Disqus - Latest Comments for chrishenry

Re: Easing the use of the AWS CLI

chrishenry — Sat, 24 Sep 2016 16:40:26 -0000

Instead of configuring the aws cli via environment, you could always mount in your ~/.aws/credentials file. You could also use something like Fugu (https://github.com/mattes/f... to take some of the pain out of that very long docker run command.

Re: How to Apply Lateral Thinking to Your Creative Work

chrishenry — Fri, 12 Sep 2014 10:48:06 -0000

I really like the point about questioning the question. It relates very closely to Spolsky's 5 whys.

Re: Our web development workflow is completely broken.

chrishenry — Sat, 29 Jun 2013 22:36:17 -0000

Excellent post! I also have to agree that tools like Brackets are super exciting.

One thing you missed in your first workflow diagram, since it didn't seem long enough. There's potentially a step to transfer the file to server, should you be working on a remote server.

Re: Tech Tuesday: Technical Debt

chrishenry — Tue, 04 Jun 2013 08:52:25 -0000

Albert, thanks for writing about such a crucial topic. A couple other points; Engineers hate seeing technical debt pile up because it creates maintenance headaches that a good product should take of by itself. So keeping up is pretty important to making sure an engineering team stays engaged.

We've found that smaller technical debt items are things that should be addressed throughout the course of dev cycles. We prioritize smaller items that streamline the team's workflows, and things that are likely to break as we scale. In order to get bigger debt items accomplished, we use a modified form Spolsky's "Don't write new code until bugs are fixed." We won't build new features on top of any system we feel is holding technical debt, making any debt payback part of building new features.

Re: How To Survive a Ground-Up Rewrite Without Losing Your Sanity

chrishenry — Sun, 05 May 2013 19:50:47 -0000

This article is, by far, the most pragmatic, reasonable, sound and sane advice I've ever seen on moving forward large applications. Particularly, the data migration bits.

Thanks for sharing!

Re: http://kylebanker.com/blog/2010/09/21/the-joy-of-mongodb-indexes/

chrishenry — Sun, 06 May 2012 17:42:17 -0000

One thing to be particularly careful of in MongoDB is that it will allow you to create indexes on things that don't exist. For example, if you fat finger part of a compound index, Mongo will still create the index. So even if you *think* you have your indexes created correctly, you should have someone else spot check them.

Also, this; https://twitter.com/#!/chri...

Re: Two Right Legs

chrishenry — Sun, 06 May 2012 17:18:51 -0000

One thing I learned while cooking chicken legs - Start skin side down. It sears better, and makes the skin crispier. Love this article.

Re: Responses To Asinine Comments About Tattoos

chrishenry — Tue, 04 Oct 2011 00:28:54 -0000

My favorite is: Did that hurt? My answer: No, it felt great when the 250 pound Goth dude jabbed a needle in my skin several thousand times.

Never gets old.

Re: The Most Important Start Up Lesson I Have Learned From Fred Wilson

chrishenry — Thu, 10 Mar 2011 14:15:04 -0000

"Always ask if they have to have it, and it has to be now." - Forcing team members to focus on what is really important can be hard. Asking questions about when and if things need to happen is a great way to allow the team's needs to self-organize themselves into your development.

Re: How to Measure Page Load Time With Google Analytics

chrishenry — Fri, 18 Feb 2011 00:29:23 -0000

The time that the browser takes to load the page isn't the entire story. When a user clicks a link, they start counting from the time they click that link, *not* when the browser downloads the HTML. The start event of this measurement is meaningless to the end user since most of the time, they don't know or care when the initial HTML payload is processed.

It seems the point of this exercise is to measure of how well the front end of the site is performing, which is important in and of itself. However, throw in some additional measurements, such as initial response, and you have a bunch of data that makes no sense.

Also, how are any issues raised *not* an IT concern? Is the marketing team going to implement a CDN to get those response times down?

Re: How to Measure Page Load Time With Google Analytics

chrishenry — Sun, 24 Oct 2010 15:58:16 -0000

Nice writeup. However, I think it's really important for you to call more attention to the the fact that you are not measuring server response time for the initial html payload, but time it takes the browser to load the rest of the assets the page requires. These two things are rather different, each with different causes and solutions. Your measurement of page load time is highly susceptible to Internet speeds, and local networks, DNS, etc., and not a recommended use of GA. I'd take these numbers with a grain of salt.

