| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Comments:
<0> tempest1: 1 is not a column that doesn't exist. 1 is a value which in a SELECT will be a column, and hence it exists. it's a calculated column that happens to not depend on any table columns.
<0> SELECT 284 AS "this is a column";
<1> arjenAU: yea, after using it in context i understand
<2> wow, arjen did it again
<0> jpm__: what?
<3> arjenAU rocks!
<2> you should win some sort of award
<4> it's for an invoice system so it's required.. but if it's possible it would save me lots of programming hours
<1> arjenAU: but select count(*) would also give you a desired result
<3> Nick2000, What is required?
<1> depending on how you wanted to script it
<0> tempest1: print out the table and having a secretary decypher the pwd also works.
<0> jpm__: so what did I do?
<4> that intstead of 200699 turning into 200700, it would go 2006100
<2> you trolled someone into oblivion
<0> jpm__: tick off a recalcitrant individual? any time. my pleasure.
<1> arjenAU: are you still under the ***umption that you started arguing with a reason?
<2> see, that seems kinda weird coming from the community person
<3> Nick2000, Why not just add a year column?
<2> but what do i know
<3> Nick2000, I can hardly think it would save "hours" of programming time.
<0> tempest1: nah not really. I just don't want to leave the lurkers here with the impression that you had a poin
<0> could cause unnecessary confusion and less than optimal implementation
<2> but he does have a point
<3> Nick2000, PRIMARY KEY( `year`, `id` )
<2> i use count(*) myself instead of select 1
<2> either one works
<1> arjenAU: I didn't have a point, but you had a ****ty way of pointing it out... waiting almost a page of retorts to get to it...
<3> SELECT 1 is better.
<2> so there's no point to be had?
<5> arjenAU: i did that but in logs i get "sql plugin: no result found"
<2> bullcrap
<4> yeah that's my last solution.. right now i got about 20.000 lines of code, and still the prefix solution would save some time :)
<2> whatever it's easiest to understand is better
<0> tempest1: I'm hopeful that ppl use their own brain to figure something out, rather than someone else's.
<2> and i think select 1 is kinda out there
<2> but again, what do i know
<1> arjenAU: then why bother responding to my question at all, if it's not to give an answer?
<0> magyar: like I said, sasl is a piece of evilness. I don't use it anymore.
<3> Nick2000, so 2006100 is the invoice number?
<3> Seems bad.
<3> tempest1, Ok cool it please.
<4> so the primary key solution wouldn't work?
<3> Nick2000, It would work, but I don't know your situation.
<0> tempest1: an answer may not be a complete spoonfeed, but rather a helpful guide to enlightenment. which, I do appreciate, is sometimes wasted and might as well be replaced with a brick.
<5> arjenAU: what cab i use?
<3> arjenAU, So, about that mysqldump and compatible=mysql40... With that flag, it leaves off auto_increment. Know why?
<1> arjenAU: your first response "<0> tempest1: nop. what's selected is irrelevant."
<5> can
<2> tempest1, don't feel bad, i agree with you
<0> jpm__: I don't have patience like Zak. actually, Zak doesn't either, really. he just makes a really good go of it.
<1> count(*) doesn't return any information about what is selected
<2> tempest1, arjen is on crack
<1> how am i supposed to think that's leading to something?
<4> ok.. What I need is for the counter to go from say 200699 to 2006100.. instead of the ordinary 200699 to 200700, follow me?
<3> Seriously guys...
<1> Sorry
<1> done
<3> Nick2000, Yes. But what is 200699? An invoice number?
<0> tempest1: COUNT(*) itself is a calculated value. why make the server do smth it doesn't need to. I'm not arguing its validity.
<4> yes :)
<2> arjenAU, it's easier to maintain imo
<3> Doesn't it seem "off" to have a different length invoice number?
<2> arjenAU, you see count(*) on that query, and you know what needs to be done afterwards
<6> Mmm... Im thinking about a table to use as map. The map would be a 100x100 square, so, 10,000 blocks, but some objects may use more than one block. Where can I read about how to 'create' it properly?
<1> arjenAU: calculate a value and return it, or store as many ones as there are results to your query in memory?
<2> you check if it is 1, then there is a match
<0> Nick2000: perhaps you need to conclude that an invoice number does not nessarily need to be generated by an auto-inc.
<2> let's all talk to arjen at the same time!
<0> tempest1: for the query in question, there's either 1 match, or none. it was a user/pwd check, and usernames would be unique.
<1> lol, is that the debate?
