TechDistortion

Pragmatic on Constellation

When Ben Alexander asked if I wanted to do a show with him as part of his new “podcasting syndicate¹” called “Fiat Lux” late last year, I had no idea the amount of depth he had with his vision for where he wanted to improve on podcasts as an art-form and media-format. (I just wanted to make the podcast I’d always wanted to make)

It started out like many other podcasts do with a SquareSpace front end, Libsyn on the backend, RSS feed for iTunes and Podcatchers to latch on to and some links in the show notes. As this was happening in the background Ben had assembled a team of great people that shared his vision for a better way to get podcast content from the creators to the end listeners including: Jamie Ryan, Sid O’Neill and Lorenzo Guddemi. In the last two months it became clear to me that a lot of work was going on in the background and I was approached by Ben with a concept.

What if there was a better way?

The two big problems that Constellation solves (from my perspective and just for the moment: more features are coming) are firstly show notes are now essentially full-form articles (for Pragmatic) with show notes embedded in the content in context: no longer are they bulletised lists of links. Secondly the player is embedded in the webpage (that’s nothing new) but allows chapters to be linked directly on the web based on the chapters and segments. This makes each episode searchable², easily referrable and linkable. For a show like Pragmatic - that’s a huge deal.

What Constellation is

Same network with a new name? Sort of:

• Constellation is the name of the network
• Fiat Lux is the company behind the network (It’s analogous to Fox News being part of News Corp)

Previously both had been the same but the leap forward the site/delivery methods have taken demanded a new name.

Frequently Asked Questions

Q: Why are there only a small number of Pragmatic episodes on Constellation? A: I prepared four episodes with Chapter indexes and detailed show notes ready for the site launch with another episode updated and due to be posted in a few days. Eventually all episodes including follow-up episodes will be posted to Constellation.

Q: Why are you updating old shows? People have listened already haven’t they? A: For people that haven’t listened already this is the new standard I want Pragmatic to conform to consistently and it will eventually apply to all episodes of the show. For existing listeners this will make it significantly easier to go back and reference content from existing episodes. Most importantly: I care about the quality of what has been produced and want to make it better. Status-quo isn’t acceptable.

Q: It’s a lot of extra effort? A: That’s not a question professor. Not as much as you may think. I already spend a lot of time creating detailed notes for my own use, so breaking them into segments and making them readable is not too much extra work. The benefits of searchability, indexing and a more polished result are more than worth that extra effort.

Q: Why has Pragmatic been late recently? A: There’s only so much time available and in that time I’ve been updating past episodes rather than making new ones and shortly will be dealing with long-standing health issues. Pragmatic will not have a new episode for a few weeks however rest assured, we’ve still got lots of topics to cover and with all the great feedback I’ve got a backlog of follow up episodes just waiting to be recorded.

Q: Will companion articles still be posted on TechDistortion? A: A slightly modified version of the show notes for each episode may be cross-posted on TechDistortion if I believe its a standalone article that fits with the site. Companion articles however will no longer be done in their current form, no.

What do you, the listener, think?

Clearly the team behind Constellation and the other hosts on Constellation believe that what we’re doing is a step fowards. Ultimately though it’s you, the listener, that matters the most. We’re very keen to hear your thoughts on the new design, show structures, whatever you like/dislike and of course suggestions. We’re doing this for you.

As always you can contact me via the Feedback form or contact any of the Constellation team with any feedback you may have.

The Critical Path

Many people bandy-around the expression “The Critical Path” but what does it really mean and how is it useful? The truth is that it’s a part of “Project Management” and this phrase carries negative connotations for a lot of people but it’s a tool to help us get to where we need to be.

Project Management is an often hated discipline for a litany of really good reasons but the reality is that it’s a very broad topic. According to WikiPedia “Project management is the process and activity of planning, organising, motivating and controlling resources, procedures and protocols to achieve specific goals.” Most people don’t realise that what every person does every single day is a form of project management. If you need get some Monroe Apples before the kids get home then there’s a list of activities you might need to perform in order to complete that task: a list of sub-tasks that need to be performed in order to achieve that result. Most people will either consciously or subconsciously perform any one or most of those tasks without stopping to think about the details too much. That’s fine if your goal is to fill up a fruit bowl whose biggest repercussion if you don’t fill it up are cranky kids getting home from school or you just have to choose something else to eat.

