Comments for A Party to the World

Comment on zsh (or bash++) by noogie

noogie — Mon, 14 Nov 2011 19:19:08 +0000

you may want to look at .inputrc modifications where you can bind pg/pg down to history autocompletetiong…

“\e[6~”: history-search-forward
“\e[5~”: history-search-backward

Comment on SEOmoz and Dev Blog Post by dan.lecocq

dan.lecocq — Sat, 10 Sep 2011 16:13:51 +0000

Sure! Let me know the next time you’re downtown during the week, and I’ll buy you a cup of that coffee Seattle-ites are all so crazy about!

Comment on SEOmoz and Dev Blog Post by Andrew Ferguson

Andrew Ferguson — Wed, 31 Aug 2011 22:43:23 +0000

Facebook stalking says “YES!” and you’re married?! I think we need to catch up….

Comment on SEOmoz and Dev Blog Post by Andrew Ferguson

Andrew Ferguson — Wed, 31 Aug 2011 22:42:34 +0000

Are you actually in Seattle though!?

Comment on Triangle Strip for Grids – A Construction by Ricardo Sanchez

Ricardo Sanchez — Wed, 10 Aug 2011 20:33:05 +0000

Dan excellent tutorial!

I am new to opengl in general and I am struggling to implement this, I start a stackoverflow topic on the subject and I wonder if you will be able to have a look and let me know and assist on my implementation, any help will be much appreciated

Here is the link
http://stackoverflow.com/questions/7011017/help-getting-vertex-indices-from-a-grid

Thanks

Comment on Command Line Stopwatch (time cat) by Matt M

Matt M — Fri, 05 Aug 2011 01:34:26 +0000

Very nifty. I’m putting that one in my back pocket for later.

Comment on Boost, typedef, #define and GCC Pain by dan.lecocq

dan.lecocq — Thu, 04 Aug 2011 22:52:08 +0000

Yeah, I was really pleased to find this one. Also, really, Effective C++ is one of the most useful things I’ve read in recent memory.

Comment on Boost, typedef, #define and GCC Pain by cwalsh

cwalsh — Thu, 04 Aug 2011 21:31:30 +0000

Brilliant! I’ll put that -E flag into my toolbox :) I’m working in a very old and very ugly c/c++ code-base; there are #defines and externs all over the place!

Comment on Critical Language Scholarship – Results by Teri

Teri — Sat, 16 Jul 2011 03:44:35 +0000

Hi Dan,

You have probably put this whole Critical Language Scholarship behind you and moved on. However, if you don’t mind me asking, why do you think you were not picked? I read your essay, and it was superbly written. You seem to have so much cultural experience and be the right candidate for such a program.

I’m planning to apply to the 2012 group for Arabic.

Comment on numerical_limits by dan.lecocq

dan.lecocq — Sat, 14 May 2011 19:13:06 +0000

I occurs to me in reading your question what you actually meant. These variables will be set on network events. For example, the program can yield control to the steering library periodically, which checks to see if a steering client has connected, and if so, if it has issued any commands.

There’s a notion of a steerable application, which can advertise that clients can change certain parameters, or issue certain commands. One use case might be that the mouse_x coordinate and mouse_y coordinate can be specified by a client (we got a demo working of a WebSocket app acting as a trackpad on a remote machine).

But the main use case we are trying to support is that you have a large-scale simulation running on a remote resource (cluster, supercomputer, etc.), and want to be able to interact with it through a thin client without too much extra work. The simulation codes we’re working with are already behemoths, and can run for 12 hours or days at a time, and researchers often find that they could have terminated early had they been getting live updates of the results. Without yielding their allocation, this steering library enables them to check up on it as they like, issue commands, change parameters, etc. from say, an iPhone, the browser, or their desktop.

Comment on numerical_limits by dan.lecocq

dan.lecocq — Sat, 14 May 2011 19:02:22 +0000

The goal is to have a class that helps to manage the interactions with outside sources (steering clients) and the variables of an application. In the end, the variable *does* get set directly, but a client might pass in a float, int, or widget to the set function for a given variable. So it’s not sufficient to know the type passed into set(), but it has to also have some knowledge of the datatype the variable is pointing to.

The necessity for the variable class is that you don’t want to just hold a void*, because that requires the class to have a little more intelligence about the type, and can make it easier for the client to accidentally misuse, or abuse the class. This arrangement offers a stronger guarantee about what types of operations can and can’t happen to the data you register to it.

Comment on numerical_limits by Andrew Ferguson

Andrew Ferguson — Sat, 14 May 2011 18:25:52 +0000

Is the goal of steerable variables to provide data validation? Why wouldn’t you just set the variable directly?

Comment on Triangle Strip for Grids – A Construction by dan.lecocq

dan.lecocq — Thu, 10 Mar 2011 13:39:58 +0000

As far as I know, WebGL doesn’t have display lists. At least in Chrome and WebKit/Safari, I didn’t find a gl.NewList(). I’d be surprised if there were, as it’s supposed to essentially follow the OpenGL ES 2.0 spec, which doesn’t include display lists.

In what environment did you find gl.NewList()? You’ve make me curious!

Comment on Triangle Strip for Grids – A Construction by John

John — Wed, 09 Mar 2011 23:16:20 +0000

Correction: it looks like gl.NewList (); is available. Was it added after this article? Feel free to skip my post or update the article if display lists are truly currently supported in webgl.

Thanks

Comment on Triangle Strip for Grids – A Construction by John

John — Wed, 09 Mar 2011 23:11:37 +0000

Thanks for trying to cheer me up. I still prefer display lists in more ways than one. It’s a shame they’re trying to bury such a flexible and powerful feature. VBOs are just a crippled version of the display list.

Comment on Triangle Strip for Grids – A Construction by Noa

Noa — Mon, 07 Mar 2011 14:55:09 +0000

Dan,

Yes I was using the FFP, pre-computing the normals per vertex on cpu. Because the orientation of the triangles on the triangle strip alternates, my normals where miss-oriented on every odd row. Because I need to compute normals per vertex, I had to modify the algorithm and add some glue.

Comment on Triangle Strip for Grids – A Construction by dan.lecocq

dan.lecocq — Tue, 01 Mar 2011 15:13:15 +0000

I hadn’t noticed any funny business with the normals, but that’s good to know! I wonder if it was because I was using my own shader — were you using the fixed-function pipeline?

Comment on Triangle Strip for Grids – A Construction by Noa

Noa — Mon, 28 Feb 2011 20:51:54 +0000

It was very helpful for me to, thanks. I implemented it, and realized at some point that the normals are inverted at every change of row (because I’m setting normals per vertex). I had to add some more “glue” every time. So in the previous explanation, I inserted “7″ twice, and got 0, 4, 1, 5, 2, 6, 3, 7,7,7,8, 12, 9, 13, 10, 14, 11, and 15. That way the normal stays in the right orientation.

Noa

Comment on Wrapping printf(1) by dan.lecocq

dan.lecocq — Sun, 27 Feb 2011 22:20:30 +0000

That’s a good approach. The having different file descriptors is akin to an approach used in some code I work with, VisIt from LBNL.

Comment on Wrapping printf(1) by Andrew Ferguson

Andrew Ferguson — Fri, 25 Feb 2011 23:12:18 +0000

Why not set different levels for the “error” function? You could set some as errors and others as notices…sort of like what PHP does: http://www.php.net/manual/en/errorfunc.constants.php with “error_reporting()”