Camelia, the Perl 6 bug

IRC log for #bioperl, 2011-03-16

| Channels | #bioperl index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:36 bag_ joined #bioperl
00:57 dbolser who's in charge of the bioperl wiki?
01:01 dbolser 403 Forbidden : error occurred when accessing http://www.bioperl.org/w/api.php after 6 attempt(s)
01:01 dbolser sucks
01:03 dbolser 403 Forbidden : error occurred when accessing http://gmod.org/w/api.php after 6 attempt(s)
01:03 dbolser sucks too
01:05 bag_ left #bioperl
04:40 andrei_ left #bioperl
07:03 cassj left #bioperl
07:19 bag_ joined #bioperl
08:08 bag_ left #bioperl
08:16 cassj joined #bioperl
09:27 ank left #bioperl
09:30 ank joined #bioperl
13:40 cassj left #bioperl
13:42 pyrimidine dbolser: the API has been like that for a while.  Let me see if I can fix it.
14:00 pyrimidine dbolser: maybe try https://bioperl.org/w/api.php
14:24 dbolser pyrimidine: works
14:24 * dbolser wonders why
14:25 dbolser pyrimidine: know about the gmod wiki api?
14:25 pyrimidine dbolser: probably something to do with permissions on the file server
14:25 pyrimidine re: gmod, that's scott cain's setup
14:25 dbolser weird... api.php should be the same perm as index.php
14:26 dbolser pyrimidine: OK, it has the same error
14:26 * dbolser queries sexy wiki data...
14:26 pyrimidine dbolser: there is a good reason to make it hard to access the API
14:27 dbolser the author of MediaWiki::API is bitching because I had teh audacity to log some bugs against his module
14:27 dbolser pyrimidine: is that a question or a statement?
14:27 pyrimidine statement
14:27 dbolser please explain
14:28 pyrimidine just common sense: if we allow anyone with a login to access the API, it can hammer the server
14:28 pyrimidine not to mention spam issues
14:28 pyrimidine (which we already have a problem with)
14:28 pyrimidine I don't think this will be a problem with OpenID now
14:28 pyrimidine but we'll see
14:29 dbolser pyrimidine: you can't edit via the api if you can't edit via the http interface
14:29 dbolser yup, you can hammer the server, but only marginally more malicously than via http
14:30 dbolser <error
14:30 dbolser sorry: <error code="readapidenied"
14:31 dbolser it's just common sense
14:31 dbolser ;-)
14:31 dbolser wow I'm bitchy
14:31 pyrimidine dbolser: sorry, not sure what you mean
14:31 dbolser sorry, about what?
14:31 pyrimidine once you have a login for the HTTP interface, then you could potentially edit a lot of pages
14:31 dbolser yes
14:32 pyrimidine we think the latest batch of spam is probably human-initiated (human creates the account, bot spams the rest)
14:32 dbolser ic, but I don't see how that applies to api
14:33 dbolser you can specifically disallow edits via the api if it makes you feel better
14:33 pyrimidine well, that opens up the ability to do edits on a much wider scale
14:33 dbolser pyrimidine: not really
14:33 dbolser most of the bots are scripted against the http
14:33 pyrimidine okay, whatever
14:33 dbolser ok fine
14:33 dbolser ;-)
14:33 dbolser sorry whine so much
14:34 dbolser the openID login is failing for me on bioperl.org
14:34 pyrimidine have you tied your old account to an openID
14:35 dbolser not yet, just clicked 'login with openID' then clicked on the google icon, I get teh following error:
14:35 dbolser The page that you requested is invalid.
14:35 pyrimidine http://www.bioperl.org/wiki/Special:OpenIDConvert
14:35 dbolser https://accounts.google.com/o/openid2/au​th?st=byKixbcSobH1q59PgCB0&amp;hl=en-GB
14:35 dbolser it should happen automatically?
14:35 dbolser pyrimidine: you've limited page views to logged in people only?
14:35 dbolser is that intentional?
14:36 pyrimidine no
14:36 dbolser previous experience with openID, login with open id, then get a screen saying 'link this ot an existing account'
14:36 dbolser pyrimidine: can you logout and try that link?
