Google Summer of Code – Page 12

One step at a time – The beginnings of CommonsNet

Post author:Agata Wiśniewska
Post published:May 31, 2016
Post category:FOSSASIA GSoC

The beginnings

If you have been accepted to a serious project like Google Summer of Code is, you can feel lost and scared. I think it’s nothing special and probably everyone experiences it. You can feel that pressure because you want to fulfill all expectations, follow your obligations and to do your best, but working in such project is something different from working on your own, private and small one. Your organisation and mentors require something from you, and they can even provide you with a detailed guideline how to behave but doubts may occur anyway.

My advice is not to give up and go through that tought period in order to experience the joy of results and sense of satisfaction, and to learn something to be better in the future. I am going to tell about my beginnings and to provide you with some tips from my own experience

CommonsNet

CommonsNet (feel free to see it) is a new project of FOSSASIA. It focuses on providing users with transparent information about WiFi they may use in public places like hotels, restaurants, stations. The thing is that for now, if you go to a new place, and want to connect to Internet, you look for a free WIFI sign and as soon as you find it you try to connect. But think about it, how much do you know about this connection? Is is safe for your private data? How fast is it? Does the Internet connection have any legal restrictions? I suppose that you answer ‘no’ to all these questions. But what if you know? Or if you can compare details of different WIFI available in a specific public place and connect to more suitable for your needs. I am sure you will appreciate it. I hope to run this project successfully and I am going to tell you more about it in next posts.

How to start?

Due to the fact that CommonsNet is a new project as I have mentioned before, and for now apart from mentor @agonarch and FOSSASIA leaders @mariobehling @hpdang, I am an only contributor, I am in a good position to tell you what are my steps. Remember not to think about all at once. It will make you crazy.

So first of all – prepare your work. Try to get to know about your project as much as possible. Follow group chat, GitHub repositories, do research in Internet about the subject of area of your project or don’t be afraid to ask your team member. That’s what I have done at first. I have prepared a Google Doc about all WiFi details. I have tried to get to know as much as possible and to gather this information in a clear, easy-to-understand way.

project details

I need it because I will be preparing a wizard form for users to let them provide all important details about their WiFi. I need to think seriously which data are important and have to be used to do it. It is not finished yet and will be changing (yes, I am going to share it with you and update you about changes!) but for now I want you to follow my view about it, how am I going to use the gathered information. wizard-ui

Next step is to prepare user stories. I think it’s a crucial point before you start to implement your project. I think there is no point of developing something until you think who will your user be. You need to imagine him/her and try to predict what he or she may expect from your app. Remeber – even if app is well coded it’s useless until somebody wants to use it. You can find many tutorials how to write a good user story in Internet. Just type ‘user stories’ in Google search. Some of them are here;

http://www.romanpichler.com/blog/10-tips-writing-good-user-stories/

https://www.mountaingoatsoftware.com/agile/user-stories

You can also see my user stories created for CommonsNet .

Furthermore, I have prepared a mockups to visualize my ideas. I think it’s also an important part of running your project. It will help you to express and concretize your ideas and let the whole team discuss about it. And there is no doubt that it’s easier to change a simple draft of mockups than coded views. You can see my mockups here: CommonsNet Mockups Screen Shot 2016-06-23 at 12.38.20

As soon as you finish all these activities is a high time to start creating issues on GitHub. Yes, of course, you probably have already started, so have I, but I am talking about further issues which help you to take control over your progress and work, discuss on specific subject and share it with other.

Lost on GitHub?

Is is possible at all? I suppose we all know and use GitHub. It’s a perfect place of working to all programmers. Its possibilities seems to be unlimited. But maybe some of you experience the difficulties which I have experienced at first, because just like me you have used GitHub so far only for your private aims and simply just pushed code and have not worried about creating issues, following discussions and organizing your work step by step . Let me to explain you why and how to follow GitHub flow.

GitHub issues let you and your team take control over your work. It’s really important to create bigger, let’s say main issues, and then subissues, which help you to divide your work into small parts. Remember – only one step at a time! Using my mockups first I have created some issues which present main tasks like ‘deploying app to Heroku’ or different pages in my app like ‘Home’, ‘About us’. And then I have created many smaller issues – subissues to present what tasks I have to do in each section like ‘Home’ -> ‘Impementing top menu’, ‘Implementing footer’, ‘Implementing big button’. It helps me to control where I am, what have been done, and what do I need to do next. And I think the smaller the tasks are, the more fruitful the discussion and work can be, because you can simply refine each detail. Please feel free to see CommonsNet issues. It’s not finished yet, and while working I am going to add further issues but it presents the main idea I am talking about.

