Haiku, Inc. accepts a second development contract in 2011!

News posted on Sun, 2011-06-12 16:44

On the Haiku, Inc. mailing list, Ingo Weinhold proposed a development contract. You may know Ingo better by his commit id "bonefish", as he recently merged his POSIX signals development branch back into trunk as r42116. His proposal is for 160 hours relating to furthering package management with a €2,000 EUR payout. That is roughly a meager $18 USD per hour!

These paid-contracts allow developers to dedicate large blocks of time, which normally would not be available for Haiku development. These contracts are not for purchasing an hour here, three hours there, nor for working on something that would not help reaching R1. The contracts put Haiku on the fast track to R1.

In his proposal, Ingo mentions "I believe package management is still the number one roadblock for R1". Here are some bullet points from his proposal:

  • Adjust the build system and tools to build and install the Haiku core system as a package.
  • Work on tools to build Haiku packages in a clean chroot environment.
  • Adjust build "recipes" for HaikuPorts packages as needed and build the packages. A tool to check the package against packaging guidelines might need to be implemented.
  • Extend packagefs to provide information about active packages to the tools that need it (e.g. the package solver).
  • Implement the package solver (as Oliver proposed based on libsatsolver).

And now for some related financial information. As you can see on Haiku, Inc.'s website, more than $5,000 USD has been donated this year. The financial balance is around $12,750 USD. Even after the successful completion of this contract, Haiku, Inc.'s funds are capable of supporting additional development contracts.

If you are interested in a contract for any work that relates to getting Haiku R1 completed, contact Haiku, Inc.. Lastly and most importantly, thank-you to all of Haiku's financial supporters!

Comments

Re: Haiku, Inc. accepts a second development contract in 2011!

That's more than great news. I'm also very happy about the recent signal stuff done by ingo. To bad, that it seems it's not included in alpha3 since I will need the pthread cancellation.

It would be great, to see stippi having also a contract related to the Deskbar rewrite.
http://haiku-os.org/blog/stippi/2011-04-07_deskbar_rewrite

Haiku in it's current state is already quite mature and useful, but i think it needs more modern look, and i think the desktop/deskbar need to become more practical and useful to use. I like stippies ideas and perhaps we could encourage him more. For a deskbar rewrite I would be more than happy to donate money.

Re: Haiku, Inc. accepts a second development contract in 2011!

Great! Being able to update Haiku core system from within Haiku will be a great enhancement when it comes to a system as heavily in development as Haiku is. Also great seeing it's one of the 'three wise ones' (Ingo, Axel, Stippi) who'll be doing this. Totally agree with ingo that this is needed for R1 to become a reality. Meanwhile I look forward upgrading to Alpha 3!

Re: Haiku, Inc. accepts a second development contract in 2011!

It would be nice to read about a summary from the first batch of money Haiku Inc. spent on the related contract, what was finished, and what still needs to be done (and if what was promised was done, or all the hours payed for were used), because the last sentence of the last blog entry regarding the last contract reads:

"That's it for now - I'll try to continue working on this stuff during the next weeks ..."

So to me, it sounds what was intended to be completed may not have been, or at least not reported. This would help the next developer.

Re: Haiku, Inc. accepts a second development contract in 2011!

Oliver did indeed use all his hours, but then continued working on the package management beyond the time he was paid for.

Ingo has apparently already spoken with Oliver regarding the progress he made, so I'm not sure there's a problem.

Re: Haiku, Inc. accepts a second development contract in 2011!

Thanks Urias. Regardless however, a conclusive report would be welcome (at least, myself as a donor to Haiku Inc., is something I expect - whether from the developer or Haiku Inc). As I said, to me, it sounded like the job was incomplete.

I really think this line should also be included under 'Development Contracts":

"Providing a written summary of how the funding helped to further the Haiku Project is required. This helps the community learn how Haiku is advancing as a direct result of donations. In turn, this helps to encourage people to donate additional funds with the confidence that it will be money well spent."

http://www.haiku-inc.org/request-funds.html#development-contracts

So, to me, I don't really see how the contract/work furthered Haiku. Can I type anything beyond installoptionalpackage x, and will this resolve dependencies? What did the $2000+ accomplish in layman's terms?? I know he did a lot of work, and I saw a lot of commits, but in the end, there's really nothing more Haiku can do now than before in terms of package management (is there? I don't know). As a result, to me as a donor, this is discouraging (as well as no report). Perhaps if I had something to read about what has been done, and how this will help Haiku, I would be more inclined to donate again.

With most bounties I could see end results and use them, i.e Wifi, Bluetooth, Printing, Gnash, an SATA driver, etc.

