- Debugger: Getting mixed signals
- 'Packaging Infrastructure' Contract Weekly Report #4
- Haiku monthly activity report - 06/2015
- 'Packaging Infrastructure' Contract Weekly Report #3
- 'Packaging Infrastructure' Contract Weekly Report #2
- GCI 2014 winners trip report (mentor side)
- TeX Live and LyX; Changes to the boot code
- 'Packaging Infrastructure' Contract Weekly Report #1
- Beginning of 'Packaging Infrastructure' Contract
- Haiku monthly activity report - 05/2015
End of contract - closing words.
Hi,
As you probably have noticed, there were no weekly report in the previous weeks. The reason for this is that my contract is currently stopped. There is currently not enough money in Haiku's treasure chest to safely continue it. So, it's time to me to get back to "real life" and a full-time job in a software development company.
First of all I want to thank everyone who made this long contract possible by donating money to Haiku. It was a great experience for me, and a lot of fun as well. I did my best to move Haiku forward towards the R1 release. Unfortunately the beta 1 still isn't there, and we currently have 57 blocking tickets. It is a small number, but only the most complex or big issues are left.
I have spent the greatest part of this month job-hunting, but I found something now (in a company with some free software culture, even). It took me some time doing all the job interviews and the like, and required quite a lot of my energy and motivation, so not much of it was left for Haiku work in the evenings. But once things have settled a bit I will make some time for Haiku in my free-time schedule again. Progress may not be as fast as it was during this year, but I'll still be around.
I'm finishing this contract with some things left on hold, so I'll document the current state of things here, for someone else to pick it up.
I still have the libbind to netresolv replacement project in my local git tree. It is still not working reliably, and it needs getifaddrs moved to libnetwork instead of libbnetapi (I tried to do that but my code wasn't acceptable as it needed too much hacks on the libbnetapi internals). WebKit is also left in a somewhat incomplete state. I think there is a crashing bug in the latest package, because of an ABI change in the network kit, but this should be easily solved by a rebuild of the webkit package. There are remaining drawing issues (both incorrect drawing and performance problems) which need to be solved on app_server side. There are also known problems left in the network backend and WebKit network code, mostly with web sockets and http cookies.
Let's have a look at the short-term future of Haiku now. The next event is Google Summer of Code. Despite the name, it starts now, in the middle of the winter, and this week is the application period for organizations. This year Urias will be our org admin, and Axel will be the backup admin. You can have a look at the ideas page already:
https://www.haiku-os.org/community/gsoc/2015/ideas. We have quite a lot of ideas there, but unfortunately we don't have much mentors around. If you think you could mentor one of these (even if you don't have commit access to Haiku), it's time to tell so :).
Another upcoming event is the GCI winners trip. This is a reward for the GCI winners (2 from each of the 12 organization) who will visit Google headquarters and enjoy various talks, and other events. One mentor from each org is also invited, and this year I'll be going, as the students both decided they wanted to see me there.
On Haiku, inc. side, there is discussion in progress for a rework of the way it works. The board of directors worked by co-opting until now, but that didn't work too well. It's basically the same people running the inc for a very long time, and they can't commit enough time to it anymore. There is work being done there on modifying the internal organization so board members are renewed more frequently and with a more transparent process. They will probably be looking for new board members at some point, so if you think you can help with the boring paperwork, tax forms, financial reports, or reporting on Haiku, inc decisions and events, you can contact the current team at contact@haiku-inc.org.
On the Haiku project side, Augustin Cavalier got commit access. He joins the developer team and will contribute in various areas, including user interface and documentation. Oliver Tappe is stepping down from his role as the main system administrator. There is quite a lot of work to do in that area, and we have a small team of people organizing the sysadmin to work more as a team effort, in order to share responsabilities and workload. This is coordinated at the haiku-sysadmin mailing list, so if you think you can help, join there: http://www.freelists.org/list/haiku-sysadmin.
Finally, even though my contract is stopped, I would like to encourage everyone to continue donating money to Haiku, inc. There may be other people interested in doing contract work (or maybe myself again at a later point), and there may be other ways to spend the money in useful ways. This is open to discussion on the haiku-inc mailing list (http://www.freelists.org/list/haiku-inc), it could include helping our GSoC students to attend BeGeistert with a (partial) travel refund, for example. Or things like buying specific hardware for Haiku devs to write drivers for it, etc. As I said, you can send suggestions on the haiku-inc mailing list, where they will be discussed and maybe adopted.
- PulkoMandy's blog
- Login or register to post comments
Comments
Re: End of contract - closing words.
A link to the remaining B1 issues for those interested.
First-rate work, Pulkomandy
As with every aspect of what you've done for Haiku, even your summary article at the end of your contract is well-done. You're such a great member to the Haiku community, and I hope you're around for a long time. I also hope we can get some cash together and involve you again on a paid basis. Thanks for your hard work, good results, and also for the nice way you help and respond in IRC.
Dane
Re: End of contract - closing words.
In Regards to donations...
Don't forget to sign up for 'smile.amazon.com' and select Haiku, Inc. as your 'charity', since they are registered as a non profit charity.
Anytime you buy through 'smile.amazon.com' instead of just 'amazon.com' and it's eligible, Amazon will donate a portion of your purchase to your selected 'charity'.
http://smile.amazon.com
Re: End of contract - closing words.
That's a great initiative from Amazon and a great tip!
That should really be given more attention. I would really encourage the website maintainers to put up a module on the frontpage with that information.
Adrien, it was always interesting reading your posts, and thanks for your work.
I'm really glad that the board at Haiku Inc. has decided to finally change their ways. I hope some of my criticism helped in that regard.
I don't want to rain on the parade, but I still can't get over these open-ended contracts. I think the web positive project was set up to fail from the beginning. It was a behemoth of a project for one man. In the end, like package management, we're left with something that, at best, half-works. Mark my words, in a year and a half, Web Positive will be little better than what it is now, and probably won't even be compatible with Haiku itself anymore.
It's time Haiku Inc. propose projects that are result oriented, or at best a mixture of both types of contracts. It's time to gamble. In my eyes, and many others, these contracts are failures. Of course, the argument has always been that open-ended contracts are more comfortable for the developers. Who wouldn't want a nice reliable pay cheque every month. But, I think it's poker time. Haiku is at the stage of irrelevancy. That said, post attractive contracts. I am unsure how much was spent on Adrien's contract -$25,000 USD? If Haiku Inc. were to put out a result-oriented contract, where you get paid for an end result, then the risk is on the developer. If that developer is instead attracted to a $25,000 contract, at his/her risk of finishing a complete contract, then wonderful! Let them do it however fast they can. If they don't complete it within the timeline, what has Haiku Inc. lost? Not the precious funds it has little of. Moving these contracts to external developers would take away the personal aspect of a possible failed contract.
