Message boards : Number crunching : efforts in reviving old users which dont get credits
Previous · 1 · 2 · 3 · Next
Author | Message |
---|---|
Feet1st Send message Joined: 30 Dec 05 Posts: 1755 Credit: 4,690,520 RAC: 0 |
I generally look to the number of ACTIVE hosts (the second chart) as the best indication of what changes are occurring. Even that is difficult to use, because a host has to be inactive for 60 days before it reduces the count. So, while a decline has a 60 day delay, an increase is reflected immediately. Add this signature to your EMail: Running Microsoft's "System Idle Process" will never help cure cancer, AIDS nor Alzheimer's. But running Rosetta@home just might! https://boinc.bakerlab.org/rosetta/ |
rochester new york Send message Joined: 2 Jul 06 Posts: 2842 Credit: 2,020,043 RAC: 0 |
which is more accurate the rosetta site or boinc? http://boincstats.com/stats/project_graph.php?pr=rosetta&view=users I generally look to the number of ACTIVE hosts (the second chart) as the best indication of what changes are occurring. |
Paul D. Buck Send message Joined: 17 Sep 05 Posts: 815 Credit: 1,812,737 RAC: 0 |
which is more accurate the rosetta site or boinc? http://boincstats.com/stats/project_graph.php?pr=rosetta&view=users Yes... :) It is the same data ... Just different aggregation periods and counting methods ... |
mikey Send message Joined: 5 Jan 06 Posts: 1895 Credit: 9,188,754 RAC: 3,104 |
i read through the thread from feet1st... are there any numbers? e.g. how many users came back? And I left, my machines simply won't download the new mini-rosetta app! So I guess between us the count remains steady. I WILL try again to download when the new app comes out, but even my 2 newest machines hang at the exact same spot in the download! One thing I may try is putting Linux on the machine and trying that way. I have finally figured out how to make the newest Ubuntu work for me and crunch Boinc projects. I JUST thought of that, so since I am off until Monday this weekend is the perfect time to try it. |
Jaykay Send message Joined: 13 Nov 08 Posts: 29 Credit: 1,743,205 RAC: 0 |
the second chart written here by feet1st is even better, 87% of all hosts are inactive ;) mikey: mike tyka had the same problem, it should be worked on :) back to topic... any ideas, feedback etc? |
Nothing But Idle Time Send message Joined: 28 Sep 05 Posts: 209 Credit: 139,545 RAC: 0 |
the second chart written here by feet1st is even better, 87% of all hosts are inactive This 87% number signifies (IMO) that contributors had bad experiences with Rosetta apps and would not tolerate the persistent error rate. Once they leave there is good chance they won't return. People do love this project's goals, but many have a low tolerance for failed tasks. That's because much of one's contribution is wasted and, further, failing to correct the persistent error rate indicates a lack of appreciation for one's contribution. I do believe many people share this view. |
Jaykay Send message Joined: 13 Nov 08 Posts: 29 Credit: 1,743,205 RAC: 0 |
@nothing but idle time: i dont know whether you read the news on R@H, but 1.54 is quite stable and has a low error rate... |
robertmiles Send message Joined: 16 Jun 08 Posts: 1233 Credit: 14,284,221 RAC: 995 |
@nothing but idle time: i dont know whether you read the news on R@H, but 1.54 is quite stable and has a low error rate... Which R@h? RALPH@home is already up to testing version 1.58, in order to lower the error rate even more, even though their home page doesn't make that obvious. Rosetta@home still uses 1.54, though, and the low error rate isn't for everyone, especially those who have set BOINC to use less than 100% of the available CPU time. |
Nothing But Idle Time Send message Joined: 28 Sep 05 Posts: 209 Credit: 139,545 RAC: 0 |
@nothing but idle time: i dont know whether you read the news on R@H, but 1.54 is quite stable and has a low error rate... Well, I don't know whether 1.54 has a low error rate or not, but it still has errors. Otherwise why do we have a thread to report errors for every single version of the app that is produced? We didn't get to Mini version 1.54 or to Rosetta version 5.98 over night; there is a long history that leads up to the 87% inactive hosts. And no doubt there will be v1.55, etc, all with errors of some kind. All I'm saying is that dedicated Rosetta crunchers have a high tolerance for errors while the inactive users probably do not. |
robertmiles Send message Joined: 16 Jun 08 Posts: 1233 Credit: 14,284,221 RAC: 995 |
@nothing but idle time: i dont know whether you read the news on R@H, but 1.54 is quite stable and has a low error rate... And the RALPH@home crunchers who try to filter out those errors before you see them must have an even higher tolerance for errors, especially if they do it on a machine with much else of importance on it. |
Jaykay Send message Joined: 13 Nov 08 Posts: 29 Credit: 1,743,205 RAC: 0 |
nothing but idle time: you are one of those who are against everything arent you? i dont know really but i guess other projects are also not 100% error-free... robertmiles: i meant rosetta, i know ralph has newer versions. but i think its a fact that the error rate i the lowest for.. i dont know.. some months? many months? could we please come back to topic? please dont discuss whether 1.54 is quite error-free or not, but whether the emailcampaign is a good idea. |
