lampstax

According to this tool here ( http://www.bx.com.au/tools/ultimate-php-error-reporting-wizard ) .. the only way I could get to 4983 .. is E_ALL & ~E_NOTICE & ~E_COMPILE_WARNING & ~E_USER_NOTICE & ~E_DEPRECATED & ~E_USER_DEPRECATED BUT error log is still showing notices .. is there any other level it could be ?

When I use ini_get to check my error_reporting level, I get a weird value ( 4983 ) that I can't find anywhere by googling. Also not on the list of possible levels here ( http://itech.hubpages.com/hub/php-error_reporting ) .. does anyone know how to determine what exactly this level is? Thanks!

Thanks for everyone's suggestion, I have been slammed at work so it took me a while to get back to this project. From playing with the demo, Margento is my front runner. It has built in user group functionality so I think i can build off that much faster than customizing OSCommerce or Hika. Will play around with that more and let you know if I run into any issues. Cheers!

Hey guys, I'm looking for a shopping cart ( preferably open source and PHP ) that supports a few uncommon features. I spent most of yesterday digging through the various php carts ( os commerce, zen cart, ect ) and didn't find much useful things. As far as I can tell, all these things will need to be customized so my best bet would be to find a cart that support modules .. almost like a Drupal or Wordpress of ecommerce. BUT hopefully I'm wrong and there is something out there already that supports: Advanced user roles controls / customization. One of the user role ( lets call them creators ) should be able to create a shopping cart list, save the cart and generate a code. Another user role ( lets call them purchasers ) should be able to input the cart code, and automatically have all items in that saved cart added to their own cart to proceed with checkout. A royalty program If a creator generated a cart code that lots of people use to purchase, we need to track this information to give them a royalty at the end of the month. Thanks for your help.

It is the URL of the page that I'm on. Agreed its not the best URL format but its what the application and I am working with. Its to link directly to a specific part of a page, however, I need to be able to detect what part that is. Any suggestions ?

Hey guys, I'm dealing a url that is similar to hxxp://domain/directory/?#{$id}. What I'm trying to do is extract $id from that. Its not available as a $_GET variable and $_SERVER['REQUEST_URI'] comes up as "/directory/?". Does anyone know how I could retrieve $id ? Thanks

Hey btherl. Thanks for writing all that out! I will give it a try tomorrow and see if I get lucky.

Hey guys, I'm running WAMP on Windows 7 and trying to pass a command to mogrify. I tried passthru(), system(), exec() but none worked. A sample command is convert -format jpeg -extract 2048x1536+0+0 +repage -thumbnail 100x75 C:/wamp/www/alpha.offernation.org/uploads/properties/property_original_1277852331.jpg C:/wamp/www/alpha.offernation.org/uploads/properties/property_original_1277852331_100x75.jpg I know the command itself works because if I run it directly from the DOS prompt it works. When I run with a simple command like 'dir': $last_line = passthru('dir', $retval); I get the correct output .. but when I replace 'dir' with the command above .. i get this: $last_line = 4 and $retval = NULL I think it could be a permission issue with windows 7, but I'm not sure how to troubleshoot. Google is not being very helpful either. Can anyone point me in the right direction? Thanks

Hi guys, Posted this in the #help IRC but getting no reply, so reposting it here. I'm having a weird issue with CURL .. hopefuly someone can help. Basically the app calls a few external url to retrieve xml and add to the db. Some API are slower than other so we should technically be using curl_multi, but for now, still just using normal curl. However, while retrieving the XML, curl seems to block all other HTTP connections to the site. By blocking the request, I mean other requests to the site ( completely unrelated portion of the site ) just get a 'loading' message until the curl finishes, then the other requested page loads. We have two domains on the same physical machines ( live env and dev env ). I've verified multiple times that CURL is not actually blocking request to the entire server. When CURL is executed on dev, live still loads, and vice versa. Google doesn't show anyone running into anything similar to this. So this might be a pretty obscure issue. But it seems like this would be a big issue if it was a problem with CURL so this might be a shot in the dark. Does anyone know if there is a curl_setopt option missing or something that causing this behavior ?

Looks like Mchl beat me to it.

Not 100% sure this is why you're getting 0 result, but try replacing the ORDER BY 'reads' DESC with ORDER BY `reads` DESC. I think the way you are quoting 'reads' makes mysql process that as a value, and not a table column.

Hi guys, I'm running into a pretty obscure problems. Once in a while, our website would go down for a few seconds and come back up. Server load would spike 20 or 30 times normal for a few seconds ( causing blank pages and 404s ) and settle back down. We have already checked slow logs and optimized the queries so I'm pretty confident that its not a single bad query killing the db. We also noticed that during most of these spikes, mysql would hit max connection so my suspicion is that because of the way its coded, each hit to a certain part of our site creates multiple mysql connection, I need a way to map apache processes ( preferably with referring url ) to the number of mysql connections opened for each process and also see how long the connections are kept by the process. Maybe the connections need to be manually closed or something like that. SHOW FULL PROCESSLIST isnt doing enough for us. Does anyone know of any monitoring program that could give us a map view of the website hit to mysql connection or more debugging data for this type of problem? Thanks ahead of time.

