Haiku to mentor 3 interns in Outreachy and GSoC | Haiku Project


#1

Hi there!

The selected students for both Outreachy and Google Summer of Code were publiched on Tuesday.


This is a companion discussion topic for the original entry at https://www.haiku-os.org/blog/pulkomandy/2019-05-08_haiku_to_mentor_3_interns_in_outreachy_and_gsoc/

#2

Which programme backs each of the three students?


#3

Preetpal is under Outreachy, the two others under GSoC.


#5

Look at students accepted in Blender gsoc projects this year. They know what they are doing !


#6

Oooh , pulkomandy , thanks for your hard work on Web+ :slight_smile:


#7

firefox is extremely needed.

for hardware in my opinion chode any small computer (raspberry pi) with x86.
all must working


#8

Yes i vote for firefox and gnash there are very needed!!


#9

That’s great for you but this isn’t a vote here :wink: It’s about the GSOC and outreachy internships.
The “porting browsers vs. Web+ development” topic was discussed in quite some detail in this thread: Haiku features I’d like to see


#10

Can you elaborate why flash would be needed? Thanks.


#11

In other words not all but most of pages are being using adobe flash player


#12

So we’ve paid $5,500 to get someone to add a preferences panel?


#13

Do you know someone who would do it cheaper? Haiku accepts contract proposals outside of Outreachy as well.

The internship (community bonding period included) lasts for 4 months (may, june, july, august). So that’s $1375 a month, assuming 40 hours a week and 4 weeks per month, it’s about $8.6 per hour. A little above the minimum wage in the US, well below the one in France, for example. It’s also about 2/3rd of the already low rate we used for previous contracts (which were all at $12.5/hour).

Also, it’s an internship, the goal is that people working in Outreachy stay with us after the summer is over and continue contributing.


#14

I am not a big fan of unified preference panel. Was not a big fan of the Zeta one either. I prefer to get to my simpler pref from tracker directly. The tracker is already there to unify things. It make it less trivial to all more input devices in the future if there is a monolithic panel with no real usability gain.

One thing i would LOVE to see is an unified file copy/move/download/upload tracker window-app.


#15

The goal here is to unify only the input preferences. Having keymap, keyboard, mouse and touchpad preferences as 4 separate applications makes little sense and gets in the way of advanced configuration in many cases (for example something as simple as using different speed settings on different pointing devices).

And who said there was “no usability gain” or that the new panel would be monolithic? The Network preferences are all in a single application (interfaces, network protocols, and services), yet many parts of it are add-ons. If needed we will go with a similar design here.

A single big unified preference app for everything is not on the roadmap, indeed.


#16

I wonder how many hours it’d take an experienced Haiku developer like you to complete this project. If it’s, say, 50 hours, then it’d be cheaper than $5,500 even if the developer charges $100 per hour.

I just finished a contract with a software company. Not only did they want to know how much I charge per hour, they also wanted an estimate of how many hours it’d take me so that they won’t go over their budget for the project.


#17

50 hours is not much. You can look at the output of our coding sprints, which tend to be more than that (since they only last a week, we can put in more than 8 hours per day).

This project involves:

  • UI/UX design, which requires some thinking if only to find a nice way to merge the existing preference panels together with their largely different aspect ratios
  • Enumerating and identifying input devices using BInputDevice
  • Visualizing the input devices in a list, and showing the appropriate settings for each
  • Allowing different settings for each device (a major refactor that will have impact in the input_server and the way it uses its settings)
  • Designing from scratch the UI for configuring or at least listing joysticks
  • Integration of new features such as PadBlocker
  • Handling of the migration of the various settings from the existing preferences to the new format to ensure settings are not lost for users who update their installations

Also this is done by someone who previously was unfamiliar with our APIs. That’s a price to pay for not using common technologies like Qt, people have to get up to speed with our tech, as well as our environment.

The project idea was designed by our developer team and put in the GSoC ideas list because we think it is a reasonable amount of work for a 3 month internship. It does not seem to be a 50 hour task for me, otherwise someone would already have spent two week-ends on it and completed it, right?

And, in this case it is a contract with a fixed number of hours, so it cannot go over budget. And if we are not happy with the progress, we can stop it.


#18

It look expensive but there is always a possibility that the developper stay for further free development.


#19

One thing i would LOVE to see is an unified file copy/move/download/upload tracker window-app.

Agreed. Using an enhanced notification service to support these background tasks status progress and notification would be great.


#20

Brakuje mi taniego tabletu. Myślę, że programiści powinni skupić się na jakiejś komórce z androidem na której dało by sie uruchomic Haiku lub jakimś tablecie.
Tani tablet, nie koniecznie z super wypasionym procesorem to było by fajne narzędzie mobilne.
Przeglądarka juz jest, vi, office, wiele osób nie potrzebuje wiele więcej.


#21

one question. How many dolars programmer needed to port haiku to other processor?
risc-v, epiphany, arm, spark, motorola …