OK, so here we go.........
Seeing Berkeley didn't bother to tell users that the main project URL had changed, and you need to enter an entirely different one now. What happens to already completed WU's that are sitting in your cache unsent, when you detach from the project and reconnect to the new URL - from the looks of it they are deleted. I have hundreds sitting on various machines and if this is the case IT IS TOTALLY UNACCEPTABLE.

I read the posts on here occasionally and am amazed at the attitude of a lot of the posters. ie.....
well done Berkeley
smooth upgrade
job well done
blah, blah

well I'm sorry but it hasn't been up until now. The software is finicky and unfinished, the servers unreliable and the whole new Boinc thing has already put off a whole load of crunchers with already mentioned crazy points systems etc
I've been looking forward to Boinc/Seti2 for a long time now, as a lot of us have, and with the Beta/Alpha taking forever as it did I expected the project when made live to be well finished and professional. ITS NOT!
From the looks of it even the servers weren't specced correctly, and couldn't take the load.
In any commercial concern if a project was run like this there would be SERIOUS questions asked of everyone involved.
Before I get flamed and shot down for critising, remember we all donate our computer time to this project ie electricity costs etc, and therefor are shareholders in this - so if we have serious concerns they should be dealt with immediately.
I'm SERIOUSLY dissapointed with all of this

I have no problem with what you are saying 888 & pretty much agree with everything you are saying. I've tried to keep my criticism to a minimum but it's hard at times. That URL change had me screwed up for a couple of days also.

It seems to me as each week goes by things are getting worse not better. More options are turned off & more people complaing about getting WU's and getting Credit. People are dumping work Units by the Boatload because they can't figure out whats going on. It's not very pretty at the moment and who knows when it will get better. :/

JOIN TEAM

Ok, here I go again. I have continued to use the same project URL since BOINC went public with no problem. They changed the schedulers once but in most cases after a few failures that autocorrected. Others learned the hardway that if you detach/reattach you lose all of the workunits, completed and ready to send or not yet touched. Fair enough. That's how people learn, and problems (potential or otherwise) get discovered and fixed. You shouldn't need to detach/reattach to get your units sent. But even so, in these early stages of BOINC, it is probably a good idea not to get too far ahead in case changes are made to lighten server load, fix bugs or whatever causing some kind of change like that. And should something that drastic happen, I should expect UCB to communicate it fairly quickly.

Don't be amazed at my attitude of acceptance and praise for UCB. I applaud them as much for the success as for the failures. And that is because thus far, they have been prompt at fixing problems once the arise. Problems that otherwise would have been difficult to find without a full load on the system. The software is relatively sound. It has it's problems, but the BASIC functionality works. Would I have liked a better product? Sure. But it's free software, from a nonprofit scientific organization with limited assets. It works. Not perfectly, but it still works. You obviously don't have much experience dealing with lower budget software projects, or even higher budget ones for that matter. You are making the same irrational complaints as have already been made before. And they still don't hold any ground. The problems encountered could not really be accurately guaged or found except under a real live action load designed to stress every aspect of the system simultaneously and in the same manner as a release.

As far as being a shareholder? I use between 3 and 5 computers at any given time (sometimes a bit more for limited periods of time). I essentially dedicate my BEST computer to processing SETI, pushing other processor intensive applications to a computer that has less than HALF of the processing ability, sometimes spending days or even upwards of a week on them just so SETI can get the most from me. I overclocked and overvolted recently for another project and elected to leave it there with a house fan focused on it just to keep the thing cool so that SETI could take advantage of it. So now it's using more electricity, creating a lot more heat (very noticeable to the temperature of my room, and noise from that stupid fan. But yet I do not consider myself a "shareholder" or "entitled" to anything more than we have recieved. It isn't my project. It is UCB's. They choose to handle it in whatever way they see fit and best, and in my opinion, they have handled things fairly well and been much more communicative than I've ever seen. I consider the personal costs of running SETI minimal when compared with the exhorbinate prices of many software programs that function very poorly and are produced by multibillion dollar companies (MS Windows Me, Windows 98 for many, even the super stable WinXP with all of its security problems).

