Lack of coders

[Xycaleth]

We've had a lack of coders working on OpenJK for a long while now, and I guess this has shown in the lack of progress. I'm devoting my time now on rend2, so that doesn't really leave anyone fully committed to work on OpenJK.

 

Just thought I would give people the heads up in case they're wondering what's happening!

[Stoiss]

rend2 deserve this opportunity to be worked

i hope to see a lot of awesome progress with this in the next couple of mouth

[Archangel35757]

We've had a lack of coders working on OpenJK for a long while now, and I guess this has shown in the lack of progress. I'm devoting my time now on rend2, so that doesn't really leave anyone fully committed to work on OpenJK.

 

Just thought I would give people the heads up in case they're wondering what's happening!

How sad... there are 33 folks online at #jacoders (not counting me)... can't some of them commit to tackling one issue?

 

@@Xycaleth - Thank you for your commitment and perseverance!

[Tempust85]

What happened to ensiform and razor? Did they leave?

[Raz0r]

I've been very busy for a while and not in the best health. Ensi is much the same.

 

Perhaps we need to break down the idea that only certain magical people have the ability to code.

Literally anyone can learn.

[eezstreet]

I've been very busy for a while and not in the best health. Ensi is much the same.

 

Perhaps we need to break down the idea that only certain magical people have the ability to code.

Literally anyone can learn.

This is my primary motivation for the tutorials here on the Hub. I am tired of people asking others to code something for them or to give them an arbitrary copy/paste tutorial. While yes, there are tutorials on the web for this, it's better to work with real/practical examples and where this knowledge can be applied.

[Archangel35757]

While I applaud your efforts and the tutorial is great... it's daunting to try and wrap one's head around how intricately the code interacts... I think something akin to the archived Code3Arena tutorials were just about the right amount for a newbie to bite off.

[Artorias]

Those tutorials are very useful. Thank you for them.

 

Coding is fun.

[eezstreet]

While I applaud your efforts and the tutorial is great... it's daunting to try and wrap one's head around how intricately the code interacts... o think something akin to the archived Code3Arena tutorials were just about the right amount for a newbie to bite off.

So...copy/paste tutorials?

I disagree. The Code3Arena tutorials didn't really explain the systems at all, they just gave you some code to copy and paste. I think we need more tutorials or guides that explain how the code interacts and less junk.

[Tempust85]

Anyone can learn, but who can actually improve/fix things without breaking it?

[Archangel35757]

So...copy/paste tutorials?

I disagree. The Code3Arena tutorials didn't really explain the systems at all, they just gave you some code to copy and paste. I think we need more tutorials or guides that explain how the code interacts and less junk.

I do agree with you... but it's a huge task to explain to a novice coder how all the parts of the code interact and work together to actaully make a solid game whilst they learn C/C++. As for myself, it would be helpful if there was a giant block diagram / flow chart that showed the major code elements and how they interact... and spit out a game at the end of the process.

 

I was so adamant about fixing the 3ds Max dotXSI exporter code (that I set aside learning modeling & animation) and took two college C++ classes and spent another year pouring over the 3ds Max SDK and the dotXSI code just to get it fixed... @@DT85 also invested himself in that effort...

 

I'm still learning the 3ds Max SDK as I write a new animatable pivot controller-- and I couldn't do this if it weren't for their released plugin examples & samples.

 

Perhaps a smattering of both approaches (with a little more in-depth explaining for Code3Arena-like tutorials)?

[mrwonko]

Anyone can learn, but who can actually improve/fix things without breaking it?

That too is just a matter of experience. Code reviews would help - you need to understand that you did something wrong and what, in order to be able to learn from it. But that's even more work.

 

Thorough documentation would certainly be useful, but that too is yet more work.

 

Maybe I should try helping part-time. I'm not very familiar with the JKA code, yet quite familiar with C/C++, so I'd be in a great position to explore the code and write about it. It's just that I'm quite busy with multiple other projects in the pipeline; I'd need to switch back and forth between those and OpenJK. (And all but one of those projects are actually JKA related, too.)

 

I'll think about it some more next week; I'm still super busy with Uni this week.

[Kessno]

Maybe I should try helping part-time.

 

On-top of your developing new modding tools and creating another JKFiles mirror? 

[mrwonko]

On-top of your developing new modding tools and creating another JKFiles mirror? 

And a new browser and a mod manager (including the web infrastructure for it) and converting an old German JKA tutorial archive to my new website, yes. And maybe improving my Blender exporter a little, although the new editor would make it obsolete (by including Ghoul2 tools) so I don't really feel like doing that.

 

As I said, I'll need to consider it a bit more next week.

[Archangel35757]

@@Xycaleth, @@Raz0r, @@ensiform et al... maybe it would help if the OpenJK project leaders could ask some of the coders lurking at #jacoders if they would volunteer to tackle just a single, high-priority issue, and if they agree then assign them one of their choice thru github. This way folks can take ownership of an issue and work to resolve it. And it would boost the morale of the 1 or 2 who are actively working on OpenJK.

[gerbilOFdoom]

I don't contribute due to a combination of lack of time and a lack of ability coding C. I'm a web programmer - I can handle scripting languages well enough but adding in all the massive features in C with my limited knowledge of programming theory and I'm just good enough to create memory leaks and bugs in all the things.

 

In terms of adding on new coders: the best thing for it may simply be to spend time teaching or tutoring. For example, knowing just which features of C I should get to know to be familiar with most of the engine structure would be a significant boon. Most C tutorials start with simple variable types and declaration when I need to be filled in on the data structures, directory/file structure, and the overall flow of the code. The last two I can work out by poking around, as much as it may turn me off to the whole idea. Data structures I can study with my google-fu or by returning to college. 

 

Then there are parts of the engine that are total black-magic to me. The netcode, for example. I couldn't figure out how it worked or how to add additional data fields. Lightsabers are also difficult. I am also relatively unfamiliar with Visual Studio so any tools it provides are lost on me unless I'm shown to them.

 

[Lockgo]

I think I found one or two things I can fix. Where can I report and make changes?

[mrwonko]

https://help.github.com/articles/fork-a-repo/

 

Issues can be reported at https://github.com/JACoders/OpenJK/issues

[ensiform]

We could use someone to fix up G2 handles usage in SP.

 

Also cleaning up module access in SP between gamecode and engine.

 

ALSO SP could use someone who knows how to serialize existing data structures and keep it compatible with base for savegames.

[Archangel35757]

We could use someone to fix up G2 handles usage in SP.

 

Also cleaning up module access in SP between gamecode and engine.

 

ALSO SP could use someone who knows how to serialize existing data structures and keep it compatible with base for savegames.

Maybe Rich Whitehouse or Gil Grib? Gil is one of the original Raven coders... who works elsewhere now... is he active on jacoders irc? I have his email I can send you.

[ensiform]

I don't know if they would even be available or willing

 

Rich is not, for sure.

 

Also @@mrwonko @@Xycaleth @@Raz0r we need to categorize some of the headers if you want them to be cleaned as far as what would go where.  So far I have q_endian, q_types, q_validatehelpers.

[mrwonko]

I feel comfortable doing most of these changes, it will just take time.