Probably the greatest legacy of the contract work for Web Positive that Adrien did were all the indirect fixes to Haiku itself - but this is not what the users wanted - they wanted a web browser (even this is arguable, as there were plenty of other tasks/projects that had higher voting percentages). This poll, by the way, is highly outdated. It's about time to again poll what's left of the community, to see what they want to see in Haiku so that the proper projects can be funded, and money starts flowing again. In the end, what the user wanted, was a browser in which he/she could log into Facebook, Gmail, use IM, see and hear YouTube videos, do online banking, visit the web's top 100 websites, all in a browser that wouldn't crash or lock the system up. I don't really think they care about all the under-the-hood fixes - I know I don't in this case. I would have rather seen a portable browser that relies on a stable framework and that just works than something that will inevitably be impossible to keep up to date and is unreliable.
Yes, the developer always has to be found for these projects, however, I don't really think enough effort was put forth by Haiku Inc. to look outside of Haiku's core developers for persons that could do the contract work. i.e posting the contract work to notable contracting sites for software development etc. The argument has always been that, we need someone that knows the inner-workings of Haiku, and can contribute code that's clean and acceptable to the core devs - 'the Haiku way'. Well, time and time again, I've seen developers pop in to the Haiku scene out of nowhere to feel right at home developing for Haiku.
I think the fact that there's little money available speaks to the community's lack of enthusiasm for Haiku anymore. They have spoken; it's the judge and jury (that cute little pig on the donat-o-meter that irregularly gets updated).
The most exciting part about this article is that the board of directors will be recycled and hopefully the election process, reporting, etc. will be in a more transparent and democratic manner so that new ideas & new ways come about to support Haiku's mandate: "a non-profit organization dedicated to supporting the Haiku Project (the "Project") and the development of Haiku® (the "Software")". With the board admitting they are 'in crisis mode', them hardly being able to put in annual meetings, minutes, and update the corporation's financial status regularly, etc. - they have absolutely failed to support that mandate, and it's ABOUT TIME this is going to change.
Excellent news.
Re: End of contract - closing words.
I think we discussed this already: for a project as big as Web+, I, as a developper, am not going to take the gamble for a result-driven contract. I spent one year on this, and yes, you can consider it a failure. There is still a lot to do. So, in your result-driven world, I would have got nothing for a year of hard work, and Haiku would have lost a developer (not nothing as you say). And personally, I would have lost my girlfriend, my home, and a few more. Do you think this is a reasonable risk? For the low price? (yes, $25000 for a year of work is a low price. I'm starting a new job next week and I'll be earning €34000 per year. This gives you an idea how low the Haiku price is for a reasonably skilled software developer - most of the Haiku devs have more experience than I do and earn a lot more than this).
I accepted this contract not because it was "a nice paycheck". I accepted it because I wanted to make Haiku go forward, and I needed a better web browser and a better system. This is my motivation. The money was just a way to keep me alive (I need a roof, internet access, electricity and some food to continue working on Haiku).
Haiku is not a company. "taking out the personal aspect" can't work. We would be working with a dev for one year (somehow he will have to integrate with the existing team) and then telling him "sorry, your work is not good enough, we keep the money!". Do you really think anyone would accept that, working at 2/3rd the rate of an usual software development job with a big "maybe" on wether they get paid in the end? Well, find a dev willing to do that, let him send his proposal to Haiku, inc and I'm fairly sure they will accept it. Haiku, inc works this way, devs send a proposal and they accept or reject it. And it's up to the dev to set their conditions and what they plan to work on.
Challenge accepted on keeping Web+ working for the coming year and a half. As I mentionned, I'll still be around in the Haiku community (this is not a bounty-driven contract from an uninterested dev doing it just for the money!). I'm keeping WebKit sources up to date and making sure they build. There should be yet another release up soon.
Also, I already mentionned that there are other ways to donate money to support Haiku. You know it, as Haikuware accepts donations. The HSA is also there running BeGeistert and the coding sprint. The amount of money going to each of these entities clearly shows where the interest of the community is and what they think works best. It is amazing that I could work on this for one year funded solely on donations. Not many open source projects can say the same. So once again, thanks to everyone for your support!
Re: End of contract - closing words.
I'm with you on this one. Is that before taxes are payd? In Sweden I would make more money unempoyd. Thanks for your work :)
I have loved reading your weekly progress :)
Nice that you found a nice job and what happend next we will see in the future :)
//FredrikRe: End of contract - closing words.
On the other side, the pay of 34000€ per year allows me to take 35 to 40 days off work (which I'll use for attending BeGeistert and other Haiku related events, of course ;)), and some other advantages (corporate or legal). As well as a much better public health insurance and probably some other similar things I don't even know about yet.
As you can see, the money I got from Haiku did not get me rich. It was enough to pay the bills, and that's it. The situation is different in other countries inside and outside Europe, but relocating wasn't an option for me. In any case, as I said, Haiku, inc. will consider any offer, so if someone wants to do better work at a lower price, I will be very happy with that as well. Nothing is better for me than seeing someone else do work from my TODO list.
Re: End of contract - closing words.
Hi Adrien,
As was shown, developers are wiling to take these sort of risks if the potential reward is attractive enough - the bounty system at Haikuware proved that (even with petty <$3000 contracts, imagine some worth $25,000), and provided several 'as defined in the terms of the bounty' results (i.e I stick my printer in a USB port it prints, I can now connect to Wifi, my SATA hard drive works, etc). These developers are usually young, unattached, maybe even still living at home, etc. They are there... Perhaps this is something Haiku Inc. should look for in future contracts; that is to ask for the person's relationship status (not uncommon), their demographics/therefore cost of living, etc. And do a rough calculation to see if the developer has too many distractions or financial burdens to produce results and not be distracted, etc. A $25,000 contract, and the money would go a lot further to a developer living in i.e Romania, India, etc. than in Western Europe or the USA (as you've rightly pointed out). Haiku should just be more cost-effective and efficient minded and realize they can't afford to fulfill a mature Western developer expectations and try to outsource equally talented students or Eastern developers. I also bet these developers would be more appreciative of the donors money and not insult them by saying that they're under-paid and that with the money it basically covers cost of living etc. That could be saved, and the money accepted gracefully.
