Finish the Job | Hacker News Reader

Finish the Job

birchtree.me

45 pointscharlieirish23 comments

My experience is that you need to be able to ship iterations very quickly. You need to have a big blue button that says 'ship' on it. If you can release often, then the last 10% actually looks like 10 little 1% jobs.

Completing a 1% job 10 times is much easier than completing 1 10% job.

Totally agree - ship what you have now, and then add the polish. Seriously if people care what your content is about they will live with lack of polish.

There is however an Internet sixth sense - that says "dodgy". A two page site with a email collection form mailchimp is waaaay more forgiveable than something that reeks of pump and dump

Absolutely. I believe these two groups of people are mutually exclusive: 1) Producers of 'Pump and Dump' sites and 2) People who struggle with the last x%.

Group 1 always produce low quality junk that triggers our 'dodgy' alarms. Group 2 always worry about producing something that is polished and finished.

The secret is to be in Group 2 but to have someone next to you shouting "Ship It!" as regularly as possible.

Some 37 signals outlet had a nice discussion on this. The key in their view was to limit the number of features rather than the quality of said features. Something about thinking of the features you want on a 2D map, and the thickness being the quality, building up your volume of work. To reduce volume you can break off pieces, but you should strive for even (and sufficient) thickness.

Too thin - pump & dump. To big volume - group 2. Small, but thick - good shit.

I guess this depends on the market you are entering though surely?

If you are entering an already pretty saturated market then you are probably going to struggle for customer attention without looking very slick.

Relevant quote by Tom Cargill, and one of my favourites:

«The first 90 percent of the code accounts for the first 90 percent of the development time. The remaining 10 percent of the code accounts for the other 90 percent of the development time»

My personal experience is that the last 10% are hard when:

  - I didn't stick to the specs
  - Added specs while coding
  - Didn't test while coding
  - Didn't fix shitty while coding

Because I think specs are very important I don't believe in too much of agile. I think the waterfall model is very useful but you have to keep the waterfalls very small. Maybe you can call this the agile waterfall model: waterfall -> release -> waterfall -> release

Agile basically is very small waterfalls. Of course you need to think up front about what you need to do before you do it. The fallacy of the waterfall model is the idea that you can accurately predict the full structure of the system before you write the first line of code.

One could picture it like Escher's Waterfall, you just keep going around: http://en.wikipedia.org/wiki/File:Escher_Waterfall.jpg

I feel much the same way about finishing and releasing. What I find interesting is that once I put something out there, opinions will vary wildly on the question whether it's way overdone and should have been released much earlier, or whether it's nowhere near ready to be released. I'm beginning to think that there is no such thing as the "right time" to release. What matters is that I have a clear vision of what I think is a great achievement that's not overdone. That vision gives me a finish line and provides me with a work rhythm. That "last 10%" that charlieirsh describes is part of that rhythm.

The last 10% has little to do with the number of remaining items, and all to do with the type of tasks and mental blocks. Project fatigue sets in.

Within the last 10% are small details pushed aside as the pace was chugging along for the other 90%. Seemingly trivial problems are hidden until the end, during which, developers/engineers are most eager to move on to the next mentally stimulating new project.

I think there is a reason why in schooling, 'A' level work requires a completion of that last 10%.

He's 90% dona and then he'll show it to the world? Ehm, what if the world doesn't like it? Aren't we pass this? We do pre-launch, research and try to spread the concept and acquire interest and real-world feedback before we go 'all-in' right?

The way I understood the original submission was that "100%" indicates the point where the thing is ready to be released to the world at large. Before that, one may well do iterations involving users and potential customers. That will of course make the actual release, or launch, less of a traumatic event. And yet, perhaps I'm old-fashioned, but to me, release is still an important milestone, sort of a finish line, or a goal I'm working towards. That "first 90% vs. last 10%" really resonates with me.

It's because he doesn't show or tell the reader what he's working on. He only talks about "my next website" and "this project". I figured if he had shown it to the world he wouldn't mind linking to it in this post at the very least... But of course I could be completely wrong. I just can't tell by this blog post.

as they say: if you finished 90% of the project, you are halfway done.

You beat me to it ... but I've found that it's really more dependent upon how accurate my idea of "done" is. The more time I put into making an accurate list of the tasks that have to be completed to call it done, the more accurate my assessment of the 90% point is.

What happens when you haven't thought your task list through is that you tend to add items that required to support the MVP's feature list. I actually find it pretty easy to avoid the features that I've decided have to wait for a later version.

Does it follow that if you are 99% of the way through, ie you've done 90% of the last 10%, you are 75% done?

Almost: the first 90% of the project takes about 90% of the total time, while the last 10% usually take the other 90% of the total time.

Maybe I was just being arrogant! I left a company feeling like I was 90% through a project after working on it for 3 months. The company released the finished project over 3 months later, and I wondered what took so long. I said "I knew that the developer they put on the project was really slow!" But maybe I was just being arrogant.

I am at 90% right now and it is painful! Need to finish and release http://get.infyouse.com

I just hope your domain name does not drag your project down. Del.icio.us is as piece of cake, compared (compared with del.icio.us because I had to chunk it in a similar way). If I were you, I would really look for another name, ASAP.

What does it mean - I cannot translate the "mashup" from German - but I get the idea you are a personal wishlist site.

When someone brags about something they intend to do, it comes off as lazy and more than a bit self-congratulatory (without actually earning it).

I guess some do it as a way of gaining attention and free advertising.

Reply on news.ycombinator.com