jacabado

Location: Lisbon, Portugal

Remote: Yes Willing to relocate: No

Technologies:

- DevOps - Prometheus, ElasticSearch, Akamai CDN, Continuous Delivery on Kubernetes, Terraform and Jenkins.

- NFT Testing - Load testing with jMeter, Gatling. End-to-End performance analysis with Sentry and New Relic.

- Dev - Java, Kafka integration. Some experience with pet projects on Python and Node.

Résumé/CV: https://www.linkedin.com/in/jracabado/

I'm interested on technical leadership or senior reliability engineering roles. Having been working as an Architect the last year I lack regular hands-on practice but I'm happy to be challenged for a role that demands it.

I value autonomy and hard problems.

Or that plans are only useful when they fail and are adjusted.

The first time I read this I could not understand it. It makes a lot of sense now seeing how plan execution is where project management really happens.

This is tangential do the OP but what would it take to send external input to a completion script?

The idea would be to have a different interface to do the auto-completion.

I'm thinking of creating a visual UI through service discovery and providing an easy shortcut sheet to various points of a system.

Nassim Taleb disagrees with you https://mises.org/library/how-time-and-uncertainty-can-make-...

You just can't do it implicitl because the compiler can't distinguish Func<string> from Expr<Func<string>>.

But you can have a generic extension method that returns one of the types and then via inference you can do it implicitly typed.

It shoud be something like this in C# 5:

  public static class Lamb
  {
      public static Func<T> da(this Func<T> f)
      {
          return f;
      }
  }

  var egocentricFn = Lamb.da(() => "im a function");

In the starting days of Reddit and then HN this really bugged me a lot.

One of the solutions I wanted to try was to partition the user base on a similar way that posts are filtered to you. That is based on the up votes you give. Now I'm not sure if it happens here but at least on Reddit you have it on your customised first page.

I believe that it isn't the user base that gets worse but that it gets much more diversified and people entrench themselves in groups and the "How bout dem Cowboys" problem arises.

Think of it as user and posts clustering, the secret sauce would be how those clusters interact, I don't have a tentative answer for that. This idea would top current aggregators by valuing discussion and intervenients in the same way that posts are valued.

This adresses a pain that I suffer:

An app to easily compare SQL execution plans.

I believe the better way would be graphical. I have needed this on MS SQL Server, SQL Sentry Plan Explorer has helped but lacks this comparison. Today I do it by diffin execution plans on XML format.

On my wildest dreams I would have a REPL accepting a DSL that would allow me to query the different DMV's (those are SQL Server data management views which give you insight on the inner state of SQL Server, Red Gate has a nice site on them http://sqlmonitormetrics.red-gate.com).

If it already exist is some form or platform please share.

Roslyn from Microsoft is also an initiative in this direction.

But it has been on beta for almost 3 years I think. I wonder if it will ever go forward.

I haven't seen it on the new features of 2012, but I have seen it on some demos of 2014 Beta: http://stevenormrod.com/2013/09/an-overview-of-sql-server-20...

Could you elaborate on SQL Server vs Redis claim? Or throw some references?

I am not doubting but just honestly asking. I've been discussing a possible Redis solution for complementing our current SQL Server.

Our problem isn't that big, we need to serve 10k req/sec of a simple query, can SQL Server do that? Isn't the connection pool the bottleneck to handle loads of this scale?

Eric Lippert says the yield construct is the most complex of the C# language[1], how hard was it?

I just tried a yield in a recursive function and the output is as simple as a non-recursive function. Are there any edge cases interesting to see?

[1]http://blogs.msdn.com/b/ericlippert/archive/2009/08/24/itera...

Oh no, didn't even know that happened.

Really nice to recognize a name here. Such a shame I won't be able to attend this, I have to return to Portugal on the 17th.

I heard some tales of you in Fullsix when I worked there and would love to get to know how it's going on SO. Best of the luck for the presentation.

Anyone knows where to get those invitation codes?

On a related note, I've developed a small browser plugin for my agency to track which questions my team consults on StackOverflow.

It's completely opt-in and anonymous and in 1 month we collected near 500 consulted questions. Right now we have a widget which show the last consulted questions and a tag cloud. But wanna improve upon the insights we can extract from this data.

What do you think about this?

Not apropos your post, but slightly relevant.

After having some involuntary contact with law, I found that I love to read laws and jurisdiction. I'm finishing my MsC in computer engineering and would love to try to pursue a career that would intersect technology and law. I've just found a LLM, a masters in Law, that doesn't requires a Law degree in the University of Edinburgh, they even have the possibility for distance learning. (http://www.law.ed.ac.uk/ahrc/teaching/llm/llminnovationtechn...)

