Could you guys try it out for me? Check it fully, hammer it with requests, that sorta thing. Just take note, I've set the delay to 10 seconds now.

Its here: http://www.blue-world.co.uk/personal/email_friend/email_friend.php

Hi Jonathon

I just tried your script. Two things to report: 1) The message was classified as spam on the receiving side - not sure why this happend; usually the spam filter is quite reliable - perhaps because it was sent from a different server than the address in the "From:" field ?, and 2) the message was sent twice. I only entered one email address in the input mask but the "To:" field in the header of each of the received emails contained the address twice (separated by a comma).

Anyway, great job, but I thought you might want to know... -H.

"Housing..." when to click through to show another codeword, were you able to find some that were more clear to you? I'm attempting to ask, if we happen to hit a set where you can't see ANY of them? Or was it roughly 50/50? Or, how much were you able to see?

I copied a bunch of the pictures and gave what I thought it said. If anyone is interested, email me at eko at u dot washington dot edu.

-E

---

1) The message was classified as spam on the receiving side - not sure why this happend; usually the spam filter is quite reliable - perhaps because it was sent from a different server than the address in the "From:" field ?

Yes, I believe that's because the SPAM filter noticed that the sender's domain does not exist (which is the primary filter that I'm aware of). That issue should rake care of itself once the function is incorporated into the Rosetta site, and the EMail is sent from bakerlab.org.

---

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/

Thats annoying Holderin, thanks for telling me. I'll have another look at it. Thanks for the information Housing... I'll talk to the image validation module's writer and see if he can help.

Just food for thought... My script has (already) in its current (hard to get to location) been under attack. Now if its anyone here testing my script's security (friendly like), now would be a real good time to tell me, before I start tracing and abuse reporting procedures. Send me an email ASAP on johnathon [at] blue ~dash~ world .dot. co .dot. uk.

Housing, I'm interested - the more I can give to the guy the more he could probbably help me. I'm just about to send you an e-mail...

---

Johnathon, wrt Hoelderlin's report that emails via the "invite script" end in the spam folder (I've just tried it and it ended up in my spam folder as well).
I've written in the past (see here) about the aspects of running such scripts in today's email environment.

The main issue is that one can't safely "spoof" (impersonate) Internet sender-address anymore, as it was the case in the past. I.e. the way your script currently operates, i.e. by setting the From/Sender field to the user-supplied e-mail, will get the email in the spam-folder or simply blocked by most ISPs (incl. big ones like AOL). And it will result in getting the IP address hosting the script blocked.

So you'd need to change the script to send from an email address that belongs to the organisation which is "hosting" the invite script.

At best, you could put the name of the inviting person in the Subject line, e.g. if John Doe is inviting Jonathon:

From: invitescript@yourdomain.com
To: Jonathon@yourisp.com
Subject: Your friend "John Doe" thought you might be interested in Rosetta@home

...content here...

But you can't put arbitrary email addresses in the From: field and expect the email to reach the recipient.

PS: I still fail to see why one would use a script, rather than just send to all people in his own addressbook, using the "teaser" content you've developed here.

---

Best UFO Resources
Wikipedia R@h
How-To: Join Distributed Computing projects that benefit humanity

you can't put arbitrary email addresses in the From: field and expect the email to reach the recipient.

Agreed. I don't believe he's doing that in his test. He's not saying the EMail came from the person's EMail address, but he does have a domain name that's not found by a browser, so I think that's the problem.

I still fail to see why one would use a script, rather than just send to all people in his own addressbook, using the "teaser" content you've developed here.

I agree, there are merits to other approaches of "telling-a-friend". But the merits of this approach are that the project team can have some measure of how many notes are going out, and even tally up how many actually sign up with the same EMail address down the road.

It also puts the project team in control of the message, and gives them the ability to adjust it over time as they see fit, so we're all sending the same message.

Other than that, the script blocks the SPAMers from using the function... which would be of limited use to them anyway, because they would still be sending the Rosetta message, not their SPAM. But, agreed, if someone wanted to defame R@H they might send the same thing 10 times to the same people, THAT would get annoying... I don't believe Jonathan has a copy of a database to track and prevent redundant sends. But it would be possible to add such later, along with producing the feedback on number of "conversions" you get as a result of the EMails that go out.

---

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/

Wow. This is great. Almost 1,000 new hosts added yesterday and total users passed the 50,000 mark. Over 23k teraflops also. I think that is pretty impressive.


