LHC@home

Since many of the most important issues and problems cannot be adequately addressed by "ten little words" that comprise a "sound bite" this is again going to be one of Paul's long posts, but it is important, please bear with me, and read it all carefully ...



In the last couple of months we have had a number of incidents and issues, including, but not limited to:



* Possible use of BOINC or a science application as a "payload" or "stealth install".

* Revelation of personal/tracking information that a participant felt might be used in a manner inimical to that participant's safety.

* Possible cross-project "cyber-stalking" of one participant by another as indicated by harassment posts on more than one project's message boards.

* Uncorrected alteration of account information.

* Possible censorship of posts.

* Failure to agree on a centralized repository of BOINC documentation that has an official standing (though I still have hopes).

* Failure to notify participants of adverse actions.

* Failure to meet participant desire for a easy and trouble-free transition from other volunteer Distributed Computing projects/environments.

* Can you think of others?



These incidents and issues are indications that I feel may be symptomatic of the lack of a true "BOINC Community" identity. The last time I brought this up the reaction ranged from indifference to outright hostility. However, the concept of total project independence carries within itself the seeds of chaos and anarchy.



Worse there are few identified or usable avenues of relief for a participant if the issue is one where the situation is unrecognized (or unacknowledged), or if it is in fact caused by a project or a project team member, or is subject to multiple "jurisdictions" because it involves multiple projects and parties.



In that the BOINC Community as a whole is best served when we can maximize our "capture rate" and "retention rate" of participants, these issues are not unique and localized to the individual projects and so the model of complete and total project independence cannot hope to meet these challenges. In part this is because we are still very much independent and project-centric and not enough BOINC-centric enough.



The plain fact of the matter is that we do have a community, we just do not extend to it recognition. Third party statistic sites, team web sites, general BOINC web sites, the vast number of forums, the unofficial BOINC Wiki and the collaborative efforts there ... all this, and more, says we have a community.



To expand on a few of the issues above ...



1a) There have been two major migrations of participants from one DC environment to the BOINC Community and in both cases our success rate of participant capture was perhaps less than desirable. In the case of SETI@Home Classic the difficulties of the project to supply work was expected but because of these issues many of the participants simply quit the project blaming BOINC as the culprit and not accepting the fact that it was a temporary hiccup by one and only one project.



1b) For Rosetta@Home the experience was with the participants from the FaD project who had difficulties with the sizes of the downloads and the three bugs/restrictions (1% bug, over-time bug, keep in memory restriction). Again, these are, we hope, temporary issues. However, one of the more common reactions of the participant was a rejection of BOINC and a failure to avail themselves of other possible projects of a similar nature that also use the BOINC Software.



2) The cases of accused censorship have occurred on a number of projects, most recently and publicly (to my knowledge) on Rosetta@Home (sorry guys), but it has occurred with less notice on other projects, and this is perhaps the more worrisome. Stealth censorship, or even the perception of it, is perhaps more damaging to confidence in the moral fiber of the community. However, I will state categorically that from my observation, and in my opinion, the accusation(s) against Rosetta@Home's moderators are essentially without merit and the imposition of some semblance of, how shall I say it, a more genteel public environment was warranted (and for my part welcomed). As a disclaimer I will note that I too have been subject to moderator/project actions including removal of posts, a thread (I think it has only been one so far), and a rather severe "talking to" by a moderator. And, in all but one of these cases applied to me, agree that the actions were most likely appropriate.



3) The issue with identity release is possibly the most important of the issues in this day where there are numerous instances where a minor breach can lead to major consequences. However, the nature of this incident, and because by the time I was aware of the problem it was in the process of resolution I do not have a complete nor clear picture of the incident. However, I do know it involved at least two projects and quite a few participants. The point is that if this situation had not been recognized as the problem that it was by either one of the involved projects then the remedy secured on the other project(s) is/are essentially rendered meaningless.



An act, or failure to act, by one project within the community damages not only that project and its reputation but the community as a whole. The use of volunteer help has partially alleviated problems local to a single project but carries with it the minor tinge of not being "official" enough for some. And, the use of such moderators in no way addresses cross-project issues except in that many of the volunteer moderators on one project also serve on other projects (and actually do a remarkable job under what can and must be very trying circumstances).



The fact that the basic BOINC System is open source means that anyone can create a BOINC project and then pursue their own agenda which may be detrimental, benign, or positive. The standard EULA states that no project assumes a stance about any other project and that the ultimate responsibility is the participant's. This ultimate responsibility will always be on the participant. But, at this time the participant has to take it on faith that the project though not asserting its opinion of other projects it does not state that this is also true for itself.



Consider this, few if any, projects have published their guidelines for moderator actions, participant redress procedures, etc. or have created and maintained the ones published. To put it mildly, not only do the projects not express an opinion of the quality and safety of other projects but also do not do so for themselves! At this time, the participant has to take much on faith, the opinions of other participants (which may be biased or inaccurate), and is essentially without recourse if the project were to change character at a later stage of its history.