Screen Shot 2016-06-23 at 12.57.48.png

I hope these tips help you to run your work and to go through harder time – easier. And remember even the longest journey starts from the first step!

Please follow CommonsNet webiste https://commonsnet.herokuapp.com/ to be updated about progress, latest news and tips how to resolve programming problems you may experience.

Feel free to follow us on social media Facebook https://web.facebook.com/CommonsNetApp/ Twitter https://twitter.com/Commons_Net

The Official Beginning

Post author:Siddhant Gupta
Post published:May 29, 2016
Post category:FOSSASIA

This blog marks the start of the coding period for Google Summer of Code. I will start by talking about the community bonding period and then my plan to begin the summer.

Community bonding was an interesting and intense period. Managing college, assignments, exams and working on the project all at the same time. I think this is the capability that sets us apart. We had started out with fixing small issues in the steam-shell. Steam-shell is a command line interface that allows the user certain limited set of commands to interact with the server. I will list down the issues I solved and the improvements I made.

My first task was to improve the display of commands like ‘look’. This command lists out the gates/files/documents/rooms/conatiners present in the current room. The old display showed one entry in one line. I improved this to make it something similar to the ls command of linux. I used sprintf(“%#-80s”,<variable>) to give the output a structure.

Next I worked on the create command. I found a bug that this command was not working for containers and documents. I fixed it by adding the required conditions in the code and then I worked upon to improve the functionality of the command as suggested by my mentor Trilok. I added the feature to enter the destination where the new object is to be created and a ‘.’ to create it in the present directory. Initially this feature was only for Rooms and Containers, later I extended this to all kinds of objects.

After this I spent a few days studying the COAL protocol. I have already written in detail about this in my previous blog. Though I was not successful in completely understanding the protocol but this gave me an idea of the client-server architecture. This was towards the end of the community bonding period and I decided to spend the rest of my time with steam-shell itself.

Now as the coding period begins officially I will be working to fix the edit command. Currently the edit command opens the file on a vi/vim/emac instance. The vi and vim opens in the same window and it misbehaves. Then I will be extending the feature for multiple documents and in the next week move on to work on COAL again.

Google Summer of Code: A new beginning

Post author:Siddhant Gupta
Post published:May 15, 2016
Post category:FOSSASIA

I am working on sTeam under FOSSASIA for GSoC 2016. sTeam is collaboration platform. It allows users to form groups and share various kinds of documents. It provides a virtual knowledge space where users can meet and interact. These spaces are called rooms, these can contain various containers to store documents in an organized manner. The rooms are connected by gates to allow the movement of users and documents from one room to another. The most interesting thing about the project is that it is in pike programming language. This is not very commonly used and I will get to learn new things and myriad of experiences.

Now moving into the technical details of my project I will be talking about the first step in my project that is implementing TLS (Transfer Layer Security) Protocol, also known as SSL. Currently all the communication between the server and the command line client is over COAL protocol. COAL is a home grown protocol developed by the original developers of sTeam. Currently COAL does not use any kind of encryption. My first task will be to integrate SSL into COAL, so just like http becomes https COAL will become COALS. After this I will be working on the improving the client interface, which I will talk about in the coming weeks.

For now in the community bonding period I have been actively participating in the discussions on the IRC and working on stash, discovering bugs in the project and getting used to the interface and the code. Recently we received the guidelines for the summer from our mentors Martin and Trilok. The guidelines include clearly mentioned milestones and the tasks we need to complete. We will also be having a team meeting everyday discussing our day’s work and issues.

I would like to thank FOSSASIA for giving me this wonderful oppurtunity to show my skills. All the discussion during this period is just making me feel more professional and getting me excited for the summers. In programmers term

printf(“let the fun begin !”);

searchQuick Apprise: EIGHT #GoogleSummerOfCode #FOSSASIA

Post author:Jigyasa Grover
Post published:August 29, 2015
Post category:FOSSASIA

The intended “searchQuick” (sQuick) is an application to enable a user to search a set of books or texts, like an encyclopedia, or some other topical book collection offline built in the open source platform Pharo 4.0.

