Hacker News Reader

Agile Manifesto (http://www.agilemanifesto.org/principles.html):

>I did not do Sprint planning meeting and ticket estimation (Agile poker cards) because people on my team loved to argue over minutiae's and I strongly suspected half of the team didn't know the details of what the other half is working on; So instead, we had a weekly "flowchart" meeting on a large whiteboard where we brainstormed and added more details to our software project (a big data pipeline lends itself to a flowchart); I found it to be a better visualization of the project than a post-it covered poster.

Principle #5: Build projects around motivated individuals. Give them the environment and support they need, and trust them to get the job done.

> The crux of the argument was Agile wasn't well suited for our team; the sprint model of Agile easily lends itself to arbitrary deadlines and building for the "demo" of the story/epic of that sprint and build-up of tech-debt

Principle #7: Working software is the primary measure of progress

>In the end, management relented and we ended our Agile experiment. Project tracking is done in Google Doc's with milestones instead

Principle #12: At regular intervals, the team reflects on how to become more effective, then tunes and adjusts its behavior accordingly.

Presuming that you're delivering more responsively and quickly to the customer than you were before, which you didn't mention, it sounds like you're doing just fine with Agile.

Thanks lr4444lr for highlighting and aligning the various principles the Agile manifesto with the minutiae details of the project I described. I definitely see your point (which I interpret as Agile is not a set of rigid techniques but one can adapt a team's makeup and culture to whatever will make software people feel motivated/empowered to do the work and have quick responsive/iterative cycles with the customer folks).

>Presuming that you're delivering more responsively and quickly to the customer than you were before, which you didn't mention

In the interest of full disclosure, I think we had a project time over-run of 2x (estimated 4 months to 8 months) mainly due to "feature creep" (is that a feature or a bug in Agile iterative cycles?); the project is a high-throughput data pipeline for genome sequencing and annotation. We overran time as a case of our customers/users wanted to add more bells and whistles to the tail end of the pipeline as they saw the end product more and more and proposed more added steps.

In the interest of full disclosure, I think we had a project time over-run of 2x (estimated 4 months to 8 months) mainly due to "feature creep" (is that a feature or a bug in Agile iterative cycles?)

That's a sign of weak leadership by whoever was leading the project.

New feature requests will always come. Agile is about prioritizing them compared to what is already in the backlog.

You appear to be conflating Scrum with Agile, avoiding doing any Scrum-like practices and avoided the clear communication problems within your team ("I strongly suspected half of the team didn't know the details of what the other half is working on").

And then you say your Agile experiment didn't work.

the sprint model of Agile easily lends itself to arbitrary deadlines and building for the "demo" of the story/epic of that sprint

"Sprints" isn't Agile, it's Scrum (See for example Kanban).

and build-up of tech-debt (e.g., build a demo for one sprint, spent next 3 sprints fixing bugs for the features of the first sprint).

That is the complete opposite of anything that any kind of agile process would recommend.

To be honest, given that you don't know the difference between Scrum and Agile I think it's fair to say you don't know what you are talking about.

I'm no fan of pure Scrum. I think it's an overly ceremonial approach to building software and the ceremonial aspects of it lead to a cargo-cult like mindset, where people concentrate on the process rather than the reasons why things are done.

But Agile as a general approach works well, and a properly run Scrum process isn't a bad place to start.

Thanks nl for your thoughtful feedback! I'll follow up to all your reply threads here.

I think you pointed out my negligence in not dealing with "communication problems" on my team, "weak leadership" for not putting my foot down (not prioritizing the backlog of the project and thus making the project delivery date late), avoiding morning standup's completely instead adapting and adjusting a good process on the fly (ensuring that everyone during their updates are succinct); and finally just my plain ignorance in what is the simple distinction between Scrum and Agile is, all points duly noted :) (And being honest, I don't agree with 100% of your criticism; but I do take to heart all of them and will try to work on improve on those specific things you brought up in terms of project management skills in the future!)

Although I think we disagree on what is a good project management methodology for software, I think at least we can agree on the fact that having a proper lead to run the project irrespective of the project methodology is really important.

Agile/scrum defense arguments are like a doomsday sects defenses for why dd didn't happen, everybody didn't believe enough. Now if all would believe...

> I did not do morning standup's because I did not think I was going to get buy-in from the rest of my team

Standup is expressly designed to function as a substitute for meetings - and to avoid the situation where employees waste time in meetings where only 10% of the meeting content is relevant to their particular domain/work.

The reason individuals are required to stand, is so that there's a slight physical discomfort in order that people report quickly (ours are about 1 minute per person).

If there's some topical convergence on a bit of work that emerges from standup that requires further elaboration and communication between members - then the idea is that those team-members will be able to self-organize a meeting outside standup.

The key benefit is that noone else on the team needs to be distracted with things that don't concern them.

It really puzzles me, that so few seem to comprehend the basic principles behind Agile workflow.

How do you do standup with a remote workforce?

Don't be overly literal, and use a conference call, Google Hangout, etc. If time zone differences prevent even that, do a "virtual standup" using email, wiki, blog, or whatever you have available.

How is that different to a meeting? I'm not an agile practitioner, so apologies if I've missed something important.

