Skip to main content

Initializing Scrum

Update: Just fixed some outgoing links.

After I began working at my new workplace, I've been gently pushing Agile here and there. We've been brainstorming how we can get agile working for our people, and so far we've adopted a few patterns and practices, but putting those plans into action has taken some time. Originally I hoped that things would go very quickly, but agile is about people, and people are hard to change.

Getting better, always.


We have a product that works really well, and pretty much all of the developers are happy. (Generally speaking, this state can be dangerous for a business. There is a danger of falling asleep, one day waking up to see that your competitors have gotten a good hold of your customers, with a better product and better marketing.) Having an environment that strives for improvement is a critical part of persisting success. And striving for improvement, well, that's what agile is.

Luckily, our team of developers do a lot of activities in the hunt for technological improvement. There is a team of architects (and by that I mean hands-on hard-core, coding Java gurus) that work across the other developers, singling out targets for refactoring and other improvements. Another practice I really dig is the weekly Java-team meeting, where one of the developer present a subject either on good programming practice, new tool/feature, or explaining a certain part of our domain model.

As well as pushing for technical excellence, we also want to push for process excellence. We want to give accurate estimates, we want to deliver fast, and we want to deliver the stuff that is most important first. This is where I (and the people who hired me) figured that Scrum would be a good try, but before I explain how the first week of scrumming has gone, I need to explain a little more about our team.

Who does what?

We've got nearly twenty developers working on our product (there are other teams in the company, but we're an independent gang for the most part). In this product there are many families of functionality, and there is no clear designation of which developer does which functionality. There is a lot of tacit knowledge on who does what: "Bob did a lot of work on the fizzerburpilator", or "Bill wrote the shackalock-editor, and occasionally does some maintenance on it" (features are given fantasy names for non-disclosure reasons).

Understandably, it is a bit fustrating for our customer-contacts to ask for new features, because often (a) they don't know which developer can do it, (b) the chief architect is too busy to spend time managing the tasks of every single developer, and (c) the developer who knows how to make the new feature is busy doing other critical stuff.

The solution to this frustration seems pretty clear: Organize into sub-teams, preferably by feature. However, we can't just jump right in and do a full-blown re-organization, we have to be careful about it, and we have been waiting for a fitting opportunity to try this out.

Taking Scrum for a spin

So, recently we had a big customer come to us with a feature request, let's call it "fizzburp hickup". It is a bit too big for one developer to do by himself, and there are some architectural challenges which would probably require a weighed consensus of developer opinions rather than the brilliant idea of a single programmer. The scope and definition of "fizzburp hickup" are pretty well described, and we have an ample time-frame, so we figured that this would be the perfect opportunity to try out Scrum with a small team of developers, called Team Espresso. If the experiment goes well, Team Espresso can continue existing after their feature has been developed, keep on maintaining it and adopting some other features as well.

The first sprint got scheduled to start this Monday. I knew that boiling up the backlog for the team would eat up alot of time, so I started the week before by chipping the requirement spec into user stories, using the customer-contact for verifying that I had gotten a good understanding of the domain, and also tried out User Story Mapping for the first time. I ran through the story cards with our chief architect and got a first-impression of estimates. Getting him to put estimates on user-features instead of architectural components was really hard, but in the end we came through with a rough estimate of the backlog-effort. We used 'tricky points' as our unit of estimation.

We then had half-a-day "kick-off" with the customer-contact and architects. First we went through the requirement spec and had the customer-contact (who unfortunately went on vacation and was represented by the, erm, chief architect) communicate the motivation for the new features. Naturally, the focus went technical early on, but I've been assured that there has already been so much communication with the customer on the "fizzburp hickup" already, that we shouldn't need any more input from them for the time being. Hmmm!

Sprint planning

After the kick-off, we gathered the team around the story map to do estimation and commitment to the first sprint (one week!). Naturally, the team zoomed down into technical discussion on every other story card (which is a good thing!). We discovered a good few open technical and functional questions that we would have to take back to the architects and customer-contacts. I thought this process was very fruitful, because we were finding challenges early that we would have to face sooner or later.

However, in the rest of the team there was a feeling of "let's just go and get started it", while I was forcing them to put tricky points on the user stories, and again it proved difficult to get them estimating features instead of technical components. Their point (as I understood it) was that in the end, we would have to deliver the same functionality anyway, and thinking in user-stories instead of components would only disturb our estimation thinking and architecture plans. That actually was a good point. My reply was "do all the architectural consideration you want to, but leave the story cards", in other words: bear with me on this. Finally, we committed to our first sprint, choosing three stories summing up to 30 tricky points.

And off we go!

On the first day of the sprint, I tried to get the team breaking the three stories down into tasks with hourly estimates. This proved even more difficult. I was inflicting even more Scrum overhead on a team which wanted to get to work. The point of breaking backlog items into tasks is to get a picture of remaining hours for sprint burndown, as well as assigning tasks to point person. Does this need to be explicitly done in such a small team for such a small project? I think so, because for a practiced team this overhead is virtually nothing. For my team in this sprint, however, we let it go.

Halfway through the sprint, the team got a bit nervous about delivering/presenting the result to the customer. Showing off that little progress would only anger and confuse the customer, making them come up with a burdening set of detailed requirements to "speed us up". So we extended our sprint to last two weeks instead of one.