Bringing up to the rear of the summers, the project was brought to a penultimate stage by achieving the following tasks:

Handling empty string searches by raising error pop ups.

searchButtonClicked
searchBar accept .
(myString isEmptyOrNil  ) 
ifTrue: [self errorPopUp ] 
ifFalse: [ 
     myString := searchBar getText asString .
          self printSearchResults
               ] .

Adding Help, About and Feedback sections to give an authentic application look and required details to an interested developer.

Inserting ScrollPane for BrowseFile list menu

browseScroll := ScrollPane new.
browseScroll scroller addMorph: browse.

Truncating BrowseFile list menu file titles to have their extent within the #MenuMorph: boundary
```
title := anObject truncateWithElipsisTo: 25. 
```

Removal of OK/CANCEL buttons from the Search results accordion widget
```
dialog buttons: {}.
```

Implementation of a Search Bar for searching via Search results accordion widget.

Categorizing methods as: accessing, initialization, submorphs–add/remove etc.

Removal of unwanted/redundant/commented code lines i.e. scrubbing dead-code.

UPCOMING: Wrap Up.

[Tutorial] Continuous Integration Automated Build for your Pharo Application

Post author:Jigyasa Grover
Post published:August 22, 2015
Post category:FOSSASIA Tutorial

reposted from jigyasagrover.wordpress.com/ci-automated-build-for-your-pharo-application

Hello Fellas !

This post aims to put forward the basics of Build Automation and also brief the steps required to put up a Pharo application on Continuous Integration, Inria which is a platform for Scheduled Automated Build.

For simplicity, Build automation is the act of scripting or automating a wide variety of tasks that software developers do in their day-to-day activities including things like:

compiling computer source code into binary code
packaging binary code
running automated tests
deploying to production systems
creating documentation and/or release notes

Various types of automation are as:

On-Demand automation such as a user running a script at the command line
Scheduled automation such as a continuous integration server running a nightly build
Triggered automation such as a continuous integration server running a build on every commit to a version control system.

In recent years, build management tools have provided relief when it comes to automating the build process.

The dominant benefits of continuous integration include:

Improvement of product quality
Acceleration of compile and link processing
Elimination of redundant tasks
Minimization of ‘bad builds’
Have history of builds and releases in order to investigate issues
Save time and money – because of above listed reasons.

A build system should fulfill certain requirements.

Basic requirements:

Frequent or overnight builds to catch problems early.
Support for Source Code Dependency Management
Incremental build processing
Reporting that traces source to binary matching
Build acceleration
Extraction and reporting on build compile and link usage

Optional requirements:

Generate release notes and other documentation such as help pages
Build status reporting
Test pass or fail reporting
Summary of the features added/modified/deleted with each new build

Considering the above mentioned advantages of automated build, the below enlisted steps will help to put up your own Pharo application hosted on github on the CI server for continuous integration/scheduled build.

1. Log on to Continuous Integration, Inria website (https://ci.inria.fr/).

2. Click on ‘Sign Up‘ at the top-right corner, enter the required details and register for CI.

3. From the ‘Dashboard‘ option located at the top most of the screen click on ‘Join an existing project‘ blue button as shown .

4. Search ‘pharo-contribution‘ in the enlisted public projects and click on ‘Join‘

5. On clicking the ‘Join‘ button, a message stating: “Request to join the project ‘pharo-contribution’ sent.” appears.

6. It might take a day or two for the request approval mail to deliver at your registered Email ID.

The E-Mail content is as follows:
Your request to join pharo-contribution has been accepted
Hi _ _ _,
Your request to join the project pharo-contribution has been accepted !
Regards,
Support team.

7. Click on ‘My Account‘ option and under ‘My Projects‘ check the status of pharo-contribution project. It should state ‘member‘.

8. Now, visit the LINK: https://ci.inria.fr/pharo-contribution/job/JobTemplate/ to create a ‘New Job‘

9. Read all the steps mentioned carefully. After going through all the points, click on the ‘New Job‘ mentioned in point 2 on the Project Job Template web page.