There’s three (maybe four) conditions that dictate when you need to project plan in a conscious and structured way:

• When the repercussions of failing at the result are extremely unpalatable, expensive or dangerous.
• When many or most of the tasks aren’t straightforward or subconsciously executable.
• When there are a significant list of tasks that are difficult to keep in your head at any moment in time.
• Potentially you could say there’s a fourth reason: Your boss tells you that you have to.

Now the decision has been made to structure our tasks, the method/tool that’s most popular in the world today is by far the Gantt Chart.

Gantt Charts

Developed by Henry Gantt in the 1910s it’s a project planning tool that displays tasks on a linear X-Axis with respect to time. It’s evolved to include resource allocations and few other features as well but mostly unchanged from its original design over 100 years ago. A little known fact about the Gantt chart is that a Polish gentleman by the name of Karol Adamiecki in 1896 developed the “Harmonogram” through his work operating a steel mill however it wasn’t published widely until 1931 and translated into English but Gantt is credited as the original inventor. These days there’s a lot of software around that can produce Gantt charts including Microsoft Project which is now called Project Professional 2013. The more serious software package used in Construction Projects I’ve been involved with is made by Oracle and is called Primavera and it’s current version is usually what it goes by: “P6”

There are several free versions out there as well such as Gantt Project for Mac.

The concept is simple enough: make a list of all of the tasks and sub-tasks then assign durations to each one of them. After that connect each task to each other task upon which it depends (dependencies) and then there’s a map of all of the tasks and how they interrelate to help illustrate what needs to happen by when. Using the earlier example such that our final deliverable milestone will be 6 Apples in the fruit bowl by 4:00pm as the kids are going to be home by 4:00pm. Now that there’s a deadline we set the starting time at 11:00am.

For this example our task breakdown is as follows:

• Empty old fruit out of fruit bowl
• Find clothes and get dressed
• Find car keys and start car
• Eat museli bar
• Drive to gas station and fill car up
• Drive to fruit shop/orchard
• Drive home
• Put Monroe Apples in empty Fruit Bowl

The sequence of driving may or may not matter but the car won’t make it home without stopping at the gas station. As forward-planning is the name of the game we’re going to opt to do that first. Also to ensure we’re not cranky when the kids get home it’s important to consume a museli bar at some point during the time allotted in order to stave-off hunger. Noting that we don’t technically have to empty the fruit bowl of the manky old fruit until we show up with the fresh Apples, still choose to forward-plan and do that first. The rest of the activities are essentially sequential. Using the free Gantt Project mentioned previously the final Gantt chart GAN format illustrates the float for emptying out the fruit bowl and for eating the museli bar and final dependancies. Note that using the example and that most project planning tools work in whole days of effort, it’s not possible to break down the tasks to real-life durations hence each activity is set to take ‘1 day’ on the chart and they’re all the same length. Alas.

PERT Charts

Project Evaluation and Review Technique (PERT) Developed in the 1957 by the United States Navy. A PERT resembles a state transition diagram with tasks referred to as ‘Nodes’ and each node contains the information about that task: Name, expected (normal) duration, early start time (ES), early finish time (EF), late start time (LS), late finish time (LF) and the crowd favourite: Slack - also known as Float. Each of the tasks is connected to each of its dependent tasks and ends up looking a bit like a web. Most of those items are self-explanatory however the last one isn’t so much. The idea centers around task concurrency whereby if a task must be completed by the end milestone but can be conducted in parallel to another set of tasks, and that task takes less time to complete then its absolute start date can “float” earlier or later in time and still meet the final deadline. Float is meaningless in a single path PERT or Gantt for that matter.

It’s not uncommon for people to prepare PERTs in convention drawing tools that are cheaper/more readily available such as Visio however both Project and Primavera have PERT analysis tools included in them. The original PERT used “activity on arrow” methodologies that placed detail about durations on the transitions between the nodes however this has fallen out of favour and now “activity on node” is the most common form of PERT. {The last PERT I was involved in was back at Nortel in 1997 and I haven’t come across them used in any projects in Australia. Doesn’t mean they aren’t used just haven’t seen them. 100 years on Gantt is still king. I’m not going to talk any more about PERTs but they are interesting if you want to know more check out the show notes.}

The Critical Path Method

You may be thinking: thanks for the high level crash course in project management but what the heck does this have to do with A CRITICAL PATH?? It’s often used as an expression:

• ‘We’re on the critical path right now!’
• ‘This is a critical path activity’
• ‘We are critical path here people!’

