DDE Does Disruption

This week we received an incident about an application that was not accessible any longer. It turned out that all the design elements where signed with my Notes ID which has (of course) insufficient rights to run in the Production environment.

The curious case here is that nobody intentionally replaced or refreshed the design of the application.

The application still had the “inherit design from template” option enabled. However the template resides on a different server and there is no replication between the Production and this staging server.

I know I have the two databases somewhere in a Working Set in my Domino Designer so most obvious option for the design refresh must have come from my workstation even though I have not worked with these applications this week .

Have you experienced something similar also? What was the cause?

Source Control disruption in DDE

In a project me and my colleague faced major complications working towards a GIT repository on TFS. In short: she did not see the changes I had made in her Domino Designer even though her GIT client had transferred everything correctly to the On Disk Project (ODP) on her local drive.

It seemed that the problem was totally focusing on her development setup. For the good order here is how our both environments look like:

  • Me: TFS, SourceTree, Domino Designer FP8.
  • Colleague: TFS, Visual Studio, Domino Designer FP8.

As a result we are facing major delay in the project because we can’t rely on the setup so we are 50% developer resources short.

So in order to test our bad experiences we sat up a small test with some other colleagues to see if we could reproduce the disruption and allocate where it occurs.

We had no particular test script in mind. We just sat up a new repository on TFS, we cloned it locally with different GIT clients, imported the ODP’s in DDE, created new NSF’s from it and we started to make changes in them.

Luckily for us a new disruption appeared fairly quickly. Changes made by two of my colleagues did not appear in my NSF but they could see each other changes.

It turned out that the changes where in my ODP but DDE was not able to transfer them in the NSF.

Here are some dumps that shows the situation. The use of capitals in the H2 element are not similar.

XPage in DDE:

XPage in ODP Project in DDE:

Same XPage in ODP on local drive:

Removing the imported ODP, removing source control with the NSF, re-importing the ODP in DDE and re-establishing source control for the NSF with the new project caused that I could see the changes in the NSF. However with the next change which a colleague made we were facing the same disruption on my machine.

For us it clear that this is not a work situation we can trust when DDE is not able to make the transition from ODP to NSF.

Have you experienced the same and have you come up with a work-around?

For us it would mean that we can not complete the DevOps chain that is set for development environments…

XPages sufficient for line of business type of applications?

Hi there, currently I am following another Angular course since it seems to have become the leading development framework at work. So back to learning all the rules within Angular.

At the moment I am modernizing a Domino application with the help of XPages which:

  1. Implements Model-View-Controller architecture, mostly inspired by the guys at Pipelia since IBM never told us to do so.
  2. Is written in Java to support the MVC architecture and to have close integration with XPages runtime.
  3. Is using Expression Language wherever possible to avoid usage of SSJS.
  4. Uses the lifecycle of JSF in XPages at the max.
  5. To cover support for different devices I am using Bootstrap as front-end framework. So I miss some native behavior which I do not tend to cover-up.

So far so good and I think I have come quiet long in my project so I still dare to call it rapid application development.

The code-base has been reduced dramatically and all exotic upcoming JavaScript libraries from the early 2000 I have been able to replace with just XPages. With my latent UX skills and extending the out of the box Bootstrap I might now even call this application ‘sexy’ 🙂

I know I haven’t touched many areas discussed in the XPages community such as:

  • Websockets (I do not see a use-case yet).
  • Writing Java servlets (please pass me a demo NSF).
  • Watson services (cloud is still a sensitive topic).
  • set up micro-services with smartNSF and consume them in my Java code with an mapper library (requires changes in the environment).
  • Integration with IBM Connections.
  • Redefining my data with the help of a Graph DB.

Either I see little usage, it is not possible or there is no-one to guide me (the information is certainly not provided by the vendor).

So now back to Angular. Learning all these rules, technologies and new tools setup I was wondering what new technical options this framework will bring me at work. Reflecting on the type of customer-orders I receive I am wondering:

Is XPages not sufficient for most of your line of business apps?

Perhaps you have a though about this?

Happy development & enjoy your summer 🙂

Where are we?

Today I received a question on a post I wrote some time ago. I had to start up my local development environment just to see if I remembered where I placed the code that was asked for.

So where are we? Or what are we doing?

Still Domino. Still XPages. But then mainly in Java. The story about application development with XPages that IBM did NOT tell you. Why? You may tell.

I was thinking about gathering my past experience and ideas and combine it into a ‘product’ but workload and a private life with small kids is preventing me from this.

OpenNTF? Naaa. Too tired.

CU at Engage? Probably not (I am a developer and development sessions are scarce).

CU at GraphConnect? Good chance!

This blog? Yes, probably some new posts soon.

Something in the pipeline? Yes, tomorrow a presentation for colleagues on using Java in XPages development.

For now: time to watch some CL football 🙂


Quick responsive type-ahead function

For an XPages form I needed a workaround for a combobox with a large data-set. At first I tried the select2 jQuery plugin so I could also have a filter option as some form of type-ahead. However it turned out that my select2 list became quiet unresponsive when an entry far in the list was selected (generating the list and filtering was fine).