Of course, 'Haiku is not a company', but it is an organization with responsibilities, legal ones, a mandate, and in charge of other people's money that was donated to it. For all intents and purposes it is almost exactly like a company, except it's not for profit.
"sorry, your work is not good enough, we keep the money!"
That's because Haiku Inc. doesn't know how to exactly define terms in a contract. If the developer fulfills all the terms in the contract, than Haiku Inc. has no choice but to pay the developer. And they would gladly do so if the developer was able to produce the result of a well defined contract.
" Haiku, inc works this way, devs send a proposal and they accept or reject it. And it's up to the dev to set their conditions and what they plan to work on."
Yes, and that's why Haiku Inc. shouldn't work this way. They are there "dedicated to supporting the Haiku Project (the "Project") and the development of Haiku® (the "Software")." Haiku Inc. should set the conditions, not the developer. Who is the paymaster here?? This is so illogical. I struggle to find examples where an employer says, you know what, you want a job, great - you tell us what to do (I am sure there are examples, but 95% of the time, it's the other way around)
As I mentioned earlier, Haiku Inc. doesn't do enough to find developers. They expect the developers to just come to Haiku Inc. It's not uncommon for organizations to spend a small percentage of their budget on advertising. How many times has Haiku Inc. used financial resources to post job ads in legitimate online portals? Well, it hasn't, because it can't come up with a serious, defined contract that states exactly what it wants from the developer. And in the end, as has been shown, we are left with software that is practically useless. I really do believe if Haiku Inc. put all their eggs in one basket for one fiscal year; i.e a $30,000 goal-oriented contract, it would find a developer easily. If the developer delivered and the end-user sees a tangible working piece of software, instead of something that's incomplete and non-working, they'd be more inclined to donate in the future. I know I would.
The biggest insult to the donors, is the state the contract is left in:
"WebKit is also left in a somewhat incomplete state."
I don't know about you, but I'd feel pretty bad about that one. So, we have to take the developers word, and be so graced that it will be fixed in spare time. As with package management, similar promises were made. It didn't happen. It was left in such an end-user useless state, that had it not been for Stippi committing all his free time to clean up the dirty work and enhance package management, it would be nothing.
Call me blind, but I just can't put a positive spin on any of this, except that hopefully the board will change and take better care of the financial resources it does have and pay for work that's defined in a contract and where an end-result is produced and that is of value to those who donated. And Adrien, I don't mean to be-little your work. It was great what you did, I just find that it's misguided (by Haiku Inc.) and didn't produce something end-users expected - unfortunately the reward is Haiku's current fiscal status.
Re: End of contract - closing words.
What happens with smaller bounties is that it drives people to work on some feature in their free time. You need to understand that: if the bounties had not been there, those people would have worked on other features instead. Yes, it is a way to drive the development team with user input, in the form of throwing money at people doing the "right" things. But it simply can't work as a full-time job, because it is too risky.
This is not what Haiku, inc. is trying to do. When people donate money to Haiku, inc., they expect it to, as you like to quote, "supporting the Haiku Project (the "Project") and the development of Haiku® (the "Software").". I think it is very clear to everyone that Haiku, inc. is designed to not be a decision making entity. The decisions on the development of the project are in the hands of the development team. When the development team (or one of the developers, or, actually, anyone else) needs money to perform one task, they ask Haiku, inc. And the organization can only decide wether this is in line with the goals.
You have to accept this: people donating money to Haiku, inc. are putting it in the hands of the org, and the org leaves most of the decision power to the Haiku project and the development team. This is by design, and on that side, things work as designed.
If you (or anyone else) don't agree with the way things are run there, it is not a problem. You can contribute to the Haiku project without Haiku, inc. You can encourage people to donate money elsewhere. Whatever works for you: bounties at haikuware as you did before, the Haiku Support Association, which will probably spend the money in a more "active" way (organizing BeGeistert, for example), create your own non-profit, start a kickstarter campaign, or hire a developer directly in your own company to work on Haiku. It's not like Haiku, inc. is preventing any of these from happening. Yet, the money was donated to Haiku, inc. and allowed to fund a contract for more than one year. This shows that a significant number of people are confident enough in that way of working.
I'm now waiting for your "let's hire a team of romanian or indian developers to finish Haiku" Kickstarter campaign. Why do you need to involve Haiku, inc. in it? You can run it yourself, you don't need official approval or anything. And it may even work.
On the fact that I will continue working on Haiku on my free time: of course I will. It is my only operating system at home, and I need it working. I would be very happy for your team of romanian and indian programmers to fix all the bugs so I don't have to do it myself, but until this happens, I will have to do it myself. Not because I want to please end-users or anything like that, but because I need it done. I don't ask anyone to be "so graced" or anything. I am very annoyed by this half-working web browser, as much as any other Haiku user. And I also had to look for a new job, which is not the most fun thing to do when you know there will be no paycheck at the end of the month. No, I'm not trying to suck money out and only spend 3 hours a day writing code to give the impression that there is some activity happening. I spent countless hours on writing code for Haiku, much more than I actually billed, especially if you include the GCI mentoring tasks, getting our GSoC ideas list ready, etc. I enjoyed it because I know it makes my preferred OS move forward and it makes my computer more useable. This is what I'm after, and for this reason I am ready to accept the unreasonably low pay and the uncertainity (several times during the contract I was not sure I would get paid for the following month).
If you think the best way to run Haiku is to crush the developers with always lower paying rates, outsourcing the project to other people if they are willing to do it at a lower price, putting as much pressure on them as possible by threatening to not pay them if they underdeliver, and so on, well. Maybe you can find sone developer willing to work for you, but it certainly won't be me. And it also won't contribute to making Haiku more widely known in a positive way. During my job interviews this month I had to explain what Haiku is and how it happened that I was able to work for an open source project for a whole year, and this is the same positive explanation I give about it at various open source events. It really helps making the project look a bit more serious when people know there is a full-time developer with a long-term contract, and enough users to fund that with only donations. As I said in an earlier comment, not all open source projects can say the same.
In a model where the goal is to reduce costs by all means, by hiring anonymous developers, you won't get the same result. You may get the requested job done (probably with shortcuts in the implementation, instead of doing things "the right way" - but it seems you are ok with that). But don't expect anything more. No promotion at open source events, no GCI mentoring, people leaving when their contract is done, and possibly some bad reputation out of it. You may get the OS further, in the short term at least, but probably not the community that goes with it.