14:36 pyrimidine ah, yes.  Let me see if I can fix that
14:36 dbolser ty
14:37 andrei_ joined #bioperl
14:37 andrei_ hey guys
14:37 andrei_ I have a problem with this code : https://gist.github.com/872576
14:38 pyrimidine dbolser: I've reactivated the login prompt
14:38 andrei_ I am trying to extract the CDS genes and print the sequence into a file along with the length start and stop
14:39 andrei_ but that code prints 2 times each gene and it also prints the source primary tag
14:40 dbolser pyrimidine: ty. can I configure it so openID doesn't auto log out?
14:40 pyrimidine I don't think it auto logs out anyway
14:41 pyrimidine at least, it doesn't for me
14:41 dbolser ok, cheers
14:42 dbolser great, that works now
14:42 dbolser pyrimidine: you want me to research the 'no edits via the api' option for you?
14:43 pyrimidine no need if the user edits are tied to an OpenID or a specific ID
14:43 dbolser andrei_: are you sure the seq_object only has one sequence?
14:43 pyrimidine dbolser: creation of new accounts is disabled, so new users have to use OpenID
14:43 andrei_ yes
14:44 dbolser pyrimidine: ok, I'm just wondering if my script will play nicely with the https, and if not I can ask for normal http access to the api
14:44 andrei_ and what it prints is exactly the same 2 times
14:44 dbolser andrei_: so there are apparently two identical features in teh seq_object?
14:44 dbolser for my $feat_object ($seq_object->get_SeqFeatures)  # <- goes twice?
14:45 pyrimidine dbolser: not sure, but I would like to know for my own reasons (want to move HOWTOs back into the dist, then have a bot sync the wiki with updates from github)
14:45 dbolser pyrimidine: certainly scriptable, but ... what happens when a user edits the howto in the wiki?
14:46 andrei_ I think that is the problem at the for loop
14:46 pyrimidine dbolser: those won't be editable
14:46 pyrimidine dbolser: but the talk page will be
14:47 pyrimidine anyone who wants to edit the docs will have to do so via the github repo
14:47 dbolser hmm
14:47 pyrimidine there was a discussion on the mail list on this
14:48 pyrimidine the general consensus is that having docs tied directly to the software is a very good thing
14:48 dbolser ic
14:49 pyrimidine but it will be a maintenance nightmare to have two copies floating around, hence us syncing the two and locking the wiki page down
14:49 pyrimidine (it will represent the latest code in master branch)
14:49 dbolser yes, one has to be locked I agree
14:49 dbolser (or deleted)
14:49 pyrimidine of course, this hasn't happened yet, so we'll see
14:50 dbolser OK, I logged in via OpenID, but the api is giving me the same error:
14:50 dbolser https://www.bioperl.org/w/api.ph​p?action=query&amp;meta=siteinfo
14:50 dbolser <error code="readapidenied" info="You need read permission to use this module" xml:space="preserve">
14:51 dbolser I'll check if api plays with openID
14:51 pyrimidine okay, then this is definitely seems like a permissions problem of some sort
14:51 dbolser else, we could create one new user with api rights
14:57 dbolser pyrimidine: for your reference, I've just uploaded my (not very well written) 'wiki manipulation' perl code here: https://github.com/dbolser/BioWiki
14:58 dbolser it uses this PM extensively: http://search.cpan.org/dist/Med​iaWiki-API/lib/MediaWiki/API.pm (good luck logging bugs ;-)
14:59 dbolser most people use the pywikipedia robot framework though (which doesn't use the api)
15:22 andrei_ can anyone give me an example of how to extract features from a genbank? I tried http://github.com/bioperl/bioperl-live/​tree/master/examples/seq/extract_cds.pl , but the link is down...
15:26 pyrimidine I love that github 404 page
15:28 kai andrei_: hang on, let me see if I have that code published somewhere
15:30 kai andrei_: the gist is grep ($_->primary_tag eq 'CDS_motif', $seq->top_SeqFeatures);
15:30 pyrimidine andrei_: not seeing that in the master branch, where did you get the link from?
15:31 andrei_ I got it from http://www.bioperl.org/wiki/Bioperl_scripts
15:32 andrei_ it is in the other section
15:32 kai so use Bio::SeqIO; my $seq_obj = Bio::SeqIO(-file=>"filename", -format=>"genbank"); while(defined(my $seq = $seq_obj->next_seq)) { my @cds_list = grep ($_->primary_tag eq 'CDS_motif', $seq->top_SeqFeatures); #do stuff; }
15:32 pyrimidine ah, I see it
15:32 kai er, sorry
15:32 kai CDS, not CDS_motif
15:32 kai cut&paste error
15:32 kai not sure if it's the nicest way to do this, but it works :)
15:33 andrei_ what is grep?