Can you tell me what should I expect to do with such degree? I was thinking something like paralegal work on technology related cases.

Besides that, what could I do to prepare myself before entering the LLM? I'm almost completely ignorant on most Law concepts/theory.

Thanks, and good luck for your startup. I had thought of something like that but for the government decrees, to increase state transparency.

I agree with your point just wanted to point to Nodejitsu node.js work that's highly relevant to it:

http://blog.nodejitsu.com/scaling-isomorphic-javascript-code

"Javascript is now an isomorphic language. By isomorphic we mean that any given line of code (with notable exceptions) can execute both on the client and the server. On the surface this seemingly innocuous property creates a number of challenges that are not solved by current MVC-based patterns. (...) In conclusion, we will explore a new pattern: Resource-View-Presenter."

They already did for some time, I am not sure if they've done in the scope of this rebrand, it is surely an welcome addition. [1]

Don't expect that it just works with any XMPP client as it uses a proprietary authentication mechanism "very similar to Facebook" OAUTH2.[1] More popular clients have already integrated it.

That's why Windows devs drink.

[1] http://blog.process-one.net/details_on_msns_xmpp_server/

Slightly related, I'm looking for a MVC Node.js framework, that is, one that runs on the server. Does anyobody know one?

Edit: Just got this one from another post on the front-page https://github.com/mde/geddy.

It's a shame Portugal is missing there. Portugal was once the biggest empire thanks to sailing and colonization. But oh well all we've left is a debt crisis to solve now :)

Technically I like it at lot, I'm looking to build a similar thing but I'm still in the conceptual phase.

I think you should define better your target market to be able to seduce them better, maybe move the 'Try it out' to the top of the screen and talk about some use cases of your target market. I'm just trying to help, I'm sure you know better of your business than me.

Wow I feel trolled by the author. Come on Google shaped the web as we know today with PageRank and AdWords, isn't it totally unfair to put things in that way? Also the author confuses search with discovery just to prove his point.

I agree it feels dumb to surf the web they we do know. But this is a case of less is more. We can think of more complex solutions but those won't solve what our current solution solves for most of the people.

Surely there is a next big thing waiting to happen regarding content discovery, the last one I saw that attempts that is Prismatic, but there is a long road to even become a plausible alternative to what we do in our current search engines.

relevant: "A collection of command-line tools for researchers in machine learning, data mining, and related fields." http://waffles.sourceforge.net/

I didn't knew how to make an enumerable with an yield.

http://csharpindepth.com/Articles/Chapter11/StreamingAndIter...

I had used it to make a state machine but didn't know its relation to IEnumerable. The article was enlightening to a C# latecomer like me.

So from what I get good blended is more expensive than good pure/single malt?

I have never been able to appreciate blended as much as I appreciate pure malt, but my sample is really limited. My favorite so far is Glenlivet 18 years.

edit: pure->pure/single

Maybe it also let's you go through China's GFW if you have some censorable content. But I'm sure the front-end developers I work with (in Portugal) are not thinking in censorship when they do that.

Thanks for the correction!

Corrected version:

$('#detect').click(); setTimeout(function() { $('.face').prepend('<img class="santaHat" src="http://officialpsds.com/images/thumbs/Santa-Hat-psd20202.png width="90px" />'); $('.santaHat').css('position', 'relative'); $('.santaHat').css('top', '-60px');$('.face').css('border-width','0px'); }, 200);

Try this in the page:

$('#detect').click(); setTimeout(function() { $('.face').prepend('<img class="santaHat" src="http://officialpsds.com/images/thumbs/Santa-Hat-psd20202.png width="90px" />'); $('.santaHat').css('position', 'relative'); $('.santaHat').css('top', '-60px');$('.face').css('border-width','0px'); }, 200);

I was just going through their archives trying to answer my question and found a great discussion on this post:

http://musicmachinery.com/2011/05/14/how-good-is-googles-ins...

There are some insightful comments from names I recognize from my research.

Looks amazing, can't wait to get access to it.

I'm starting my thesis on music information retrieval, just studying the related work for now. If anybody has any suggestion on the directions I could follow would be really welcome.

My initial idea would be to focus on playlist generation taking into account user's history and usage. So far I've seen a lot of related work exploiting song similarity, some cool work on music mood and some on assisted playlist building. I'm also not ruling out recommendation or discovery.