We also changed around the planned story items. The architects found the time to do some smart fixes that allowed us to grab and do some of the more important user stories off the backlog.

At the end of the sprint now on friday, we had completed only one of the three committed stories, plus one extra because of the architecture fix. We had made alot of progress on the other items, but not completed them.

Retrospective

So, in order to help us reflect on how we should continue in the next sprint, let us sum up what practices we have tried and given up:

* Daily standup - stopped this, because we are already doing daily standup with the rest of the developers (I might write a post about how it was to get people adopting daily standup later).
* Story planning, we will continue this, although we are aware that it doesn't give us that much value without an on-site customer.
* Task tracking - stop this, too much overhead for such a small scope.
* One week sprints - go over to two week sprints instead.

I hope that we will be able to resume some of these, the 1-week-sprint in particular. There is a big organizational challenge which is making it difficult for us, and that is a quite common one in organizations adopting Scrum: We are part of two teams: On the one hand, we are Team Espresso, an agile group dedicated to deliver the "fizzburp hickup" functionality. On the other hand, we are part of the larger developer team, supporting and maintaining all the bits and pieces of our software, getting ready for our coming release ("fizzburp" is scheduled to be released in the next one after this one). Juggling these two on an hourly basis is proving quite difficult for the team.

I'll try to post back with an update after the next sprint. It could be that my expectations for how smoothly this would go were a bit too high, but hopefully I will be able to post back with a bit more uplifting results by then.

Popular posts from this blog

Encrypting and Decrypting with Spring

I was recently working with protecting some sensitive data in a typical Java application with a database underneath. We convert the data on its way out of the application using Spring Security Crypto Utilities. It "was decided" that we'd be doing AES with a key-length of 256, and this just happens to be the kind of encryption Spring crypto does out of the box. Sweet!

The big aber is that whatever JRE is running the application has to be patched with Oracle's JCE in order to do 256 bits. It's a fascinating story, the short version being that U.S. companies are restricted from exporting various encryption algorithms to certain countries, and some countries are restricted from importing them.

Once I had patched my JRE with the JCE, I found it fascinating how straight forward it was to encrypt and decrypt using the Spring Encryptors. So just for fun at the weekend, I threw together a little desktop app that will encrypt and decrypt stuff for the given password and sa…

Managing dot-files with vcsh and myrepos

Say I want to get my dot-files out on a new computer. Here's what I do:

# install vcsh & myrepos via apt/brew/etc
vcsh clone https://github.com/tfnico/config-mr.git mr
mr update

Done! All dot-files are ready to use and in place. No deploy command, no linking up symlinks to the files. No checking/out in my entire home directory as a Git repository. Yet, all my dot-files are neatly kept in fine-grained repositories, and any changes I make are immediately ready to be committed:

config-atom.git
    -> ~/.atom/*

config-mr.git
    -> ~/.mrconfig
    -> ~/.config/mr/*

config-tmuxinator.git  
    -> ~/.tmuxinator/*

config-vim.git
    -> ~/.vimrc
    -> ~/.vim/*

config-bin.git   
    -> ~/bin/*

config-git.git          
    -> ~/.gitconfig

config-tmux.git  
    -> ~/.tmux.conf    

config-zsh.git
    -> ~/.zshrc

How can this be? The key here is to use vcsh to keep track of your dot-files, and its partner myrepos/mr for operating on many repositories at the same time.

I discovere…

Always use git-svn with --prefix

TLDR: I've recently been forced back into using git-svn, and while I was at it, I noticed that git-svn generally behaves a lot better when it is initialized using the --prefix option.

Frankly, I can't see any reason why you would ever want to use git-svn without --prefix. It even added some major simplifications to my old git-svn mirror setup.

Update: Some of the advantages of this solution will disappear in newer versions of Git.

For example, make a standard-layout svn clone:

$ git svn clone -s https://svn.company.com/repos/project-foo/

You'll get this .git/config:

[svn-remote "svn"]
        url = https://svn.company.com/repos/
        fetch = project-foo/trunk:refs/remotes/trunk
        branches = project-foo/branches/*:refs/remotes/*
        tags = project-foo/tags/*:refs/remotes/tags/*

And the remote branches looks like this (git branch -a):
    remotes/trunk
    remotes/feat-bar

(Compared to regular remote branches, they look very odd because there is no remote name i…

The Best Log Viewer Ever

This is what it looks like when I want to have a look through the logfile, to see what a user did on one of our machines one day:


Read the whole story about how it works on the Viaboxx Systems blog (and upvote on DZone!).

Microsoft ups their Git efforts another notch

This week Microsoft announced first class Git support embedded in the coming version of Visual Studio.

Now, it's not completely shocking. We could have seen it coming since Microsoft started offering Git repos on CodePlex, and more recently offering a Git client for TFS. In any case, these are some big news. Scott Hanselman weighs on some features and some more background here.

For those who are a bit unaware of what the Git situation on Windows looks like these days, I've dotted down these notes:
Some explanation on these:

msysGit has long been The Way to use Git on Windows. It's basically a port of Git itself, so it's a command-line tool.GitExtensions (includes Visual Studio integration), TortoiseGit, Git Shell, posh-git and most other tools are powered by msysGit.libgit2 is a native library for doing Git stuff. It is developed completely separate from Git itself. The above tools could (and should) probably use libgit2 instead of hooking onto and around msysGit.Github…