Re: End of contract - closing words.
"Whatever you think, I highly doubt we would find someone to accept a $25000/year contract with a result obligation. This is too much of a gamble for anyone."
Well, $25,000 was already gambled on you, what's the difference? The $25,000 is gone, that's the difference - and further, WebKit is left in an incomplete state.
That's the problem though. 'You highly doubt'. Well have you or Haiku Inc. ever tried?
"I'm now waiting for your "let's hire a team of romanian or indian developers to finish Haiku" Kickstarter campaign. Why do you need to involve Haiku, inc. in it? You can run it yourself, you don't need official approval or anything. And it may even work"
I have already proven it works, so it's really a no-brainer for Haiku Inc. to try it.
I shouldn't have to do what Haiku Inc. is mandated to do though. Haiku Inc. has already shown that package management and web positive contracts were failures, so they could try something different. Part of the problem with me doing anything is reputability. I am just a single person. A non-profit organization has legal obligations for money donated. People trust a Haiku Inc. more than an average Joe starting a kickstarter campaign. Haiku should realize, leverage and take advantage of this. I really am sick of arguing these things that are so obvious. Put me on the board of directors and I will help Haiku strategize new and different ways of achieving results; perhaps even a kickstarter campaign.
"You may get the requested job done (probably with shortcuts in the implementation, instead of doing things "the right way" - but it seems you are ok with that). But don't expect anything more. No promotion at open source events, no GCI mentoring, people leaving when their contract is done, and possibly some bad reputation out of it. You may get the OS further, in the short term at least, but probably not the community that goes with it."
I would take a perfectly polished web browser or package manager over 1 or 2 developers that dwindle off and don't support the community any day (this happened with mmlr's contract, bonefish et. al anyways, and will happen with you too). I'm sure the potential users, community supporters, and donors Haiku wins in return for contracts that deliver would be several orders of magnitude higher.
'May, Probably', etc. It's all conjecture. The only way Haiku Inc. will know is by doing - and they don't. There's no harm and no financial risk in trying. There is in open-ended contracts.
I wish you all the best at your new job!
Re: End of contract - closing words.
Have you tried? Otherwise it is also all conjecture on your side. The only way we will know is by doing. Prove Haiku, Inc. wrong, and maybe they'll change their mind, or if they still don't, they will run out of money as everyone donates to your new org instead.
I don't consider either contract a failure. The package management system works well enough for daily use and is much better than what we had before. So it is for the WebKit port (in current nightlies it is broken, but I fixed this and I'm building a new package which I'll upload later today - yes, in my free time!)
I see you don't appreciate the efforts I put into this to their right value anyway. The fact is a web browser is a complex software project and it takes much more effort than you seem to think, even just porting one and keeping it up to date. And the work I did had a much broader effect, making apps such as BlogPositive, HaikuWeather or HaikuTodo possible, fixing problems in GoBe Productive and ProjectConceptor, and also a few other things. This is needed if we want Haiku to succeed, with only a great web browser we wouldn't have much to offer (our goal isn't to replicate ChromeOS). Thanks to my efforts in the service kit, it is now quite easy to write a native app which gets or puts data on the Internet. This is where Haiku can shine and compete with other systems. The next step in that direction is embedding HTML views into native apps for a seamless integration of 'web' and local apps. This is where the future is, I think.
Re: End of contract - closing words.
Adrien, it isn't conjecture on my part. I have already pushed through results at Haikuware with goal/result driven projects. Haiku Inc. has never tried this - and it makes no sense not to. Further, I have tried to offer my help to Haiku Inc.
You don't seem to understand that I am not responsible to anyone. Haiku Inc. is. I don't have to prove anything, even though I already have. It is Haiku that should have to prove that goal-oriented contracts don't succeed (at no financial risk to the organization). Anything less is negligence.
"I see you don't appreciate the efforts I put into this to their right value anyway. The fact is a web browser is a complex software project and it takes much more effort than you seem to think, even just porting one and keeping it up to date."
Then you didn't read the part where I said it's a behemoth project for one man, and was set to fail from the start. I do appreciate the work you've done, it's just all over the place; and when you decide x or y needs work, you work on it instead of other relevant things.
Anyways, the goal of a contract is to produce something that works fully at the end of the contract and not 1.5 years later, and what donors also expect. It's clear we differ on these issues. I gave this analogy before. If I hire a contractor to build a house, I can't live in it without the roof. It's incomplete. Only someone with no sanity would pay 100% for the incomplete house. I realize that's not what the open-contract offered, but that's why goal oriented contracts need to be definitively defined and would better serve the community.
Re: End of contract - closing words.
I like what you did with Bounties in haikuware, I also like that there are a lot of app there. Some working some not but that are more up to the developer who did them and if no developer are around the they will not be updated. :)
I also like the way the Haiku Inc contract went, both of them. I hated the PM in the beginning and it was not finished but what we have now is lot better I think and that would never have been without the contract.
When It comes to having a goal oriented project we will need a lot more money than 25.000 and for a lot longer than one year and no "perhaps longer than a year if there are any money left."
When it comes to building houses the builders will not build a house for small pocket changes and the workers will stop building if they don't get paid and you will still be billed for the time they have spent building on the house.
Re: End of contract - closing words.
The Haiku Code Drive was an extension to the Google Summer of Code, allowing to pay students to work on specific features with a result obligation. It is exaclty what you suggest: result obligation, and low pay.
It didn't work that well. For this reason, from 2010 on the funds were instead allocated to long-time project members. This page summarizes what has been accomplished since then: http://haiku-inc.org/funded-development.html
You can't say they did not try this, they did, they failed at it, and switched to something else. Please don't try to rewrite the history.
Re: End of contract - closing words.
So Haiku Inc. tried a bounty-like program for, at best, one year (and with students at that) and consider it 'tried and tested'? I hardly consider this a legitimate effort... Why discontinue them at all? Does it cost Haiku Inc. more money to leave an article on their website or close this opportunity? Thanks for bringing this to my attention though, I stand corrected (I think).
Re: End of contract - closing words.
Karl, I find it quite disturbing how you present a distorted version of reality or even blatantly lie.
At no point have Oliver, I, or Haiku Inc. claimed that package management would be finished within the time of the contracts. It was clear to everyone that package management is very big project. Pure spare time development would have taken many years, which is why no one even started with it (except for my initial proof-of-concept packagefs implementation). The purpose of the contracts was to chip away on the big block of total work, so that it would become smaller and more manageable, and to get the basic PM system to a point that would allow other developers to contribute to it.
