larsberg

39 ms·

331.
by larsberg 15y ago
The soon-to-be-released data retention policies for the NSF's CISE (basically, the arm of the National Science Foundation that funds all research) will most likely require complete and free access to not only the code for your implementatio
332.
by larsberg 15y ago
I was recently talking with Prof. Dave MacQueen about this (he ran a group at Bell Labs for around 20 years before the Lucent debacle). The most amazing thing was the management style of that lab. Once a year, you had to write an "I am smar
333.
by larsberg 15y ago
Not the software development teams in Redmond, at least as of 2007, though some of the support and administrative services certainly were. The goal was to give all dev team members their own private office. Interns and the more junior peopl
334.
by larsberg 15y ago
That's what I did after a nice stint at Microsoft, and it's amazing --- working on compilers is even _more_ fun than you'd think! But, there's always the gloom & doom at the end. Post-doc positions are easy to find, but there are very,
335.
by larsberg 15y ago
Please, learn from the experience of Office. Menus that change themselves confuse and infuriate every type of user --- experts, developers, novices, etc. People build up muscle memory, and breaking it unexpectedly leads to very angry users.
336.
by larsberg 15y ago
I found the "would you take a 10% paycut" matched my experience precisely. Back when I worked at Microsoft on developer tools, I would sometimes get contacted by people who were in the finance industry with salaries 2-4x what we could offer
337.
by larsberg 15y ago
It's funny that you mention the keyboard shortcuts! When I talk to old hackers on Linux about moving to OSX, I tell them, "well, if you move, you don't have to configure your power management and display drivers and it has Emacs keybindings
338.
by larsberg 15y ago
I worked for Microsoft in the Seattle area for ~6 years in the early 2000s. My wife and I (both US citizens) went up to Richmond and Vancouver, BC every few months to spend a weekend eating great chinese food. Every single time, the trip in
339.
by larsberg 15y ago
Of course, the Java equivalents would be slightly different. For example, "More than 4 pattern names in a single class identifier."
340.
by larsberg 15y ago
FYI, if you are looking for mind-blowing amounts of money and don't mind either Chicago or NY, your background is loved by folks in the financial industry.
341.
by larsberg 15y ago
You didn't talk about where you are, and location matters a lot. Also, I hate to sound like reddit, but your friend might want to talk to an employment lawyer if they have issues with this kind of clause. Many of my friends work in finance,
342.
by larsberg 15y ago
Two reasons: 1) I'm often copy/pasting from publications, and the PDF-rendered glyphs don't turn into proper unicode either (OSX, Preview, paste to Firefox, at least) 2) There are conventional differences between the different arrows such a
343.
by larsberg 15y ago
binary-trees is rough to beat C on. It involves allocating and traversing some pretty gigantic data structures, and the C code is optimized by using region-style allocation to avoid all of the individual calls to free (or garbage collection
344.
by larsberg 15y ago
I've also noticed that very few of my CS academic colleagues are on anything social other than Google+. The biggest feature it needs is LaTeX escape formatting. I'm sick of seeing "$\mapsto$" in posts instead of a nice arrow.
345.
by larsberg 15y ago
All scientists start out without a PhD. Then, they do some research. At some point, they either have published enough novel papers or have done a single piece of novel work that advances the art and warrants granting of a PhD. If these stud
346.
by larsberg 15y ago
It's still not trivial even with massive parallelism. Performance is easily killed if you: 1) Completely saturate the memory interconnect between the processors and RAM chips 2) If your parallel collector threads are touching the same memor
347.
by larsberg 15y ago
Even if you assume the compiler is perfect and only making semantically valid transformations, changing performance characteristics often changes your odds of hitting race conditions.
348.
by larsberg 15y ago
This law seems to depend on the mailing list and the age/status of the language. I wonder how it went during the "Successor ML" conversations a while back? Today, though, it seems as if almost every language conversation we have about Stand
349.
by larsberg 15y ago
Yes, certainly the vector instructions. There are a few keys issues many C compilers seem to run into today: - Loop unroll identification is really bad. For example, ICC will unroll and turn a single-level loop with a very obvious body into
350.
by larsberg 15y ago
Indeed, we see this today with our compiler (Manticore). On our 48-core AMD and 30-core Intel boxes, we now routinely have flattening speedup curves due to the small sequential portions of even our supposedly "embarrasingly parallel" benchm
351.
by larsberg 15y ago
The only thing harder to hire than an experienced developer is a manager experienced at turning junior or early-career developers into senior ones. Seriously, few things can sink the ship faster than taking on inexperienced people and hopin
352.
by larsberg 15y ago
I wonder if this increased responsed time is uniform across all tags. I work on ML implementations (SML/NJ and Manticore), and I've basically given up on answering questions on SO because every time I go there every question has been answer
353.
by larsberg 15y ago
Indeed, and thanks for the correction!
354.
by larsberg 15y ago
I'm pretty shocked you got it for so little. I've been offered more than five times that for my "first name only" domain, and it's just been from other people with the same first name. Though they're right about the lowballs and not even re
355.
by larsberg 15y ago
No. Not even if you send it to your kindle via name_num@free.kindle.com. But, Calibre is now much better at generating real TOCs and remembering their position when you go to/from sections & articles. The Economist (free!) and Reuters N
356.
by larsberg 15y ago
WiFinder
357.
by larsberg 15y ago
While my hiring manager experience predates GitHub, I would be more concerned in verifying that they actually did what they claimed than that I agreed with the specifics of their commits. If they got the the tasks completed, that's a Good S
358.
by larsberg 15y ago
That doesn't matter. Back when I wrote my first iOS app (which used private APIs extensively), I had long conversations with several reveiwers (e-mail and phone) where they said it was fine. We all know how _that_ played out...
359.
by larsberg 15y ago
I think the author is pointing out that the Kindle version of Dune is terrible. Horrible typos, bad formatting, etc. Unfortunately, many middle-aged books have this problem, and unless you can wade through the Amazon reviews (which are not
360.
by larsberg 15y ago
I think a much better post may have been tell people to check for their apps in the marketplace first because there are a huge number of common apps. But, they might not have _your_ apps, if your apps include Pandora, Uber Cab, or Bump, for

More ›