Quote:
Originally Posted by Dan..
P.S. I fixed that clock_gettime() issue. Thanks for the report!
|
No problem! Every fixed bug is a good bug
Quote:
Originally Posted by Dan..
- visual logging (HTML log) - Yeah sure, because a pile of shit (HTML tags) is always better. Let's also take into consideration that there are many browsers that can open a .HTML file that's over 100mbs.
|
Using the HTML log to log debug messages is sure silly.
Users prefer such visual logging (as far as I saw) because it is easier to find code mistakes with it.
Quote:
Originally Posted by Dan..
- multi-threading - I lol'd in real life when I read this. The way I handle multi-threading is the real way. Your plugin fetches malformed results. Anyway, you can manually implement a pool in Pawn or wait for me to do it in v2.7.
|
Why should the results be malformed? Before being such an arrogant smart-ass you should really consider reading the MySQL documentation.
Quote:
Originally Posted by MySQL documentation
Many threads can access different result sets that are retrieved with mysql_store_result().
|
Source:
http://dev.mysql.com/doc/refman/5.1/...d-clients.html
Quote:
Originally Posted by Dan..
? ORM-like system (R33+) - proof or gtfo :hurr:
|
Are you 14 or something?
Quote:
Originally Posted by Dan..
+ y_inline support - I didn't have any plans on implementing it, but now that you mention, it's there.
|
Good job.
Quote:
Originally Posted by Dan..
+ unthreaded queries - People don't have to put much effort to convert their old scripts and get rid of bugs. In some cases (especially in OnGameModeInit and OnGameModeExit) unthreaded queries are required.
+ PostgreSQL support - Cool or what?
|
Yes, I already mentioned these in my previous post.
Quote:
Originally Posted by Dan..
+ "live" (a.k.a. non-cached) queries - Especially when a large amount of data has to be loaded.
|
But I forgot to mention this one, sorry!
Quote:
Originally Posted by Dan..
+ striking resemblance with SQLite api - It comes in handy for users that want to convert their old scripts.
|
Some macros and define's could also do this. Okay, the natives of your plugin and the SQLite api are similar and thus not every time a macro/define can be used there. Well, I'm restricted to the current natives, I can't just change their syntax.
Quote:
Originally Posted by Dan..
+ proper result handling
|
What's so wrong with the way I handle results?
Quote:
Originally Posted by Dan..
+ parameters can be pushed by reference - Great, ain't it?
|
In what scenario can this be useful?
Quote:
Originally Posted by Dan..
+ multiple result sets
|
Again, as I said (indirectly) in my previous post, my plugin also offers this feature. I didn't mentioned features our both plugins have.
Quote:
Originally Posted by Dan..
Your plugin might be faster (I still don't believe your results), but mine offers reliability.
|
Why shouldn't you believe my results? Like I need to lie about some benchmarks.
Yes, your plugin seems stabler than mine, but I'm always working on the stability.
Quote:
Originally Posted by Dan..
I also did a quich benchmark, output:
Code:
GO!!! (Dan) | now = 45693129 (start)
connected | now = 45693134 | now - last = 5 | now - start = 5
scheduled | now = 45693136 | now - last = 2 | now - start = 7
done | now = 45693137 | now - last = 1 | now - start = 8
executed | now = 45702784 | now - last = 9647 | now - start = 9655
----------------------------------------------------------------------------
GO!!! (BlueG) | now = 45729908 (start)
connected | now = 45729913 | now - last = 5 | now - start = 5
scheduled | now = 45729915 | now - last = 2 | now - start = 7
done | now = 45729919 | now - last = 4 | now - start = 11
executed | now = 45742206 | now - last = 12287 | now - start = 12298
|
As I posted my results before, R31/32 is pretty slow ATM.
You seem like a good plugin developer, but with a really bad personality. Instead of being so rude/aggressive you should start being productive and help everyone (not just the users of your plugin(s)).