Project information |
Message boards : Wish list : Project information
| Author | Message |
|---|---|
|
As you may have noticed, the project is understaffed at the moment, and, though we would love to implement all suggestions like a PS3 application and a GPU application, and fix problems with existing applications, we do not have the manpower to do this at this moment. | |
| ID: 9099 | Rating: 0 | rate:
| |
As you may have noticed, the project is understaffed at the moment, and, though we would love to implement all suggestions like a PS3 application and a GPU application, and fix problems with existing applications, we do not have the manpower to do this at this moment. One suggestion might be to ask for help from us! We could for instance point out, or even in some cases, show you how to make fixes, making your job as simple as checking it out and physically doing it. That way you don't have to both find the problem, figure out how to fix it and then fix it too. Users can be great, if you just ask. We can be a lazy needy bunch of bums too, but there are those among us that want to see projects get better! Glad to see you are here and working on the issues! THANK YOU!!!! ____________ ![]() | |
| ID: 9101 | Rating: 0 | rate:
| |
One suggestion might be to ask for help from us! We could for instance point out, or even in some cases, show you how to make fixes, making your job as simple as checking it out and physically doing it. That way you don't have to both find the problem, figure out how to fix it and then fix it too. Users can be great, if you just ask. We can be a lazy needy bunch of bums too, but there are those among us that want to see projects get better! Glad to see you are here and working on the issues! THANK YOU!!!! If you are talking about pointing out problems and show fixes in the source code, that's what I had in mind with making the application open source. You can't fix problems in the application unless you have the source. ____________ "Life is short and meaningless, unless you make the best of it." ![]() | |
| ID: 9102 | Rating: 0 | rate:
| |
|
The fpops_est for ABC is way too low. I know it will never be very accurate, that's all been discussed long ago, but the current estimates are way too low and it's causing scheduling problems for newcomers to this project (they download far too much ABC work and tasks go into hi-priority mode, deadlines get missed, etc.) until their DCF gets adjusted up. Could you fudge the estimate upwards in a few small increments, please? There was a discussion about this problem in this thread. | |
| ID: 9103 | Rating: 0 | rate:
| |
One suggestion might be to ask for help from us! We could for instance point out, or even in some cases, show you how to make fixes, making your job as simple as checking it out and physically doing it. That way you don't have to both find the problem, figure out how to fix it and then fix it too. Users can be great, if you just ask. We can be a lazy needy bunch of bums too, but there are those among us that want to see projects get better! Glad to see you are here and working on the issues! THANK YOU!!!! No I was thinking more of the things we as users see and don't seem to work. Or tweaks to make what we do see more user friendly. ____________ ![]() | |
| ID: 9107 | Rating: 0 | rate:
| |
Regarding making the source open.... bad idea. We all want more speed from the app but doing a proper job takes precedence. If the source gets out of the project's control then the app can send erroneous results. If 2 erroneous results from 2 third-party apps verify then the list of ABC triples will not be accurate. The danger is not that the list of triple will have "ABC triples" that are not actually ABC triples because, in the end, the project can run the list through a program and verify that each triple we have found is indeed an ABC triple. The danger is that some ABC triples will NOT be found. There is no other way to find them other than to crunch ALL the work units again. I was thinking we could disable the alternate platform mechanism, so you can only run BOINC with the official apps, and only release the source code so people can optimize/port it, test it and send it to us, but I'll bring this up at the meeting to see if this is possible. We are already validating the tasks by checking whether the ABC triples found are indeed ABC triples. ____________ "Life is short and meaningless, unless you make the best of it." ![]() | |
| ID: 9109 | Rating: 0 | rate:
| |
Regarding making the source open.... bad idea. We all want more speed from the app but doing a proper job takes precedence. If the source gets out of the project's control then the app can send erroneous results. If 2 erroneous results from 2 third-party apps verify then the list of ABC triples will not be accurate. The danger is not that the list of triple will have "ABC triples" that are not actually ABC triples because, in the end, the project can run the list through a program and verify that each triple we have found is indeed an ABC triple. The danger is that some ABC triples will NOT be found. There is no other way to find them other than to crunch ALL the work units again. I suspect somebody could make their app look like an official app. and only release the source code so people can optimize/port it, test it and send it to us, If you release it to the general public then you have no control over what they do with it. Though they might intend only to optimize or port, they might unintentionally change the algorithm too. If the algorithm changes then it might not find ABC triples where it should. Releasing it to just a few trusted people is different and not as "dangerous". We are already validating the tasks by checking whether the ABC triples found are indeed ABC triples. I thought you probably would be. It's the ABC triples that are missed that concern me. Those cannot be found unless all the work units are run again. | |
| ID: 9110 | Rating: 0 | rate:
| |
Message boards :
Wish list :
Project information