So my next option was to go for a type-ahead approach. First hit on Google was a NotesIn9 vid about a fancy type based upon the original post of Tim Tripcony (hello 2009).

Due to the large data-set instead of the getAllEntriesByKey method I decided to go for a getEntryByKey method and from the found view entry continue with createViewNavFrom. This turned out to be extremely responsive!

Further I noticed in the original code a hash object is used so that does not guaranteed me a correct order of my result list so I added an additional array, just for a returning a list with correct sort order.

I like especially in this approach the option to control the returned markup so think here about Bootstrap Media object lists. And of course the ease to set the value that you want to have returned in the edit box.

This is not my first blog-post on delivering search function to your application so now you have just another solution available for your toolbox.

The script is available as a gist.

Happy development =)


Get rid of “Infinity” as the result of a @DbColumn in a combobox – Quicktip

Who has not experienced the following:

XPages has a 32 or 64 ( can’t remember which ) limit for the lookup formula whereas in Notes the limit is only for the whole of the formula e.g. after @unique is applied.

It makes it very hard to have meaningful drop downs for filtering data.

Instead you can use SSJS e.g.:

var lookupView:NotesView = database.getView(“<LOOKUPVIEW>”);

This is the equivalent to:

@DbColumn( “”,”<LOOKUPVIEW>”, 1 );

and returns you the full result without the limit.

But this also brings me to the following problem:

When I enable Select2 on a loooong combobox the list is populated the type-ahead / filtering works fine BUT when I select a value long down the list the response of the select2 combobox becomes really sloooow.

Does anyone has an approach to this?

Display JSON pretty in a textarea control

In an application I am using heavily JSON default as the data-format. To make debugging a bit more easy I am working on a toolbar to be able to check the content of the JSON objects.

Normally I use JSONLint most of the time to prettify (and validate) the JSON so it becomes a bit easier to read and check. But this copy and paste habit becomes more and more obnoxious so there I wanted to have a toolbar (like the debug toolbar) to have a good overview of the JSON objects.

However if you write the JSON object to a computed field or text area field (inputTextarea control in the XPages world) it is just plain, unformatted text. Not pretty for the eye and not much helpful for the brain.

So I tested a simple approach which turns out to run just fine. I display the json object in a textarea input control and then with CSJS I format the content. Here is some sample code:

<xp:inputTextarea id=”input-fieldx” defaultValue=”#{javascript:App.currCollection}”>
<xp:button value=”Make Pretty” id=”button1″>
<xp:eventHandler event=”onclick” submit=”false”>
<![CDATA[var name = x$(‘#{id:input-fieldx}’).attr(“name”);
<xp:scriptBlock id=”scriptBlock1″>
<![CDATA[function prettyPrint(id) {
var obj = dojo.byId(id);
var ugly = obj.value;
var json = dojo.fromJson(ugly);
var pretty = JSON.stringify(json, undefined, 4);
obj.innerHTML = pretty;

Here is what it looks like:


The content how it looks initially.


The content how it looks like after running the code.

Note I had to include the [ ] signs in my code to make the JSON valid due to circumstances in my code, I removed it from the sample code above.

At least now I understand my JSON objects again and I can avoid some copy and paste to JSONLint 🙂

This is probably my last blog of this year so I would like to wish you some wonderful Christmas days. Take care and best wishes for a healthy 2017!



Nice reversible cards in a IBM Verse style

In an application I am displaying objects with cards. Because the type of objects may differ in type and amount of data (text, images, tables, dates) the cards to display them differ.

Some cards have interaction (popup, dialogs, reversible) and some are just static. In this post I will discuss the reversible card type.

The reversible card has as features:

  • a front card with a folded corner, if you move over the fold an info icon appears which you can click to see the card’s back.
  • when you flip over to the back or front it is all nicely animated.
  • on the back of the card there is another icon to switch back to the card’s front.

The purpose of this reversible card is to save space in my UI, and to provide layout for secondary information, which is more important if you want to look at more detailed information.

It could also be a placeholder for data that, when placed in the front of the card, ask for too much landscape (e.g. a chart).

Things become much clearer with images so here are a few:


Image: Default presentation.


Image: Expand folding (animation) when mouse-over.


Image: animation when flipping over a card.


Image: the back of the card.

I use the cards within a repeat control and with Bootstrap they align nicely and are responsive:


I have no idea when Bootstrap 4 will be released with it’s card component but according to my experience cards are … less useful without interaction.

