> Linear-regression estimates show gene name errors in supplementary files have increased at an annual rate of 15 % over the past five years, outpacing the increase in published papers (3.8 % per year).
This still doesn't actually tell us if the problem is getting worse. Or if it does it is badly worded. Even assuming this 3.8% is derived from their own data, you need the number of papers published that contain genelists (I would imagine this has probably risen faster than the number of papers overall).
In other words, the authors should have plotted the error rate over time rather than the number of errors over time.
> b) it's an old timer who has never gotten with the times.
As someone who knows the author of the post, that is mostly true ;)
In all seriousness, while I generally agree with your sentiments I don't think that Perl is dead within Bioinformatics. Not least since major data providers, e.g. Ensembl, write their tools in Perl.
Oh I know it still exists, but while this is admittedly anecdotal on my part (and likely heavily biased by the institutions I've been a part of) I haven't seen a single new person who defaults to Perl in a very, very, very long time.
And to be clear, I'm not trying to rip the author or the article, it just seems ... oddly timed, like writing a contemporary article on how to use COBOL in the business computing space.
Author of the post here. I think my comeback would be Perl 6 is a completely new language, that's only just becoming usable from a long design and development process. A lot of people dont think about it this way because they have zero idea what Perl 6 is about. If you don't like sigils on your variables you aren't going to like it though :P I use Perl 5, Python and R on a daily basis. I try to avoid BASH in favour of Perl, as it genuinely is just less crazy and is perfect for the same tasks. Python isn't a great shell script replacement. I like Perl 6 mostly because it lets you write nicer Perlesk code that isn't quite so disposable and gross. At the moment Perl 6 isn't completely viable for daily use, just because of performance. I also know Java and C quite well but rarely work on anything that needs the perf of these languages given the overhead for someone who is a researcher rather than fulltime programmer.
Fair point. I haven't looked at Perl6 since the early aughts and assumed it was basically exactly what I saw then. I didn't realize that it was just now starting to blossom.
Lol not sure I would go as far to say its just started to blossom. Some flowers are weeds to one person and a pay-for wild flowers to another. Each to their own!
I think the suggestions is that Perl 6 is a very different beast to Perl 5. Or to put it another way, you could read the title of the post as "Bioinformatics and the joy of Foo Programming Language".
Perl 6 is probably best described as a dialect. The look and design of the language are very Perl-ish, but some of the key sore points from Perl 5 (OO, concurrency, etc) are addressed and it has added a number of killer features. This is not Python 2->32; it is a major overhaul of the language, with no backwards compat beyond a suggested perl 5-compatible layer (I
believe this is called 'v5').
The article touches upon a few (Grammars for instance), but I personally think the concurrency work will also be a real draw.
The other key difference is that Perl 6 is actually a specification with an official test suite and Grammar (STD). I believe the specs indicate that anything that passes the test suite can be deemed to support 'Perl 6', which really opens up the use of various backends. The Rakudo Perl 6 implementation has support for three (MoarVM, Parrot, and JVM).
On a similar minor note, it would be nice to have exact matches to the 3 letter station codes appear at the top of the list, as they do on 'official' sites.
This is a great site though and potentially very useful, great job!
I really did look around at OEM laptops when I was deciding what to buy. I wanted a laptop which would run Unix, but besides of Dell I could not find any. I was afraid that Chromebook would not have provided the hardware I wanted and I did not to pay extra for Windows license, which I'd never use anyway. After that, I think there were still Lenovo Yoga 2, Samsung Series 9 and rMBP on my list. However, after reading articles about the laptops, I found that Lenovo had bad battery life and buying Series 9 or rMBP would basically pay me the same. At this point Apple's aesthetic product won my choice, given that Series 9 could have problems with Linux drivers and it still had the unused Windows license shipped with it.
So now I have Windows/Ubuntu on dualboot on PC and a Mac laptop. I feel like the Mac is basically the Unix I wanted, without the crappy window manager of Ubuntu. However, the single thing I've loved so far has not been a hardware or aesthetic manner, but rather how Mac opens windows as it boots up from where I left. I never even knew that was possible, but I feel like it has increased my productivity a lot.
When I was shopping for a new labtop I initially settled on the Series 9, but when the 2013 rMBP's came out they really trumped the Series 9 in the hardware department so I returned the Series 9 and got a rMBP. The good thing is that the Series 9 was just as good build quality, and had excellent battery life. Also it was priced well and I think its part of the reason the 2013 crop of Macbooks had to come down in price the way they did.
I think shortly we will see the monopoly that Apple has on high end well designed laptops will come to an end.
I was in a similar situation a year ago and decided to go with the Series 9 with Full HD screen instead of a MBPr.
The weight advantage of the Series 9 and the screen are nice, but battery life, keyboard and trackpad are way behind Apple's quality. My next laptop will come from Apple and i don't see any competitor coming close any time soon. Samsung will drop out of the laptop market in 2015 alltogether btw.
This still doesn't actually tell us if the problem is getting worse. Or if it does it is badly worded. Even assuming this 3.8% is derived from their own data, you need the number of papers published that contain genelists (I would imagine this has probably risen faster than the number of papers overall).
In other words, the authors should have plotted the error rate over time rather than the number of errors over time.