The thing about the "standups" in Agile, (and Scrum in particular) is that they're supposed to be about intra-team communication. They're not status reports for management, etc. in particular. The goal is to share with your colleages what you're working on, what you're about to work on, what issues you're having, etc., in order to make sure everybody is on the same page and to eliminate misunderstandings.

If you have a remote team and people can't meet F2F, you can accomplish the same end using something like I mentioned above. You lose a little something if people are in different time-zones because it makes the whole thing more asynchronous, but you do what you have to do to accommodate your circumstances. Remember, one of the tenets of the Agile Manifesto involves not having rigid adherence to strict ceremony for ceremony's sake (I'm paraphrasing a bit).

"Figure out what works for your team, and do that" is kinda how I'd summarize the whole thing.

Thanks julian_ for your explanation of the purpose behind morning standup and how they improve people's work life.

I have to respectfully respond that I've not experienced such successful implementation of morning standup's (and btw not discounting at all your own very positive experience at all). I'll give you a few counter-examples just to offer the other perspective,

Morning standup's used as a pulpit box by the tech lead or team members who like to talk; e.g., "I was working on this ticket yesterday, and I learned something new about Java and garbage collection, let me tell you guys..." (+5-10 minutes/per person)

Morning standup's that become quick resolution and priority meetings for project managers for either early or late tickets; e.g., "Mr.PM, I finished this ticket; and now my queue is empty, I am not sure which ticket I should work on next ... Let me see, Mr.Developer, let me take a look at this large post-it board to figure out the priorities and we can go and forth and gauge your interest on several tickets I'll propose for you next..." (+5-10 minutes/per ticket)

Morning standup's that combine two teams or force separate people who have been working on completely unrelated projects for quite some time but report to the same PM, listen to the unrelated details. You can imagine how eyes glaze over for people not involved at all on those projects every morning during stand up. (perceived waste of time).

However I even agree that morning standup's are great for tech shops with very well-defined workflows and teams that has buy-in. But I also have to be respectfully honest and say that I haven't met any person who enjoy standup's except the project manager (but it might also just be a selection bias since I don't favor standup's, people who only hate standup's confides to likely-minded people aka "non-management material").

I'll give you a few counter-examples just to offer the other perspective

So.. fix those problems? None of them are hard for a good Scrum master (or whoever has the job of keeping standup moving) to fix: "Ok, we'll take this offline. Moving on".

I very much understand that there needs to be developer buy-in, that management can and will attempt to derail the process, and that dominant personalities who don't understand their purpose may use them as a pulpit.

I would still argue, that those particular issues will create worse outcomes under other management approaches. The distinguishing feature of scrum is that it is a bottom-up approach created by developers. itil and prince look ridiculous in comparison - and in terms of their rigidity in trying to formulate a set-of-rules that can be used to cover all software management processes.

At the very least scrum is a baseline of principles, created by developers for developers that can be used to argue against unproductive intrusion by management.

> Morning standup's that become quick resolution and priority meetings for project managers for either early or late tickets;

How it works for us :-

In our standups, no questions are allowed, and no dialog.

If PMs and a team-member need to convene it's strictly after/outside standup so no-one else's time is wasted.

If something urgent (production server is down) needs to be prioritized, it goes on the board as a VIT ticket.

No team member is ever compelled to interrupt what they are currently working on to pick up a VIT ticket. This forces management to correctly estimate the value of automated fail-over, redundancy, one-click deployment/roll-back systems etc.

The problem with standups is there's never a good time to schedule them. You can't schedule them for right at 9:00, because then half the team will miss it because they're dragging in, and if you schedule it for a little later on, like 10, 10:30, 11, you're basically destroying flow for the entire morning.

Throw in the fact that they are completely useless. It's like Twitter - if you've only got such limited space to talk, you can't actually say anything worthwhile.

As a general rule of thumb, if a meeting can be replaced by an asynchronous group chat or email, it doesn't need to happen.

You schedule them right before lunch, I've always felt

Aren't standups mostly a way for management to say "now the standups are at 9:30 and everyone have to be at work at 9:30" ? :-).

Also it's clearly micromanagement: Why does everyone, including your boss, have to know what you did yesterday ?

> Why does everyone, including your boss, have to know what you did yesterday ?

To identify blockers. If a member is blocked by some external factor over which they have no control, then the team leader is required to intervene/escalate to solve it to allow the member to progress.

Formulating what you did yesterday, what the goal of work is today and identifing blockers in a few short sentences (literally 30s-1m) is less management overhead than anything else I have experienced. Certainly beats filling in reporting documents.

Just because you weren't doing SCRUM™ doesn't mean you aren't being agile.

The whole point of Agile is to be able to respond to change and make better software.

It's good to read the agile manifesto every once in awhile to remind you of what agile really means. http://agilemanifesto.org

Beat me to it! It's amazing how easily we can get so mired in the jargon and products built around a certain conception of "agile" that we lose sight of the core principles, even to the point of violating them completely.

Thanks RandallBrown for teaching me the distinction. SCRUM is the methodology of morning standup's, poker playing cards while Agile is the ideal to which the techniques aspire to achieve.

The first place I was introduced to Scrum was at Transgaming. They managed to have a daily scrum across two locations, using only speakerphone. So ... the meeting itself is doable. Not sure about the rest.