Perhaps even more importantly, if you follow the twin concepts of complete project independence along with no responsibility/liability for events within another project's domain to their logical conclusion you must accept the fact that this implies that activities within another project's domain cannot be used a part of the basis for an action on any other project. If the project maintains that the community does not exist you arrive at this same conclusion. The fact that most projects take a more common sense approach to this is gratifying, but then that logically means that there is at least a tacit acknowledgment that we are in fact a community.



The discussion of this issue at Rosetta@Home I think has been, to date, the most comprehensive and did also contain this gem. Not only is there obligations on the part of the project, but there are equally important and reciprocal obligations on the part of the participants! And we have failed to delineate these ...



In summary, we have no recognized community, no community standards, and no way of enforcing or measuring compliance with the community standards. To effect a change for the better there needs to be two shifts in the mindset of both the projects and the participants:



1) The projects are going to have to surrender a minor amount of sovereignty so that cross-project issues and project/participant conflicts can be fairly and independently adjudicated if they cannot be resolved within the project.

2) The participants are going to have to recognize that though they have certain rights these are not unlimited and that they must comply with a certain minimum level of behavior. And that improper behavior on one project may lead to sanctions on several projects.



Doom-sayer predictions aside, if we do form a formal BOINC Community life as we know it would almost certainly not change significantly. The basic right of a project to do as they please within the confines of their project would remain absolute. But, if they wish to be an acknowledged member of the community that they must allow for external mediation and agree to be bound by the results of the mediation. The same would hold true of the participants.



I do not envision a massive bureaucracy or huge intrusions, but we would have to:



1) More formally define the relationship between the participants and the project.

2) Create a way to mediate cross-project and project participant issues that are not solvable within a project.

3) Projects and participants would be accepted as members of the community and upon failure to comply with the standards of the community would be removed. Participants would have more of an assurance that they are "safe" by checking if a project was in "good standing" while the projects would likely most gain from a potentially larger pool of participants.



We should note that history has shown that the powerful accepting limits upon their use of that power and extending rights towards and protecting those without power have led to greatness. Unchecked power also has been shown by that same history to inevitably lead to tyranny.



I have an initial "go" at a list of rights and responsibilities. Included at the bottom, named "Participant Rights and Responsibilities" this could be the starting place. Initially cribbed off the US Bill of rights then heavily modified. I in no way assert that this is a complete, or even a good start, but it is a start. My suggestion is that we put this up for debate on the projects, collect and collate the best ideas through a couple rounds of debate and refinement. Projects that wish to be part of the formalized community would have to agree to abide by their side of the extended definitions (and I think the primary concession is only that of allowing external and neutral arbitration), and for them to select an initial governing board.



For that board I would suggest that the projects each have one primary and one alternate. These members would form the pool of mediators (a mediator from project "A" could not of course, mediate a dispute involving their project) and aside from this would maintain the list of projects in good standing and maintain and evolve the "Participant Rights and Responsibilities". Other activities I am sure would make themselves known in due time.



I think the biggest limitation my proposal suffers from is the fact that I have primarily a participant's perspective. As a "neutral" party and as a participant in many projects I have represented to participants to the projects, and the projects to the participants (usually earning the ire of the party on the "opposite" side in those discussions) over many issues and have tried to be fair and balanced in the initial draft. However, what else could you use from us, the participants?



One last caveat and warning. Other DC networks and projects have, through their actions, "contaminated" or damaged their reputation to the point where derivative projects still suffer from participant aversion. Are we to suffer the same fate? I would submit to you we are already part way down that path.



I do not have all of the answers, but I do have the willingness to seek them out. Does the rest of the you have the same will? In that regard, I would suggest that those participants that are active in other BOINC forums, such as team sites, copy this proposal. For those of dual language capability if there is interest please translate this and make the proposal known in that language also (SIMAP, boinc.dk, etc.). In addition, if you know someone that has left BOINC, encourage them to participate in the poll. If we participants cannot solicit a reasonable level of return then we shall have little room in the future to press a demand upon a project, and will well and truly deserve what treatment we become subject to.



Note: registration is required as is used to prevent stuffing the ballot box.



So, a way to express that will is to goto the CPDN forum and vote in these two threads:

* Participant's Rights and Responsibilities

* Participant's Rights and Responsibilities - Discussion

---

Hi Paul,

I like what you are trying to achieve here, and do not have any specific details that I disagree with.

I do wonder if the tone is pitched right. This reads like a contract or a constitution, and personally I think the same degree of safety for all can be achieved in a lighter mode: that of a set of guidelines or expected norms.

Most people do the decent thing when they understand what the decent thing is. Those who don't, will not be any more likely to do the decent thing due to the existence of a document of either type.

So I'd endosre every one of the points you make here, but with that change of emphasis. Human nature being as it is, with many people resisting legalistic frameworks and even rebelling against them simply because they look like rules, in my opinion guidelines are actually more likely to be effective.

I'd rather have this than nothing; it is a great step forward on the current situation where expectations are unclear. With poeple coming in from different cultures, the expectations need to be spelt out. I would prefer *even* *more* to have guidelines than rules.

Thanks for the work you put into this.
River~~

---