[ot_207]

Originally Posted by joaquimandrade About set_pev/entity_set the results say that it is basically the same thing.

Ok then Emp` is right!

Originally Posted by Emp` I think there will need to be more testing on the actual fakemeta natives vs the engine natives. As for now the only argument has been fakemeta conversions (stocks that emulate the engine natives), but http://forums.alliedmods.net/showthread.php?t=65427 just puts it to fakemeta, there is not any fakemeta conversions so you can't completely say that fakemeta is slower in that situation.

[joaquimandrade]

Originally Posted by ot_207 Ok then Emp` is right!

He's right about what? Using fakemeta is never faster, even not being engine always clearly faster.

Basically:

All this years everybody (Except Hawk) thought that fakemeta way was faster. Fakemeta way is never faster and can be much slower.

[ot_207]

Originally Posted by joaquimandrade His right about what? Using fakemeta is never faster, even not being engine always clearly faster.

He said about natives functions like pev,set_pev,create_tr2,get_pdata_int are the same. He told us that stocks that emulate functions of engine use a lot of cpu. For example fakemeta util stocks use too much cpu usage if some of those stocks can be found in other modules than it is important to use those modules instead of the stocks.
Example:
Insead of fm_set_rendering it is best to use set_rendering from fun.

Anyway the test shows a slight difference.

[joaquimandrade]

Originally Posted by ot_207 He said about natives functions like pev,set_pev,create_tr2,get_pdata_int are the same. He told us that stocks that emulate functions of engine use a lot of cpu. For example fakemeta util stocks use too much cpu usage if some of those stocks can be found in other modules than it is important to use those modules instead of the stocks. Example: Insead of fm_set_rendering it is best to use set_rendering from fun. Anyway the test shows a slight difference.

No. He just said that more tests should be made. Here they are and fakemeta is never faster.

[ot_207]

Originally Posted by joaquimandrade No. He just said that more tests should be made. Here they are and fakemeta is never faster.

Oh, didn't read right.

[joaquimandrade]

More. (I'm assuming that pev_valid(id) is equivalent to is_valid_ent(id)). If they aren't, the results are invalid.

[SchlumPF*]

the more results i see the more i think that i really should convert my plugins to use fakemeta only if there is no way to avoid it except for huge codes which can be done by a few fm lines :/

[ConnorMcLeod]

Originally Posted by joaquimandrade More. (I'm assuming that pev_valid(id) is equivalent to is_valid_ent(id)). If they aren't, the results are invalid.

is_valid_ent only returns 0/1 when pev_valid returns 0,1 or 2.

/**
 * Checks the validity of an entity.
 *
 * @param entindex		The entity id to check.
 *
 * @return 			0 on invalid entity
 *						1 on entity is valid
 *						2 on entity is valid and it has private data (safe to use pdata natives on).
 */
native pev_valid(entindex);

[joaquimandrade]

Originally Posted by ConnorMcLeod is_valid_ent only returns 0/1 when pev_valid returns 0,1 or 2. /** * Checks the validity of an entity. * * @param entindex The entity id to check. * * @return 0 on invalid entity * 1 on entity is valid * 2 on entity is valid and it has private data (safe to use pdata natives on). */ native pev_valid(entindex);

Ok. So the comparison is not fair.
A question: is it normal needing the "return 2" case?

[SchlumPF*]

never used pev_valid(id) == 2 but i guess it is not really needed.
if you set some ents pdata you should already know whether it has pdata. but it may be usefull if you loop through different kinds of ents (classnames)


a new question which came to my mind
is there any difference between
client_preThink( plr ) and register_forward( FM_PlayerPreThink, "fwdPlayerPreThink" ) ?