Like all good expressions that encapsulate a good idea they tend to get overused, usually by people that don’t really understand them, which then devalues the expression overall such that when they are used in the correct context people tend to roll their eyes and go “buzzword bingo!” Two other examples of similar expressions recently that grew tiresome include “Proactive” and “Disruption”. The correct usage from a historical perspective is The Critical Path Method and the method was developed in the late 1950s by Morgan Walker from DuPont and James Kelley Jr. of Remington Rand. It appeared around the same time as PERT. Another little known fact is that the early form of the Critical Path Method was used and partly attributed to the success of the Manhattan Project in the second world war.

Essentially the method is the idea of mapping the path between the longest paths between tasks to the end deliverable/milestone, and the earliest and latest that each activity can start and finish without making the project any longer. It’s not so much the path but the idea of iteration that matters. Once a project is planned are there ways to add addition personnel to execute paths in parallel for example? In our Apple quest example assume that we have a helpful friend living in the house and they are able to assist. By increasing the number of people it’s possible to perform more activities concurrently. Hence whilst person A is getting dressed person B is finding the keys and starting the car in parallel. In addition when person A is out procuring the Apples person B can empty the fruit bowl. This has reduced the overall project schedule by 2/7ths of the total duration and changed the critical path. On the Gantt chart from earlier the greyed out boxes indicate the critical path of tasks. On many tools red is used to highlight the critical path. This process of adding resources is often referred to as “Fast Tracking”.

Being A Useful Project Manager

There are two big issues (although there are many more than that in total) with effective Project Management and so often drive negative behaviours and responses in people affected by a schedule or a timeline. This then leads to a mistrust or lack of faith and belief in the usefulness of schedules, PERTs and Gantt Charts and their associated project managers:

• Incorrectly broken down tasks/durations and dependencies
• Using the schedule as a motivational tool vs as a planning tool

The biggest problem with project management is the disconnect between most PMs and what it is they’re managing. In construction projects there have been numerous occurrences where the PM does not appreciate the difficulties with software development as they have experience in civil construction and once the concrete is set, you don’t break it up and try again. Iteration is unheard of unless there are major dramas and that leads to bad scheduling, bad planning and a bad end result. Another issue with the “throw more people at it” fast-tracking mentality is that there are genuinely activities that can not be completed any more quickly. My favourite example of this is the idea from the software-famous Mythical Man-Month book and is sometimes referred to as Brookes Law where the Author Frederick Brooks states: “Nine women can’t make a baby in one month.” It’s a beautifully simple way to illustrate the problem that not all tasks can be fast-tracked.

The other major problem is that schedule for some PMs ceases to be about planning for the future and becomes a negotiation tool for applying more pressure to those performing the tasks in question. Rather than helping that ends up hindering the project completion and builds animosity. Ultimately Gantts, PERTs and applying the Critical Path Method are all tools to help understand the best way to plan activities to get the end result in a known period of time. Just like any tool they can be abused if incorrectly wielded so proceed with caution or the wrong conclusion will inevitably result.

I’ll Take The Gold-free Extra Oxygen Cable Please

We use cables for everything: power cords, speakers, ethernet data, fiber optics and the list goes on. Two of the more common consumer cables however are speaker and HDMI cables. Ever wanting to sell more expensive cables, companies promote adding gold plating to HDMI cable plugs and with speaker cables they peddle OFC (Oxygen-Free Copper). Warning: Hardcore audiophiles may be upset by the contents of this article.

Oxygen-Free Copper

OFC cables are typically found in audio applications for stereo systems for connecting the amplifier to the speakers themselves. The configuration of these cables is often referred to as “Figure 8” which is a play on the appearance of the cables cross-section that looks like the number 8. The examples I’m considering specifically are this OFC Cable and contrasted with a Standard Cable. Making Oxygen-Free Copper perhaps not surprisingly requires an environment with an inert-gas during production that is, well, Oxygen-free. Technically speaking though it’s still valid to count copper as being Oxygen-free even if there are 10ppm or less of Oxygen contained in it. The truth is that OFC copper only presents a 1% improvement in resistance over standard copper and the whist OFC cables are associated with greater reliability this has nothing to do with the OFC nature of the copper but rather the general construction of the cables. OFC cables use high core counts to acheive the same Cross-Sectional Area (CSA) of equivalent cables and due to Skin Effect this often results in lower effective AC resistance and better performance in that regard only. Claims of improvements to noise performance with OFC over standard copper are unfounded.