Re: Haiku, Inc. accepts a second development contract in 2011!

But, that's the thing - the contracts are not a promise to "complete" anything specific. It's just a payment for time spent. We only ask that progress be blogged (which doesn't often happen for regular development).

Oliver blogged his progress at the completion of hours that he was paid for, so it's hard to ask for more... he did then continue to work on things after his contract was up, but did not explicitly blog his further efforts.

If you feel Haiku didn't get it's money worth from that particular contract, you're certainly entitled to that opinion. Package Management isn't a small task, however, and Oliver only committed to doing research and proposing solutions. He left a summary here:

http://haiku-inc.org/funded-development.html

Keep in mind that much of the research and work he did was exposed on the mailing list, rather than on his blog.

Re: Haiku, Inc. accepts a second development contract in 2011!

Ok, I understand the contracts aren't a promise to complete anything. How efficient they are in obtaining results against bounties (for example) is a whole different argument, and I'm in the opinion they aren't.

I wasn't aware of that small summary. In any case, I don't think a one paragraph summary is sufficient. It doesn't really explain anything at all in regards to what was accomplished - how what was done furthered Haiku.

As a donor, I need to be convinced my contribution was worth it. One paragraph does nothing, and I'm not about to scour the mailing list looking into how what he did was beneficial to Haiku. I need something easy to read and find. Something that will make me feel warm & fuzzy inside and convince me to open my pocketbook again! If the contract didn't accomplish much, then let the record show that - I appreciate honesty too.

Re: Haiku, Inc. accepts a second development contract in 2011!

We can't promise you'll agree with everything Haiku, Inc. does with the funds.

You can go read the (completely bikeshedded) mailing list articles if you want to get a better idea of what Oliver did.

I wouldn't be surprised if he simply avoided blogging due to the high amount of negative feedback he received.

Development can't always produce "visible" results for end users - you have to take it on faith that a couple months of commits actually furthered the project rather than did nothing... that's how software development works.

Re: Haiku, Inc. accepts a second development contract in 2011!

I'm not arguing about how or what Haiku Inc. does with the funds or how software development works - although I did make my thoughts known and commented about it.

I'm asking Haiku to make things easy for donors and show how whatever was accomplished with a related contract helped Haiku. If the mailing list articles contain the whole story as to what was accomplished, then that should really be referenced in the one paragraph result. I'm just in the opinion that I, or anyone else as a donor, shouldn't really have to hunt on third party websites for information as to what was really accomplished. Perhaps more people will donate then.

Re: Haiku, Inc. accepts a second development contract in 2011!

Contrary to what has been mentioned, I personally consider a contract some kind of promise to deliver what has been agreed to be part of the contract. If you remind yourself what has been said in the announcement of my contract, you'll find that I made no promises (or even mentions) of actual coding. That contract was really all about getting the package-management work started and aimed for getting a clear picture how Haiku's package management system should look like.

During my work, a lot of discussion have taken place in my blog, and that discussion (and the respective companion discussion that took place on the mailing lists) are part of the result of the contract work.

I only started to write code because my contract wasn't over when I had a pretty good idea of what should be done (and continued to code for a while after my contract had finished). While I agree with you that my last, summarizing blog entry is a tad short, at the time, I was indeed in no mood to summarize the whole thing yet again, and preferred to continue coding instead. Looking at it now, I should have forced myself to write up that summary, never mind if it'd just fire up the discussion in the blog again.

Concerning whether or not we (Haiku Inc. and the developer taking a contract) should be honest about the contract not accomplishing much: it was obvious to me that people would expect contract work on package management to produce a (coded) result and I was sure not to be able to implement the whole system within one month. So I opted for starting the groundwork instead, something which I thought could be delivered in that timeframe. Sorry to hear that the result of my work has failed to meet your expectations, but since I never promised what you expected, I am not to blame for that.

Re: Haiku, Inc. accepts a second development contract in 2011!

Hi Oliver,

I'm sorry you think that I think your work failed to meet my expectations! This is not the case, and my comments don't reflect this. All I wanted was a detailed summary, and I couldn't find this. Everyone else posts their progress and summaries here at the Haiku website, and then your summary was at Haiku Inc's website. So I didn't think to look there. Then when I did see the summary, yes, this did fail to meet my expectations - not your work though! So, from what I read at your last blog entry, it seemed the work was incomplete. Basically then (code aside which was above and beyond the contract), from the money allocated, the $2000 produced three blog entries and a one paragraph summary. Forgive me for saying, but from those entries and the one paragraph summary, I'm still very unclear about how package management will be handled in Haiku and what route will be taken. So I hardly see how one paragraph can lead Haiku 'aimed for getting a clear picture how Haiku's package management system should look like.' The last blog entry didn't tie together anything in the summary, and doesn't leave me to any clear picture (IMHO). What are your suggestions in regards to package management within Haiku??