I made a small video to demonstrate the reversible cards feature:

If you want to implement the card yourself here is the code:


Most of the chemistry is performed with CSS so you also want that:


.flip {
-webkit-perspective: 800;
perspective: 800;
position: relative;
text-align: left;

.flipped {
height: 100%;
-webkit-transform: rotateY(-180deg);
transform: rotateY(-180deg);

.card {
height: 100%;
-webkit-transform-style: preserve-3d;
-webkit-transition: 0.5s;
transform-style: preserve-3d;
transition: 0.5s;

.card .face {
-webkit-backface-visibility: hidden ;
backface-visibility: hidden ;
z-index: 2;

.front {
/*position: relative;*/
background-color: #FFF;

position: absolute;
width: 100%;
z-index: 1;
border: 1px solid #e3e3e3;
box-shadow: 5px 7px rgba(0, 0, 0, 0.15);
transition: background 0.4s ease;

.front-content, .back-content {
padding: 20px;
.front-content:before {
width: 80px;
height: 80px;
float: right;
.corner-tip:before, .corner-tip:after {
background-color: #FFF;
position: absolute;
display: block;
z-index: 2;
border-top-right-radius: 60%;
width: 50%;
height: 50%;
content: “”;
.corner-tip:before {
right: 100%;
top: 0%;
background: -webkit-radial-gradient(-180% 200%, circle, rgba(255,255,255,0) 80%, rgba(0,0,0,.2) 100%);
.front:hover .corner-tip:before {
border-right: solid 1px #fff;
.front div.corner:hover .corner-tip:before {
border-right: solid 2px #fff;
.corner-tip:after {
top: 100%;
right: 0%;
background: -webkit-radial-gradient(-250% 320%, circle, rgba(255,255,255,0) 80%, rgba(0,0,0,.2) 100%);
.front:hover .corner-tip:after {
border-top: solid 1px #fff;
.front div.corner:hover .corner-tip:after {
border-top: solid 2px #fff;
.corner { /* edit these sizes for the default revealing corner size */
height: 20px;
width: 20px;
right: 0;
top: 0;
position: absolute;
overflow: visible;
.front:hover .corner { /* edit corner size (First animation, when the whole page is rollovered) */
height: 30px;
width: 30px;
.front div.corner:hover { /* edit corner size (Second animation, when the corner itself is rollovered) */
height: 50px;
width: 50px;
.corner:before {
position: absolute;
top: 0;
right: 0;
content: “”;
display: block;
width: 133%;
height: 133%;
.corner-contents:after {
position: absolute;
top: 0;
right: 0;
content: “”;
background: -webkit-linear-gradient(45deg, rgba(255, 255, 255, 0) 37%, #DDD 62%, rgba(230, 230, 230, 0.1) 64%, rgba(255, 255, 255, 0) 67%), -webkit-radial-gradient(-50% 150%, circle, transparent 74%, rgba(0, 0, 0, 0.2) 74%, transparent 81%);
display: block;
width: 133%;
height: 133%;
.corner-tip {
position: absolute;
top: 0;
right: 0;
content: “”;
background: -webkit-linear-gradient(45deg, #ddd 17%, #dfdfdf 18%, #f5f5f5 30%, #f8f8f8 34%, #eee 39%, rgba(200,200,200,0) 41%);
display: block;
width: 100%;
height: 100%;
.corner-button {
position: absolute;
top: 0;
right: 0;
color: #fff;
text-align: center;
display: inline-block;
font-size: 1.6em;
.corner-contents {
width: 125%;
position: absolute;
display: block;
overflow: hidden;
-webkit-mask: -webkit-linear-gradient(45deg, transparent 49%, #000 53%);
top: 0;
right: 0;
height: 125%;
._corner-contents:before {
content: “”;
position: absolute;
top: 0;
right: 0;
content: “”;
display: block;
width: 100%;
height: 100%;
background-color: #eeeef4; /* Match this background color to #fpc_effect-back */
.corner, .corner-contents, .corner-tip {
-webkit-transition-property: all;
-webkit-transition-duration: .3s;
-webkit-transition-timing-function: cubic-bezier(0, 0.35, .5, 1.7);

.back {
width: 100%;

-webkit-transform: rotateY(-180deg);
transform: rotateY(-180deg);
border: 1px solid #e3e3e3;
box-shadow: 5px 7px rgba(0, 0, 0, 0.15);
transition: background 0.4s ease;


.card blockquote {
border-left: none;
margin: 0;

.card blockquote img {
margin-bottom: 10px;

.card blockquote p:before {
content: “\f10d”;
font-family: ‘Fontawesome’;
float: left;
margin-right: 10px;

Okay, you are now all powered to impress your customers. Happy development 🙂

I am a certified Graphista

Sweden has many IT solutions to be proud of. I assume you are using/have used Spotify or Skype.

Perhaps you are not aware but another “Swedish” product you might have been using is Neo4j, a graph database management system. Graphs are all around us. Typical use cases are recommendation systems, fraud detection and social networks.

You can also make your IBM Notes data more ‘connected’ by using the Graph DB capabilities available in the OpenNTF Domino API.

I chose to walk the easiest path to learn about Graph data modelling and that was to learn the probably most documented, described, leading Graph Db: Neo4j. After a while I felt comfortable with all the terminology, concepts and programmatic use so I felt confident to take the test to become a Neo4j Certified Professional.

And voila!, here is the result:


The test was harder than expected, the time pressure was significant and some subjects out of scope of my experience (administrator experience in a Production environment) but with logic and wishful thinking I came to a positive result 🙂

So far I have not read too many use-cases where IBM Notes data was used within Neo4j or applications written in XPages using the graph db of Neo4j. But what is not is yet to come…