I don't know about you, but I'd feel pretty bad about that one. So, we have to take the developers word, and be so graced that it will be fixed in spare time. As with package management, similar promises were made. It didn't happen.
Besides that no such promises were made, we actually did continue and package management is in a usable state.
It was left in such an end-user useless state, that had it not been for Stippi committing all his free time to clean up the dirty work and enhance package management, it would be nothing.
I wonder what "dirty work" you think Stephan cleaned up. All he did is work on a GUI package manager application (with additional features). It was always the plan (and clearly stated) by Oliver and me that we'd work only on a CLI package manager and leave the GUI as a third-developer opportunity. Consequently Stephan started with HaikuDepot *before* our contracts were finished.
"I'm now waiting for your "let's hire a team of romanian or indian developers to finish Haiku" Kickstarter campaign. Why do you need to involve Haiku, inc. in it? You can run it yourself, you don't need official approval or anything. And it may even work"
I have already proven it works, so it's really a no-brainer for Haiku Inc. to try it.
We discussed your delusions regarding the success of the Haikuware bounty program at length on the haiku-inc mailing list. As a reminder: The program did not attract developers from outside the community and it probably didn't get any work done that wouldn't have been done anyway (respectively only diverted resources from other work). In case you forgot, please re-read the testimonial of one of the (successful) bounty takers.
I would take a perfectly polished web browser or package manager over 1 or 2 developers that dwindle off and don't support the community any day (this happened with mmlr's contract, bonefish et. al anyways, and will happen with you too).
I would take a perfectly polished Haiku over a horde of volunteer developers any day. Alas I live in this world, not in dreamland. I don't know what the "support the community" part of your statement even means. Apparently several months of unpaid work after the end of the contracts is something else.
Re: End of contract - closing words.
Hi Ingo,
Karl, I find it quite disturbing how you present a distorted version of reality or even blatantly lie.
I didn't claim/distort/lie that you said the package management would be complete by the time the contract would be done. I argued that a goal oriented contract should end up with a completely usable end-user product at the end. Most end-users I know don't even know how, or are afraid to use the CLI, so to them, only a GUI on top of the package manager is something they would use. Yes, I realize this wasn't part of the contract. I was just arguing that from an end-user standpoint, a complete functioning GUI package manager would have been more desirable 2 years ago (and part of a goal-oriented contract).
We discussed your delusions regarding the success of the Haikuware bounty program at length on the haiku-inc mailing list. As a reminder: The program did not attract developers from outside the community and it probably didn't get any work done that wouldn't have been done anyway (respectively only diverted resources from other work). In case you forgot, please re-read the testimonial of one of the (successful) bounty takers.
I don't have a budget to advertise over the internet (as Haiku Inc. should do) to attract developers from outside the community. Nevertheless, you are incorrect (I won't go so far as to say lie/distort reality or peronally insult you by saying you're delusional, but that's on your conscience and public record). Two bounties (out of 8 - 25%) were completed by external developers Adrian Panasiuk (Gnash), Ciprian Nesidan (Document Viewer). I also highly doubt things like a document viewer 'would have been done anyways'. Even bluetooth, it has received close to little attention since the bounty. It's arguable about who contacted who or who was already known to the community for other programs first; Oliver Dorantes (bluetooth), Colin Guenther (wifi). I mention arguable, because you and the rest of the community assume, that for e.g Colin was already known to the Haiku community before applying for the bounty, well guess what - this just is not true. I received an email from Colin the exact same day he introduced himself to Haiku's mailing list (albeit hours earlier applying for the bounty).
Colin's email
Colin's self-introduction
So the arguments you make, and claims that I am delusional really have no ground.
In case you forgot, please re-read the testimonial of one of the (successful) bounty takers
Wow, one single testimonial. I don't want to dig into further personal emails, but I do have several from bounty takers where they appreciated the program and found it a positive experience. To counter your sole 'testimonial' here's food for thought (by the way, Axel's bounty was incomplete, as in not completed as he himself out-lined):
http://haikuware.com/blogs/entry/wlan-stack-sum-up-on-going-and-future-work
I would take a perfectly polished Haiku over a horde of volunteer developers any day.
That's because you're a developer. End-users all over the world claim Haiku is stable enough, even in its current state, to use day to day. One of the main things users do, that don't code, is surf the internet. Without a modern browser, the OS is practically useless. That's why it's a shame that there isn't a reliable end-user ready browser available for Haiku today, and after $25,000-$30,000 spent. I am sure most users would agree, they would like to have a working stable browser like Firefox or Chromium on top of Haiku as it currently is than close to nothing at all.
Re: End of contract - closing words.
I don't want to provoke more long winded rants, what with your illness and so forth... but doesn't the fact that Colin had analyzed the WLAN stack situation of Haiku and picked it as subject for his master-thesis suggest that he would have done the work with or without a bounty? I'm sure he was happy to earn that bounty while working on his thesis, but saying that the bounty was what made him become a Haiku contributer sounds to me a bit far fetched.
Regards,
Humdinger
Re: End of contract - closing words.
logically, perfectly possible and not far fetched, because it's not unlikely that he noticed the possibility of earning money while doing his thesis. I know many that combine the work with the thesis. I guess, it was even his main aim to find a project that brings him money, and being able to do the thesis and contribute to open source. So, it was at least a combination of factors, and the money could have been even decisive.
I was also planning to make a much better documentviewer, but seeing that just about 200 Euro was collected (and seeing it's stagnation) I came to the conclusion that my effort won't pay off, not even in the smallest part (then also haiku didn't run on my new laptop, so, using a vm would have implied even more lost time/effort). So the money plays indeed a role, and it might be even the decisive factor.
But in any case, if things are not working well, then you need to do a change, to try something new. And little bountries could be that change, to attract open source fans, who think: "I won't get paied like a normal programmer, but at least I get some money and I can contribute to open source). In any case something needs to happen, otherwise haiku is falling asleep. One needs a big goal, so that a big part of the community is motivated to contribute for that big goal, and some kind of safety to make the community believe that the money/effort will have the desired result. (for example paying just partly the bountry, not everything, till the target is not reached). If Haiku would run on ARM I would be considering starting a kickstarter project of bring an e-ink ebook reader to marked based on haiku.
I think Haiku needs now again a new goal (or to refresh the old one) that brings back the passion to the haiku-community.
