This article states that YC is now valued at $13.7 billion. Actually, the cited source credits pg recently saying that is the current valuation of all of the YC startups, not of YC itself.
That would value YC at around $400m, but isn't the mean dilution insufficient, since YC would be diluted the most for the highest valued startups?
A very interesting question: once it's all said and done (let's say in 30 years), do you think YC will have a greater valuation than any single YC startup?
Unlikely given the power law distribution of startup outcomes. But if you set a threshold defining a big hit and we funded 33 of them we'd end up being one ourselves. It's conceivable we could fund 33 big hits in 30 years. Depends how much we grow.
> "Then, with enough money saved, [pg] would quit the job and devote his time to his real love—art and painting—until the money ran out, and then he would scramble for another job."
PG, is that what you do between YC batches? When will we get to see some of your paintings? :)
No, what I do between batches is work on YC. In the very beginning we considered time between batches as time off, but YC is so big now that there's always something that needs to be done.
It is unfortunate that you can't decompress for a month or two in Florence, or a similarly beautiful place. If you ever plan to visit Italy again, please let me know; there's a bunch of very interesting people to meet in Florence, including the mayor, Matteo Renzi (candidate for Prime Minister in 2014) and Paolo Barberis, founder of Dada.it, now founder of the accelerator nana bianca. And others.
It was first according to PG's knowledge, which is incomplete. Amazon.com was incorporated in 1994, and sold its first book online in July 1995 [1], a month before Viaweb. EBay launched in Sept 1995 [2], a month after Viaweb, and so was probably independently under development. Geocities was active and allowing user-contributed content by July 1995 [3], a month before. WebCrawler was the first general-purpose Internet search engine, and launched in April 1994 [4].
Viaweb was the internet's first web application. Its customers used Viaweb to create and manage their own e-commerce sites. Like Shopify (or rather, Shopify is like Viaweb).
Viaweb was not an e-commerce site like Amazon or Ebay. Its customers were merchants rather than consumers.
I'd define "web application" as "something that lets you perform a task via the web". All of the companies I linked count under that definition; Amazon let you buy books; EBay let you buy & sell odds & ends; GeoCities let you build a website; WebCrawler let you search for things.
If you define it more narrowly as "something that lets you create a site", Geocities still counts.
I think you'd have to narrow the category to "first online store builder" to make it accurate, which is still quite an accomplishment, but certainly isn't the "first web app".
Over the years, evolving in its own way, Y Combinator has continued to grow at an astounding rate. It is valued now at $13.7 billion, with the clear potential for further growth.
Ignoring the mistake of valuing YC at $13.7B instead of its companies, here's a crazy idea: I wonder when we'll see YC go for an IPO?
For me, it wouldn't make sense taking YC itself public. I can't see much of an upside to the partners for taking the company public. Please, correct me if you see otherwise.
What would be fantastic though, is for YC to incorporate their holding of each individual year (older than, say, 3 years) into a fund, and then take those public (and continue to do so). An annual YC ETF.
I am disappointed by this article. It made such a big deal about PG hating to work with Windows, and reduced the innovation of web applications to a way to not work in Windows.
Really?
Clearly the whole point was to get away from developing desktop applications, and realizing the single-source, rapid development nature of web apps.
Please don't tell me that developing client apps on Windows was so much worse then doing it on Linux. Its bullshit. Overfocusing on this aspect of the story feels like platform fanboyism, misses the real point, and just overall kind of cheapens PG's stellar image.
> Please don't tell me that developing client apps on Windows was so much worse then doing it on Linux.
- Many Windows-only concepts (registry) hold key configuration data.
- Lack of a proper command line environment (somewhat covered by Cygwin, but not ideal). PowerShell is interesting, but, again, is very alien to non-Windows tools.
- No select-middle-click copy/paste (it's an incredible time-saver).
- Limited number of available languages and tools and the difficulty of integrating them (you have to add every install directory to PATH - unless you are running Cygwin).
- Case-insensitive filesystem (and some subtle bugs that it allows).
- It's almost impossible to open a tempfile and delete it while you still have the file handle.
- No package management (you can't just "yum install" what you need)
- No package management (you have to bundle every dependency with your installer)
It's not bullshit. Unless you are developing for Windows only, developing under Windows is quite painful.
In 1995 it would have made no sense to develop a commercial desktop app for anything other than Windows.
If you didn't make Windows apps at the time, it probably seemed really complicated. It really wasn't. Most of the issues were with C/C++, and the complexity that brings. VB and Delphi were decent alternatives -- there were some 4G languages that worked ok as well.
Making viaweb a web-app instead is obvious in retrospect. But not because making Windows apps is so hard -- it's the deployment that the web brings, not the ease of development. In 1995, there was no super-easy way to make web-apps -- PG and RTM had to invent that too.
Well, actually I used to develop UI intensive MFC applications for quite some years, and absolutely loved it! Not that different from developing for Android, and as I remember it the Visual Studio was (may be still is? haven't used it for years) a far better development environment back then than the Android Studio is today. But may be I just remember being 10 years younger :) Disclaimer - I'm still a Windows user.