That aside, did any of your research focus or estimate what kind man hours are required to complete this project? It seems like this task is very large and is very desirable to be completed by the developers. Since this is the second time funds have been allocated towards package management, I as a donor would like to know how much funding is estimated/how many hours are required to complete package management.

Thanks

Re: Haiku, Inc. accepts a second development contract in 2011!

Yes, iirc Oliver's contract only covered working out and documenting how Haiku package management would work. Actual code was a bonus.

Re: Haiku, Inc. accepts a second development contract in 2011!

One aspect that you might be overlooking is that package management is not being developed in the main development line (trunk), but rather a development branch. This is is how major features are implemented -- ingo's signal branches, adrien+oliver's locale kit, the posix locale kit, are just a dew that come to mind. Major features like that simply need to be developed outside of trunk So, unless someone checks out that branch, builds, and install it , then they won't be able to see it running.

Basically, even though package management isn't in the nightly images nor the r1a3-release candidate images, it is still useful work that has been contributed to the project and brings us closer to the big R1.

Re: Haiku, Inc. accepts a second development contract in 2011!

Thanks Matt. I have no doubt that whatever was done was useful and a solid contribution.

I have overlooked your suggestion about branches, because I didn't know that.

I'd just like to read from the developer himself, how the work they did contributed to the project. If there's specific instructions to build, test, and use their work, then that would be appreciated too. I just don't see any of this in the little summary Urias pointed to. I have really no idea what was done (conclusively) and as a donor, would appreciate if it were all summarized and easy to find.

I would have never thought to look for a summary at Haiku Inc's website and am not sure the average donor would either.

Re: Haiku, Inc. accepts a second development contract in 2011!

Karl, have you actually read his last blog entry: http://www.haiku-os.org/blog/zooey/2011-02-01_package_management_present... ? It really looks like you don't even made an effort to follow all of what he has done (which was quite a lot for just a month of work), before complaining. And then even in so useless and hurtful comments like "those $2000 have resulted in three blog posts". Well, yeah. Not even one of the numbers is correct.

So, instead of spamming/trolling like this, please at least inform yourself next time.

Re: Haiku, Inc. accepts a second development contract in 2011!

Of course I've read the post. I'm not complaining, trolling or spamming. It's really regretful hearing the president of Haiku Inc. hurl insults when a donor questions what was actually completed and offers suggestions. Can't you reply to a comment without calling my posts useless? Perhaps you could be more diplomatic and say, Karl, I think you've misunderstood or haven't read things correctly (if that's in fact the case). How would you like if I said I found your work on the wireless encryption bounty useless? It really is a little too rich hearing these comments directed at someone that spends a lot of time and raises a lot of money to further Haiku.

I suggested that there should be an easy to read summary, because the last sentence in the blog entry sounded like whatever was being worked on was incomplete. The goal of the contract was to provide a clear idea what package management should look like in Haiku. So, is the last blog entry what Haiku package management should look like? I don't know, it doesn't say that anywhere.

""those $2000 have resulted in three blog posts". Well, yeah. Not even one of the numbers is correct."

The coding work was outside of the contract. So, there were three blog entries and a summary or am I missing something? The contract involved research and writing only. You're right though, at least one number was incorrect, the amount spent: $2622; sorry.

My point was even if there were more documentation (i.e on Freelists), it's scattered and disorganized. So to me, to provide a clear picture of what package management should look like, a conclusive detailed summary should have been provided. And I'm sorry if people are offended by what I thought wasn't a clear message, but there's no need to attack people.

Re: Haiku, Inc. accepts a second development contract in 2011!

I wouldn't put it in quite as harsh words, but I have to agree with Axel's notion here, when reading all the comments. Bascially at first you complain how Oliver's work was not documented, i.e. you couldn't tell what he actually did. When in fact, he posted rather many blog posts about his progress. Even disregarding all the mailing list discussions, you could get a rather clear picture of what Oliver was up to only reading his blog posts. Since the task was never to actually implement anything, let alone finish package management, of course there is no summary of what works now in Haiku because of the contract work. Later in your comments, you do shift your argument towards having expected a nice and easy to understand summary. Oliver said he should have written one. Your initial argument about lack of documentation of his progress is, however, completely unjustified (which is probably what rubbed Axel the wrong way).

