All posts by vesterli

Why Outsourcing Doesn’t Work

40 years ago, Fred Brooks told us in his book The Mythical Man-Month why full outsourcing couldn’t work. Since outsourcing was rare and difficult back then, nobody took note. Today, advances in communications and technology make outsourcing much easier. That doesn’t mean it will work.

The reason is that IT work is not uniform. There are some easy tasks (rapidly getting automated) and some hard tasks that take expertise and judgment. And most organizations are outsourcing their work to regions where IT professionals haven’t had the time yet to develop expertise and judgment.

Cost vs CompetencyIn a mature IT market, a wide range of skills exists, from basic to very advanced. As you need more advanced skills, the cost goes up, because there are fewer IT professionals with the requisite number of years of learning and experience.

In a new IT market, you can get basic competency cheaper. But because most IT professionals in these markets are relatively inexperienced, advanced skills are very rare, very expensive, and might not even exist.

The outsourcing fallacy is to think that you can move an entire, complex IT operation offshore. You can save money on moving simple tasks to regions with lots of competent but inexperienced IT people. But advanced skills won’t be available. So unless you can very cleanly separate simple tasks from advanced tasks, the communication overhead necessary to ensure that the right people get the right task will eat up any saving.

Think you can save money by outsourcing? Maybe you can. But many IT organizations have found they couldn’t. Get in touch if you need help figuring out the right level of outsourcing for your tasks and your organization.

Oracle Cloud Strategy Fail

Oracle rose to database dominance by making their software freely available. Anybody can download a $100K+ enterprise edition database and use it for personal learning as long as he likes.

The Oracle Cloud offerings, on the other hand, are strictly limited. You need to provide both a mobile phone number and a credit card number in order to get a miserly 30-day trial. Once you’ve spent your 30 days, you’ve used the one chance you get in this lifetime to learn Oracle’s 50+ cloud offerings.

Reality Distortion Field - Steve Jobs had it...

Contrast this with the approach taken by Amazon: A free tier without time limitation, and a generous 12-month trial for many of the other services. They took a page from Oracle’s playbook, offered free access and became dominant in the cloud space.

Oracle defends their stinginess by saying that it’s too expensive for them to offer free trials. And apparently, they believe they don’t need to offer good trials because their cloud is faster and cheaper.

Unfortunately, the ability of Larry Ellison to distort reality is limited. Oracle has a negligible market share in IaaS and PaaS and since they won’t invest a smidgen of their $60 billion cash hoard in better trials, they are unfortunately likely to remain a bit player in this space.

I’ve used my own phone number and credit card, and my wife’s phone number and credit card, so I’m now out of options for learning more about the Oracle cloud. But I’m learning AWS.

 

This is a post from the OraToolWatch newsletter. Don’t miss the next one, sign up

Cash is King

Medicins Sans Frontieres have an annual collection here in Denmark, and I was one of the volunteers going house-to-house to collect donations.

IMG_5599

I experienced quite a few people who didn’t have cash but still wanted to contribute. This is in alignment with recent surveys who show 21 percent of Europeans rarely use cash.

However, the belief that the cashless society will be a boon is utter techno-arrogance. It takes the average user approx 5 seconds to drop a few coins into my collection jar, and 10 seconds to fold a bank note and insert it. But nobody managed to complete an SMS transfer or mobile payment in less than 30 seconds.

It might be in the interest of shops, banks, and the tax collector to get rid of cash. But does it justify wasting 4,000 years of time globally every day? Consider the total cost to everyone in money, time and effort before you add technology to a process.

 

Don’t Let Developers Ruin Your Reputation

Developers hate it when pesky users raise bug reports against their wonderful creations. I’m a developer myself and have sometimes found myself mystified why a specific piece of code didn’t work in a specific case.

But don’t ever let developers tell the users that the code works fine, and the problem must be with the user.

I communicate with a lot of people and have been using Contactually to keep a central record of all my contacts. One nice feature of this tool is that it can use IMAP to connect to my mailbox to include the emails I’ve sent in the overview. For some reason, this IMAP functionality stopped working, and after some back-and-forth with support, I was told that the problem was with my password.

This is a rather disingenuous excuse, as the software already gives me an error if I enter an invalid password. It reminds me of the Beavis and Butthead episode called “Customers Suck“, where the two idiots don’t want to serve customers and can’t even be bothered to come up with a good excuse.

However, the poor customer service employee had no choice but to pass this lame attempt at blame-shifting to me. I had to cancel the service.

Make sure you are not allowing your developers to shrink their responsibilities and ruin your reputation with customers internal and external.

Why I Now Hate Apple (Hint: Chatbots)

In a famous Monty Python sketch, John Cleese tries to return a dead parrot to the shopkeeper where he bought it. However, the shopkeeper is impervious to reason and claims the clearly dead parrot is still alive.

Monty Python Dead Parrot sketch

I just had a “dead parrot” moment with iTunes support. Their support used to be excellent, but the humans have now been replaced by imbecilic chatbots. This is a serious miscalculation.

What really annoys customers is when they are not listened to, and not listening is the one core competency of today’s chatbots. Being served platitudes about “we understand you are unhappy” doesn’t make me happier…

If you are considering chatbots for some aspect of your operation, make sure to offer an option for the customer to give feedback. Apple doesn’t, and Tim Cook probably thinks their support is brilliant. It isn’t.

