Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Zimki, the world's first Platform as a Service that belonged to Canon (porter.run)
103 points by sungrokshim on Aug 19, 2021 | hide | past | favorite | 32 comments


Simon Wardley has continued to evolve and refine the insights that led him to conceiving Zimki, and it is well worth any tech entrepreneur's time to take a look at the current iteration of "Wardley Mapping" to better understand how technologies (and platforms) evolve:

https://learnwardleymapping.com/book/

If Blank and Ries transformed our understanding of startup tactics, Wardley is improving our understanding of strategy.

Simon's account of the Zimki episode is in Chapter 5, if you just want a bit more detail: https://medium.com/wardleymaps/the-play-and-a-decision-to-ac...


I have to admit, while I like Wardley mapping, the fact that someone talks about his own (seemingly) failed technology revolution in hindsight, makes me a bit sceptic.


I find just as many lessons in failures as success, although these are pretty arbitrary labels. Even the greatest success stories are littered with small failures overcome or even ignored.


I'm not saying we can't learn from Zimki.

I'm just a bit weary of people claiming they "invented the future" way back before the future actually came.


Simon doesn't claim he invented PaaS, but he did predict it.

As far as learning from Zimki, I think the lesson Simon says he learned was basically (paraphrasing) "make sure you don't run out of political capital" and watching out for misaligned incentives.


Zimki was SO far ahead of its time - it was at least five years too early. Server-side JavaScript, Heroku-style PaaS, with its own object storage - back in 2006.

I was at the 2007 OSCON where Simon Wardley resigned on-stage after Canon changed their mind on open sourcing it. That was quite a moment!


Must have been incredible to see that unfold in real time... I can only imagine!


I was the conference chair (selected keynotes, MCed). When his talk finished I shook his hand, looked at him and asked something like "did you just resign from the stage?" because I couldn't believe what I'd just heard. And he said "yes", and it looked like it was only just sinking in.

Amazingly powerful moment because everyone there empathised with the experience of being on the right technical track to do something amazing but management not getting it. His frustration was under control, but palpable. And ("anticlimactic" reply in this thread notwithstanding) it was clear we were witnessing something we weren't going to see again.


It was a bit anticlimactic. Think about it, there was supposed to be a major new (and Open Source) platform announcement, and instead we got a resignation with very few details.


Same here.


I interviewed with them (I knew a few people there) and had an interesting half-hour discussion with Mr Wardley about the implications of wide-spread availability of 3D printers (he had a 3D printed model of an XBOX controller) - little did we know it would be nearly another decade before they were becoming widely available.

[edit: also one of the best technical interviews I had. They sat me down with a Mac and a list of stuff to do and left me alone for, I think, an hour to work through them as best I could. No restrictions on looking stuff up, no gotcha questions, no algorithm implementations, etc.]


Heroku and Zimki are fundamentally different. We quickly evolved into a Polyglot product, with wonderful primitives and invented a nice interface for extension, namely buildpacks and our add-OBS market.

There were lots of examples of companies doing things Zimki did, most WebObjects projects felt that way IMO.



> " One of the very first PaaS offerings in history actually belonged to Canon ... until it was abruptly killed by Canon itself"

There are dozens of examples of people being first to market and not winning (Canon). As so, what's so interesting about this article?

- Facebook won, when MySpace and others were first to market

- Google won, and it was crazy late to market over Yahoo/Altavista/Excite.

- Starbucks won, when plenty of other national coffee chains existed

etc


In general, the examples you cite are of early entrants being outcompeted by later entrants. Zimki was shut down despite (or because of?) it's success and experiencing he initial stages of hypergrowth. The successful PaaS entrants didn't show up until after Zimki had been defunct for a while.

You could also characterize this as "innovative startup strangled by acquiring behemoth or clueless greedy investors" which brings to mind a different set of examples, but that's not much more interesting.

What makes Zimki interesting IMO is that Simon Wardley successfully "timed the market" of technological change, clearly explained his thinking, and has since applied the same logic to make other correct predictions (or give good strategic advice).


> What makes Zimki interesting IMO is that Simon Wardley successfully "timed the market" of technological change, clearly explained his thinking, and has since applied the same logic to make other correct predictions (or give good strategic advice).