I'm not saying there aren't problems that need to be fixed. I'm saying don't complain that there are problems. Make the problems known so they can be fixed and try to be nice and understanding about it. The folks at UCB have been putting in a lot of time on this and need all of the help they can get.

And with that, I am SERIOUSLY disappointed with the incredible lack of open-mindedness of people donating computer time to a project designed to find extraterrestrial life, something that I thought most closed-minded people would scoff at.

Chris

> OK, so here we go.........
> Seeing Berkeley didn't bother to tell users that the main project URL had
> changed, and you need to enter an entirely different one now. What happens to
> already completed WU's that are sitting in your cache unsent, when you detach
> from the project and reconnect to the new URL - from the looks of it they are
> deleted. I have hundreds sitting on various machines and if this is the case
> IT IS TOTALLY UNACCEPTABLE.
>
> I read the posts on here occasionally and am amazed at the attitude of a lot
> of the posters. ie.....
> well done Berkeley
> smooth upgrade
> job well done
> blah, blah
>
> well I'm sorry but it hasn't been up until now. The software is finicky and
> unfinished, the servers unreliable and the whole new Boinc thing has already
> put off a whole load of crunchers with already mentioned crazy points systems
> etc
> I've been looking forward to Boinc/Seti2 for a long time now, as a lot of us
> have, and with the Beta/Alpha taking forever as it did I expected the project
> when made live to be well finished and professional. ITS NOT!
> From the looks of it even the servers weren't specced correctly, and couldn't
> take the load.
> In any commercial concern if a project was run like this there would be
> SERIOUS questions asked of everyone involved.
> Before I get flamed and shot down for critising, remember we all donate our
> computer time to this project ie electricity costs etc, and therefor are
> shareholders in this - so if we have serious concerns they should be dealt
> with immediately.
> I'm SERIOUSLY dissapointed with all of this
>
>

I'm saying don't complain that there are problems.
==========

Sure Chris, I think I'll go outside and smell the Roses everythings fine and dandy in here I guess ... :/

This is a volunteer outfit not a commerical business, and they can drop the switch and shut it down fast. I can not believe the way people complain about something they are not paying for, if it so bad why do you stay, why not go somewhere else where your time will be better spent....Shareholders!! now that is funny.

Bah!!! I am in it for the money

Bah!!! I am in it for the money
==========

:O When do they send out the first Paycheck ...

I don't think everything is necessarily all "fine and dandy." There is work to be done. Constructive criticism is much better than moaning about how bad things are or that nothing works. What's bad? What doesn't work? Those are the questions we should be answering for the Dev Team. You can complain and say how much things suck until your tongue fingers are bruised from the keyboard and NOTHING will change. But when people calm down, explain your problem clearly, and stop complaining about how nothing is going right, others can start helping them or explain what is going on while the Dev Team can go find the error and fix the problem there if possible, and communicate to us the cause/status/solution of the problem and any other relavent information.

I don't see why we should jump down Berkeley's throat about this. It's pointless and doesn't solve anything.

Chris

> I'm saying don't complain that there are problems.
> ==========
>
> Sure Chris, I think I'll go outside and smell the Roses everythings fine and
> dandy in here I guess ... :/
>

> OK, so here we go.........
> Seeing Berkeley didn't bother to tell users that the main project URL had
> changed, and you need to enter an entirely different one now. What happens to
> already completed WU's that are sitting in your cache unsent, when you detach
> from the project and reconnect to the new URL - from the looks of it they are
> deleted. I have hundreds sitting on various machines and if this is the case
> IT IS TOTALLY UNACCEPTABLE.