Re: Pages Weigh Too Much

chrishenry — Thu, 21 Oct 2010 14:48:39 -0000

I too, rarely have the patience to wait for a page to load on my Droid. I blame the ubiquitousness of javascript libraries. Especially with Google hosting many of them, devs have a tendency to write off their weight.

@WesWareOnline : For wordpress users there's this plugin:
http://www.bravenewcode.com...

Re: With Great Power Comes … Great Danger (Command Line)

chrishenry — Wed, 06 Oct 2010 10:10:02 -0000

I think most devs who spend a lot of time doing anything like sysadmin work have at some point nuked a system, db, etc. I know I have. http://l.chr.ishenry.com/mi... But that experience has only made those vague reasons to have backups very, very real.

I always use my work on the command line to inform me of what needs to be automated. If I find myself doing a task over and over on the command line, it's a sure sign I need to build it into my app. However, there are always going to be instances where the command line will win. It's definitely *not* in my best interest to spend my day automating tasks that are quick and easy to do.

And command line will always, always win when you need to quickly diagnose something. More often than not, when trouble happens, tail -f with the right log reveals all.

Re: MongoDB: Agile + Scaling

chrishenry — Mon, 20 Sep 2010 00:29:20 -0000

I found exactly the same thing when building 140ctr.com. Any data that I want to store or generate just gets added to a document, and storing tweets via Twitter's JSON API is a almost stupid simple. The one thing I do need to gripe about is how much work it takes to generate aggregates. I had to write a decent amount of MapReduce code to do what MySQL can do in a single line.

Re: Gmail actually gets something really wrong.

chrishenry — Thu, 09 Sep 2010 00:47:03 -0000

While I'm torn on the matter, I can't say it benefits all. The purpose of clipping was surely to provide speed, bu it does so at the cost of damaging the reputation of anyone attempting to provide a rich, beautiful experience via email. While we *are* somewhat of an edge case, our broken newsletter could be considered an exception to what Gmail is trying to achieve. Our users, who appreciate great design, lost out. We, as a company, lost credibility.

Re: (Updated) Downtime At Rackspace Cloud

chrishenry — Tue, 03 Nov 2009 23:14:46 -0000

While I'm no fan of downtime, I really can't criticize the response of Rackspace. Their support was extremely responsive during and after the incident. The outage caused MySQL replication to break on 50% of my cluster, and I was completely restored within a few hours. I was called by my Account Manager the morning after, and she made sure there was a thorough postmortem.

All hosting companies are going to have downtime. There is no silver bullet for 100% uptime. Rackspace's SLA guarantees that, and when they don't deliver, they make good on it, no questions asked. The day they don't stand behind it, I won't work them.

Re: Pake: PHP project build system

chrishenry — Sat, 22 Aug 2009 16:29:39 -0000

Is there svn support for this? Or just git? One of the reasons I've never personally ventured into any established PHP build systems is that they're so convoluted. I actually find it a lot easier to use PHP's exec() to export any number of svn repos and rsync them into place.

Re: What makes a good programmer?

chrishenry — Sun, 02 Aug 2009 20:17:21 -0000

Depending on your business, doing "Just Enough" could be exactly what you want. When launching something new, you have absolutely idea what "enough" is. Do "too much," and you can wind up sinking time and energy into a failure. Or go "too far" in the wrong direction.

By doing "Just Enough," you get out into the real world, get feedback, and iterate.

Re: Yes%2C+Rackspace+Is+Down+And+So+Are+Many+Of+Your+Favorite%26nbsp%3BSites

chrishenry — Tue, 30 Jun 2009 18:24:28 -0000

Rackspace is definitely the best when it comes to the day-to-day issues of managing servers. They do need to work on their crisis management skills. Customers need to be notified when things go down. ( I heard about this on cnet, and when I called, it turned out the monitoring servers were down, too. )

However, the fact of the matter is that datacenters are going to go down. If people are so concerned about 100% uptime, get a hot copy in another location.

Re: OpenX Keeps On Growing, Raises $10 Million More

chrishenry — Fri, 29 May 2009 02:57:15 -0000

I'm a big fan of the OpenAds / OpenX platform. I've been using it for a few years now, and have an installation serving ~500k impressions a day without breaking a sweat.

I'd love to migrate to their hosted solution, but I can't get over how ridiculously slow their backend is. Every time I try to migrate, I get frustrated and give up. Hopefully this latest round of funding will go towards fixing that.