AV-Portal 3.23.2 (82e6d442014116effb30fa56eb6dcabdede8ee7f)

Open Source Meets the Commercial World

Video in TIB AV-Portal: Open Source Meets the Commercial World

Formal Metadata

Open Source Meets the Commercial World
The opening session
Title of Series
CC Attribution 3.0 Unported:
You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor.
Release Date

Content Metadata

Subject Area
Bambi Meets Godzilla: They Elope The world of open source and the world of commercial software intersect in profound and sometimes surprising ways. In some sense the two are like oil and water, but in other ways they can build on each other. This talk describes some of those differences — and ways they can enhance one another — using the sendmail open source mail transfer agent and Sendmail, Inc. as a case study.
Word Slide rule Computer program Game controller Computer animation Open source Personal digital assistant Natural number Multiplication sign Right angle Open set
Slide rule Trail Presentation of a group Open source Software developer Mathematical singularity Tape drive Device driver Punched tape Internetworking Term (mathematics) Hybrid computer Energy level Utility software Macro (computer science) Physical system Metropolitan area network Email Trail Relational database Tape drive Physical law Electronic mailing list Plastikkarte Bit Device driver System call Open set Macro (computer science) Process (computing) Kernel (computing) Numeral (linguistics) Computer animation Software Integrated development environment Crash (computing) Hybrid computer Quicksort Game theory
Point (geometry) Computer program Group action Open source Code Scaling (geometry) Multiplication sign Decision theory Student's t-test Mereology Code Number Data model Internetworking Endliche Modelltheorie Metropolitan area network Email Inheritance (object-oriented programming) Point (geometry) Projective plane Mereology Open set Single-precision floating-point format Proof theory Process (computing) Internetworking Computer animation Universe (mathematics)
Overhead (computing) Open source Multiplication sign Real number 1 (number) Ellipse Total S.A. Medical imaging Forest Analytic continuation Form (programming) Area Socket-Schnittstelle NP-hard Projective plane Planning Line (geometry) Word Computer animation Personal digital assistant Order (biology) Endliche Modelltheorie Right angle Freeware Pressure
INTEGRAL Confidence interval Multiplication sign Set (mathematics) Open set Icosahedron Mereology Strategy game Different (Kate Ryan album) Enterprise architecture Email Software developer Gradient Fitness function Streaming media Bit Staff (military) Lattice (order) Open set Product (business) Category of being Digital rights management Process (computing) Sample (statistics) Chain Website output Right angle Whiteboard Physical system Writing Slide rule Server (computing) Freeware Overhead (computing) Open source Software developer Virtual machine Translation (relic) Code Product (business) Number Revision control Goodness of fit Software Normal (geometry) output Overhead (computing) Cellular automaton Projective plane Code Computer network Computer animation Software Integrated development environment Logic Personal digital assistant Video game Pressure
Functional (mathematics) Freeware Service (economics) Open source Distribution (mathematics) Code Multiplication sign Knot Disk read-and-write head Total S.A. Number Product (business) Revision control Fraction (mathematics) Data model Software Computer hardware Cuboid Endliche Modelltheorie Office suite Extension (kinesiology) Position operator Stability theory Service (economics) Graph (mathematics) Scaling (geometry) Software developer Electronic mailing list Port scanner Limit (category theory) Open set Category of being Digital rights management Computer animation Software Personal digital assistant Endliche Modelltheorie Fiber bundle Freeware Fundamental theorem of algebra Extension (kinesiology)
Classical physics Open source Code View (database) Source code 1 (number) Water vapor Parameter (computer programming) Average Shift operator Rule of inference Power (physics) Number Element (mathematics) Mathematics Different (Kate Ryan album) Term (mathematics) Googol Computer network Endliche Modelltheorie Data structure Exception handling Social class Metropolitan area network Enterprise architecture Shift operator NP-hard Closed set Projective plane Electronic mailing list Drop (liquid) Term (mathematics) Port scanner Open set Product (business) Computer animation Software Quicksort Whiteboard
Point (geometry) Metropolitan area network Pairwise comparison Computer animation Open source Different (Kate Ryan album) View (database) Projective plane Maxima and minima Similarity (geometry) Open set
Computer program Open source Code Line (geometry) Gradient Multiplication sign Direction (geometry) 1 (number) Student's t-test Mereology Formal language Product (business) Endliche Modelltheorie Quantum computer Metropolitan area network Service (economics) Real number Gradient Projective plane Feedback Code Planning Bit Student's t-test Multilateration Line (geometry) Open set Product (business) Computer animation Personal digital assistant Quicksort Figurate number
Point (geometry) Open source Multiplication sign Streaming media Student's t-test Event horizon Product (business) Sound effect Roundness (object) Computer configuration Liquid Exception handling Self-organization Projective plane Sound effect 3 (number) Database transaction Student's t-test Line (geometry) Hypothesis Process (computing) Event horizon Computer animation Video game Self-organization Right angle Cycle (graph theory) Quicksort Physical system
Area Multiplication sign Chemical equation Graph (mathematics) Data storage device Knot Field (computer science) Spreadsheet Goodness of fit Computer animation Personal digital assistant Office suite Quicksort
Point (geometry) Metropolitan area network Scaling (geometry) Open source Survival analysis Survival analysis Term (mathematics) Protein Open set Mathematics Process (computing) Computer animation Radio-frequency identification Term (mathematics) Different (Kate Ryan album) Videoconferencing output Quicksort output Videoconferencing
but in the case of OK right to year very clearly and yet for some reason this started the nature get published in the program which is element of that that is my fault knowledge not pointing any fingers but this is a topic that I wrote originally for euro BST commonly in Warsaw and Robert Watson who is actually supposed to give that know couldn't because of his previous control of suggested this topic to myself this kind of interest interplay between open source and commercial interests and that's something I happen to to know something about and so I'm going to be talking about open source companies here I do not mean by that people who take in and source is pretty much everyone does employment people who actually you contribute back to the community and just a curiosity how many people understand the then the Godzilla of reference was much better than in Poland who goes on to I well I don't believe and legally allowed to play it but I do actually have it in my slide decks so if we have time I'll maybe toss that in an I had to throw this 1
and it was appropriate in Poland because that is in fact where I was I was down in South Africa and got a very small amount of Internet and had these e-mails saying can you do this and so I had to leave this slide in but because it's the most interesting slide the entire presentation so a little
about my background I started writing open-source software background 1975 with the call of that and of just software and I worked on numerous relational database system starting when I was an undergraduate this was back when we were with in relational databases didn't exist yet so it was it was new and shiny back and the process of matter of syslog and tear macros which some people believe around still use game called track which leaving out some people still play a bunch of other BSD utilities and best known for Sendmail you probably all know that already but I am not a kernel guy if you want that talk to Kirk I've always stayed kind of user level although I did write a few have a device drivers all blessedly defunct now and who uses punched tape cards and nine-track tape anymore asking them but I think probably the most important thing I did was I convened Berkeley notably Bill Joy to start using CCS and hence history begins on the day I got him to do that the few and other jobs in academia 1 of commercial environments research sort of all over the place of Don all kinds of crazy things and I started Simulink in 1998 it was 1 of the very 1st attempts to merge open source hybrid there were some companies were trying to be pure open-source companies but none of them actually did very well because her turn that it's really hard to make money when you're giving everything away and we did survive detect rational may is better than about 90 % of the other startups out there on however I will say that companies are should be added to the list of things you don't want to see made that includes sausages laws and companies it's an ugly Proc and I'll tell you a little bit about that I am now officially retired from sendmail as of September I am however still on good terms with them and hopefully after this talk still will be also central is 1 of the
1st similarly the program that the company started as 1 of the very early open source projects which is action part of BST but it was a classic example of scratching your own picture problem at Berkeley and so it was easier to write code them to do it by hand and so I wrote code and then it turns out of course other people had the same problem and it went on from there it went through various growth spurts there was a time when I abandoned this for a while because I was bored with it I get bored very easily that turns out and then I came back to it and it was originally built to solve a single local problems but that generalize due to community in need of got caught up in the Internet explosion which was an interesting process by itself and pretty much remained community supported in 1 way or another and through most of its history I did use the benevolent dictator with trusted henchman model which is pretty much where we this is done the Internet Dilog's flowed
and I turned into a success disaster of the communities feeling pretty much collapsed was spending all the time I have available were members of a full time job at this point which is not send and I was constantly getting these questions all which of course well 99 % of which were answered in the documentation but it's easier for some people to send e-mail than it is to actually read which appears to be a unique concept and some projects they're used the RTF and read these from fine manual and which can also be phrased as you're on your own and it does however require fairly sophisticated users who are capable of reading which as I said is sometimes not of good decision and a final manual to read which turns out to be a major problem with a lot of open source projects you can't tell them to go read the manual when there's no manual parade I assert here without proof that all successful large open source projects get some side out kind of outside support that sometimes this can be done necessarily mean money but I mean in our employers they're willing to let you work on this stuff you know part time you know if you're a student of maybe get to get some funds from university or or your parents are paying for you to go and do whatever you want but whatever it is but I want to get back to doing coding of started doing this support stuff which as I said was take all my time so I had this fantasy I would hire a support person and they would answer all these stupid questions and I could go back to doing what I wanted to do which is write code OK to do that I had to ditch the day job and I had to get some funding so I ended up starting a company it now there's a number of ways
you can get money coming in to a project and 1 of the popular ones to start a foundation and get donations and there's a lot of these Mozilla clips Apache Free BSD all have foundation it turns out it's very hard to set up a foundation requires very special skills especially for the 501 c 3 and not for profit of which has a lot of legal messy you have to get lawyers involved and so forth so in order to become a not for profit you have spent a lot of money and you can find a patron will shower you with money which works great if you're like pocket Mozart or 1 of the final forests and then you have to live in Vienna and right century but these days the whole it is having a very rich uncle who indulges you and I don't you can sell yourself to a company that has deep pockets and this can be very effective in some cases however keep in mind that they don't necessarily have your best interests in mind I can almost guarantee you that their image of what they wanted is not fully aligned with what you want to run in have to make some compromises and your leverage is extremely limited if you your brain is the only asset and and there are other assets if you have a large enough community user base whatever you might actually be able to leverage that but if you don't have something large enough to be of interest so basically reassign you the 1st time things come along I and you can start your own company of Red Hat started before sendmail what was initially trying to pure open-source place and there was 1 of the very 1st that explicitly in the business plan said we want to combine the 2 and somehow get them to talk to each other I actually tried something along the lines of foundations really more of a consortia where I went to a bunch of vendors who were all shipping sendmail and said if you put in a small amount of money each year you don't have to have people working full time they all had at least 1 person working full time and send and I'll do that for you and to 1 they said that sounds like a great idea as long as we know when it's done and that wasn't acceptable I think so just a quick word
about foundations because they are so popular are they do insulate you from the day-to-day pressures of corporations and they're not necessarily while they're not looking quarter quarter which is 1 of the standard things that corporations do these days but they do not prevent me from being pressured in other words if 1 of your major funders comes along and says I cut your funding next year unless you do this or don't do this or whatever other then you're stuck and they do take a lot of work to start and keep running they typically all have a half time executive director so there's an overhead of bureaucracy of all this especially if it's a not for profit is all the tax forms that have to be done and so forth and you may lose some of the good things that you do get from a real corporation and marketing and is not fundamental to to area foundations and I should say 1 of the things that foundations require is but continuous grant proposals which are harder right and the ongoing schmoozing with your funders because they all think that they're you even if they just give a hundred dollars think they're really special so I'm going to
assert that open source actually needs commercial input and my next slide is going to say why this is a radical case you don't know already but in the stock developers in fact are seldom the customers in the old days you know what I wrote sent I was my own custom right and have to go and have somebody explain to me what was customer needed because I was the 1 who needed it and other things like fine for a long time and very successfully because for the 1st time people were writing the tools that they actually wanted to get as close to having some corporation tradition of whatever they had down their throats however when you start to get into non-developer software it all just use consumer software in general as a stand in for that the designs tend to run from a really pretty mediocre to sometimes just unbelievably bad news site people up so late at nite trying to figure out how they can screw this up and that's us folks we don't think like a normal human beings that doesn't mean by the way world loners some are some people are quite social but they just think well this is a logical set of conclusions and so it must be obvious to everyone and it's not people just don't work that way and in companies you have these things called product managers Integrated Product Manager is somebody who has a certain amount of technical background is also very good talking to the customers and doing the translation between the 2 so you know good product manager most of the job is just being a translated however it is also true that not all product managers are great product managers in fact fairly few of them unfortunately I so there are examples of other than 50 get from a corporations soft items that from corporations actually pay people write documentation and some of the some open source projects have been pretty good about that but it's not 1 of open source is traditional strength shall we say and front-line support to unburden the developers so in sleep them a little bit in you know overhead you know at home if I want to upgrade my network I have to do it myself work somebody does it for me but there is this the
tension between open source and commercial ventures our open source is might loosely be described as being about building things you know creating things sharing things with other people flexibility to do what it is you want what think you need to making the world a better place in some sense a although we can debate some of those solving interesting problem is often a big part of it sometimes personal development hey I need to learn about this all right the codon and experience is that way all these kind of things which ironically are almost a little bit touchy-feely commercial software development is about making money make no mistake about that that's what it's about is not about great technology is not about making something beautiful it is not about making the world a better place money is the life blood of companies and that's 1 of those things were when I started the company I have to admit that I was a little hippy dippy and was kind of thinking no we're going to have a different kind of company is going to you birds will be singing and so forth and it's not that the sales guys don't understand how to make money by giving the part that I was multiple times before I had his fit in a staff meeting accused of being a communist by the cells people and they were serious I mean like your fuckin on like look I started the company I don't think I mean stock company he and there is an immense pressure toward feature creep to keep people coming back and buy more it's not good enough for them to just download your software installed they have to keep coming back year after year and look at something you know a monstrosity like Quicken every year there's always pressure the by the new version of Quicken this strategy now is it just shuts down after 3 years will work at all and so you're forced to buy a new version which has new features that I don't want typically known as feature creep which can be defined as doing more and more or less and less well 1 the other thing that is very important in companies is if you miss payroll your you know I in my personal life and willing to sort of take out a loan against my next month's income or something like that and that's OK not true in companies when they hit the wall it's and it can happen very very suddenly and very very surprising so when talk a little bit about markets and who wants to buy open source stuff OK folks who just want for free really really hard for to make money off of people whose top priority is that they get it for free businesses for what is a business mean what sites I mean are you talking about small businesses are you talking about enterprise markets medium-size business carrier-grade is 1 of the things I like to talk talk about the with things like e-mail servers Curia grade is we're all Comcast or something on and consumers well consumers are very fickle they need a lot of Polish and Polish is what I when we 1st running x 11 at Berkeley there were these toolkits coming out that really loaded down your machine and they did things like they made the corners of buttons around I was like why the fuck would throw away my machine to make buttons around as well in truth you need rounded corners on your but that's just the way it is square buttons look clunky it's it's a kind of crazy things I guess I should say that this this is by the way is that they don't just buy chronically by trust in they by no confidence whatever and you know the what that means is not necessarily that the codes going to be greater than even like that that they trust the company behind it they believe they're going to be supported so for that's very very hard to do in open source environment where you can at any time staying on board with this and then go on they want something that they know hasn't uh vested incentive in staying alive in open source does tend to commodities market and we saw that very clearly with MTA's there were a number of times a very expensive MTA's open wavefile some had something called the and what not and as and became free MTA's became more ubiquitous and they just couldn't sell that and we certainly knew that going into some so we had buying something else to do and generally that means moving up this these supply chain the future of making venture stuff the and might I want to emphasize that a lot of founders start off thinking that they've got this really great technology and so they should start a company really great technology has very very little to do with starting a company and so on you know if I sound bitter maybe that's because I am a little bit I learned a lot of so there's a number of
models and you can use to actually make some money off of Open Source What is the people your source free and you sell something else this is kind what right had tried to do this so you know support this sold services people customization that kind of thing but the code was free and they you also consult stability and twice say stability here I don't necessarily stability of the code but once against that trust that I was just talking about the the belief that you will be in business you know next quarter next year but but they're limited economies of scale and that's the most everything it's in there requires as you have a bad customers you have to add employees and the nice thing about things like software is it doesn't really cost you very much at all for somebody else to do another download so it's the same you know if you have 10 customers a million customers and another approach is to keep it free but sold bundles of this kind of fundamental thing behind a lot of appliances and you know they should them a box and they it looks like they're buying the hardware but they're not really what they're buying is all this stuff on hardware cost next to nothing typically I must have special backplane and things like that some distros can often be this way you get this distro and you drop it in and it just works if you've ever tried to configure x 11 on Free BSD for example but not PCB-ST with previously you'll know why you want to that I can have 3 basic technology and with non open-source add-ons and this generally works best when you have a fairly clean extension model and or somehow have somewhere you can wrap the open source software and commercial software if you actually have to go in and do a customer version of the open source software that has the commercial extensions then you're not really selling your source more so an open source was plus or minus minus as the case may be and this generally supersets the sell something else I includes everything in that category we certainly don't know so out an appliance and not have support phone number for example the sendmail by the way was the 3rd style so we need some sense pioneered this and there's the 4th 1 that has been tried a couple times which is you do the technology graph to take the open source software you say OK that's frozen in time we're never going to support again if you want the supported version of the by the commercial once and people tend to vote with their feet and those things tend to just go away and ironically even things that they might pay for they walk away from just because he took away from I have a whole talk on
starting a company and I'm not going to give it today but I've said before starting companies has very very little to do with technology and a little technology is a good thing but what it is about finance you starting from the 1st day you go out of your head and hands trying to get somebody to give you money that's about sales or marketing and it's about support that's services should about management there but it didn't and the and there's some engineer and engineering is in most companies is a tiny fraction of the budget I have to confess that I started off I did know most of these things did and I now I do and I I'm not sure I'm better off for it keep in mind the 3 most powerful positions in a company or the CEO the VP of sales and the chief financial officer those who want to run the company notice VP of Engineering is not in the list so what can we observe well-financed is obviously about money you know you can get people to give you money you know initially the investors later on customers so for them avoid having other people take it away sales obviously about money and what else marketing that's a little more indirect but it's basically about making it easier for people to buy your product so it's about getting money into the company support charge for support you doesn't scale as well as software sales but you know it's still definitely about you don't like give support away because it's a very expensive thing to give away services is is supported services tend to differ because services more customization and things like that and charge of bond for services that's really that's the you know 3 thousand or they can solve and engineering 0 who that's a cost you know engineers don't go out and get people to send the money they in fact sit around going we need more money for us to do our development still only major function that really looks like a major cost etc. as opposed to a profit center so just little word about
corporate culture and debated whether I would put this in because this you can do dive on this really easily the major 1 is is the company that the engineering or sales marketing driven there's a number of new ones in here some companies try to avoid asking the culture question at all you know the usual argument is well you know it these have to be equal and if you know you're even asking the question then there's something wrong with your fundamental model and when that happens sales and marketing when there's almost no the company that is in fact engineering-driven a major 1 I can think of is and Google it's definitely an anomaly that a lot of very unusual things and starting up the company and certainly if you think you're going to start the next rule I look at the financial structure of at least as much as the technology because the financial structure is fascinating i investors prefer the sales marketing driven approach and they're on the board really the investors were on the board but generally speaking the founders are not the major investors so you don't really have that power now pure sales and marketing driven and leads to some obvious at some aberrations but it turns out it's very hard to avoid this in a fiscal crisis and it comes down to do we support the people who are bringing money into the company or the people who were spending money on a fiscal crisis the ones bringing the kind of money in always when they remember money is the lifeblood of company and a fiscal crisis by the way always comes along sooner or later it doesn't matter what company you are how well funded or whatever it it will happen in 1 way or another the possible exception of this is when you're sitting on a cash pile of billions of years zillion dollars which means you're Apple Workgroup is in there the only companies around the like that but the big difference between these 2 concepts is the short-term versus long-term view companies these days is not always been historically true but in the last 40 years 50 years of companies tend to be focused on the next quarter actually next quarter is the long-term view this quarter of for short term view next quarter's long-term engineers like to think a year out 5 years that sort of thing and this changes in very deep way is the way you think about almost all problems so assuming you're trying
to convince your company that they are to be doing some open source there are some justifications you can make this is not an exhaustive list by and wonder kind standard 1 is all we do this we can you get people contributing to our project we get community but they'll write code for us to our costs etc. etc. and that does happen sometimes not always licenses can be a show-stopper here so pick your license very carefully but it can be useful as a recruiting incentive if you're for whatever reason needs really hot engineers for the project you're working on it and they tend to really like working on open source stuff and of the I think that's a little sneaky there of it can be useful as a recruiting incentive but it can have unintended well unintended to them maybe very tended to you having culture shift because people who work on open source software tend to have different priorities than even engineers that work on closed source software now 1 of the things that is really interesting is and some of the CEO's get this is you can disrupt the market and if you can disrupt the market in a way that your competitors are not expecting that you are you can get an advantage over them that way it's a little tricky to play off because of course it makes it hard for you to make money as well but but at least you are not stuck with the element of surprise and and it can make certain customers more comfortable these customers are not consumers or anything like that the big corporations that have their own you know highly technical debt people in the house because what happens is if for for whatever reason you go under but they have the source code they're not dead in the water they can still do things and for a certain small but fairly large class of small but large small classic customers but those customers are large enterprises and this can actually be a good thing this is not an exhaustive list said that just a few I could think of so I want
do some lifecycle comparisons here this is actually definitely Robert suggestion it's kind of interesting I start talking about open source projects and research on the same non-military non-proprietary research for this doesn't apply to pharmaceutical companies for example we have a whole different view of what researchers and the company and the similarities on this all try point out and there's also some spectacular differences so everything starts off
at the initial inspiration and the idea like people would have an open source is usually scratching this is very broadly defined can me I have a problem with me and I know somebody came up and said boy if true would be neat if we had a lot a research project is starts by asking questions what we mix these 2 chemicals together I wonder what would happen to the companies start by seeing a revenue opportunity simple as that
time so the next step is kind of making it possible you got the idea now you want to flesh it out so open source you start by seeing if it's already been done over what's actually she peoples tend to go the opposite direction far too often somebody else did it but I like the way they did it so I'm going to do it again OK it's not clear that the world gets this value out of that you do architectural design and some people sometimes open source often sort grows this this whole thing called agile programming that is very popular right now which can work very well but for a lot of people it's what it really means is I don't want to think about the design now worry about that later Africa code or something wrong with that you choose your language tool and tools that is that very early in open source compared to the other and you start writing code as simple as that research project you start off kind of the same as open source were search for literature it's already done so what the outstanding problems are you gotta get funding and open source projects typically don't have to start off by thinking about getting funding so in this case researchers more like companies and you start to line up grad student when the company you write a business plan business plans actually can be very useful when you try and justify these things even to yourself sometimes the open up you line up your investors which is time to go into that investors offered animals cells but that and I figure out for corporate culture we as I mentioned earlier and that's an optional step a lot of companies skip it entirely and uh that tends to run create problems later but it can be done fire team once the investors common which the birthing the baby
and making it actually have open source you do you're here early releases 0 x and start building your community getting early uses the ones who would like to live on the leading edge research projects to start writing the code doing the actual research and start writing right up that go into certain kind of conferences and that sort of thing maybe the stuff to the press will start writing about you know the cool new quantum computing thing you're working on whatever and companies start building your product so in that sense once again it's kind like a research project line of early customers always want have some early customers in part to get feedback from them on what it is they want by you really lucky you can get them to pay for the product before you built and that's part of the financing models and start doing trade shows that was going you want people to anticipate your product when comes out so making it real open source you do release 1 . 0 0 1 and you discover you have a support problem this seems to always happen even if you think you've anticipated and see figure out a way to deal with that and you discovered by the way the docks if you wrote docks at all they're not good enough so this is where the real and problems start to come out in open source in a research project well it's all about publish-or-perish start that structuring those papers that to conferences etc. friends at Berkeley grad students and seems like every 3 weeks there submitting another new paper and I don't know how they do it in a company to get your 1st release of the product of a little bit of sales and marketing and so forth but now you need to start to scale those up make it a little more real the next obvious step is you need to grow the thing so
an open source well if you don't have users don't have a community you're over you're done otherwise you write your Riley but this is very very important step is in the open source right to and you try and avoid second-system effect although this is a purely optional step and come up with release 2 research projects I you know research projects you don't really grows on you sort of winding down and move on so ceases time your students slaves start to graduating and you write the the hard Journal you transactions on kind of papers the company you your 2nd release at this point you really need and get profitability and so forgetting to profitability almost always involves doing away off so this is your 1st or maybe 2nd and possibly 3rd round of layoffs and I probably get another investment round so the idea is we wanna become profitable we cut our expenses down by laying off everyone we don't possibly need we go out to investors and say 0 look at how their books are you should give us more money and then you continue on from there and I'm sorry but that's the way it really works and then the next steps a lot of what you do open source is actually it's a more complicated problem we might think you might just throw to the winds notice by the way that most people work on open-source projects don't do it for the rest of their life some do there are exceptions but mostly want go off and do something else the exciting thing is writing stuff it's not supporting it and so you know brought to the winds Italy if somebody else will pick it up if it's important another line handing it over to somewhat larger organization you go and try and get the Apache Foundation to pick it up this can include selling into a company can be commercialized it start your own company around it and maybe just keep going to say well I'm having fun and not just keep working on it there's absolutely nothing wrong with this in the open source world remember you don't have to be doing getting and continuing income stream because there's no such thing to begin with in a research project the next step is to ask another question which is often suggested by the previous cycle accompany this kind of 3 outcomes you have what's called a liquidity event which is where the investors can take their money out and then you continue ongoing and that becomes an ongoing project you have a liquidity event this is usually by being purchased by a larger company and you get assimilated into the company and you go away and that may or may not mean that you're and engineers in India employees going on you can be assimilated and just kill off your product and there's just no givens here mean it becomes the center of their product lines and of course there's bankruptcy and which and once again you don't have the option of just throwing it out to the winds and let somebody else supporting if you died
so I'm going to finish up here there's kind to big mistakes that founders make which almost always cause the founders to be forced out and sometimes fairly since founders are always surprised that they're out of a job within 3 to 5 years generally sometimes less than that 1st is assuming that
they know everything is by far the most common I know exactly the way I want this company to be I'm going to control it I'm not gonna let anybody else try and convince me otherwise this is the way I want it and you don't know everything a special about areas you don't know anything about and making other people's lives miserable will involve a palace coup in you will be asked it's just the way the 2nd
1 is assuming it's the kind of the opposite assuming that everyone else knows more than you do and and they have no hidden agendas everyone has hidden agendas I have hidden agenda and the where people who try and tell you that their field is so complicated that they can't possibly explain it to you when you're smart people you can probably actually understand a lot of In the case of sendmail we start off with a very very good VP of marketing which was a complete shock to me because I've never met the VP of marketing before that and and he was every time I asked a question was like eager to explain it he really wanted the engineers understand what marketing is all about and listen to the engineers and and work things together and so forth he left we got another 1 in who preferred walking his office store not letting anyone and he believed in quantitative marketing so he spent all this time looking at spreadsheets and graphs as opposed to being out talking to people about it is a very bad thing and that's when I started be told well yeah you you you don't understand that and I made the mistake perhaps of falling back to 4 going OK you know I don't want to it's not that I thought that I couldn't understand it is that I didn't wanna be getting in the way so for so obviously some sort of happy balance is needed that's very very hard to so some
conclusions there is some good news I don't have any doubt that commercial input open source has allowed it to take on much larger problems than it could without the commercial input and it's both in terms of the scale of things that it has been able to approach and the kinds of problems it's been able to approach and that's a big big change from 1975 I can assure you the bad news of course is that open source in this process is sort of lost its innocence and you know the current corporations emphasize the short-term survival over technological beauty and so forth and so on the open source at some point has to neutralize some of that and that's not always a great thing so finally there is a 3rd mistake that founders make this is my made in abundance but we then we'll be able to get back to protein sort of
fat sold B this kind of thing I've got a letter from question or 2 working from the video you at a at so maybe I can