Re: End of contract - closing words.
So, there's another supportive 'testimonial' from a bounty taker themselves, and evidence I'm not delusional (that even though the money collected wasn't attractive enough, which really wasn't much and other bounties had attracted a lot more, the idea is sane). Out of curiosity Cipri, what do you think would have been a suitable amount to continue your work and produce a stable final release?
I wasn't saying that the bounty was the sole factor they did the work, but I was alluding to the idea that it probably was seen as an attractive 'bonus', 'plus', 'cherry on top', and as Cipri said, may have a been a decisive factor in doing the coding or not. Do your masters for nothing, or have a $3000 bonus on top?... seems a no-brainer to me, but something Haiku developers/the board are so resistant to, it's unbelievable. The email shown was just to confirm that Haiku and Haikuware were both contacted at the exact same time to discuss the project, so I think anyone can deduce from this, both were a decisive factor in taking on the project itself. And I posted the wifi bounty application email because many have said Colin was long known to the community before applying to the bounty, which of course, was false.
As a director, this is some of the creative thinking I would come up with to further Haiku's development (but will actually do it just to see what the response is, and then approach the board if there's a positive outcome):
I will write the university where there's already a relation (where Colin studied), and in particular, the professor that oversaw Colin's master thesis. I will ask him if he would be interested in letting his master students know of potential projects that Haiku Inc. may propose to offer stipends of $5000/successful masters project (similar to what Colin did). These would be based on:
https://www.haiku-os.org/community/gsoc/2014/ideas
or a project the students propose and are mutually agreeable to by Haiku Inc., student, and professor.
It would also probably make sense to approach the University of Auckland and offer stipends for masters students:
We, Christof Lutteroth and Gerald Weber from the Department of Computer Science, would like to continue this tradition of utilizing Haiku as a platform for academic research. We are looking for potential postgraduate students and visitors to seize this opportunity to further their own education, as well as Haiku.
https://www.haiku-os.org/news/2011-11-19_university_auckland_scholarship_awarded_call_studentsvisitors\
In fact when another institution is looking for postgraduate students and offering money themselves (and has already produced results), this opportunity could be posted on Haiku's frontpage (similar to the GSOC module), one shouldn't have to dig into the news archives to find that information - and Haiku wonders why it can't find external developers...
Again, these ideas cost nothing to Haiku Inc. (by asking or making the projects available), and may potentially provide quality students and quality results (and could be attempted annually, not just once). As shown, student developers (and developers in general) may be more inclined to take on a project when there's a financial reward attached, as opposed to none at all... In Colin's case $2100, in Alex's $3700 (USD$), the concept works (even with meagre funds) and there's absolutely no reason in my mind for Haiku Inc. or anyone else to keep rejecting the concept and be ignorant to its proven success - if there are please speak up and discuss this, maybe I am missing something.
Re: End of contract - closing words.
To give an outside view, $25 000 is far less than I'd expect to hire a junior person into my team. And for someone capable and experienced enough to manage themselves, tackle hard problems rather than punt them to me to fix and so on I would expect to spend in excess of $70 000. And I'd expect to give them a benefits package that's pretty significant on top of that, company pension contributions, private medical insurance, long vacations, that sort of thing which Haiku Inc. can't efficiently offer.
Whilst obviously I'm not about to give Adrien unalloyed praise here after numerous criticisms elsewhere I think that Haiku Inc got amazing value at $25 000 for a year's work by someone competent to do what is ultimately not (for most people) terribly fulfilling work. I very much doubt that you can find someone competent to do the same work for less money as an actual job.
And although I do know people who love risk enough to take on a $25 000 all-or-nothing contract none of them would have these crazy ideas about the value of $25 000. They'd assume you intended a 4-8 week contract with maybe some little bits of sub-contracting, and they would eyeball the concept on that basis. If you said you expected a year's work for $25 000 on this risk shared basis they would laugh until they cried because that won't even pay their rent.
Re: End of contract - closing words.
Karl,
I see that:
1) You don't agree with the way Haiku Inc. spends the donation money in general (among which there is money from you, I gather).
2) Don't agree with the fact that Adrien has been working on the Webkit more specifically
3) Think the programming pace of Adrien was too slow.
About points 1 and 2: this is something you have to take up with Haiku Inc. And I'm not sure wether this forum is the best way to do it. Certainly the way you posts are formulated, you give the impression of a personal attack on Adrien, rather than have a disagreement with Haiku Inc.
About point 3: I know that you didn't literally state that, in your opinion, Adrien has been wasting a lot of time, but consider this. You made the total of all the money spent on Adriens contract and stated that if it was presented as a (big) bounty the work would have been done. That certainly implies that others could have done much better, and thus that Adrien has been 'lazy'. Hence the 'personal attack' thing I mentioned earlier.
If you have been following Adriens progress week, by week. If you have followed what he did outside his contract, if you see the progress that has been made in those area's he worked on, I don't see how you can say that (Adrien being too slow, that is).
If you see, by contrast, the amount of pay he received for it (1500 net per month), I think the money was well spend.
I, too, am often frustrated by the glacial pace of the progress of the Haiku OS. But with the contract of Adrien there was finally progress. Finally the big issues like package management en web support were tackled. We're so close much closer to the beta release now, that I don't feel like your pessimism is valid.
One more thing: putting the web kit as a single 25 000 bonus was never a real option. Adriens contract was initially for a few months. It has been steadily renewed. I think this is because the increased pace of development spurred extra donations. So it seems, to me, that his work was appreciated by those who donated.
Re: End of contract - closing words.
Thank you for the hard work
Re: End of contract - closing words.
I really want to say a big "Thank you", for all of your efforts and dedication to the Haiku project.
I'm only a final user, but from my point of view, the project had a big boost since Adrien started to participate. Not only in WebPositive, also in a lot of different fixes and general support in the BugTracker, and the forums).
I only had words of gratitude toward you.
My best wishes in your new work, and hope you still be around.
Thank you Adrien!
PD: sorry for my bad english! :-)
Re: End of contract - closing words.
Thanks for all your hard work for over a year, Adrien! Haiku is much less buggy and more polished in about every aspect thanks to you.
It's really too bad that the money finally ran out. I wonder if a donation rally could have eked out a few more months, if it had started early enough. Unfortunately the funds running low wasn't announced anywhere. A shame, because that could have brought us the Beta release "on time", whereas now it seems to have sunk back into the murky waters of uncertainty.