Users (last day ) :
50,202 (+518)
Hosts (last day ) :
107,101 (+954)
Credits last 24h :
2,347,374
Total credits :
256,265,321
TeraFLOPS estimate: 23.474

---

Founder of BOINC GROUP - Objectivists - Philosophically minded rational data crunchers.

The automated installer of Rosetta@home, are you thinking doing it in the way BBC CCE works (client then download apps) and not a full install including science app (as it'll probably have been updated by the time it is installed).

(Windows)
So basically Install BOINC client as single user, have it automatically run the wizzard with rosetta at home (you may need to build your own boinc manager or pop into the developers list as there was somehting similar to this within the past moths, maybe related to BBC CCE/CPDN)
They then sign up for a new account there and then it's done.


You should check the [url=http://www.bbc.co.uk/sn/hottopics/climatechange/[/url] website for some idea's and maybe some of the work is partly done for you (videos faqs etc..)

---

Team mauisun.org

Just one suggestion:

Create a roadmap with milestones and short-to-midterm targets. People like to bring down clear-defined "mountains" of computation work and like to see the mountain shrink. You have something like an accellerating countdown. I think this is very thrilling and gives you a short-term goal understandable to everyone.

Furthermore a roadmap with different goals gives you the feeling the project is well-planned and has well-defined objectives to reach. I think this is one reason of the success of Einstein (200 day targets with milestones like S3,S4, S5 etc.).

While the goal (cure cancer etc.) is clear and honorable the link how this project contributes in achieving this goal is weak (gaining understanding, comparing proteins, folding and refolding them, improving models and predictions etc...). But a roadmap and short-term-goals (folding 200.000 proteins in stage 1, leading to this progress in the science) would bring a closer connection between the long-term goal and the actual contribution of this project.

---

Create a roadmap with milestones

I certainly agree. But, to try to keep things organized on the boards, I'm going to call that thought "off topic". Your idea will bring more users or more of their resource share, and so is more appropriate on another thread we've also been using. See if you think I'm on the same wavelength in that original post where I talked about

Show some guage of completion (entire project): It's kinda nice the way Einstein shows the total number of work units there will be, and the number remaining to be completed. It throws down a challenge to complete them.

fluffy
My present thought is still in draft/pending/"to be organized/reviewed" status, but I'm thinking I'd prefer to save the user that initial 5+MB download of the R@H application, and have it on the CD. So, initially the only thing they'd have to download would be a WU (i.e., you're only 5 minutes away from crunching your first WU!). Your point that this obsoletes your CDs rather quickly is well-taken. We'd have to devise a plan to distribute each batch of CDs produced quickly so they don't age on a shelf, if my "CD includes Rosetta" idea is the final implementation.

I'm thinking the easy-to-install CD would have an easy-to-use user's guide, along with some descriptive animations or how Rosetta works, project information, and a BOINC overview diagram which you review before it installs anything. I'm also thinking perhaps it should prompt for your EMail address, and an optional team name, which might be written on the face of the CD by the school or person that gave it to you. It would also have sections for advanced information on using BOINC, message boards, technical level science perhaps.

It would also have a PC tester, which would assess your Windows version (I'm thinking to stick to the Windows audience initially to help reduce the complexity of the CD, and yet target the mass audience), your CPU speed, installed memory, and internet speed and it would then bring up a preferences advisor for you. If you have a slow network, it would advise longer WUs for example. And, ultimately, it may advise against installing R@H on a given machine that doesn't meet the guidelines.

I have joined the BOINC developers list, but haven't had the time to orient myself on all of the information available there. Any pointers welcome. I see I could place an XML file in the BOINC install which would register Rosetta, but you'd still have to download it, and I think you have to place a valid user ID in that file. This would make it difficult for a team to assure that the CDs they distribute will enroll people in their team (unless you could on each group altering the CD, which I'd prefer to avoid). So, I'm still in doubt on some of the specific mechanics of how it will all work.

It appears that users can be created, and preferences or team changed via simple CGI calls to the website. So, I'm hopeful that an install client can be created which would take over those initial functions from BOINC, and use alternate means of getting the info. and registering it with the project for the initial setup. Then when BOINC starts for it's first time, it would look just the same as when an existing user starts it, and there's no WUs on their machine.

---

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/

(Borrowing from dag...)

If you're a gamer, and it fits within the message board's usage guidelines, please encourage Rosetta amongst other online gamers.

---

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/

You may want to base your install of trying sooner rather than later on th latest boinc client (in dev 5.3.31 which is available and listed on the boinc client download site) I say this as apparently 5.4.x will be out soon-ish which will be an officail release.

