I can’t believe an open source project thinks in term of retention, market share, and attracting the double click.

Wasn’t Firefox about providing the best user experience?, as Kari pointed out here: http://techiteasy.org/2007/08/27/webkit-or-of-frameworks-and-browsers/

There’s definitely something wrong about Firefox.

From this post we cant really tell that you have such a “meager” experience. This is full of insight and unshakable views.

I didn’t really understood this was not addressing SaaS. I’m so used to thinking of new apps as web ones. There are enough hints in here to find it out so I guess I was a bit quick to reply. My mistake.

I still believe that developping client software is going the hard way though (distribution, deployment, maintenance and user support etc …)

As far as RoR is concerned, well this is a very young framework. And as such he doesn’t have a lot of support and tools around. But thanks to the open source community slowly but surely tools are coming (RadRails, CruiseControlRb, etc …) for development side.

It may still be an issue for the operating side of thing but I’m pretty sure it wont take long.

As far as J2EE is concerned (I dont know or .Net so I wont talk about it) I think the trend is rather to have both working together. As a hint : sun bought JRuby.

My guess we will have more and more front end solution using RoR and connecting to Java Back End using REST or XML-RPC. You then have best of both worlds.

We did this at IN FUSIO and it worked pretty well.

First, I loved your article on AlphaGeeks. Although I believe we still have lots to do to show masses that doing software development as a job is really cool. Especially in Europe.

Second, I could’ve stated that these different ‘tips’ reflected my meager experience and that I don’t believe I’m right. This is how I feel it: it’s an opinion, not an exhaustive one, and there can be different approaches to this problematic.

Third, most of my points make a lot more sense when talking about packaged, client software. It doesn’t all apply to SaaS. I should’ve made that point clearer in my post. My approach to online services and software (OSS ) would be very similar to yours.

About 37 Signals, your main reference: I’m amazed by they approach, I find their apps thrilling, I loved their book, and I believe they deserve their success. Great Internet Software company.

I’m fond of agile methodologies by the way. Let’s talk about that later.

And I love the Ruby on Rails framework. I think you save a lot of time (less lines of code) coding, and it’s amazing how interactions with the database included in the programming language boosts productivity. However, Ruby on Rails is great for small web apps but no match for an industrial software approach. It lacks the tools, the support, the documentation. RoR makes the packaging process complex. In short, RoR is great for long-tail soft of web apps but it is not and will never become J2EE or .NET.

About the competition: I think that you don’t start a company to loose. You’re there to win. And nice, gentle, kind, not competitive people don’t build Google, Oracle or Microsoft. They just stagnate. In other words, watching and reacting to your competitors’ moves doesn’t antagonize focusing on your product: you can still do both. But if you’re to make an impact on the high tech industry, you shouldn’t play pussy. Ruthlessness should be a state of mind. Again, maybe I’m wrong, that’s just my opinion.

Many thanks for your interesting remarks Cecil. A pleasure to answer to you.

PS: btw, Firefox does operate a money-oriented lock-in. Think about Firefox as the browser of Google and you’ll get my point…

Woops I missed that one : Think interface. Agree with this one. It actually is definitely my number one.

Just that I think you miss something regarding coders. I made a post on this (http://ceciiil.wordpress.com/2007/06/05/respect-to-the-alphageeks/) I tend to agree with Tim Bray when he says that the key decisions are now made by the programmers. People like David Heinemier Hansson (creator of ruby on rails, working for guess who) are now considered as priceless assets of the global economy thanks to their ability to envision the way to build our digital world. They are not mere work force anymore. I’m sure that’s not what you meant but that’s what may be understood.

5) I would rather talk about methodology. Agile methodology which helps keeping a great control on what’s going one while having everybody feeling comfortable with the project. I shall do a post on Scrum methodology shortly

6) This is a definite no-go. Thinking about things like scalability and flexibility at the early stage of the project will only slow you down and complicate the software. I give you an example : when Twitter started the whole thing (by the way they have a RubyOnRails server) they didn’t have a clue about scalability. Their position : if we have problem with scalability then it means we have success, therefore money : we shall be able to address the issue when it arises. Which they actually did a few months ago.

7) Think lock-in : definitely agree on this one. Note that for firefox this is not a contract or money oriented lock-in at it would be with Microsoft product

cool

9)I dont agree with that one. The only thing you have to keep in mind is your user and how comfortable and confident she will be when using your product. Ignore competition just focus on your customer and your app. Check out kathy sierra (headrush.typepad.com) or hugh mcleod (www.gapingvoid.com) on this one.

10) cool

11) Not at the beginning. Think of your user.

Jeremy this is a tremendous work you’ve done with that. I disagree with mlost of the points but that’s only my position.

May I suggest you to look at the getting real manifesto from 37Signals - I dont know if I’ve mentionned them already Sorry to insist with these guys but they really set the new standards for the start up 2.0.

Many points where I diverge here. I will base my comments mostly on the success of 37 Signals which really should be an example in terms of success story.

Small business (7 or 8 people spread in different locations) doing web based enterprise (CRM, PM) tools. Business Week best of the web awards 2005/2006 and top 10 vendors winning the SMB online experience.

For those who dont know 37signals, check out their page : http://www.37signals.com.

1) Think Strategy : If you ask them they dont know where they will be in 10 or even 5 years time. The only thing they are sure about : “we’ll be in the business”.

2) Think intelletual property : 37Signals invented Ruby On Rails back in 2004, the new web development framework all enterprise Gurus (Martin Fowler, Bruce Tate) are migrating to. This is obviously open source technology.

3) Think world class engineer. Question Jeremy : how do you rate they are world class ? With their CV ? their reference ? What I ve read the most from successul hiring in start-up : Hire enthusiastic and passionate people. Because this cant be faked. And if you dont have a position for them, create it. As far as techies are concerned, their contribution to open source project is a great hint for their passion.