I'm not sure exactly how to say it, but the article didn't really take us through the thought process of pg as he "made yc possible".
Two things I'm interested in hearing more about:
1) What was the thinking behind viaweb? yc? How did pg arrive at the decision? Was it as simple as the insights that were mentioned? Were alternatives considered? (Before I started http://www.collegeanswerz.com/, I had a list of hundreds of ideas that I went through before deciding on a website with better college reviews.) If so, what made viaweb/yc better than the alternatives? What aspects of viaweb/yc were forseen, and which just happened because of luck?
2) What was pg thinking he'd do with his life before viaweb? yc? Before viaweb, it says that he was basically flip flopping between consulting and art. What did he think he'd do with his life at this point? Any plans? What was he feeling? Excited? Anxious? Bored? What about the 7 years between viaweb and yc? Thoughts and plans about what he'd do with his life?
I know this stuff isn't directly related to viaweb and yc, but it's useful to know thought processes, rather than just touching on the main insights. You can't really learn from insights. You can learn from thought processes.
44 comments
[ 3.7 ms ] story [ 110 ms ] thread[edited for clarity]
A very interesting question: once it's all said and done (let's say in 30 years), do you think YC will have a greater valuation than any single YC startup?
(3% is our estimate for the biggest successes.)
PG, is that what you do between YC batches? When will we get to see some of your paintings? :)
http://www.paulgraham.com/first.html
http://ycombinator.com/viaweb
[1] http://en.wikipedia.org/wiki/Amazon.com
[2] http://en.wikipedia.org/wiki/EBay#History
[3] http://en.wikipedia.org/wiki/GeoCities#History
[4] http://en.wikipedia.org/wiki/WebCrawler
Viaweb was not an e-commerce site like Amazon or Ebay. Its customers were merchants rather than consumers.
If you define it more narrowly as "something that lets you create a site", Geocities still counts.
I think you'd have to narrow the category to "first online store builder" to make it accurate, which is still quite an accomplishment, but certainly isn't the "first web app".
Ignoring the mistake of valuing YC at $13.7B instead of its companies, here's a crazy idea: I wonder when we'll see YC go for an IPO?
What would be fantastic though, is for YC to incorporate their holding of each individual year (older than, say, 3 years) into a fund, and then take those public (and continue to do so). An annual YC ETF.
PG, has an annual YC ETF crossed your mind?
Really?
Clearly the whole point was to get away from developing desktop applications, and realizing the single-source, rapid development nature of web apps.
Please don't tell me that developing client apps on Windows was so much worse then doing it on Linux. Its bullshit. Overfocusing on this aspect of the story feels like platform fanboyism, misses the real point, and just overall kind of cheapens PG's stellar image.
Uh, what Linux? No, pg worked on Lisp Machines before, and that's why Windows were hurting him so much.
- Many Windows-only concepts (registry) hold key configuration data.
- Lack of a proper command line environment (somewhat covered by Cygwin, but not ideal). PowerShell is interesting, but, again, is very alien to non-Windows tools.
- No select-middle-click copy/paste (it's an incredible time-saver).
- Limited number of available languages and tools and the difficulty of integrating them (you have to add every install directory to PATH - unless you are running Cygwin).
- Case-insensitive filesystem (and some subtle bugs that it allows).
- It's almost impossible to open a tempfile and delete it while you still have the file handle.
- No package management (you can't just "yum install" what you need)
- No package management (you have to bundle every dependency with your installer)
It's not bullshit. Unless you are developing for Windows only, developing under Windows is quite painful.
If you didn't make Windows apps at the time, it probably seemed really complicated. It really wasn't. Most of the issues were with C/C++, and the complexity that brings. VB and Delphi were decent alternatives -- there were some 4G languages that worked ok as well.
Making viaweb a web-app instead is obvious in retrospect. But not because making Windows apps is so hard -- it's the deployment that the web brings, not the ease of development. In 1995, there was no super-easy way to make web-apps -- PG and RTM had to invent that too.
Is it true that "distaste for Windows" was a motivating factor behind Viaweb?
For me, Windows was an enormous itch. I scratched it with UNIX. Thank you, Redmond, for the motivation.
Two things I'm interested in hearing more about:
1) What was the thinking behind viaweb? yc? How did pg arrive at the decision? Was it as simple as the insights that were mentioned? Were alternatives considered? (Before I started http://www.collegeanswerz.com/, I had a list of hundreds of ideas that I went through before deciding on a website with better college reviews.) If so, what made viaweb/yc better than the alternatives? What aspects of viaweb/yc were forseen, and which just happened because of luck?
2) What was pg thinking he'd do with his life before viaweb? yc? Before viaweb, it says that he was basically flip flopping between consulting and art. What did he think he'd do with his life at this point? Any plans? What was he feeling? Excited? Anxious? Bored? What about the 7 years between viaweb and yc? Thoughts and plans about what he'd do with his life?
I know this stuff isn't directly related to viaweb and yc, but it's useful to know thought processes, rather than just touching on the main insights. You can't really learn from insights. You can learn from thought processes.