Mod.Sense Volunteer moderator Send message Joined: 22 Aug 06 Posts: 4018 Credit: 0 RAC: 0 |
...it still has errors. Otherwise why do we have a thread to report errors for every single version of the app that is produced? So, creating a central place for people to report and discuss issues with each version of the application is proof of something? Yes, it is proof that the Project Team wants to know if problems exist so they can address them. There are 81 posts to the v1.54 thread. Now count how many actually report a problem, and consider how nearly a quarter of a million results were completed just in the last 24hrs. All the charts of inactive users can really be said to prove is that new host IDs are created, and old host IDs become inactive. There are many reasons that host IDs change on BOINC client machines. And the old ID becomes "inactive" and will never become active again. Regardless of the feelings of the person that owns that machine, and regardless of whether that machine is still crunching for the project or not. Rosetta Moderator: Mod.Sense |
robertmiles Send message Joined: 16 Jun 08 Posts: 1233 Credit: 14,284,221 RAC: 995 |
All the charts of inactive users can really be said to prove is that new host IDs are created, and old host IDs become inactive. There are many reasons that host IDs change on BOINC client machines. And the old ID becomes "inactive" and will never become active again. Regardless of the feelings of the person that owns that machine, and regardless of whether that machine is still crunching for the project or not. Note that you appear to be explaining why old host IDs become inactive, but not why users become inactive. I've used private messages to bring a few people back, but mainly those who are already members of my team. |
Mod.Sense Volunteer moderator Send message Joined: 22 Aug 06 Posts: 4018 Credit: 0 RAC: 0 |
Note that you appear to be explaining why old host IDs become inactive, but not why users become inactive. Noted... but then how many EMail addresses have you let expire due to spam or a switch in ISPs? And I suspect that many people just create a new user account with their new EMail address, not necessarily even aware that the existing account can be changed, and go on from there. Other users have multiple accounts, and participate on multiple teams. So, it is very difficult to gleen much concrete information from the available data. Hard to disagree that there are "old users" that could use "reviving". But impossible to know how many really exist. As has been pointed out, extensive efforts have already been attempted, and the results were barely measurable. And, as with all of the above, impossible to directly attribute to the effort. And so I would encourage you to pursue the other side of the coin, which would be efforts to recruit new users. There are numerous approaches to this idea. And very few of them have been attempted yet. One idea that surfaces over and over is to get publicity. But after the many printed articles and the huge splash, including being slash dotted, by Fold.it didn't have any immediately decernable impact on Rosetta@home either. And so, I feel a more direct word-of-mouth campaign is the untapped approach. But how do you motivate people to implement it? Rosetta Moderator: Mod.Sense |
Nothing But Idle Time Send message Joined: 28 Sep 05 Posts: 209 Credit: 139,545 RAC: 0 |
nothing but idle time: you are one of those who are against everything arent you? i dont know really but i guess other projects are also not 100% error-free... Please get your facts straight. I never said anywhere in this thread that I was against something. In this case you refer to the effort to "revive" inactive hosts/users. I merely opined on why users become inactive and said previously that your efforts would best be served by trying to get new users and not try to coax back into service folks who have decided not to participate. So in that respect Mod.Sense agrees with me. I have nothing further to say on this topic, but feel free to get in the last word, most people do. |
mikey Send message Joined: 5 Jan 06 Posts: 1895 Credit: 9,188,754 RAC: 3,104 |
One idea that surfaces over and over is to get publicity. But after the many printed articles and the huge splash, including being slash dotted, by Fold.it didn't have any immediately decernable impact on Rosetta@home either. And so, I feel a more direct word-of-mouth campaign is the untapped approach. But how do you motivate people to implement it? One idea would be to bring it up at Conferences etc. Bring it up in talks, in meetings etc. Basically everywhere people of Science get together and talk about stuff. This would involve the Scientists not only talking ABOUT what they do but ACTIVELY requesting assistance in WHAT they do. If we could reach out to the Teachers, Scientists in Schools not only colleges but all levels, they could reach a vast multitude of people that have no idea about Rosetta, or some of the other really good projects! When the scientists write a paper do they say HOW they got the info, or just the results? Usually the how is discussed but not the nitty-gritty details. IF the paper said, for example, this breakthru