Tried it too ways: DELIMITER | CREATE TRIGGER set_default_options AFTER INSERT ON authors FOR EACH ROW BEGIN SELECT SLEEP(2); PREPARE stmt_name FROM "UPDATE authors SET issued = '9' WHERE authorID = ?"; SET @test_parm = "2"; EXECUTE stmt_name USING @test_parm; END; | DELIMITER ; and even this didnt work DELIMITER | CREATE TRIGGER set_default_options AFTER INSERT ON authors FOR EACH ROW BEGIN SELECT SLEEP(2); PREPARE stmt_name FROM "UPDATE authors SET issued = '9' WHERE authorID = '2'"; EXECUTE stmt_name; END; | DELIMITER ;

Actually if I can get this trigger to work, it would solve my problems too, but I'm have some syntax issues DELIMITER | CREATE TRIGGER wpmu_new_blog_set_default_options AFTER INSERT ON wp_blogs FOR EACH ROW BEGIN SELECT SLEEP(2); SET @var := CONCAT('wp_', NEW.blog_id ,'_options'); UPDATE @var SET option_value = '0' WHERE option_name = 'option_x'; END; | DELIMITER ; I feel like I'm sooo close, but its not working. I"m not too familiar with triggers. I bet its something small that I'm not seeing. The MySQL error that I see is : ERROR 1064 (42000): You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near '@var SET option_value = '0' WHERE option_name = 'option_x';

You have an extra = there. It should be <?php $select = "SELECT * FROM jobs WHERE assigned_to like '%u%'"; ?>

Is it possible to create something in MySQL ( trigger ? stored procedure ? something else ? ) so that when any table is created on a DB thats has a name fitting a certain structure .. lets say like "user1_data", "user2_data" .. to insert a row into that newly created table? Ofcourse this structure is a pretty ugly db design ( not mine ) and it could be done easily a php / java / ruby script, the situation dictates that no script is used. I'm stumped. What about you guys?

Wow, thanks for the detailed reply. I think thats a good idea, almost like db 'versioning', which is really exactly what I need. I'm surprise that there's nothing out there like that already (commercial or open source). Might be a bit too time consuming at this stage of the project though, so I'll see if we can actually allocate some time and put some version of that tool on the to build list. Thanks

Hey guys, I have three db servers ( live and beta and dev ). Obviously as the name suggest, each serves a different purpose, and while the db structure are very similar, all three db have differences in them. Some more obvious than other. A new table in dev thats not yet in beta or live is easier to spot than a int value in some table has been changed from smallint to medium int. I need to find a way to compare two dbs or alteast two tables and find all the differences among them (atleast structurally, but data differences too would be just great). I could spend time building my own code that queries information_schema, but something tells me that I would be reinventing the wheel. Could any of you fine gents point me to a script / tool that does this ? Thanks ahead of time.

Hi, I'm designing a db for a new application that involves some real estate stuff. Its alot of data, and I was originally going to split them into 5 identical MyISAM tables: `property_condos`, `property_singlefamily`, `property_multi`, `property_land`, `property_others` and a MERGE table `properties_all` to rejoin them. I couldnt simply set INSERT_METHOD because its restrictive to first / last table only. I wanted to insert property into the table that correspond to its type. Such as condos all go into the `property_condo` table. So I relied on my script to do the job of selecting the correct table and doing the insert instead of inserting into `properties_all` and letting MERGE figure it out. Now my issue: Each table has an auto increment field `propertyID`, so if theres a row in each of the `property_x` table, there would be a `propertyID` = 1 in each table. Essentially I end up with 5 rows on the merge table with `propertyID` = 1. First time really working with MERGE tables so I didnt expect this. Is there any way to fix this? Perhaps by `linking` primary auto increment keys across table? Or is this a flaw in my db design?

I'm studying from the mysql test book and doing the exercises and I run into this question. ( see attached image ). The book claims that this table is in 3NF, but to me, it just seems a little off. Here's the exact text of the answer: - The City table is in the first normal form (1NF) because it doesn't have repeating groups within its rows. - It's in the second normal form (2NF) because it's in 1NF and all its non-index values depend fully on the primary key value (the ID column). - It's in the third normal form (3NF) because it's in 2NF and all its non-index values depend directly on the primary key (and not on some other non-index value). For example, since the only index is the primary key, the id number, I dont see how the value of Country Code 'AFG' depends directly on the ID, and not the Name column. I'm new to normalization so maybe thats why this doesnt look 'normal' to me.

Hey guys, I'm trying to create flow map (not sure if this is the right way to describe it) of a mysql database. All tables are connected somehow through referenced key, although not enforced references, since this most tables are MyIsam. I'm trying DB Visualizer, but I must be doing something wrong because I can get the tables to show up in their object view, but no references. Maybe because its MyISAM tables ? Or perhaps its just me being a newb to DBV. Here's a screen shot from the DB Visualizer site of what I'm trying to accomplish. http://i44.tinypic.com/2meu0ck.gif There are almost 100 tables. Any quick way to do this ?

Very very cool link Mchl. Thanks!

Hmm .. very likely thats what they meant. I was thinking along the line of buggy MySQL only issues, random made up example of a quirk would be something like: MAX() returns all rows when used in a complex joined with 2 LEFT JOIN and 3 RIGHT JOIN and 1 UPSIDE DOWN JOIN

Hey guys, I just blew an interview because of a question: "Please name some quirks specific to MySQL. What problem would they potentially cause and how you would prevent them?" Needless to say, my idiotic stumbling response costed me the interview. So hopefully we can get a list of quirky wierd MySQL behaviors to help other future interviewees or just save someone alot of time and headaches. This post is mainly for MySQL gods to chime in while us mortal worship and learn. LOL. So please, if you know of quirky complex join behavior specific to MySQL and not other dbs, post them up, but please no question if you just want to ask how to do LEFT JOIN on two tables (unless there's quirkyness to your query!).

Yeah. That makes sense. If it was a join or view .. I would have to add the new table into the join .. simple enough .. but it could make a huge although simple sql statement .. and i would have to make that change everywhere .. with Merge, i can just modify the merge table definition once a year! thanks!