15:33 kai http://perldoc.perl.org/functions/grep.html
16:21 andrei_ left #bioperl
16:55 pyrimidine dbolser: got it working: http://www.bioperl.org/w/api.php
16:56 pyrimidine dbolser: was a setting in LocalSettings.php that was off
16:56 dbolser pyrimidine: cheers dude!
16:56 pyrimidine :)
16:57 dbolser the wiki will now be part of the 'activity cloud' here: http://bifx.org/wiki/BioWIki
16:57 * dbolser goes to pester scottc
16:57 dbolser pyrimidine: feel free to fill in details here: http://www.bioinformatics.org/wiki/BioPerl_Wiki
16:58 dbolser (click 'edit with form') .. btw... can you install smw on bioperl.org?
17:10 pyrimidine dbolser: I changed the info about Jason (don't think he wants his new email linked to the wiki, and he's not as active as he used to be)
17:10 pyrimidine dbolser: what is smw?
17:43 mzgrideng joined #bioperl
17:43 mzgrideng left #bioperl
17:45 mzgrideng joined #bioperl
18:00 andrei_ joined #bioperl
18:43 andrei_ left #bioperl
18:50 blossom joined #bioperl
18:51 blossom hi
18:51 blossom is there like a TEXT::Wrap for a fasta object?
18:53 pyrimidine blossom: do you mean, set the line length for the sequence data?
18:53 blossom yes
18:54 pyrimidine Bio::SeqIO::fasta has a setting for that, let me see
18:55 blossom yes found it
18:55 blossom width
18:55 pyrimidine yep
18:55 blossom thanks
19:11 blossom ->spiced_seq->seq() returns which object?
19:32 pyrimidine blossom: spliced_seq returns a Bio::Seq, seq() from a Bio::Seq returns a string
19:33 blossom ok so I cannot use width on it
19:53 blossom left #bioperl
20:01 pyrimidine ah, blossom left.  There is a difference btwn Bio::Seq and Bio::SeqIO :)
20:07 rbuels these people.  irc proxies.
20:38 pyrimidine http://fuckyeahnouns.com/robert%20buels
20:38 * pyrimidine is carrying over a #moose meme
20:50 rbuels o_O
20:50 rbuels it's very blurry, whatever that is
20:50 rbuels somebody's poster about sgn?
20:50 rbuels pyrimidine: ^^^
20:52 rbuels weird
21:00 cassj joined #bioperl
21:04 cassj left #bioperl
21:18 driley left #bioperl
21:21 driley_ joined #bioperl
21:23 driley_ left #bioperl
21:28 driley_ joined #bioperl
21:36 driley_ left #bioperl
21:36 driley_ joined #bioperl
21:36 driley_ left #bioperl
21:37 driley joined #bioperl
21:37 driley left #bioperl
21:38 driley joined #bioperl
21:39 driley_ joined #bioperl
21:39 driley left #bioperl
21:39 driley_ left #bioperl
21:47 driley joined #bioperl
21:50 driley_ joined #bioperl
21:54 driley left #bioperl
21:57 driley_ left #bioperl
21:58 driley_ joined #bioperl
22:04 driley_ left #bioperl
22:04 driley_ joined #bioperl
22:26 driley_ left #bioperl
22:30 driley_ joined #bioperl
23:05 ank left #bioperl
23:06 ank joined #bioperl
23:28 mzgrideng left #bioperl
23:52 dbolser thats very weird
23:52 dbolser pyrimidine: if you want to limit the powers of the api on the bioperl wiki: http://www.mediawiki.org/wik​i/API:Restricting_API_usage
23:53 dbolser e.g. $wgEnableWriteAPI = false; # In LocalSettings.php
23:54 dbolser for a crash course in setting MW perms, see this note:
23:54 dbolser http://www.mediawiki.org/wiki/Ex​tension:ReCAPTCHA#Configuration
23:55 dbolser pyrimidine: SMW is what you use to create forms and queries in the wiki

| Channels | #bioperl index | Today | | Search | Google Search | Plain-Text | summary