10. Enter the ‘Project Name‘ in the ‘Item Name: ‘ box and choose ‘Copy from existing item‘ option and fill ‘JobTemplate‘

11. After clicking OK, You will be directed to your project configuration.

12. Fill in the description of the project in the desired box.

13. Fill int the configuration details of your project like:

* Maximum number of builds
* Link to GitHub Project
* Source Code Manager
* Build Triggers
* Schedule of build (@hourly, @daily, @weekly, @fortnightly, @monthly, @yearly etc.)
* Configuration Matrix (User Defined Axis: Name && Version Values- stable, development etc.)
* Build environment options
* Post-build actions
* Report regressed tests

14. The main task is to carefully write the commands in the ‘Execute Shell‘
The default commands are as:

15. After saving and applying the changes, the application is all set for automated build.

16. Each build’s ‘Console Output‘ can be used to analyse the steps and highlight the weak areas of the project.
For instance: The below output is of a project whose stable version build was successful.

TIP: Keep a regular tab on the build results and analyze each line of the Console Output with utmost care.

Hope this post was able to help you start with the automation build process of Pharo Application.

Do like if it was worth a read !
Post queries/suggestions as comments 🙂 Looking forward to them.

UPCOMING: Next, I plan to share experience of putting up my Pharo application searchQuick on CI Inria for automated build. I intend to detail about the various configuration settings applied along with the Execute Shell commands utilized for a GitHub project 🙂

Introduction Accredits: Wikipedia
Resources: Build Automation and Continuous Integration .

searchQuick Apprise: SEVEN #GoogleSummerOfCode #FOSSASIA

Post author:Jigyasa Grover
Post published:August 18, 2015
Post category:FOSSASIA

The intended “searchQuick” (sQuick) is an application to enable a user to search a set of books or texts, like an encyclopedia, or some other topical book collection offline built in the open source platform Pharo 4.0.

After the chief tasks of search functionality and automated build were done with, the next undertaking included working on finer details and embellishments.

Embedding Jenkins automated build status icon in GitHub markdown file
Relative widget re-sizing by using ‘World extent x‘ and ‘World extent y‘ co-ordinates instead of hard coded co-ordinates
Modifying the Accordion Widget by addition of ‘Search Bar‘ at the top
Checking for duplicates in the ‘Browse Files‘ menu, thus reducing the CPU consumption
Equalizing the sizes of all the windows to bring uniformity
Addition of ‘Scroll Pane‘ in accordion search result display list
Multi-line search result display by extending the Expander Title Morph and use of new line character in labels (otherwise not supported by default)
Truncating file content to first n characters for neater look in Expander Title

Latest Screenshot of Accordion Widget:

UPCOMING:

Removal of OK & CANCEL buttons (present by default in Pluggable Dialog Window) from Accordion Widget
Implementing of Search via the result window as well
Relative re-sizing of background images (Image Morphs)

searchQuick Apprise: SIX #GoogleSummerOfCode #FOSSASIA

Post author:Jigyasa Grover
Post published:August 8, 2015
Post category:FOSSASIA

The intended “searchQuick” (sQuick) is an application to enable a user to search a set of books or texts, like an encyclopedia, or some other topical book collection offline built in the open source platform Pharo 4.0.

The main task achieved was putting up the application up on Continuous Integration, Inria for automated build. It was indeed a beneficial idea as it helped me keep a check on the builds and work on issues.
Being a newbie, this work was cumbersome initially but with the help of my mentors and the #pharo community, I was able to accomplish it. To assist fellow Pharo-ers, I have compiled all the information regarding CI Automated Build for yout Pharo Application and published the same on my blog-spot. Kindly go through it for a complete understanding 🙂

Other tasks completed as of now include:

Putting up the project for automated build on https://ci.inria.fr/
Successful ‘stable’ and ‘development’ version builds
Accessing resource folder via MCGitHubRepository, Removal of manual download option
By default full screen system window open
Removing redundant code by creating open argument methods
Abolishment of hard-coded font family and font point size
Categorization of methods & classes
GUI Embellishment with background colors, borders etc.

Upcoming:

Dynamic widget re-sizing
Multi-line search result title
Putting up Help and About sections
Removal of old configurations

searchQuick Apprise: FIVE #GoogleSummerOfCode #FOSSASIA

Post author:Jigyasa Grover
Post published:August 1, 2015
Post category:FOSSASIA

The intended “searchQuick” (sQuick) is an application to enable a user to search a set of books or texts, like an encyclopedia, or some other topical book collection offline built in the open source platform Pharo 4.0.

