This Bugzilla instance is a read-only archive of historic NetBeans bug reports. To report a bug in NetBeans please follow the project's instructions for reporting issues.
Should attempt to hide email addresses that appear on nb.org from spam harversters etc. Eg use the common format of : user at host dot domain in place of a full working email add. All pages on www are parsed already in the publishing system, should be easy to add an address obfuscater.
I don't think the dot matters a lot. There are lots of dots in the web. About the '@', I suggest '()', like: asrail()netbeans.org You can add spaces or something: asrail () netbeans.org But I prefer without spaces. If the spammer desire he cab configure the program to look for whatever he wants, as the most looks for *'@'*, it's the most important to get rid of. I know people with mails with a lot of dots, so it's not very nice: someone at provider dot subprovider dot domain dot subdomain Just my one cent. ;)
OK this is mostly done. I've settled on a combination approach : - encoding email addresses in a mix of decimal and hexadecimal encodings, which are interpreted and displayed correctly by a browser, but appear as html entities (and hence not immediately recognisable as email addresses) in the source; - additionally using javascript to further hide those encoded addresses, where javascript is available. If a browser does not support javascript or has it turned off, the "plain" hex/dec encoded version is used. Since a harvester (probably, so far) does not have a javascript engine it will not be able to interpret the js in the source. Note that this wont really decrease the amount of spam anyone is currently getting, since your address is already out there on spam lists or CDs or whatever. However it should stop any more addresses from being harvested from netbeans.org. Neither of these approaches are guaranteed, but hopefully enough to make it not worth anyone's while to decode. At least for a while. I'll start committing updated html tonight.
This is a little tricky :-) I am doing this during the automated testwww -> www publishing process, which pareses each html file for publishing. There are a few cases that I can't catch, or rather are too resource-intensive to be worth catching IMO : 1) mailto's with newlines, eg <a href="mailto:someone@somewhere.com">Mail me about it</a>. 2) mailto's or email addresses inside html comments, like <!-- <a href="mailto:someone@somewhere.com">Mail me about it</a>. --> To catch and correclty handle these significantly increases the time taken to process the page. I would rather find the places those exist and manually correct them on testwww. So - I'm commiting the changes to www, but slowly. Pls let me know if anyone notices any problems.
This is now done and live. All email addresses on www are now hidden. I checked pretty carefully and as far as I can tell all is working. If you notice anything not working pls let me know. Unfortunately addresses on testwww still appear in plain text :-/ I'll work on that. I think www is the higher priortiy site so I'm dropping priority.
We recently moved out from Collabnet's infrastructure