Oracle Critical Patch Update

Oracle has released the latest quarterly critical patch update (CPU). The database gets off lightly this time with two moderate severity vulnerabilities in SQL*Plus and the Oracle JVM. On the other hand, Oracle Secure Backup is not very secure with a bug that can be remotely exploited without authentication. Bad.

The Fusion Middleware stack gets 31 fixes, of which 20 are in the bad group of remotely exploitable without authentication. There is a lot of WebCenter stuff as well as some WebLogic and little Oracle Service Bus. Read the notes and update your environments.

Almost all of the Oracle applications (E-Business Suite, Siebel, J.D. Edwards) are also vulnerable, many through the critical Apache Struts 2 vulnerability (CVE-2017-5638). Oracle has fixed everything related to this Struts 2 bug in this CPU, but if you are running anything else based on Struts 2, make sure you update to a non-vulnerable version.

Do You Need to Learn a New Technology?

TL;DR: Yes.

Developers often ask me which language or tool they should use or learn. I have definite opinions on good and bad tools for various tasks, but the most important tip is to continue learning new technologies. This gives you joy in your life, prevents burnout, and provides a platform when the time comes to move on from your current technology.

For your day job, you want a language that is stable or increasing. It doesn’t matter if it is outside the top ten in rankings like the TIOBE index. For example, Oracle’s proprietary PL/SQL database language has been hovering around place number 20 for many years, and PL/SQL programmers are not likely to be out of a job anytime soon.

But you still need to continually add to your skill set. People who keep doing the same thing lose the joy and wonder of making something work, which is often what got them started in IT in the first place. To prevent burnout, carve out time to work on something new every week.

Don’t expect your employer to give you this time. In some organizations, you might be able to use allocated training time to learn something on your own, but even Google’s famous “20% time” for side projects is 20% on top of the 100% you already work.

You should be grateful you have the privilege to work in IT. To keep that privilege, you should invest time in yourself and your life.

This is an excerpt from the monthly Spiritual Programmer newsletter. Don’t miss the next issue, sign up here

IT Entering the Dark Ages (Again)

Historians have described the period following the collapse of the Western Roman Empire (400 to 1400 AD) as the “Dark Ages.” Existing knowledge was lost and society regressed to a more primitive organization and technology.

In IT, we do not learn from history. We routinely throw away existing knowledge to start over, constantly emerging from each dark age only to enter a new one.

I was just reminded of this unfortunate tendency when I opened The Economist on my iPad. I used to read the magazine in traditional form on dead trees (aka paper) but moved to their iPad app to get my magazine on the publication date and not two days later. Their first iPad app reproduced the magazine layout with several narrow columns of text, re-using centuries of typographical knowledge. But in the new version, the clueless digital natives have decided to make the text one wide column with the lines way too close together, which makes it much harder to read.

Next time you get the bright idea to change something that has worked well (a page layout, a business process, or an IT framework), reflect on whether the change will really make it easier for the system to fulfill its promise.

Why Does Oracle Exist?

At an Oracle Partner event this week in Croatia, I got the latest updates on Oracle’s products. And it hit me that nobody, not even Oracle, understands why the company exists.

I looked at their website but was unable to find a mission statement or discern any coherent vision. It seems Oracle exists simply because it does.

Back in their database days, they wanted to manage all the world’s information. But in their current incarnation, their vision more cloudy than cloud.

Why on earth does an enterprise software company like Oracle dabble in chatbots? Why are they building an IFTTT clone? Why are they running “Oracle Code” events and talking about everything but Oracle software? Why are they coming from a sub-one-percent IaaS market share and announcing their intention to rule the IaaS world?

What Oracle should do is:

  • Build on their strength in SaaS. I believe they are on track to living up to Mark Hurd’s vision of being one of the two SaaS vendors  with 80% of the market (and no, SAP won’t be the other one)
  • Provide PaaS trials limited in power, but not in time. Nobody can figure out how to use Oracle PaaS offerings in meager 30-day trials
  • Concentrate on real differentiators like Application Builder Cloud Service (and stop trying to provide their version of every cloud service in the universe).

Oracle is a great software engineering company. I hope they figure out why they exist.

Oracle Stock Rises on Cloud Surprise

Stockbrokers were taken by surprise by Oracle’s Cloud revenue when Oracle announced quarterly results last week, and Oracle stock duly jumped by seven percent. It has fallen back somewhat since but is still up three percent.

ORCL

(source: Yahoo Finance)

Oracle Cloud revenue is up by 63% and now makes up 13% of Oracle’s $9.3 billion quarterly revenue. It is not clear how much of this is the “cloud credits” that is reportedly bundled into renewal and new on-premise deals. It will be interesting to see if customers find a good use for these credits and will buy more once they are used up.

As an ERP and database company, it would make the most sense for Oracle to push their strong SaaS and PaaS offerings. SaaS and PaaS currently make up 85% of Oracle cloud revenue, but they have decided to try to muscle into the already-crowded market for commodity computing services. With $195 million of IaaS revenue, it doesn’t make much sense for Oracle to try to catch up to Amazon’s $3.5 billion.