Disqus - Latest Comments for jamtoday

Re: Private: How to install Ethereum on OSX (and test mining!)

jamtoday — Sat, 18 Jan 2014 20:08:48 -0000

Running on Mac OS 10.8 I'm getting the following error running the "go get -u -t github.com/ethereum/go-ethe..." command:

Undefined symbols for architecture i386:
.....
ld: symbol(s) not found for architecture i386

This is returned while retrieving the src/github.com/obscuren/secp256... package.

Has anyone else run into this? I googled around and tried reconfiguring gmp with certain params but nothing so far has worked.

Re: One Person Profitable

jamtoday — Mon, 22 Feb 2010 19:09:52 -0000

I'm compiling a list of founders who have achieved OPP. There are the obvious ones like Markus Frind and Gabe Rivera, but I'm sure there are many others out there who have also successfully bootstrapped as single founders.

Please comment if you know of any verified one person profitable startups...

Re: Shock Artistry: Lady Gaga’s Social Factors of Fame

jamtoday — Sun, 31 Jan 2010 17:39:01 -0000

"If there was an absence of free channels to hear Gaga, would her album sales be higher?"

The great thing about a powerful multi-soure analytics app like Trendrr is it helps you simulate the answer to these questions, like a magic 8 ball for business insight. If you're able to segment the market into who is buying albums and who isn't buying them, then it should be possible to extrapolate this to a fairly accurate degree.

Re: From Scans To Listens

jamtoday — Sat, 19 Dec 2009 15:11:11 -0000

"3. We can overlay revenue numbers to see which spikes in social music activities actually translate to dollars in the bank and thus, which career building activities were worthwhile."

Or even perhaps just which social music activities lead to higher CTRs. The end result being the same kind of detailed analytics that you can get from Mixpanel or Google Analytics, except for music activities instead of webpage activities, or music activity embedded in webpage activity.

Some of the best value I've gotten from analytics products has been through the customizability offered by APIs, making it easy for me to create custom reports, etc. I'm curious about what the differences would be in a music-specific analytics API, and whether it be possible to incorporate A/B testing or other forms of hypothesis testing.

Re: They're almost here! - Nick's Blog

jamtoday — Wed, 02 Dec 2009 04:32:38 -0000

A bit belated, but I've been running into a task queue dilemma that would be perfect for a more in-depth examination, since it's probably somewhat common.

I'm working on an application that makes heavy use of deferred tasks, and there are some tasks that need to be much more high priority than most other tasks, for real-time update type functionality.

Of course, that's what queue.yaml is for, right? But I've had some difficulty getting the desired results, because my task queue is always backlogged and even the tasks in a high-priority queue I've setup with a large bucket size and high rate often take a while to get executed.

Tasks can be more tricky than they appear at first glance. Especially for those of us not too familiar with token buckets, the question of bucket size can be mystifying. And some tricks for task queue introspection would be really helpful...on that note, I'm crossing my fingers for better real-time task queue debugging through the admin console in one of the impending SDK releases!

Re: So here is the issue with Android

jamtoday — Wed, 07 Oct 2009 16:55:51 -0000

Come on FSJ, you should know more than anyone that the market will force a conspiracy among all the hardware manufacturers to figure this out. One year from now, this will be well on its way to being a solved issue.

Re: Playing with Fwix and Twilio (and an App Engine CacheCompress utility)

jamtoday — Mon, 05 Oct 2009 13:23:50 -0000

The pattern I use often is to do a db.get() on a list of keys, and
then cache the entities. But all it takes is one deprecated key that
results in None being returned from the datastore, and you've got
yourself a heterogenous list.

Come to think of it, it would be nice if something like this utility
were to be made available as a memcache method like
memcache.compress_data(arbitrary_data), because it's usually not safe
to assume that you're working with a homogenous list. And in the
future there might be some other types of pre-pickle data compression
or trickery that could be done in addition to model_to_protobuf.

Re: Playing with Fwix and Twilio (and an App Engine CacheCompress utility)

jamtoday — Mon, 05 Oct 2009 12:08:15 -0000

Thanks for your feedback, Nick!

Good point about the pickling. As you mentioned in your post, pickling
entities is sloooow.

I've changed that admittedly kludgey first 5 conditional statement to
"if any(isinstance(x, db.Model) for x in data)" as you suggest. any()
is one of those useful methods I always forget about.

When you say that ProtoBufObj is unnecessary, I'm curious what the
alternative is, since the binary object returned by model_to_protobuf
is a string, and is therefore not distinguishable by type.

Re: Any programming n00bs want to learn Python with me and Julie?

jamtoday — Wed, 23 Sep 2009 23:10:17 -0000

If you like interactivity, I'd recommend Crunchy: http://code.google.com/p/cr...

Re: DNS progress, another wish for a miracle (Scripting News)

jamtoday — Sun, 20 Sep 2009 00:57:02 -0000

Yes, I agree. This is my sticking point. Tornado looks really interesting, but I'm wondering whether it's crucial. The best thing would be a library that could be used on any python server.

Re: DNS progress, another wish for a miracle (Scripting News)

jamtoday — Sat, 19 Sep 2009 17:04:34 -0000

I'm in the process of moving all of my projects (even the old PHP ones) to App Engine (my experience with GAE has been nothing but positive so far).