I also ran in to the problem where when the URL was changed my machines could no longer contact the schedulers - so they were unable to update to the new URL. It is possible to edit some of your files and directory names to get to the new URL without losing your WU's. Below is the text from a previous post of mine on the subject - hope it helps :

> Does anyone know what needs to be done in the XML files to "force" the BOINC
> client to look at the new location? Is this even possible?
>
> JHutch
>

Yes you can, I did it on several of my computers. I tried with one and didn't have any luck, so I detached
and re-attached that computer and then did a comparison of the files on that computer to the other computers
and edited them accordingly. I should have written everything down, but I didn't - so here's what I found
as far as I can remember :

- shut down Boinc
- back up your BOINC directory first (of course) so you can recover if need be
- rename "account_setiboinc.ssl.berkeley.edu.xml" to "account_setiathome.berkeley.edu.xml"
- in the "projects" directory, rename the directory "setiboinc.ssl.berkeley.edu_sah" to "setiathome.berkeley.edu"

- edit the following files, replacing "http://setiboinc.ssl.berkeley.edu/sah/" with "http://setiathome.berkeley.edu/"
these addresses are usually associated with the tag "master_url" :

in the main Boinc directory:
- "account_setiathome.berkeley.edu.xml"
- "client_state.xml"
- "sched_request.xml"

in each of the "slots" directories (if you have more than one):
- "init_data.xml" needs to have "C:/Program Files/BOINC/projects/setiboinc.ssl.berkeley.edu_sah" replaced with
"C:/Program Files/BOINC/projects/setiathome.berkeley.edu" to let it know of the name change you did to the directory
- "result.sah"
- "setiathome_3.08_windows_intelx86.pdb"
- "work_unit.sah"

I also deleted the files "stderr.old" and "stdout.old", and edited "stderr.txt" and "stdout.txt" to remove all data
from them so that when I re-started Boinc I knew it would start off without any log data, just to be sure.

Fire Boinc up and do a project update, it should work unless I forgot something.

I could go for a dollar per cobblestone. ;)

Then maybe I could get some of my debt paid off and pay for school.. :) And then I could buy new computer equipment to get more cobblestones even faster ... ahh I better stop before I set myself up for too much disappointment :)

> Bah!!! I am in it for the money
> ==========
>
> :O When do they send out the first Paycheck ...
>
>

This is kind of interesting. The instructions for setting up BOINC said to use http://setiathome.berkeley.edu when it went public. And that is what people have always been told to use on here when they've had problems. I didn't realize anyone was actually using another URL or that another URL was even working. But changing the server in the top of the XML file definately seems the way to go in order to not lose workunits. I also know that if you flat out delete the scheduling server (AND its identifier tags) it will automatically update with the main URL to retrieve the location of the scheduler. If you don't delete those scheduler identifier tags though it assumes the url of the server is well.. nothing and doesn't work.

Chris


> I also ran in to the problem where when the URL was changed my machines could
> no longer contact the schedulers - so they were unable to update to the new
> URL. It is possible to edit some of your files and directory names to get to
> the new URL without losing your WU's. Below is the text from a previous post
> of mine on the subject - hope it helps :
>
>
> Yes you can, I did it on several of my computers. I tried with one and didn't
> have any luck, so I detached
> and re-attached that computer and then did a comparison of the files on that
> computer to the other computers
> and edited them accordingly. I should have written everything down, but I
> didn't - so here's what I found
> as far as I can remember :
>
> - shut down Boinc
> - back up your BOINC directory first (of course) so you can recover if need
> be
> - rename "account_setiboinc.ssl.berkeley.edu.xml" to
> "account_setiathome.berkeley.edu.xml"
> - in the "projects" directory, rename the directory
> "setiboinc.ssl.berkeley.edu_sah" to "setiathome.berkeley.edu"
>
> - edit the following files, replacing "http://setiboinc.ssl.berkeley.edu/sah/"
> with "http://setiathome.berkeley.edu/"
> these addresses are usually associated with the tag "master_url" :
>
> in the main Boinc directory:
> - "account_setiathome.berkeley.edu.xml"
> - "client_state.xml"
> - "sched_request.xml"
>
> in each of the "slots" directories (if you have more than one):
> - "init_data.xml" needs to have "C:/Program
> Files/BOINC/projects/setiboinc.ssl.berkeley.edu_sah" replaced with
> "C:/Program Files/BOINC/projects/setiathome.berkeley.edu" to let it know of
> the name change you did to the directory
> - "result.sah"
> - "setiathome_3.08_windows_intelx86.pdb"
> - "work_unit.sah"
>
> I also deleted the files "stderr.old" and "stdout.old", and edited
> "stderr.txt" and "stdout.txt" to remove all data
> from them so that when I re-started Boinc I knew it would start off without
> any log data, just to be sure.
>
> Fire Boinc up and do a project update, it should work unless I forgot
> something.
>

