error: skipping "/var/log/interchange/error.log" because parent directory has insecure permissions (It's world writable or writable by group which is not "root") Set "su" directive in config file to tell logrotate which user/group should be used for rotation.
perl use on RHEL and CentOS.
Shockingly, this has been broken since this commit: commit 13dcf43 Author: Mike Heins <firstname.lastname@example.org> Date: Fri Sep 13 20:46:21 2002 +0000 But Perl didn't notice it until version 5.18.0, where it throws many errors like this at Interchange startup time: Calling UI...UserTag 'content_modify' subroutine failed compilation: Global symbol "$table" requires explicit package name at [...]
module you have. Modify sha1 filter to use this. New globally available routine &Vend::Util::sha1_hex.
…lity as [warnings]). Useful for localization. Example: [error name=username set="That was a bad username. Please contact us at %s or %s." param.0="(800) 555-1212" param.1="(512) 555-1212" ]
New body tags for [more-list] can be used to "salt" more links with SEO-driven content. These tags can turn a link like: http://foo.com/scan/MM=[md5x:x].html?mv_more_ip=1&mv_nextpage=results&pf=sql into http://foo.com/scan/leisure-bikes/traditional-bikes/page-2/MM=[md5x:x].html?mv_more_ip=1&mv_nextpage=results&pf=sql While the new tags will work inside any more list, they are really only valuable in conjunction with the use of Permanent More, so those higher-value links cached by search engines actually lead to viable resources. See release notes and documentation on Permanent More if you are unfamiliar with it, or you can see my blog post on it at: http://blog.endpoint.com/2012/01/interchange-search-caching-with.html The new body tags are: * [more-pretty-url] This container tag includes the data of interest you'd like to plug into the more-list URLs to make them "pretty". The value can contain anything matching a-z, 0-9, /, _, and -. If any characters other than those described are included, they are converted in contiguous blocks to a single -. Any contiguous multiples of - (after conversion) and / are collapsed into a single one. Finally, any - and / are stripped from the head and tail. What this does is allow a flexible vat into which you can throw meaningful data (without having to test or sanitize at all) which will produce a path fragment of relevant terms that is put into the more-list URL between the "scan/" and "/MM=..." paths within overall URL. Example: [more-list] [more-pretty-url]men's clothing/pants/bermuda shorts[/more-pretty-url] ... [/more-list] Produces a more link with the following format: http://foo.com/scan/men-s-clothing/pants/bermuda-shorts/MM=a52a53...html?... Even though the body of a [more-list] is interpolated on reparse, the contents within [more-pretty-url] specifically are interpolated. It is expected that you would have some algorithm you would use to determine the pretty parameters and those would be rendered via ITL within the [more-pretty-url] tag. * [more-incl-pageno] Additionally, to provide more context and uniqueness to the URLs, you can optionally request that the page number be included as the last path segment of the "pretty" URL. This can either be the default (which is "page-N"), or can be any customized string you define, provided that you include the sprintf %d integer interpolation flag so that the page number will be included. Unlike [more-pretty-url], the contents of this tag are *not* pre-interpolated. Expanding the above example: [more-list] [more-pretty-url]men's clothing/pants/bermuda shorts[/more-pretty-url] [more-incl-pageno][/more-incl-pageno] ... [/more-list] Produces a more link with the following format for page 4: http://foo.com/scan/men-s-clothing/pants/bermuda-shorts/page-4/MM=a52a53...html?... Or a customized page path with: [more-list] [more-pretty-url]men's clothing/pants/bermuda shorts[/more-pretty-url] [more-incl-pageno]search page %d[/more-incl-pageno] ... [/more-list] produces new URL for page 4: http://foo.com/scan/men-s-clothing/pants/bermuda-shorts/search-page-4/MM=a52a53...html?...
Perl and we should not add new change based on that.
Now supports PayPal Express Checkout as well, without requiring any additional payment modules. Supports Payflow's fraud screening service. Sends additional order details to Payflow, including shipping, tax, basic item details, etc. Brings module in line with general Interchange payment module coding practices, such as gen_order_id(). Adds support for delayed captures. Raises timeout to practical level, akin to other payment modules. Sends new request-id for each backend transaction (capture, void), so as to avoid "duplicate transaction" errors.
… versions of Safe.
to separate, presumably insert-only, database table. Designed to allow administration personnel to look at passwords, without allowing access to web-connected systems. Or perhaps more properly, to check prior MD5-encrypted password values for repeat passwords. Docs in UserDB.pm POD.
…eins. from: http://www.icdevgroup.org/pipermail/interchange-users/2011-February/052657.html "Looks like it would be a good thing to do in some cases. However, if as with more and more catalogs, you use Web 2.0 features that call the database via IC, then you will potentially break those functions." Set Pragma set_httponly in catalog.cfg and you would be good to go.
Basically a row_hash() function that uses a foreign key column instead of primary key. Example: code ticket type 1234 9876 request 1233 5555 response my $rec1 = $db->row_hash('1234'); my $rec2 = $db->foreign_hash('ticket','9876'); Each of the above would return the same record hash. Of course the foreign column has to be unique, or an array will be returned.