FD should increase size of the team developing ED

[Ralph Vargr]

Elite Dangerous does not need more programmers.

Programmers *are* needed to write the second version, which fixes all the current version's issues.

 

[lenna]

...

Errm.... wrong on many counts.

Both in terms of general software complexity and on the SQA assessment front... sounds like you have bought in too much to the MMO is complex and the pro-megaserver kitsch.

Whether we are talking about ESO/GW2/WOW or any other predominantly 2D map based product (yes - I know GW2 is not 100% flat but it is still predominantly 2D map based in overall approach to motion and design) the general level of system complexity is clearly far lower than that of ED regardless of any perceived content comparisons. I could also point out various other flaws in your obviously biased perspective but all this is pretty moot where the matter at hand is concerned.

ED is very obviously a highly complex product and from a high level of personal experience with it and others (inc. GW2) from a user perspective, and developing various software products of various types over the years I would say you have no true idea about what you are talking about and are probably regurgitating some misleading spiel from one social media warrior or other who almost certainly knows only as much as you if not less.

 

[stephen_usher]

Yes, but as I'm sure you are aware, the mythical man month is essentially an application of 'the law of diminishing returns' (or 'too many cooks') after a certain, optimal, size of team is reached. Unskilled/unfamiliar staff hamper progress and clear lines of communication can be lost. However, *Before* you get to that point, *if* the staff are already familiar with the code base, or add a needed specialisation, you *do* get greater output. If I have to do a simple 'rote' task, e.g. picking crops from a field, you can bet I'm not going to want to do it all by myself. Likewise, if I need to do something out of my capabilities, like doing some plumbing, I'm going to get better results be getting a specialist in. Dammit Context! Stop being King!

So the question really should be 'is Elite sufficiently resourced'? And only Frontier's management and HR department can answer that one. And their answer will be different to what we, the unwashed masses, think should be the answer to yield the outputs we desire, to the timescales we yearn for, in our selfish, avaricious, hungry little hearts.

Not quite, there is the point of management overhead potentially increasing at a higher rate than the number of people added to the teams, especially if the teams can modify code interfaces which affect the code from other teams. This problem has been hinted about by Sandy as he has remarked that the Elite Dangerous code "is a beast" where changing one thing can cascade issues to many others, nothing is simple any more.

It should be noted that due to the performance imperatives that games code has to be a compromise between maintainability and "speed hacks", so it's somewhat understandable that the normal code isolation techniques used on highly maintainable code bases would be compromised.