Guys,

let's not forget that SAH and BOINC is first and foremost a *research project*. It is not an implementation of a commercially available, off-the-shelf product. A university environment is not the same as a commercial software development shop where you have experts for most things. I've worked in both environments. I was a scientist at Carnegie Mellon for over ten years and now work in private industry. In a university environment you often have folks doing the jobs of three or more people for very little compensation because funding simply isn't enough to provide adequate manpower or expertise. You often can't just "order new servers" when you find that your current one no longer meets requirements. You have to make do with what you have and in the absence of unlimited funding learn to be creative about it. It is this creativity and learning to make do with what you have that often leads to new discoveries. Let's also not forget that in science failures can also lead to new discoveries.

If the Berkeley folks are anything like the dedicated people I used to work with at Carnegie Mellon, they are doing everything they can and more to make this project a success. Instead of complaining about things we should be congratulating them on doing a great job under difficult circumstances and supporting them when things aren't going as well as expected.

I, for one, am very excited at the opportunity to be a part of this project and am more than happy to endure the ups and downs. After all, the goal is more than worth the effort.




Olli
Geek und Kuhtreiber

http://setiboinc.ssl.berkeley.edu/sah/ is the URL I am connecting to right now, and it's the URL address the BOINC GUI says I'm connecting to, it's where I'm getting my work from also ... We shouldn't ever have to go through what bfarrant is saying to do.

I am probably capable of doing that, but why should I have to to reconnect with SETI, and how many people are out there that if you gave them 10 years to figure it out they still wouldn't be able to do it.

> http://setiboinc.ssl.berkeley.edu/sah/ is the URL I am connecting to right
> now, and it's the URL address the BOINC GUI says I'm connecting to, it's where
> I'm getting my work from also ... We shouldn't ever have to go through what
> bfarrant is saying to do.
>
> I am probably capable of doing that, but why should I have to to reconnect
> with SETI, and how many people are out there that if you gave them 10 years to
> figure it out they still wouldn't be able to do it.
>
>
>
The correct URL to use when attaching to the SETI project is "http://setiathome.berkeley.edu". That has been the URL to use since SETI BOINC went public. Once you are attached the GUI connects to "http://setiboinc.ssl.berkeley.edu/sah/".

You should read what both Bill and I said. If you followed the instructions for setting BOINC up then you shouldn't need to worry about changing stuff around. The only real thing you should have to change though is the server in the XML (delete the scheduler) and rename the server name folder to the correct server name. Or perhaps if you are able finish your workunits, turn them in, and then detach reattach with the correct one. These and the instructions he gave are simply for retaining workunits.

Bill: For the sake of accuracy, the server that BOINC connects to (the scheduler) is actually http://setiboincdata.ssl.berkeley.edu/say_cgi/cgi :)

Chris

> http://setiboinc.ssl.berkeley.edu/sah/ is the URL I am connecting to right
> now, and it's the URL address the BOINC GUI says I'm connecting to, it's where
> I'm getting my work from also ... We shouldn't ever have to go through what
> bfarrant is saying to do.
>
> I am probably capable of doing that, but why should I have to to reconnect
> with SETI, and how many people are out there that if you gave them 10 years to
> figure it out they still wouldn't be able to do it.
>
>
>

Well I tried for a day and 1/2 to connect and re-attach with the http://setiboinc.ssl.berkeley.edu/sah/ address, finally I did a google search and the ssl url came up and I got back into the site and reattached as soon as I used that address & I've been using it ever since...