I've been aching for a chance to get an rssCloud server working on App Engine, because I could make it an open relay up until it activated billing, and even then it would only start out as a few cents a day. And afaik long polling works just fine over regular ol' HTTP.

Are you dead-set on Tornado, or would you be willing to settle for App Engine?

Re: Introducing Balloons: Free multimedia overlays for bloggers

jamtoday — Thu, 30 Jul 2009 18:13:55 -0000

Balloons look fantastic. It would be even better if developers could access this feature with a javascript library.

Re: http://9astronauts.com/articles/javascript-tutorial-create-a-bar-chart-plug-in-for-jquery/

jamtoday — Tue, 28 Jul 2009 04:10:04 -0000

Is there a way of accessing specific bars within the chart? For instance, if we want to customize the background color for each bar in the chart...

Re: Dumb XML question (Scripting News)

jamtoday — Tue, 21 Jul 2009 18:23:28 -0000

"It's because I'm developing an app where, at first, everyone who cares about
it will be people like you and me."

If this is true, people should be able to handle viewing the source, or using Firebug.

Re: Heyzap story Q1-Q2 2009 :)

jamtoday — Mon, 20 Jul 2009 20:02:01 -0000

Incredible.

I remember the first day you guys were setting up your RoR server on a Macbook. You showed me the app running in a browser and I'm pretty sure it was a blank screen but apparently you and Jude saw something there that I didn't see.

And indeed you did. You guys have totally artfully dodged the blue shell...or are you the blue shell? Either way, you guys have zoomed ahead to first place.

Re: A more low-tech approach to ping hubs (Scripting News)

jamtoday — Sat, 11 Jul 2009 17:16:25 -0000

There aren't many examples of truly real-time apps running on Twitter, but that's largely because there's no push notification service provided. It's a bit of a chicken and egg problem.

Re: A more low-tech approach to ping hubs (Scripting News)

jamtoday — Sat, 11 Jul 2009 02:58:00 -0000

Unfortunately, up-to-the-second really does make a difference, for lots of the real-time focused apps (like multiplayer games) where every second counts.

Re: One Protocol to Ping Them All

jamtoday — Fri, 12 Jun 2009 22:09:10 -0000

Josh,

Yes, I am! I should posting again soon about my PSHB activities.

Re: Google Squared Is No Freebase-Killer

jamtoday — Fri, 05 Jun 2009 17:00:09 -0000

We have a naive classifier processing the wikipedia category graph. It might
be better than the current algorithm.

Is there a way we can interact with this estimation data programatically,
and perhaps offer new estimations through an eMQL adapter?

Re: Google Squared Is No Freebase-Killer

jamtoday — Fri, 05 Jun 2009 16:21:11 -0000

Interesting post. Thanks for the heads up!

Re: twdsc.us: @jamtoday. "http://twdsc.us is great but i think it would have much more potential if shared with developers. -- http://bit.ly/...

jamtoday — Sat, 30 May 2009 03:56:24 -0000

So for instance, I could write this Disqus comment at www.tweetmeme.com...

Re: Discussions in Twitter, day 2 (Scripting News)

jamtoday — Sat, 30 May 2009 03:52:19 -0000

If you were really a mensch you might offer these disqus widgets in an iframe or js script (with a link back to twdsc.us) , and take a tweet id as an argument.

I'd gladly prepare it on the publishing side, so long as I could consume it...

Re: Is there value in being on the SUL? (Scripting News)

jamtoday — Sat, 30 May 2009 03:24:36 -0000

Revolving door rates?

What if we just go all out and say that the goal of Twitter is to create as successful users as possible. Successful meaning you post frequently and engage in trending topics or hashtags, and gain lots of followers.

It wouldn't be hard - well, it wouldn't be impossible, at least - to figure out which Twitter users are the best ones for the SUL based on the effect they on average have on newbies.

In other words, the most motivational users rise to the top. And we begin to realize what success really consists of on social platforms: the call to action.

Maybe, just maybe, the call to action is the new PageRank.

Re: Twitter as coral reef, cont'd (Scripting News)

jamtoday — Thu, 28 May 2009 15:37:53 -0000

#Chutzpah -- "The big announcement is Google Wave. How much you want to bet in 5 years it'll be as famous as OpenDoc is today."

Re: Why Isn't Paypal More Successful?

jamtoday — Tue, 26 May 2009 12:45:34 -0000

It's funny you write about this now, since just a few days ago I had a *terrible* Paypal experience.

I paid my roommate some rent money at the end of last week, and he told me the next day that he received the payment twice. Sure enough, it turns out that two identical payments were made - within 9 hundredths of a second. As a developer, I instantly recognized this as a double-submitted POST form, and I was somewhat astonished that Paypal would fall for such an amateurish problem.

I filed a complaint, clearly explaining what had happened. I expected a quick resolution, and surely enough, we got one. My roommate just received a call that his account is being suspended while they investigate his account history. Even though, of course, the problem is not in any way his fault.

I don't know about you Fred, but I'm really hoping there's a better option than Paypal when it comes to reinventing the way payments work