|
|
|
|
|
|
by joshuafcole
2782 days ago
|
|
|
I won't speak for the rest of the team, but walking away from Eve is unquestionably the hardest thing I have ever had to do. We were given a once in a lifetime chance to reinvent the way we build software, with the only constraints being our imagination and the clock imposed by our runway. I think by the end we were tantalizingly close to success too. Many of the tools we made were significantly more productive for certain domains of work. Unfortunately, close doesn't cut it in a startup, and despite many, many sleepless nights pushing we weren't able to tie a single cohesive product together that was both better enough and general enough. You might argue that we should have settled on one of our iterations that was close rather than continuing the search for something better. I can't really disagree with that except to say that each of the latest prototypes was such a large improvement over the last that it felt irresponsible to stop short. Unfortunately, we've had to find work that would keep a roof over our heads, but that doesn't mean we've buried the work we've done. We've open sourced Light Table and every iteration of Eve that was remotely useful, and we've worked hard to share the most important things we learned on the journey with those who might continue it through essays, speaking, and thorough commenting in the most interesting sections of our code. Nothing would make me happier than having someone take inspiration from our attempt and build the future of programming before I have the financial freedom to return to working on it. Like everyone at Kodowa, I just want to live in the world where those tools exist. I am sad that we weren't able to complete them in the short time that we had, but I am not sad that we tried. |
|
|
For all the effort invested to not be in vain, it is important for the community to be able to learn (generalizable) lessons from these failures. Hence the typical focus on putting out a product, even if not complete, and publishing a paper, to close out a line of research.
When I try to look for an understandable summary of lessons learned from LT/Eve, I get the feeling that the tidbits are a little scattered, and not quite in a form manifestly usable by the community. Have you solicited suggestions about how this could be improved? For starters, it would be great if there could be a centralized repository listing all the various blog posts, talks, etc during the evolution of LT/Eve and the retrospectives. If that already exists, it would be great to publicize it more, and the community could organize around that and discuss how to take it further. There's a lot of excitement about the direction LT/Eve pioneered and it would be good to harness it for the next round of attempts :-)
Also, I don't mean to imply that the LT/Eve team has not done that -- eg: thanks for open sourcing LT! I'm just thinking out loud about how to better pass on the less tangible learnings.