Perl 6 - the future is here, just unevenly distributed

IRC log for #bioperl, 2014-03-17

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

All times shown according to UTC.

Time Nick Message
10:53 Jayk97 joined #bioperl
17:21 carandraug joined #bioperl
17:21 carandraug joined #bioperl
18:27 github [bioperl-live] cjfields pushed 1 new commit to v1.6.x: https://github.com/bioperl/bioperl-live/comm​it/7d31bd5626b02c7fcc920738aa17c2e010e2b00f
18:27 github bioperl-live/v1.6.x 7d31bd5 Chris Fields: Bite the bullet, require Clone....
19:38 github [bioperl-live] cjfields pushed 1 new commit to v1.6.x: https://github.com/bioperl/bioperl-live/comm​it/0711754236110ac3406780012462f0474cbe6e60
19:38 github bioperl-live/v1.6.x 0711754 Chris Fields: use Clone only, and require it
20:06 eliasp joined #bioperl
20:06 eliasp hi
20:07 eliasp just setting up an Ensembl environment for one of our devs … does anyone know, why Ensembl requires such an old BioPerl version (1.2.3, 11 y old)?
20:08 eliasp have there been major changes in BioPerl which keep the Ensembl devs from upgrading?
20:20 trs eliasp: I don't know the answer to your question, but are there tests for ensembl?  if so, try running them on a newer bioperl and see what happens.
20:21 eliasp trs: k, might try that… set it up according to their requirements for now, but don't want to keep this extra environment around all the time
20:39 carandraug eliasp, the best thing is to ask ensembl really. Most likely they have just decided to test on one version and don't bother with it. And they didn't even released it on CPAN ><
20:39 eliasp carandraug: yes… that bothers me as well…
20:39 carandraug I really don't like about it
20:40 eliasp such projects make our lives simply hard for nothing
20:40 carandraug but unless users bother them with this type of stuff, and let them know that's now how you develop software, they have no reason to change their ways
20:40 eliasp then there's no proper Makefile.PL/Build.PL … manual build/installation required etc.
20:41 eliasp carandraug: in case our devs decide to stick with it, I'll definitely do…
20:41 eliasp carandraug: if it's just a one-time thing, I won't put that effort into it
20:41 carandraug they developed a framework and then make it really hard for the users. Yes, they provide a virtual machine that works out of the box, but that's just a workaround for poor software development.
20:41 eliasp carandraug: exactly
20:42 eliasp integration is something they've obviously never heard of
20:42 carandraug eliasp, so bitch about it to them. They won't change it for just one person, so you use NCBI, and when publish your results, leave a comment on the results that you used NCBI instead of ensembl because of this. Eventually they'll get their shit together
20:43 carandraug they'll only hear you, when you publish that they're crap, because that's when their bosses will read it and come make them questions
20:44 eliasp well, I won't publish anything… I'm just a former Perl-Dev, now doing mostly sysadmin stuff here… but in case our Bio devs will publish something based on this project, I'll definitely make sure these information + all the details around it go into the publication
20:46 carandraug eliasp, it's not you. It's the biologists. And if on their projects they say that analysed data obtained from the NCBI servers, they may choose to explain why. At least I have been asked that. That's a good time to explain it. I already explained it on my, but my supervisor will most likely play it down a bit (apparently I'm a bit too acid)
20:47 carandraug eliasp, at the moment, this is what I have http://bpaste.net/show/D8IKLHwkX0v9mp97PkIe/
20:48 carandraug eliasp, if you have to report to a biologists, you may find it useful
20:48 eliasp carandraug: perfect… would have written something similar to your last paragraph and forwarded it to our biologist
20:49 carandraug eliasp, one thing I forgot to note there. is that while NCBI provides a SOAP interface, Bioperl never got around to implement it. It actually just uses cgi
20:49 eliasp carandraug: for now, I'll finish the Ensembl setup here, but I'll talk to our biologists tomorrow, whether NCBI would be an alternative
20:49 eliasp plain CGI? ugh
20:50 eliasp what would be a reason to use Ensembl in the first place at all?
20:50 eliasp at least from my technical point of view, there are only downsides to it so far
20:50 carandraug eliasp, of special note is the lack of lucene server. With NCBI, you can have your program make the exact same search done by their servers. In ensembl, you need to download all the data and set up the search server yourself
20:51 carandraug eliasp, well, for starters there is the EU vs EUA thing. And it's more than just nationalism. They have different looks about what to do with data
20:52 carandraug NCBI is all about hoarding data. They collect all the crap, no distinction made. Ensembl does a better job at curating their data but also ends up with less
20:52 eliasp hmm, EU/EUA? any pointers for a non-biologist, what this referes to?
20:52 eliasp ok
20:52 carandraug eliasp, NCBI runs in EUA servers. Ensembl is Europe
20:53 eliasp ah, ok
20:53 trs EUA = USA?
20:53 eliasp EUA as in Europe/America
20:53 eliasp and EU = Europe only?
20:54 carandraug eliasp, yes. Sorry EUA is Estados Unidos da America (I'm Portuguese). I meant USA
20:54 trs http://en.wikipedia.org/wiki/EUA helped out here :)
20:54 eliasp aaah, ok ;)
20:54 carandraug so, USA vs EU
20:54 eliasp ok, now it makes sense… never heard EUA before ;)
20:54 carandraug but that should not matter al all!!!
20:54 eliasp sure
20:55 carandraug what matters is the level of curation and amount of data they have
20:56 carandraug then NCBI provides only access to their servers. Emsembl API also provides a library to handle the data which in my opinion should be independent of where the data comes from. But why would they bother to create a framework that handles other servers?
20:57 carandraug eliasp, lol! I found the original thread where I asked about the lucene server on their mailing list (I was just learning how to program back then) http://lists.ensembl.org/piper​mail/dev/2011-July/006486.html
20:59 eliasp :)
22:53 KCL joined #bioperl

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