Has Wardley mentioned why they didn't pursue the start-up route (like Eric Yuan at zoom.us did; or Benoit Dageville, Thierry Cruanes at SnowflakeDB did, to give recent examples)?


Not so far as I know.

I always assumed that restrictive non-competes prevented moving quickly, and that Google's launch of AppEngine then sucked the oxygen from the room.


> Facebook won, when MySpace and others were first to market

Facebook had a valuable userbase and a polished product (it looked good, no templates and music on auto-play)

> Google won, and it was crazy late to market over Yahoo/Altavista/Excite.

Google had a better product, plain simple.

> Starbucks won, when plenty of other national coffee chains existed

Starbucks won partly because it had "Italian coffee". Nobody else had espresso machines in as many locations as they did.


> Starbucks won partly because it had "Italian coffee". Nobody else had espresso machines in as many locations as they did.

Yes exactly. Here in Melbourne (where we have a culture of European style espresso everywhere), Starbucks has barely been able to get a foothold in the market.


Starbucks won because they recreated a friendly living room.

I always felt their coffee sucked, but their employees have always been nice, and friendly.

I enjoyed getting out, people watching, and feeling zero pressure to buy anything, or leave.

(I've noticed slight changes, like codes to the bathroom, but it's still my favorite place to hang. American perspective.)


> I always felt their coffee sucked

Most independents will be better, if there is any around.


What's an example of being first to market and winning a big market?

I can't think of any. First to market isn't the advantage that people think it is. It might be more important for smaller markets, perhaps.


The first commercial power plant in the US became Consolidated Edison which is still going 140 years later with a market cap of $27 billion.


Bell Telephone Company was first to the market and won. It's now known as AT&T.


> It's now known as AT&T.

... and Verizon, and a few others.


Intel for personal computer? It seems that Intel 8008/8080 based (hobby) PCs were very early one. Intel is the king for PC market until recently (so about 50 years?), but they are now losing market by ARM and AMD.

Though maybe Intel didn't intended 8008/8080 is used for personal computer.


Netflix, Uber, Amazon, PayPal, etc?

It's hard to exactly define "first to market." Was Google first to market because they were the first to have a search engine that functioned in the way theirs did, and should we put more primitive search engines in the same category?

Was Dropbox first to market? Did they win the market?

It often feels like a company is first to market, and then they win that market... and then maybe five or ten years later, an innovation comes along from a new company and they lose the market. That's just the way the market is supposed to work, and natural, but there's still an advantage to being first to market in having those years of market dominance.


> Netflix, Uber, Amazon, PayPal, etc?

Not sure if Netflix was first to DVD by mail, quite possibly; they were early, and I don't know if VHS rental by mail was ever a thing.

Uber may have been the first to internet hail livery vehicles, but Sidecar was the first to do internet ridesharing as an app, and Lyft was the first to call unlicensed taxi service ridesharing. Uber is winning in the marketplace at the moment though and Sidecar is dead; of course, winning in this market still means burning money.

Amazon was early to selling books on the web, but e-commerce catalog sales were available on pre-Internet information services. AWS defined a new category of managed hosting, but lots of other companies did similar things before.

I think BillPoint may have slightly predated PayPal. Being early here seems to have had staying power, as Yahoo's PayDirect didn't do well enough to stick around. Otoh, Venmo was good enough to buy, Zelle seems popular, and other easy credit card processors like Square and such seem to be winning over that side of PayPal's market. There's a moat, but it's not very deep.

> Was Dropbox first to market? Did they win the market?

No, there was XDrive and Yahoo Briefcase and probably more that launched and shutdown before Dropbox. Not sure if Dropbox won, OneDrive and Google Drive are contenders still.


It depends on how you define the market. I think it's pretty hard to say Google was first to any market with any reasonable definition. They're an example of how you don't have to be first if you are better. Even if you define the early Google market as a "natural language web crawler search engines", they weren't the first.


? A classical "the winner takes it all" business.

Example: ebay

Counterexample: Flight booking


Wow - Never knew. That's amazing, and sad.


Wow - interesting story.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: