option because

I wouldn’t have known the outcome until I went for it. I let my curiosity lead my way.

To my surprise the talk went well. People appreciated my effort and gave me constructive feedback too. They were very supportive of the fact that I even attempted to present at such a young age.

That day I learnt the following:

1. Don’t give up before trying. No one will judge you!
2. Always believe in yourself. Fears won’t go unless you go for them.
3. People in open source conferences are super cool :). They’re always there to encourage you.

It’s not only about speaking

Open source conferences are a great forum to meet, engage and discuss with new people. Networking is the key! I learnt a lot from other talks, discussion sessions and interactions with fellow speakers. So, if you’re planning to attend any conference this year, don’t shy away from joining any discussion you’re crossing by or starting one on your own.

A note to fellow PyLadies

On the second day of PyCon, I attended a PyLadies lunch. That event was a full house! They got an overwhelming participation not only from women but men too. It made me realize that a lot of people are working to promote diversity in the tech world. They care about the cause and are trying to provide an equal platform to us. All we need to do is step up and contribute towards it. Times are changing and we have to become a reflection of it. Let’s be a part of the change!

Dear readers, I’m going to attend PyCon this year too and if you happen to be there too do stop by to say hi! Also, feel free to give feedback on my talk by posting in comments on the post.

What is Django?

Django is a high-level Python Web framework encouraging rapid development and pragmatic, clean design. A web application framework is a toolkit of components all web applications need. It’s goal is to allow developers to focus on the parts of their application that are new and unique to their project instead of implementing the same solutions over and over again. This open source framework takes care of a lot of the hassle of Web development through ORM, authentication mechanism, Django channels etc.

What is Pycon?

PyCon is a conference for the Python community, organized by its community members. It’s held annually in multiple countries for promoting and developing the Python programming language. It is for Python enthusiasts of all experience levels, from new users to core developers. Last year it’s India chapter was held in Hyderabad.

What was my talk about?

Through my talk I shared my experience and learning from working on Django in my Outreachy project. In that internship I developed a web application on Wikimedia’s community server (called Toolforge) using Django. At that time I was new to Django and even web development. Hence, I thought my learning might help others who are struggling to get started too.

My talk was divided into three parts:

Part I: What is Django? Where and why is it a nice fit.

Part II: How caching can help in making web apps responsive.

Part III: What can one do to save time used in database queries.

Slides from the talk can be found here. The video isn’t out yet but will update the post once it’s available.

What did I learn?

Let me start this way — it was an enlightening experience. It was the first ever conference that I was attending and I was a speaker there. Sounds overwhelming, right? Believe me it was.

I applied for the conference just like that. I didn’t really give it a thought that how will I speak in front of so many people. Maybe because I had never expected that my proposal will get accepted. But now when I sit and look back at it I’m glad I didn’t think much and just applied. Had I given it much of a thought I’d have never taken up this opportunity.

In Pycon I learnt a lot. All the talks made me learn something or the other. Few were about some new cool thing in Python and few were about what is next in Python. I didn’t only learn from the talks but also from people. Meeting so many new people and engaging in discussions in with them was exciting.

But most I learnt was from my talk itself. Speaking in front of hundred odd people taught me to never give up before trying. Had I never tried I would have not known what it takes to get on that dais and speak. But now I know the answer, it’s self-belief.

Believe in yourself because no one else is going to do it for you.

Other than this I learnt how can I improve as a public speaker, how can I better structure my talk and how can I make it more engaging.

Too much for one conference, huh? That’s why the word enlightening!

In the end I’d just suggest that if it can end well for me then it can for you too. Believe in yourself, go out and seize the opportunity you’ve been holding on. People will appreciate your effort and initiative. Best of luck!

The MVP contains all the necessary functionality of the tool like OAuth, creating worklist, searching a worklist, adding articles and petscan query ID to worklist, searching articles in a worklist, choosing an article to work upon and updating its progress etc.

All the other functionality is add-on to this backbone of the tool, and we want to get early feedback from users about the tool because of which I’ve been pushing hard to release MVP within first month of my internship. Hopefully it’ll happen soon.

You can try out the tool here. If you’re interesting in giving feedback, drop me a mail at meghasharma4910@gmail.com, I’ll let you know once the MVP is released.

Stay tuned for more updates :).

I spent some time setting up the environment, database for the tool etc for a smooth development experience. I did lot of research about React, its efficiency and when its useful. In these two weeks, I also managed to set it up, along with Django on Toolforge. It was hell of a task to make React work with Django along with Webpack, Babel etc. Issues kept on coming and I kept resolving them! The whole Webpack world seemed pretty complicated to me, but I managed to survive 😅.

Having spent lot of time playing around with React and Webpack and setting them up to work with Django, I’ve now shifted my focus up on the MVP release of my tool. MVP of the tool will contain all the basic functionality like creating a worklist, adding articles to worklist, finding a worklist, claiming an article in a worklist etc but with minimal UI. The goal of the MVP release would be to gather initial feedback from the users about the tool and make necessary amendments in our plans for this tool. I hope I’ll be able to share the results of MVP release with you in the next chapter. Stay tuned!

I’ll cherish that moment forever when I saw my name in the list of selected interns for GSoC-2018.

I’d been selected by Wikimedia to build a “worklist” tool for campaigns and in-person editing events. After the successful completion of Outreachy Round-15, this was icing on the cake! The dream of getting selected in GSoC had come true!

Being a previous intern with Wikimedia for Outreachy Round-15, I was very happy to have got another chance to work with the amazing community. My experience of working with the people at wikimedia was amazing and I’m sure this GSoC journey won’t be any different.

I’m looking forward to work with my mentors and build an amazing tool in my GSoC-2018 journey. Stay tuned for the chapters in this journey of mine!

Today when I’m writing my last blog of the series, I’m both nostalgic and happy. Nostalgic — on collecting all the moments by traveling down the memory lane; Happy — because I was able to pull it off successfully!

When my mentors looked at my implementation plan for the first time, they were of the opinion that it wouldn’t fit in. Simply, because it was too much! And as with time, it became more and more complicated, I joined the same opinion-group too.

But one thing that kept me going was the dream of seeing it live. I always dreamt of the proud feeling I’ll get when users will start using it. It kept me going even at 2 am in the night or when I had already worked for 12 hours but still had one bug left to resolve. And see 3 months later, I’m here with the tool complete! Yes, it works! And it is beautiful. (This much of hard work deserves a little bit of dramatization :P.)

Reflecting back upon the journey, I realize how this internship has transformed me from a reckless college student to a super-patient developer. Every bug has taught me ‘to stay calm and code’; every complex problem has taught me how important optimization is and every meeting has taught me that a way always exists.

Hence, it won’t be wrong to say that this journey has been hell of a learning experience for me from both technical and non-technical perspective.

So today with a heavy heart, I formally say goodbye to Outreachy Round 15 on a good note.

Wait, it doesn’t mean I won’t be writing anymore. There are a lot more blogs to come. So, do stay tuned! :)

But the love for this tool and support of my mentors have kept me going. For past few weeks, I’ve been drowned in implementation of the proposed features. For more details about the status of the project, you can refer to the workboard on Phabricator.

Looking at it, it seems that the wide scope of the project made the coding phase this gigantic, but that’s not the whole truth. Performance was an equal contributor. To tell you the extent, I will share a small story with you -

As a part of the Graphs module, I was implementing a graph which showed Impact of an editor with respect to time. The original proposed formula for impact was —

Impact = [(Average page views per day)*(contribution by editor till time x)]/(total contribution by all users till time x) summed over all articles in which the user has contributed

Looks like a simple scenario, right? But when it comes to implementing it and that too in such a way, that it loads up within the response time, it becomes hell of a task! Why? Because Wikipedia has sooo manyyy users that calculating the total contribution of all the users in real-time, with respect to time, is next to impossible! To give a bit of statistics, it took approximately an hour to load for an average editor. Just imagine what will be it’s performance when it would come to regular editors!!

Ah, I know you might be wondering how did I solve it? Well, by achieving the same thing in a different way! In other words, by acting smart :). I pre -computed average page-views per day and total contribution by all and cached it for user’s all pages. Then this constant factor was used along with the variable factor of user contribution for generating real-time graphs.

By now, you might have realized the ‘why’ behind the title. Anyhow, it was fun implementing plethora of features and I’m looking forward to new twist and turns :)

Stay tuned!

How did you decide upon the community?

How is Wikimedia for a beginner?

But when yesterday the count struck 101, I knew it’s time that I write it down and help as many as I can.