<0> tempest1: the complete resultset is irrelevant. you just want to check the # or rows. either 0 or 1.
<0> and you can do that without retrieving them, there's an API call for it
<1> right, if you read what i have said earlier it depends on how you want to script the check
<4> you're right.. should have thougt of that earlier, but now I'm kinda stucked with a lot of code depending on a column id's
<1> you might want to know if you have more than one person with a username and p***word, what another piece of information is to distinguish them
<0> Nick2000: change now, or regret later. your choice.
<7> being stucked kinda ****s
<1> i've seen a few sites that don't require unique usernames when the username isn't involved with a type of community, and just authentication to a system
<4> hehe.. yeah you go a point :) but.. anyone knows if it's possible with that prefix?
<3> You're kidding...
<1> no, it's not to say i agree with the design
<3> Nick2000, Use a YEAR.
<3> Not hard to add it in. And then CONCAT it. Make a VIEW :)
<3> Got 5.0?
<4> no 4.0.24
<8> views are cool
<6> Mmm... Im thinking about a table to use as map. The map would be a 100x100 square, so, 10,000 blocks, but some objects may use more than one block. Where can I read about how to 'create' it properly?
<7> tee hee, every site is that way with bugmenot :-)
<0> Nekusagi: you can use a table as a sparse map. only block that exist need to be stored.
<1> Nekusagi: x,y coords and object ids
<0> Nekusagi: so you get an X and a Y column, and other content.
<0> quite.
<6> that's a way, indeed...
<0> you can enhance that so coordinates may contain multiple things. whatever.
<0> Nekusagi: no. it's THE way.
<8> spatial data?
<0> but feel free to suggest another way which you believe might be better.
<0> bobfield: probably not relevant, and wouldn't change the design
<2> maybe the gis stuff?
<2> 100x100 square, that reminds me of google map :)
<3> arjenAU, http://hashmysql.org/1641. The second one has auto_inc while the first doesn't...
<0> Nekusagi: an object in more than one block should still be stored in a way so that you can check properly whether an object occupies a block. so, you store a reference to it for all blocks it belongs to.
<0> you could of course GIS the whole thing and then you can use the special GIS searches as JPM said. but that's probably overkill for the purpose.
<6> Im thinking about it now... Im trying to make a very weird single player 2D RPG game, web based, but there are a lot of things that Im still not sure about how to do ^^
<1> Nekusagi: also, if each object is equal in content you could have a content table and a mapping table so that you don't waste space repeating the content in mapping
<2> hey cool
<0> will: auto-inc as part of a composite does not use the regular auto-inc logic, it lcosk the table and does a MAX()+1. it's not what you want, probably.
<1> Nekusagi: it would be a cool experiment in AJAX
<8> MySQL Battleships
<2> mysql vs. pgsql?
<3> jpm__, kickban :)
<6> Yes, I'll be using Ajax, PHP too
<0> jpm__: now who's trolling
<9> did you guys keep talking about trucks?
<1> jpm__: postgre wins, but who's counting
<9> sorry I left
<3> firewire, No :(
<10> so ajax seems ot be sweeping the php world
<2> arjenAU, what! he said battleships
<3> arjenAU, How does that mode get sent to the dump?
<1> omnipresence: it's a buzzword, overused
<6> omnipresence,: only to those that don't know all the potential of php
<3> How would MySQL 4.0 use it?
<0> will: what mode.
<1> and it really has nothign to do with PHP itself
<1> just server side languages
<3> arjenAU, That auto-inc for composite keys
<3> Ajax is not server side... it's more client-side.
<0> will: it's a simply consequence of what you do. see manual if you wish. any version
<3> Javascript...
<1> well, as far as XML from the server
<3> eh
<1> it only depends on a server side language, not necessarily PHP
<3> ehh...
<0> I hear Ruby does nice ajax integration
<1> i've smoked alot of dro, give me a break :-p
<3> You should stop...
<1> probably
<9> arjenAU: ruby is the new fad
<3> arjenAU, So it's a MySQL 4.0 limitation? SInce the regular dump has auto_inc fine.
<9> you do her?
<3> Let's say... not a chance.
<9> why not?
<3> uh... I rather do Jake Gyllenhal
<9> I'm going to just stop right there
<3> I hear Jarhead was pretty good...
<0> will: what are you talking about?
Return to
#mysql
or
Go to some related
logs:
iptables use dash
yum remove bad mirror
#gimp
#suse
emvs ubuntu
WHAT IS EMISLO
neomagic 1024x800
#perl
#css
xircom realport ubuntu
|
|