Oliver's task was to make progress with the package management. He did go much beyond his initial contract (research) and made a lot of additional progress in form of code. Package management is just not simple if done properly. There is a lot to it and if you can't get a clear picture from his existing blog posts, it isn't really something you can blame him for, although as said, there could have been another summary. For Ingo however, or anyone else picking up where Oliver has stopped, all the information, thoughts and ideas are documented in the blog posts, mailing lists discussions and the code.

Even though you would appreciate an easy to follow summary, I hope you can agree that the progress was definitely accomplished, even properly documented, and your donation was well spend.

Re: Haiku, Inc. accepts a second development contract in 2011!

"Later in your comments, you do shift your argument towards having expected a nice and easy to understand summary. Oliver said he should have written one. Your initial argument about lack of documentation of his progress is, however, completely unjustified (which is probably what rubbed Axel the wrong way)."

By later in my comments, do you mean my very first sentence? "It would be nice to read about a summary from the first batch of money Haiku Inc. spent on the related contract"

Let's not distort things. I didn't shift my argument or even complain about a lack of documentation. I did say the documentation was scattered and disorganized. I also said the last blog entry looked like work was still being done, and then all of a sudden, you find (if you're told where to look) a one paragraph summary on Haiku Inc's website! So, I'm sorry, the work was to provide a clear picture of what package management should look like/how it should work; to me the summary didn't do this. From Oliver's last blog post, if that's how package management should look like, then say that in the summary and include all references. Often work changes, other avenues are taken. The one paragraph summary didn't really follow-up on the last blog to address final decisions/recommendations or leave me with a clear understanding.

I would like to agree that the money was well spent, but honestly I'm left with a very unclear picture, so I can't pass judgement. As a donor, I want something easy to read that sums up everything, something that will help me to consider donating again. I can't read code and don't want to (shouldn't have to) spend hours researching (that was the contractor's job) on the mailing list scattered threads about what was done - I would have thought (for $2622) this would be provided in an summarized easy to read, easy to find article; that's all. You have to look at this from a donor's perspective. The donors financed this work. Make it easy for donors to get an impression about what was conclusively done (with their money) so that they may consider donating again. I'm speaking as a donor, and I'm saying/giving Haiku the advice that I didn't think the research (the final result, the decisions, the recommendations) was laid out in a way that's favourable to digest (and even Oliver agrees with this).

Re: Haiku, Inc. accepts a second development contract in 2011!

Yes, reading again what you wrote, you did request a summary right from the start. Somehow, by reading all your posts and the replies in one go, the impression stuck that you as a donor were left clue-less about what his work was and how it furthered Haiku. Oliver did document his progress via blog posts, discussions in the comments and later the mailing lists, since the comments were going around in circles. Package management seems to be a touchy subject for a lot of people and many don't take the time to read all the thoughts that lead up to a certain solution and vision. I fully agree that it's hard to digest all this information and conclude what the vision for package management is after the contract work. Sorry for mistaking your argument that Oliver didn't document his work at all.

Re: Haiku, Inc. accepts a second development contract in 2011!

Thanks Stephan, sorry if I offended anyone.

Re: Haiku, Inc. accepts a second development contract in 2011!

Arguing about arguing is not at all productive. We have a R1 to release. You wanted a better summary for what you donated. Noted. Case closed?

Re: Haiku, Inc. accepts a second development contract in 2011!

Agreed, case closed. But, I don't like being attacked and insulted for saying it.

Re: Haiku, Inc. accepts a second development contract in 2011!

Taking a step back, as a small but regular financial supporter, I'm excited to see another development contract undertaken! Any steps taken to bring package management on Haiku closer to reality is huge.

Re: Haiku, Inc. accepts a second development contract in 2011!

Im looking forward to the items within Ingo's proposal - another well deserved contract.

Re: Haiku, Inc. accepts a second development contract in 2011!

It's great news that we will have further progress with the package solution for Haiku !!!!

This project with implementation of the earlier work done on research, will be great..

I don't want to bring up the old, long discussion earlier, but I agree that if we had more information, summarizing all the great work done on haiku development, more of people that are not heavily involved (like me) would more easily open up the wallet and donate money to Haiku.

It's sometimes hard to follow all progress done, because it's so spread in different fora.

I know this is not an easy task, and I am not saying that... I am just saying that if we could get more of it, there would probably be more money as well.

Re: Haiku, Inc. accepts a second development contract in 2011!

What a waste of money for such a useless feature

Re: Haiku, Inc. accepts a second development contract in 2011!

Really??..... And how much have you contributed to the project?

Re: Haiku, Inc. accepts a second development contract in 2011!

What kinda question is that. I guess you had no counter argument, so you try this route :p