The answer to the first question is straight and simple — accidentally. I know such decisions are put through rigorous rounds of thinking; but for me it was sheer luck when I stumbled upon the right project at the right time. Eh, don’t worry! My second answer is going to be better :P

My journey in the Open Source world has been very short but still I had the audacity of writing such a bold title. This itself conveys how happy I am to be a part of Wikimedia community.

Firstly, it’s huge. It has multiple chapters and within which there are multiple projects. So yes, it can be overwhelming at first look. But don’t worry! take your time to settle in and understand the community. Their super descriptive documentation will come to your rescue.

Secondly, it is very welcoming and supportive. No matter how trivial the problem is, people here will go an extra mile to help you. I have experienced it multiple times and hence can confidently vouch for it. I even remembering asking my mentor to help me find relevant resources and he literally documented everything available! Not only this, I had a video call with a head of a team just to understand how to approach the problem statement. So, the crux is that they are ever ready to help.

Thirdly, the code base is well-maintained and well-documented. Plus they’ve a whole together separate environment for all the experimental development. And there language is no bar, so you can have a nice time understanding things around.

Lastly, their are so many different kinds of projects that you’re bound to find at least one of your interest. Believe me, every project will turn out to be a new learning experience. Starting from writing clean code to addressing the issue of scalability, you’ll experience everything and evolve along with it. And yes, the big name it carries further adds to your contributions.

What else do you need from a community? It’s the right place to be in as it is big, welcoming and simply awesome!

Hope this helps!

But eh, what is Toolforge?

Toolforge is a hosting environment for developers working on services that provide value to the Wikimedia movement.

During the evaluation phase of Outreachy I made 3 tools on Toolforge which are listed below —

1. tool to show recent edits by a user
2. tool to show the rank of a user based on number of edits in Hindi Wikipedia
3. tool to show the contributions of a user in terms of number of pages edited and created in the present year.

I felt that once one has a grasp on Toolforge, it becomes very easy for the developer to develop, deploy and test the tool.

This is the main page where you’ll get all the information about how to create a tool and how to set up environment for the language you’ll be using for your project. Information and common issues related to server can be found here.

And for all the project management tasks phabricator is the place. You can also create and manage the workboard of your project there.

Now, for the code reviews, you can use github. You can create a Pull Request everytime you want your mentor to review your code and once the pull request is accepted, you can merge the development branch with master branch. I’m sure you’ll be able to find good enough Github tutorials on the internet, so I won’t mention any here!

But development and coding is all about facing challenges, getting stuck and winning!. Its better to ask for help whenever you’re stuck. If you need help, the best place is the #wikimedia-cloud IRC channel. Or else, you can always drop me a mail at meghasharma4910@gmail.com describing what steps you took and where you got stuck in detail. I’ll be happy to help :)

Hope this helps!

Yes, you got it right! I’ve enjoyed this period so much that I can’t digest the fact that it’ll come to an end in just one and a half months. Ah, but reality always hits hard.

Jumping to what progress I’ve made in the past 2 weeks — I’ve started with the coding phase. Finally, I’ve landed into my comfort zone.

I’m working on my first module which is writing the backend for the user-summary section of my design mockup. This includes information like — since when user is contributing to Wikipedia, which all user groups he/she is in, in how many articles is he/she a major contributor etc etc.

I was able to calculate most of the metrics quite easily except one — number of articles in which user is a major contributor. This metric really brought me to my knees! Why? Because every approach I went for was either too slow or didn’t solve my use-case perfectly.

I was really amazed at my capability to think of one approach after another and code each one of them from scratch just to know how well it performs. So, this exercise made me realize —

A roadblock only makes you push your limits and discover a better you.

So currently, each passing day is a challenge which makes me struggle, learn and evolve!

Coming to the last part — key learnings in the past 2 weeks:

1. Learnt how to setup a Continuous Integration pipeline. Since my project is hosted on Github, I’m using Travis for CI.
2. Learnt how OAuth works and how to integrate it with the project.
3. Lastly, don’t give up! Exhaust every possible alternative. You might not get it right every single time but you’ll at least get the contentment of trying till the end.

Like all others, these 2 weeks have been amazing and exciting! And yes, I’m looking forward to more challenges. Otherwise how would I get the content for an engaging post, right? *wink*