Gold Plated Connectors

Primary reasons for gold-plating connectors come back to a desire to improve the contact resistance and corrosion-resistance. There’s an excellent article called the Golden Rules (PDF) which is from an connector vendor discussing all of the advantages of Gold plating for connectors. The problem is that all metals will erode with every connection/disconnection cycle of the plug and the socket. How much they degrade (how much is scraped off) will vary based on the hardness of the metal in question. In 1812 a German geologist and mineralogist Friedrich Mohs developed a scale of hardness that is still used today although it has been refined further it at least illustrates the point. Any plating used will eventually erode as the harder of the two surfaces gouges/scrapes off the other surface with the HDMI standard for example requiring a 10,000 total connect/disconnect cycles for compliant connectors. Ultimately contact resistance improvements are minimal but measureable but the cost of adding gold to a connector is often over-assumed by consumers. A gold-plated connector of a HDMI cable will have less than 0.1 grams of gold worth less than $5 USD and yet such cables sell for anywhere between $10 USD to $50 USD more than a standard HDMI cable.

Real-world Consumer Applications

Ultimately resistance is the enemy with speakers as the more resistance there is between the amplifier and the speaker the less total power is delivered to the speakers. Given a typical household cable run of 10m for a typical 25W bookshelf speaker, comparing the two cables mentioned previously we end up with a total reduction in power output of only 0.3dB of volume. Given that 3dB is half the power output, this is essentially barely measurable by a sound pressure level meter and would be barely noticable by the human ear at normal volume levels. Regarding HDMI there are both 5V and a I2C (aka IIC) bus on a standard HDMI cable and given the typical length of a HDMI cable is about 5m we can conclude that in order to drop below the minimum “Digital True” voltage a cable of extremely high resistance would need to be used. Looked at another way, given that both gold and non-gold plated cables exist and yet both work, the gold plating does nothing to improve the quality/reliability of the signal transfer for HDMI.

What Makes a Good Cable?

Good speaker cables don’t need to be flexible or made using OFC in order to perform well. Save the money spent on cables and spend it on a better amplifier and speakers. Good HDMI cables are well shielded. On balance neither gold-plating nor OFC is worth the additional money spent for essentially all consumer applications.

Service Request Metrics Killing Customer Satisfaction

Although I’ve run a very small IT “department” once it was nothing on the scale of the companies I’ve generally worked for. As scale increases and the number of people in support increases the measure of time from job opened to job closed is a metric that many management types enjoy using. The idea is that the so-called Service Request is made by someone needing help, the support person responds and helps (hopefully) and then the SR is closed.

Over the years I’ve been on the receiving end of support calls many times as I’m sure most other people have in one context or another: telephone companies, internet service providers, IT departments and human resources just to name a handful. Recently though there have been two incidents that have made we somewhat disenchanted with the concept.

Without going into too much detail, the SR was raised automatically by the system when the initial email was recived by their support system. From that point it sat until someone picked up the SR and then they responded by email with an answer but they ended their email with: “SR is now closed.”

Really? I never said it was closed. The information provided in one case didn’t solve my problem and in the other case didn’t even respond about the original subject! It simply can not be solely up to the SR responder to close the SR assigned to them without the originators consent. Essentially this allows the service provider to close out whatever SRs they choose to improve their own metrics whilst simultaneously solving no problems whatsoever.

On a separate occasion I was asked upon near-completion of a phone call, “Have we successfully answered your reasons for calling today?” The answer was actually no since the action was for them to kick me across to a different department in the same company. Upon saying this the response was, “They will assist you further but we have successfully answered your reasons for calling today.” No longer a question. I guess my opinion didn’t matter?

Of course their company clearly tracked SRs differently through different departments hence “New Accounts” let’s say and “Technical Support” would have their own SR tracking systems that quite possibly didn’t even talk to each other. The problem with that is that the individual department managers only care about their own SR closure rates and from a customer satisfaction perspective we’re interested in how the company performed overall in responding to our query.

If you’re going to use SRs in a large organisation and you don’t want to kill customer satisfaction then A) For multiple departments track SRs with a top-level SR that traces the total amount of time across all departments and B) Only allow closure of SRs with customer consent.