LOL....well it does look like my comments stirred a few things up.

Perhaps the term shareholders wasn't quite right. The point I was trying to make - and obviously failing - was that we contribute a lot of time, effort and money to this. We, as users provide the vast majority of the hardware that runs this project, and yes it is volountary and we can take our hardware elsewhere, but without this donation of computer time this project couldn't exist as it does.
And Christopher, I do deal with different levels of projects - actually in a University - which is why I made the point of saying "in any commercial concern", as educational priorities are obviously different. I also spec and build institutional servers, and with reasonable technical knowledge it is possible to estimate the server/servers neccessary to efficiently run given applications for different levels of network bandwidth.
Dont get me wrong, I really am saddened that this project has disheartened me as much as it has.

The project URL to use is:

http://setiathome.berkeley.edu

That is what the install instructioins say to use, it's what I use, and it works. The one you use may work, but who knows if it will continue to work or not. I generally find it best to use the server that they post.

Chris

> Well I tried for a day and 1/2 to connect and re-attach with the
> http://setiboinc.ssl.berkeley.edu/sah/ address, finally I did a google search
> and the ssl url came up and I got back into the site and reattached as soon as
> I used that address & I've been using it ever since...
>

Having some idea of the priorities that education facilities employ, then you would probably have some idea of the incredible amount of work that the Dev Team has been putting into this project. They have been very prompt in responding to the various problems and outages in order to stabilise and appease people. I've actually been fairly impressed by it.

On the server side, to some degree server load and bandwidth can be estimated. I'm not sure that bandwidth was ever really a problem. What was a problem though was that extremely high workload that some of the various server-side software applications exposed weaknesses and problems not previously detected under lower loads. In some cases the hardware either failed because it was bad or overloaded or whatever. Some of that could have been predicted and accounted for in advance, but the Dev Team in a previous thread stated that due to the way BOINC was released before officially going public, their prediction models were botched. So couple that with software and hardware that has many different components that all depend on each other and the complexities of the distributed computing, and you're going to have a lot of problems. But hey, it works. Not all of the features are enabled or work like they are supposed to, but on the basic level, it does work. And things are getting better with it. Patience. Let the bugs be known so that they can be fixed.

But aside from that, I am tired of people complaining essentially for the sake of complaining. It seems to have become the big thing around here and it doesn't improve anything. I'm not saying you were doing that, and based on your response I don't think you were. But the only way things are going to improve is by letting USB know what needs to be fixed. Help them help us help them.

Chris

> LOL....well it does look like my comments stirred a few things up.
>
> Perhaps the term shareholders wasn't quite right. The point I was trying to
> make - and obviously failing - was that we contribute a lot of time, effort
> and money to this. We, as users provide the vast majority of the hardware that
> runs this project, and yes it is volountary and we can take our hardware
> elsewhere, but without this donation of computer time this project couldn't
> exist as it does.
> And Christopher, I do deal with different levels of projects - actually in a
> University - which is why I made the point of saying "in any commercial
> concern", as educational priorities are obviously different. I also spec and
> build institutional servers, and with reasonable technical knowledge it is
> possible to estimate the server/servers neccessary to efficiently run given
> applications for different levels of network bandwidth.
> Dont get me wrong, I really am saddened that this project has disheartened me
> as much as it has.
>
>

All of you, no matter which side of the fence you may be on with this less than perfect transition to SETI2, are most certainly entitled to your opinions and are correct in all of them.

I seems to me that The University of California has assigned the best and the brightest of the bottom of the barrel to this project. If I was grading them on their effort I would give them a C-. But, enough of that.

What’s really missing with SETI2 is a sense of accomplishment and a competitive spirit for the little guy. No longer do you belong to a Class grouped by sign-up date, a small group of comrades to compete with. If you’re not at the top of the list for all users your accomplishments, great or small, go unnoticed altogether. I believe that in the long run this issue alone will cause major attrition to some other project that will provide this necessary function.

Just my opinion.

</p>