L2J Server

SpooKNoF wrote:

afk5min wrote:The style is called 'Defensive Programming'. Has its benefits, especially when many people are working in parallel on the whole project – instead of each being assigned a subset of features to implement or maintain.

By the way, to determine if this null check is redundant, all you need to do is open up the call hierarchy (view references also works). The worst part is, you will have to do it on each new revision.

SpooKNoF wrote:if you are so good help l2j to start with small things and then you will find more serious

I beg to differ, that is not quite how it works. Surely, from a purist's PoV, every incorrectly working feature should be fixed. Even when it is a custom system message (identical) instead of client-stored message. The thing is, small things (such as sending initial UserInfo without clan data, to make sure retail server behavior is replicated) take time that can be spent better elsewhere, such as implementing new features, or reestablishing quick hacks/patchwork code.
Unfortunately, most people who try to bring attention to small details are not able to deal with more serious matters... And never really manage to learn.

P.S. Sure, you start 'finding more serious' stuff later on. But in the big picture, the whole l2j gameserver codebase is shit. Mostly because it became acceptable to rush feature implementations and then forget about them, reusing old code exclusively by copy-pasting. That wouldn't be so bad, but it only takes one person to make a mistake in his code, and that mistake will quickly propagate everywhere, but not by design, but as separate copies.

why you bring more people to proove what? IS NOT USSULES NULL CHECK?

lord_rex wrote:by the way, before anyone would ask that "why I am wasting my time with that idiot", I am just doing it because I had a nice and peaceful discussion with Zoey and Jurisics before that idiot was breakin' our discussion with his stupid "null check thing".

lord_rex wrote:by the way, before anyone would ask that "why I am wasting my time with that idiot", I am just doing it because I had a nice and peaceful discussion with Zoey and Jurisics before that idiot was breakin' our discussion with his stupid "null check thing".

SpooKNoF wrote:java = ram and cpu at last...

SpooKNoF wrote:

lord_rex wrote:by the way, before anyone would ask that "why I am wasting my time with that idiot", I am just doing it because I had a nice and peaceful discussion with Zoey and Jurisics before that idiot was breakin' our discussion with his stupid "null check thing".

so if anyone have opinion is stupid intrepid right? because you are the intrepid now
and you have private project? i hope to see in your pack 100 null checks for the same methods because we talk for java java = ram and cpu at last... doAttack if you see what biggboss say called and from npcs how many npcs also we have speacialy on h5

afk5min wrote:P.S. Sure, you start 'finding more serious' stuff later on. But in the big picture, the whole l2j gameserver codebase is shit. Mostly because it became acceptable to rush feature implementations and then forget about them, reusing old code exclusively by copy-pasting. That wouldn't be so bad, but it only takes one person to make a mistake in his code, and that mistake will quickly propagate everywhere, but not by design, but as separate copies.

Zoey76 wrote: @SpooKNoF remove the check, once you get an NPE, check how the reference got nulled, fix the issue from the root and share a patch after it's tested, that way you would remove a useless null check without causing NPE flood in server logs.

Zoey76 wrote:I'm not agree on that, we spend almost two years reworking the core and we let features development in the back sit to make the core better, if you are going to say something try to keep it real

Zoey76 wrote:

afk5min wrote:P.S. Sure, you start 'finding more serious' stuff later on. But in the big picture, the whole l2j gameserver codebase is shit. Mostly because it became acceptable to rush feature implementations and then forget about them, reusing old code exclusively by copy-pasting. That wouldn't be so bad, but it only takes one person to make a mistake in his code, and that mistake will quickly propagate everywhere, but not by design, but as separate copies.

I'm not agree on that, we spend almost two years reworking the core and we let features development in the back sit to make the core better, if you are going to say something try to keep it real

@SpooKNoF remove the check, once you get an NPE, check how the reference got nulled, fix the issue from the root and share a patch after it's tested, that way you would remove a useless null check without causing NPE flood in server logs.

SpooKNoF wrote:dude just check how many null checks we using for the same code its at last 5-6

afk5min wrote:

SpooKNoF wrote:dude just check how many null checks we using for the same code its at last 5-6

> The style is called 'Defensive Programming'.
> Defensive Programming
> Defensive

afk5min wrote:

SpooKNoF wrote:dude just check how many null checks we using for the same code its at last 5-6

> The style is called 'Defensive Programming'.
> Defensive Programming
> Defensive

SpooKNoF wrote:

Zoey76 wrote:

afk5min wrote:P.S. Sure, you start 'finding more serious' stuff later on. But in the big picture, the whole l2j gameserver codebase is shit. Mostly because it became acceptable to rush feature implementations and then forget about them, reusing old code exclusively by copy-pasting. That wouldn't be so bad, but it only takes one person to make a mistake in his code, and that mistake will quickly propagate everywhere, but not by design, but as separate copies.

I'm not agree on that, we spend almost two years reworking the core and we let features development in the back sit to make the core better, if you are going to say something try to keep it real

@SpooKNoF remove the check, once you get an NPE, check how the reference got nulled, fix the issue from the root and share a patch after it's tested, that way you would remove a useless null check without causing NPE flood in server logs.

dude just check how many null checks we using for the same code its at last 5-6

Zoey76 wrote:

afk5min wrote:P.S. Sure, you start 'finding more serious' stuff later on. But in the big picture, the whole l2j gameserver codebase is shit. Mostly because it became acceptable to rush feature implementations and then forget about them, reusing old code exclusively by copy-pasting. That wouldn't be so bad, but it only takes one person to make a mistake in his code, and that mistake will quickly propagate everywhere, but not by design, but as separate copies.

I'm not agree on that, we spend almost two years reworking the core and we let features development in the back sit to make the core better, if you are going to say something try to keep it real

@SpooKNoF remove the check, once you get an NPE, check how the reference got nulled, fix the issue from the root and share a patch after it's tested, that way you would remove a useless null check without causing NPE flood in server logs.

lord_rex wrote:

afk5min wrote:

SpooKNoF wrote:dude just check how many null checks we using for the same code its at last 5-6

> The style is called 'Defensive Programming'.
> Defensive Programming
> Defensive

Leave him, as I did and continue our peaceful discussions. He knows only the "Parrot Style".