was conceived because of a result from the Rosetta Boinc Project that gave us the idea to do "whatever". Obviously the "whatever" is the reason for the paper. Then other scientists and chemists and "men of science" would know and begin to realize the possibilities of Distributed Computing and how it could be used. Yes some people speak of it now, but go into a conference and ask how many people have heard of Boinc and you could probably count the hands raised on one hand!! Start with the people at the top, there are much fewer of them, then let them encourage the students to go forth and make a difference in the World and support Science. THEN you will see a hundred, heck maybe even a million, fold increase in Distributed Computing participation. YES, ABSOLUTELY I tell people what my farm of computers is doing, but they have no scientific interest in it. They get a 2 minute speel on the projects I contribute to, and then they move on. We need to get people BEFORE they get hardened in their beliefs that they cannot change the World. As a kid we ALL believe we CAN change the World and make things better for everyone else! As we get older we get more jaded in those beliefs. With Distributed Computing we have that belief again!! At least I hope we are making a difference!! People in the US just voted for a black man, Barack Obama, to be President! That is a vote for a change to a new way of doing things and away from the old establishment way of doing things. Well the first thing that comes down the pike...Democrat Congressmen loading the bills with perks and pork and anything else they can! Why...because THEY haven't changed, just the President has changed. Can anything be done about it, not until enough people stand up and vote the people that do that kind of stuff out of office! Will that happen, not until people THINK they can make a difference! The whole point is that people get jaded in their beliefs and to change them is really very hard. The status quo is not scary, change can be. Distributed Computing is asking for their help in solving some of lifes serious problems, that is change! If we could change that thought process, we could lick any of lifes problems! Did you see where Bill Gates released a bunch of mosquitoes at a conference? He was talking about the problems of Malaria in Third World Countries. He later told them the mosquitoes he released did not carry Malaria. In the story I did not see anything mentioned about http://www.malariacontrol.net and their fight against Malaria. That is the kind of thing I am talking about, get the word out that there ARE things people are doing and how others can help!! |
Jaykay Send message Joined: 13 Nov 08 Posts: 29 Credit: 1,743,205 RAC: 0 |
@mod.sense: thanks, your posts helped me a lot. mikey: ES, ABSOLUTELY I tell people what my farm of computers is doing, but they have no scientific interest in it. They get a 2 minute speel on the projects I contribute to, and then they move on. i can second that, only without the farm, i only have 2 computers and only one which is crunching :) i also read a article about that conference. it was the only moment i liked bill gates, the mosquitoes were a great idea :) but youre right, nothing was mentioned about malariacontrol. i know this is getting annoying, but another question: would it make sense to send only mails to users which have lets say more than 5000 credits? this would eliminate those who tried rosetta out and just got tired of their running fans or whatever. also those who have a high rac in other projects could be eliminated, most of them had also a good reason to leave rosetta. the remaining users maybe just stopped rosetta e.g. because of high failure rates, but in general they may be interested in distributed computing (because they ran rosetta for a while) i hope you dont get tired of this discussion, i liked your feedback very much :) |
rochester new york Send message Joined: 2 Jul 06 Posts: 2842 Credit: 2,020,043 RAC: 0 |
too bad he did not mention rosetta@home the project could use the publicity @mod.sense: thanks, your posts helped me a lot. |
robertmiles Send message Joined: 16 Jun 08 Posts: 1233 Credit: 14,284,221 RAC: 995 |
too bad he did not mention rosetta@home the project could use the publicity Especially since Rosetta@home is still more active than malariacontrol; I haven't seen any workunits available on malariacontrol for the last month. However, it's hard to persuade people to participate in any BOINC project that's both memory-hungry and has a high error rate. I've tried, though. |
mikey Send message Joined: 5 Jan 06 Posts: 1895 Credit: 9,188,754 RAC: 3,104 |
too bad he did not mention rosetta@home the project could use the publicity That was my point...he didn't mention ANY project! It was his chance to ask people to donate their spare cycles towards a program he obviously feels passionate about! AND he wrote computer programs!!!! Money can go along way towards solving some of the Worlds issues, but so can our computers! And that is not near as much cost to each one of us!! I wish I had a working email address, that he actually reads, so we could advise him on this. |
Message boards :
Number crunching :
efforts in reviving old users which dont get credits
©2024 University of Washington
https://www.bakerlab.org