As the rudimentary structure of the application is sewn up, embellishment of GUI and rigorous testing are the major part of course of action.

On eMBee’s ( +Martin Bähr ) suggestion to build up an accordion widget to display the search results, various trials were conducted to design a similar one in Pharo.

The task of developing the accordion widget in Pharo was achieved using Expander Morphs. Looping through the search results array, a #newExpander: was added in each #newRow: of the modal built.

A challenging chore was to add a scroll-able content on the click of the desired search result expander. Sundry experiments with #newLabel: and #newText: in #newScrollPaneFor: {i.e. adding text model and labels in scroll pane} had no effect. Eventually, #newTextEditorFor: did the trick and the desired look was created.

Next on the cards is putting up sQuick for automated build on the CI Server, as suggested by +Sean DeNigris for its various advantages which include:

Improvement of product quality
Acceleration of compile and link processing
Elimination of redundant tasks
Minimization of ‘bad builds’
Have history of builds and releases in order to investigate issues
Save time and money – because of above listed reasons.

For simplicity, Build automation is the act of scripting or automating a wide variety of tasks that software developers do in their day-to-day activities including things like:

compiling computer source code into binary code
packaging binary code
running automated tests
deploying to production systems
creating documentation and/or release notes

UPCOMING:

To achieve Build Automation for sQuick, I have already registered on CI and configured sQuick.
Next endeavor is to look into the red signal in the build evaluation.

Stay tuned for more….Post any queries, will be happy to help 🙂

Dockerizing sTeam

Post author:Jai Pradeesh
Post published:July 31, 2015
Post category:FOSSASIA

I am currently working with sTeam collaboration platform as a GSoC dev under FOSSASIA umbrella.

sTeam has a lot of depencencies. A lot! One major issue faced by developers was version conflict between dependencies. Docker seemed to solve this issue. Docker is a great image distribution model for server templates. It uses btrfs (a copy-on-write filesystem) to keep track of filesystem diff’s which can be committed and collaborated on with other users (like git). It also has a central repository of disk images that allow you to easily run different operating systems and shares the host kernel.

In this post, I will explain the workflow of containerizing sTeam with Docker.

Prepend sudo

Install docker on host system, start & enable the docker.service

dnf install docker
systemctl start docker
systemctl enable docker

Pull a base image from docker hub

Note: It is upto you which image to use. I am using Ubuntu as base image.

docker pull ubuntu:latest

List the images to verify the pull

docker images

.. should display ubuntu latest xxxxxxxxxxxx x months ago x MB

Build your own image

docker build -t="dolftax/steam"

Lets run bash to install sTeam itself and its dependencies

docker run -t -i dolftax/steam /bin/bash

Install the packages and its dependencies.

In my case, sTeam server. Installation steps –https://github.com/societyserver/sTeam/wiki/Installation-steps#manual

Grab latest container ID. The first one would be the recently closed container.

docker ps -a

Commit the changes so that you don’t lose the data when container exits

docker commit [container-id] dolftax/steam:v1

You should get a long hash as the success message

Note: v1 is a tag Don’t use latest tag. Don’t be tempted by it.

Create a docker hub account and login

docker login

.. which is self explanatory.

Push the image to docker hub (Before that, create a docker hub account anddocker login)

docker push dolftax/steam

Done!

.. after containerizing with docker, sTeam installation is as easy as

docker pull dolftax/steam:v1

Try containerizing your project and you would love it.

Intriguing Read (Bootstrapping Reflective Systems: The Case of Pharo)

Post author:Jigyasa Grover
Post published:July 12, 2015
Post category:FOSSASIA

Hiya Fellas 😀

So, while I was browsing through various developments in Pharo, I came across a very eye-catching paper on Bootstrapping Reflective Systems: The case of Pharo .

For beginners, Bootstrapping is basically the process of writing a compiler or an assembler in the source programming language in which it is intended to compile. Applying this technique leads to self-hosting compiler that proceeds without external input.

Though, I am still a newbie and in the process of grasping all this ingenious stuff.
But I am sure that this paper ( Poli12-BootstrappingSmalltalk-SCP ) by G. Polito , S. Ducasse , L. Fabresse , N. Bouraqadi , B. van Ryseghem would make for an interesting read.

Do scan through it and post below your ideas and queries !