Blog - Comments

dewd comments on Nitrogen, Riak, and 1,000 Lines of Erlang

dewd — Fri, 05 Mar 2010 03:08:42 +0000

Rusty,
I tried a prior version of Riak (v0.6), and everything worked like a charm. I'll put a note about version dependencies on the github wiki...
thanks again -- great (!) learning aid :-)
--dewd

dewd comments on Nitrogen, Riak, and 1,000 Lines of Erlang

dewd — Fri, 05 Mar 2010 01:02:49 +0000

Rusty,

I love what you're doing with Nitrogen & Riak and wanted to use SlideBlast as a starting point for learning. Unfortunately I'm having difficulty getting it running, wondering if you can point me in the direction of the problem.

it looks to me that my install is failing on the riak start within caster (I've tried both the mochi & inets versions), though I can manually start & test riak from the command line.

the install recipe was from the README.markdown, though I've subsequently tried a full download / build of riak.

macosx 10.6
erl

any suggestions?
any forum that you would point me to (other than your blog?)

keep up the great work!

----- trace ----

=CRASH REPORT==== 4-Mar-2010::19:52:57 ===
crasher:
initial call: application_master:init/4
pid: <0.170.0>
registered_name: []
exception exit: {bad_return,
{{caster_inets_app,start,[normal,[]]},
{'EXIT',
{undef,
[{riak,start,[["riak.config"]]},
{caster_inets_app,start,2},
{application_master,start_it_old,4}]}}}}
in function application_master:init/4

Evie24Ma comments on Rod Hilton Loves Pair Programming

Evie24Ma — Wed, 24 Feb 2010 02:55:42 +0000

People think that the term paper format writing seems to be the most time wasting issue. But, students rely on the paper writing service help every time when it is needed.

joe comments on Submodules and Subrepos Done Right

joe — Fri, 12 Feb 2010 11:45:30 +0000

thanks,

thats exactly what i needed.
if i looked at the script, i thought its ingenious simple.

David Dabbs comments on gproc - Erlang Global Process Registry

David Dabbs — Tue, 12 Jan 2010 16:15:01 +0000

Hello. Thank you for the informative post. I'm new to Erlang, but I think qproc is what I'm looking for. The latest qproc version notes that

. . .In its current state, the global registration facility is broken and should not be used. It will be migrated over to a new version of gen_leader. This work will be done with low priority unless people express a strong urge to use this functionality.

Does this mean one should not try to distribute the process registry across nodes? I only need the process lookup (for a session store) in one node. Also, there seem to be multiple _gen_leader efforts, including one to merge and update them into one codebase. Are any of these needed to run qproc, and if so, which> Thanks.

Ulf Wiger comments on gproc - Erlang Global Process Registry

Ulf Wiger — Thu, 31 Dec 2009 08:09:09 +0000

The bugs that Thomas is speaking of were not found in gproc, but in a predecessor of it. Gproc handles those cases differently. Work is underway to use the same QuickCheck methods for testing gproc. So far, we have at least been able show that it doesn't suffer the same problems as its predecessor.

Rusty comments on gproc - Erlang Global Process Registry

Rusty — Wed, 02 Dec 2009 05:08:44 +0000

@Andreas - Thanks for the heads up, fixed the bug.

Rusty comments on Nitrogen at the Erlang Factory across the Pond

Rusty — Wed, 02 Dec 2009 05:07:38 +0000

Thanks! I used Prezi to create the presentation. Interesting software, but requires much more upfront planning than a plain old slide deck.

Kevin Dewalt comments on The Bilski Case and Software Patents

Kevin Dewalt — Wed, 02 Dec 2009 03:40:08 +0000

First, you are one talented hombre and I'm honored to call you a colleague. If the world TRULY knew how smart and diverse programmers are the Dilbert stereotypes would vanish. How many people could become Erlang experts and take an interest in the legal implications of creativity?

Second, I'm glad you touched on pros and cons in this debate. I also see both. In the end...I guess I feel like the whole software patent process is so truly broken that it needs to be tossed. Software should probably mirror copyright law and not patent law. It just moves too fast and will ultimately move faster. Plus...for what ever reason... it seems to me that 'patents' are primarily a club of the entrenched competitor and not the innovator.

In my last startup I totally punted on the patent issue. I just didn't have the time or bandwidth to focus on it. I was too busy focusing on the product/market fit. Which is really, really, tough, tough enough to take the risk that someone with 'intellectual property' would use the courts to stop me from helping customers. Something about such a system creeps me out.

But it is a tough issue with a lot of sides...