Anyway, I'll keep donating in the hopes we find another (or the same) excellent developer that can push us forwards. At least as long Haiku Inc. sticks to the contractor scheme. "Bounties" have obviously failed; why else has Haikuware closed shop so many years ago?
Anyway... I'm happy you found a job and hope you'll like it there. Even more I hope they don't tax you too much so you'll find time and motivation for Haiku! :)
Regards,
Humdinger
Re: End of contract - closing words.
Re: End of contract - closing words.
I will have to spend 3 hours per day in public transports, and I have a laptop. Let's see if that works :)
3hrs on public transit, yikes
Re: End of contract - closing words.
Re: End of contract - closing words.
Re: End of contract - closing words.
Thanks for all your hard work for over a year, Adrien! Haiku is much less buggy and more polished in about every aspect thanks to you.
It's really too bad that the money finally ran out. I wonder if a donation rally could have eked out a few more months, if it had started early enough. Unfortunately the funds running low wasn't announced anywhere. A shame, because that could have brought us the Beta release "on time", whereas now it seems to have sunk back into the murky waters of uncertainty.
Anyway, I'll keep donating in the hopes we find another (or the same) excellent developer that can push us forwards. At least as long Haiku Inc. sticks to the contractor scheme. "Bounties" have obviously failed; why else has Haikuware closed shop so many years ago?
Anyway... I'm happy you found a job and hope you'll like it there. Even more I hope they don't tax you too much so you'll find time and motivation for Haiku! :)
Regards,
Humdinger
I think it is time to knock on doors, big doors, who need a 503c, but it is going to be a tough sales pitch, unless there is something to offer in return.
Re: End of contract - closing words.
Anyway, I'll keep donating in the hopes we find another (or the same) excellent developer that can push us forwards. At least as long Haiku Inc. sticks to the contractor scheme. "Bounties" have obviously failed; why else has Haikuware closed shop so many years ago?
Hi Humdinger,
Haikuware hasn't closed at all - it continues to serve the community. More software, more unique Haiku software (i.e not common unix ports/dependencies) is uploaded there than made available by the package manager. The bounties haven't been continued because of #2 & 3 below; additionally, I believe Haiku Inc. with more people involved and more cash available could and should do the same and with more legitimacy/trust behind it (non-profit org; although I wouldn't say I was untrustworthy, I kept my word on all occasions). In this sense, Id hardly call the bounties a failure, many successful projects were completed. There were 8 successful bounties (in the 4-5 years they were active), how many successful contracts has Haiku Inc. had?
I might add, I haven't made a cent off any of the website, as a matter of fact, it's a continued expenditure that I gladly support (although I wish more would donate to help it). There are a couple of reasons, much like Haiku's own website, it hasn't been updated (besides maintenance and CMS updates) in >5 years.
1) The website works. people upload, people download (including Haiku developers who back stab and defame it).
2) I have some physically limiting medical issues that greatly restrict me from even the most simplest of daily activities that many wouldn't even cross their mind or take for granted.
3) Many have lost interest in Haiku (this is really evident in traffic/search statistics from 4-5 years ago as compared to today); myself included (i.e Haiku's declining popularity.. Haiku is an incomplete state - for years. Like updating Haiku's website, updating Haikuware would be a huge undertaking (for one man); the same for Haiku's website. It's probably the reason both haven't been updated in some time - they work, they serve their purpose, updating isn't worth it.
I am not complaining that Haiku is taking too long for a final release. I realize it's open source and develops at its own pace. I just didn't expect it to take this long for R1 to come out, as i am sure much of the community didn't either, and have moved on. I have and will continue to maintain the website, it's not going anywhere! And if/when R1 approaches (and I am still kicking), there'll be a nice new shiny interface awaiting for you to all look forward to ;)
Re: End of contract - closing words.
Thanks a lot for your efforts Adrien. Reading your blog entries and seeing Haiku improving on all areas has been a real pleasure to follow. Big respect for investing a big chunk for your life to make Haiku better for all of us. All the best for the future!
For people who are unhappy about certain goals not being met: STfU and code. Your pathces are your only statement, I'm eager to see them happen.
Thanks again Adrien!
Re: End of contract - closing words.
"For people who are unhappy about certain goals not being met: STfU and code. Your pathces are your only statement"
You do realize the ONLY reason this contract was made possible was because people opened their pocket books correct? So, you're incorrect. Patches are not the only statement.
Re: End of contract - closing words.
pitching in I think is a very nice thing indeed but only a substitute for the Real work which is the programming and research involved. Thru work i've seen a great number of outsourcing projects going to shit and all things considered I feel the cash Adrien received was well spent.
No work, no voice.
Re: End of contract - closing words.
Right, the voice is produced through work, which is funded. If you appreciate Adrien's work and are ignorant to the fact the money was at the base, then you have no argument. Without the funding there would be no voice.
As Adrien states:
"I'll do my best to keep ip up and running. I can't commit to anything (I don't know yet how the new job will be like)"
Funding, cash and contracts make the work predictable (to an extent, better with goal-oriented contracts) and the resultant 'voice' possible. Otherwise it's all flaky, I can't commit to this, that, and package management and web positive may be end-user ready through volunteer work in 10 years (when they're out-dated and there's something better available).
Re: End of contract - closing words.
Re: End of contract - closing words.
Dear Kvdman,
I have not donated money to the Haiku-OS project. So perhaps in your world, I am not allowed to comment. But hey, it's not your world. I have had the intention to for years by the way, I've just not had spare cash that often - things are looking up for me currently however, so I plan to make a few moderate donations during this year, that's by the by though.
I have followed this project since the very beginning of the Openbeos days with a keen interest. I played around with early beos on mac and pc, zeta was probably the last i tried successfully; Haiku alpha never installed for me :( and not that i didn't have time; but I don't really have an interest in running via virtual machines and so have held off waiting for stabler times to arrive - which I'm sure they will, I expect the Beta will install for starters! ;-) 3D accelaration I'm not holding out hope for yet but..mm. . Anyway - my point is not that I wish to claim a right to any opinion due to ostensibly having NOT been any active part of the Haiku community whatsoever - that would indeed be fanciful.
I'm here to politely remind you that if you're not Trolling - which is how you're coming across to me. But if you're not, you're at least off the mark. Personally I think you have stepped, many times even just in your last few comments, into the territory of Ad Hominem attack on Adrien. Personally I think the webmasters or mistresses should look into that and if they feel justified, ban you and block your IP. But maybe that's off the deep end(?) .
