2009-11-04:
[10:19] <p3k> just read yesterday?s messages about helma 1.7 and added some older and newer bugs i found to the target.[10:20] <p3k> obviously, these are some more than what zumbrunn communicated yesterday (and they were even without my latest additions today).[10:20] <p3k> but feel free to move them around to another target if necessary.[10:20] <p3k> http://dev.helma.org/bugs/buglist.cgi?query_format=advanced&short_desc_type=allwordssubstr&short_desc=&product=Helma&target_milestone=1.7.0&long_desc_type=substring&long_desc=&bug_file_loc_type=allwordssubstr&bug_file_loc=&bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED&emailassigned_to1=1&emailtype1=substring&email1=&emailassigned_to2=1&emailreporter2=1&emailcc2=1&emailtype2=substring&email2=&bugidtype=include&bug_id=&chfieldfrom=&chfieldto[10:20] <p3k> fieldvalue=&cmdtype=doit&order=Reuse+same+sort+as+last+time&field0-0-0=noop&type0-0-0=noop&value0-0-0=[10:22] <p3k> fsck hopbot chokes on the long url, so exceptionally here is a tiny one: http://tinyurl.com/ylx4s3r[10:23] <p3k> nevertheless, i too would be pleased to hear about the plans for 1.7 ? it has gotten a wee bit quiet around it lately.[10:40] <simono> thx for the overview. i additionally tagged #689 for 1.7.0 release[10:41] <Triotex> hi tobi! i talked with hannes about antville's isUrl and isEmail regex. maybe this would be also an idea for 1.7, if it's just a simple change in the String module[10:41] <simono> p3k: regardnig #686 the macro default problem[10:42] <simono> didn't try but me thinks it works if you write <% debug.macro | debug.filter default=foo %>[10:42] <simono> made sense for me[10:43] <simono> nope i'm wrong[10:43] <p3k> bug #689: yes right, you told me about that. but there is no bug report, is there?[10:44] <p3k> bug #686: if i remember correctly, i tried that, too[10:44] <p3k> but i would expect the default parameter to belong to the filter function then, anyway...[10:46] <simono> 689 how do you mean no bug rep?[10:46] <p3k> sorry, that was meant to be a reply to triotex[10:47] <simono> 686 hm.. i wouldn't have expected default to be passed through to filter, unless the macro outputs the default[10:48] <simono> maybe i'm looking at it wrong[10:49] <Triotex> @p3k i'm writing on the bug report for isUrl and isEmail and will add some examples[10:50] <p3k> excellent, triotex! thanks.[10:56] <Triotex> btw. just found another isURL-related bug: http://tinyurl.com/ygg77t8[10:56] <Triotex> this line should NOT return an URL, or? :-)[11:12] <p3k> afair it does not need to return the url but it tries to create a url and if that fails it returns false. code i would not write nowadays ? probably :)[11:23] <p3k> woohoo, hannes is at the bugs![11:27] <simono> yey hannes is squashing them left&right[11:28] <Triotex> i will look for some good test cases from other projects to get really bad but valid urls[11:33] <p3k> triotex: the longer i think about it the more i am about to think that the return of the url in isURL is by purpose. i bet the original method ? which probably did not use java.net.URL ? did this, too.[11:47] <simono> p3k: i can not reproduce #655 with the testapp[11:48] <simono> for me its the same output with and without repository.1 (your demo app)[11:48] <simono> especially if i change "Site#main" to "site#main" nothing gets rendered at all (as you would have expected)[11:49] <p3k> well, i would have to test it myself again. it is almost one year since the bug was filed... maybe it got already fixed in the meantime?[11:49] <simono> seems that way, hm[11:50] <simono> or might be a macos specific problem? but that would be very strange[11:52] <p3k> well why not? os x is using a case-insensitive fs by default[11:52] <simono> oh didn't know that, i thought it does it the unixy way[11:59] <Triotex> hmmm, just looked through the java.net.URL source and it performces just a basic URL check with some indexOf() searches which will also allow invalid URLs[12:00] <Triotex> drupal has the same bug report: http://drupal.org/node/124492[12:02] <Triotex> Boah checking if there is a valid URLs is like checking if there god exists :-)[12:03] <Triotex> 1) there has to be an IDN -> ASCII-URL conversion[12:03] <p3k> simono: just ran the test on my machine again and could not reproduce it either. looks like the bug is fixed.[12:06] <simono> p3k: yeah![12:11] <p3k> triotex: the IDN stuff will be interesting for antville because we currently have an issue with site names like http://mot?rhead.antville.org ? der browser will dann http://www.antville.org/xn--motrhead-p4a/[12:30] <Triotex> hmmm, if you create a java.net.URL("http://<iamcool>example.com"), what would you expect as return value?[12:31] <Triotex> java will strip the <iamcool> and will return just http://example.com :-)[12:38] <Triotex> there are some good methods in java 6, but helma does not require the newest java version. which i think is a bad behavior[12:38] <Triotex> java 5 reaches EOF soon[13:02] <Triotex> i filed the isUrl()-bug under ticket 690[15:31] <hannesw> simono p3k: if #655 works for both of you should I close it?[15:38] <simono> imo yes[15:39] <hannesw> k
In the channel now:
Logs by date: