Perl 6 - the future is here, just unevenly distributed

IRC log for #crimsonfu, 2015-03-05

crimsonfu - sysadmins who code

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

All times shown according to UTC.

Time Nick Message
01:34 pdurbin bah. just lost all my SMS history. stupid Android
01:35 pdurbin all technology hates me today: https://twitter.com/philipdurbin/status/573142546836275200
02:18 searchbot` joined #crimsonfu
02:22 JoeJulian joined #crimsonfu
03:20 pdurbin skay: talking about you at http://irclog.greptilian.com/sourcefu/2015-03-05 with sivoais
03:23 sivoais :-)
03:25 sivoais pdurbin: re: the SMS history, I have an app I installed that redirects all my SMS/call logs to my e-mail
03:26 sivoais I used to have something that sent them to me via XMPP too, but that stopped working for some reason O_o
04:40 skay o/
04:42 skay The runmycode.org group didn't want to collaborate on an open source version, and one of the PIs had me and a postdoc do work to make an open source reboot -- we didn't get to the point where we could spin up VMs for people though
04:46 sivoais didn't RunMyCode come out of academia? Columbia Uni. iirc
04:47 sivoais hehe, interestingly enough, I was complaining about wanting devops for science in #pdl the other day...
04:47 skay sivoais: Columbia and two orgs in France
04:47 skay sivoais: yes! science ops
04:48 skay I spent a lot of time ranting at my PI about devops trying to say how runmycode is like travis-ci
04:48 skay or could be
04:48 sivoais Too many times have I downloaded code and it just doesn't even compile.
04:49 sivoais I just added Travis-CI support to the PDL project! That's why it's been on my mind. :-)
04:49 skay she had me and her post doc take papers from a sample of Science publications to see if we could download materials and get them to run
04:49 skay I don't know when that will be published
04:49 skay I decided not to stick with the job and didn't renew after a year
04:49 skay It's a lot easier to get things done in an industry setting
04:50 * sivoais has never worked in the ... real world... open-source is the closest I've gotten to doing software engineering
04:50 skay I went in to industry instead of grad school.
04:51 skay I was interested in cognitive science and got psychology and cs undergrad degrees.
04:51 skay but got a programming job
04:52 sivoais nice! I did my undergrad in neuroengineering and now I'm almost finished with a CS masters
04:52 * skay needs to turn in early
04:52 skay neato!
04:52 skay o/
04:52 sivoais oh, we'll chat later then! :-)
04:52 sivoais o/
04:52 skay I gotta a meeting at 7 am
04:52 skay timezones!
04:52 skay o/
13:37 bene runmycode looks like a project in search of a problem
13:50 skay bene: it allowed people to run code on a CERN cluster
13:51 skay bene: to generalize it, they should have done another prototype that allowed anyone to glue together something like a repo and travis or jenkins and a generation of a DOI at the same time
13:51 skay bene: while also providing an archive for all of the data and results being used
13:53 bene it doesn't actually mention this cluster option anywhere on the main page or faq AFAICT
13:54 bene and the data/code/paper hosting thing has a million fragmented platforms already
13:54 skay bene: while being able to run things on bare metal or vms
13:54 skay bene: no kidding
13:55 skay bene: NO KIDDING
13:55 bene i was chatting with some old labmates and the lab i was working for at HMS is now setting up a repository for structural biology data..
13:55 skay bene: one reason why I left the job
13:56 bene yeah
13:56 skay bene: there are already repositories like that. can't remember off hte top of my head.
13:56 bene yes, i know
13:56 pdurbin bene: it's for science. reproducibility of results
13:57 skay yeah, it glues together things that developers have already had for a long time but with strict packaging of everything that went in to the computation
13:57 bene pdurbin: no, it's just fodder for grant funding and the resulting egoboo/departmental standing that results from bringing in grant dollars
13:57 * bene is pretty jaded about academia at this point
13:57 skay btw, the guys who owe runmycode.org didn't collaborate with my PI and wouldn't even share all of the source code
13:58 skay I had to get my PI to fight with them for 3 months. the entire academic environment of that group was toxic. I set up a meeting to resign after that 3 months, and my PI talked me out of it
13:58 skay and basically told me to JFDI a new platform
13:59 skay bene: a lot of this is an attempt to apply a technical solution to a social problem.
13:59 skay bene: are you familiar with all of hte news across all disciplines for the lack of reproducibility in research?
13:59 bene yes
14:00 bene NIH/NSF is handing out money trying to get people to solve that problem
14:01 bene so it was a regular topic of the technical lab meetings
14:06 skay I like the efforts of software-carpentry for actually training people to understand technology versus people creating brand new tools in some NIH syndrome
14:07 skay scientists should learn to how to use what already exists and understand why and understand how to create tools that are maintainable enough for their needs
14:08 skay there was too much ego where I worked
14:09 skay there were some technology that could have been actually useful instead of some wanking
14:09 skay but I was the wrong person to hire for it
14:10 skay sometimes I was filled with incandescent rage
14:19 pdurbin bene: someone needs a hug, sounds like
14:20 skay not me, I'm good
14:20 skay I'm just trying to convey how things were.
14:20 pdurbin skay: don't mind bene. he's usually not this grumpy :)
14:21 skay I think that group was an outlier for a bad place to work in academia
14:21 skay do I sound like I'm comisserating? I hope so
14:21 skay if not, I will quieten down
14:21 skay sometimes commiseration is a nice thing
14:22 skay it sounds like I'm not as jaded yet
14:28 bene i'm good too
14:28 bene the problem with software in research is that the toolmaking is generally not valued
14:28 bene the incentives for building/maintaining the tools are all messed up
14:29 bene i.e. NSF will give you money to build a new tool, but getting them to fund maintenance/bug-fixing is "not novel" and much harder
14:38 bene not to mention that software citations are usually relegated to the paper addendum or only available in the online version
14:38 bene if they get cited at all
14:39 bene the software developers in structural biology all ruefully joke about how "methods developers" don't get much credit for their work unless they also make some huge scientific discovery along with the method
15:14 bene anyway, they are hard problems and worth working on
15:16 bene but primarily social and _not_ technical
15:22 skay http://video.open-bio.org/video/1/a-history-of-bioinformatics-in-the-year-2039
15:22 skay borrows a slide from The Life and Death of Yavascript
15:23 skay numfocus talk https://www.youtube.com/watch?v=j9ydiVbassQ
15:23 skay mentions the social problems too
17:18 bene http://bostinno.streetwise.co/2015/03/03/startup-culture-beers-on-tap-at-boston-tech-companies/ <- more proof that you need to care about the beer in your keg if you're using it for recruiting
17:19 skay we have a bar with hackerspace brewed beer at our hackerspace. <3
17:22 bene i was kibitzing with someone here about office beer as a recruiting tool last week
18:54 semiosis i'm on a mac!
19:00 pdurbin semiosis: did it call you a slacker yet?
19:00 semiosis well, i did just install slack
19:01 semiosis ooh textual made a noise
19:01 semiosis this is neat
19:23 tfhartmann joined #crimsonfu
21:09 tfhartmann joined #crimsonfu

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

crimsonfu - sysadmins who code