To my eyes, Adrien has been MORE than polite and gracious in even giving your plenty of air time on this really. He owes you nothing! As a donator you are but a patron would be to an artist. If you don't like the art work, it's a case of 'tough!' really ; draw your own picture, or sculpt your own WebPositive if you will. You don't get to demand a new art work I'm afraid. We get paid in this world for our Time! in the main. Your Bounty argument if fallacious ; if t'were a simpler task perhaps the analogy could be nearer that of having Tradesmen bidding to build your (non-leaking, picturesque yet highly functional) two-storey extension with conservatory out the back -- but I fear, and I didn't do any software programming since 94, so I'm WELL out of practice - but nevertheless I fear some tasks you're Ideally asking of Pulkomandy are nearer having one man build one city, or at least one suburb.. in less than 5 yrs; perhaps not realistic. Maybe you just need some perspective. or maybe to grow up. or maybe I'm wrong.
My first comment on here. I'll try not to comment again until I have something useful to say. Could be another 10 years. I suggest you do the same
Re: End of contract - closing words.
I for one donated to Haiku Inc so that this very thing could happen. A developer working on Haiku. So, just for the record I consider the money well spent on this contract. I am sad to see it end. Best of luck with your new job. I do hope you are able to continue with improving the web surfing experience in Haiku. It is already way above where it used to be. I would hate to see it regress.
David
Re: End of contract - closing words.
Re: End of contract - closing words.
Thanks for doing the contract work for so long! I really appreciate the many things in Haiku you fixed, enhanced and brought forward during that time.
Re: End of contract - closing words.
Where I put my signature to help?
Re: End of contract - closing words.
Great work! Will miss it!
Most I liked the weekly report about your work!
This is the part I will mostly miss if you stop writing reports anymore!
Maybe someone else could fill this gap... writing weekly reports!
For less money! Someone who knows about developement and Haiku history...
If there is a weekly report... Maybe peopel will more likely return to
the Haiku-Webpage every week?!
Many anoying bugs are not resolved in this year! I liked the way Adrian was
working! If there was a problem somewhere else... he tried to solve this as well...
Unfortunately some of the bugs remain unsolved...
(Network preferences are not stored until now!) for example...
The big blocker bugs are not touched for more than a year...
At least it seems so to me...
It would be great if the developers could try to fix this blocker bugs first...
And then showing/teaching users how to use Haiku.
Thanks Adrien...
Re: End of contract - closing words.
As for the "blocking bugs", I solved everything I could. It is not easy because I am not skilled and familiar with all parts of the system and fixing all the bugs alone would take me a lot of time to understand and learn the existing code. Axel started to work on the network preferences, so I think the situation there will improve further.
Re: End of contract - closing words. - Bounty vs. Contract
The end of the WebKit porting contract with unresolved deficiencies is a bit of a disappointment. In a number of comments, it is suggested that completion of the task would have been achieved had this been done through a bounty.
However, given that a freelance/mercenary programmer generally desires to minimize efforts deployed to earn the bounty, it is unlikely that there would have been all those "Weekly Reports" written to document the internals of WebKit and Haiku and how to approach the interface between the two.
I believe that such documentation will provide a solid basis not just for the completion of the port but also as a starting point for porting other major applications. The current annoyance with the remaining unresolved deficiencies may be just a small price to pay for the long term goal of making Haiku more appealing for daily use.
Re: End of contract - closing words.
Thanks for all the hard work you put into this project over the past year. I'm really going to miss the weekly reports which helped me to gain an understanding of the inner workings of Haiku and the issues that you were facing. At times I felt maybe you were spreading yourself a little thin by deviating from the browser development to fix other unrelated issues. However, if you had not, then those issues would probably still be unresolved. Either way, Haiku is better because of your efforts. Good luck with the new job.
Re: End of contract - closing words.
I've been donating every month for a long while, best of luck on your new job. There's only one commenter who came across as needing a smack across the face with a big wet fish for confusing appreciation & politics. Everyone else wishes you "All the Best !".
Re: End of contract - closing words.
This contact was a raging success from my point of view. The weekly updates were great and I enjoyed reading them. Haiku improved in many areas other than just Webkit.
Many bugs in Haiku were uncovered and fixed along the way, improvements were made, and the steady activity helped keep things vital and alive.
What Haiku really needs is some big donors or corporate sponsors so efforts like this one are not rare but the norm.
Keep up the good work.
Re: End of contract - closing words.
I just wanted to add my voice to the many others who say:
Thank you for your great work. Haiku is a much better system. A pitty that the contract didn't hold out long enough to get us to beta (and that is not your fault, that's just how it is).
I also wanted to add that, while I'm not a regular dweller on these blogs, I do follow the discussions in general. I must say that through it, I've seen you grow as a developer. Both in how you attack your work and in the way you do your communication in the different discussion threads.
This comforts me in the sense that apart from a small wage you clearly got a lot of experience out of this contract that allowed you to grow as a developer and as a professional. I'm sure this will help you advance your career in the way you want it to go.
Good luck and thanks for staying around to help in your free time!
Re: End of contract - closing words.
haiku works so much better as a result of your work. thank you! it's been a joy, also, reading up on your progress.
anyone hung up on the sate of webkit would have to ignore the improvements it necessitated in every other part of the operating system, and also the fact that the last few months of the project were no longer webkit development exclusively but coordination towards haiku beta. some crowds can't be pleased.
Re: End of contract - closing words.
I think, or better said, I guess, the more important problem could be, that haiku is loosing its attractivity. 5 Years ago I was unhappy with both windows and linux, but now in the meantime it seems that also the linux desktop is fine and we are also promised new developments (KDE 5.2, wayland, Mir, Material design/Quantum Paper for linux (Quantum Os) ) which makes it more exciting, while haiku has really not bad technology under the hood, visually it doesn't make advances, it's not bringing the excitement. I think it's loosing the cool-factor, and in this situation I guess it doesn't have big potential of gaining new fans (except keeping the die-hard-old-beos-fans, which are not that numerous). Karl has good points as does Adrien. I think in most cases a developer with experience will not risk (why to risk, especially if risk/gain is big? A developer with experience has enough other opportunities, but in fact there are also good programmers, that probably would work for a lot less, than the normal programmer).
As was discussed pro and contra outsourcing, I have the impression one should proceed like galileo, instead of making long philosophical discussion, just make a simple experiment. One can try first with small amounts like 500 dollar, to test how it works out.