Thinking about it including ALL the mahor files, especially that large bbdep02 nd dunbrak files even with the current client would be great (as on dialup it's a pain in the arse)
My thinking is these have never changed, (bbdep and dunbrak) but even if the sci-app changes what is the harm. It's either up to-date (no download) or it needs to download a new version, compared to it always needing to downlaod it :-). So that solves that one, just need to make it easy to alter the files which they are anyway.

As for teams....
That's what is causing me the most problems thinkin about. I'll need to setup a new account to see how boinc-manager currently does it, but I don't think it speicifies a team (or at least it never used to).
So you could use the 'join this team link' each team has.
- but this adds more hassle to a user and requires custom CD's for each team
- or using you enter team name written on CD, you could use the Team number as it's easier to script and use said link..


Anyways, this is about the xml file thingy
http://boinc.berkeley.edu/win_install.php
though I see nothing about teams

of to have a look at .... err .... just hope to find somthing...

---

Team mauisun.org

As for teams....

I think the Set account info RPC call should be doable for setting the team. You'll also see "Create account" there as well. These are just rpc calls. Any sockets application could be used to perform them. So, I'm picturing a simple entry screen which invokes these RPC calls and places the resulting account info. into the appropriate .XML files that it installs on the PC. Then you start BOINC and you're off! (...I hope :)

We're many months from having such a CD with all the materials I've outlined, and the specific BOINC version available at the time should be of little consequence to how the CD is done. Unless the Dutch Power Cows complete one with similar materials.

If anyone reading this is interested in helping with any portion of the easy to install CD project, please join the Yahoo! group mentioned here.

Cureseekers? How well does my description match the CD you are working on?

---

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/

Yep, Dutch Power Cows, it was created for the upcoming stampede, to help non-dc freaks easy install the rosetta client :)

Can anyone tell me how to reach the DPC? Or add a link to their automated installer? I don't speak Dutch, so didn't get far in following the team links.

Better yet, could one (or more!) of those in-the-know about that automated installer come join the Yahoo! volunteers group?

I'm here ;)
The automaticaly installer is currently in development, so...
I'm leaving for a week, but when I return, I'll see what I can do
(if someone else already didn't. My teammembers are aware of it, so...)

You dont have to speak dutch...all you have to do is be able to say MOOOOOOOOOOOOO ... ( just kidding) BTW the Cows are in the middle of a fight with the Xtremes and the gentle bantering that is being generated among the two rivals is having the effect of both teams increasing their crunching power either by adding more members or adding more hardware into their set ups. ( The levels of trash talk mong the teams are rip roaring hilarious) .

So if you are memeber of a team convince your team to challenge a rival to a "match" and let the fun of the challenge beging. If the Xtreme's BBQ VS The Cow's Stampede "grudge" match is an example...the level of WUs produced and people participating will increase all of this while having fun.

---

This and no other is the root from which a Tyrant springs; when he first appears he is a protector.”
Plato

Then post both links so we can all partake of the fun... :-)

---

Founder of BOINC GROUP - Objectivists - Philosophically minded rational data crunchers.

Maybe all of your ideas are beginning to work. According to the Rosetta home page, the TeraFLOPS estimate has jumped nearly 50% to 29.925! It was hovering around 21 yesterday...I hope it's no fluke.

---

Maybe all of your ideas are beginning to work.

At THIS point, that would be impressive! Because other than the Rosetta Wiki revamp, (which is GREAT WORK, I just don't think it's intended to line 'em up at the door) I don't think we've really done anything :) Could just be due to the issueance of the credit for the work lost over the weekend due to WU errors. But this chart shows a nice trend! Seems a lot of people are adding more PCs to existing accounts (number of new "hosts" is typically double that of new "users"). And perhaps increasing resource share.

---

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/

Maybe all of your ideas are beginning to work. According to the Rosetta home page, the TeraFLOPS estimate has jumped nearly 50% to 29.925! It was hovering around 21 yesterday...I hope it's no fluke.

That's a fluke due to the credits granted for errors. The interesting graph is this:

http://www.boincstats.com/stats/project_graph3.php?pr=rosetta&table=credits

and there you see only a modest increase in average credit over the past 30 days.

---

Ok, I think I've worked out a couple more of the bugs in the script. Can you test it for me please?

http://www.blue-world.co.uk/personal/email_friend/email_friend.php

Once I know the script is functioning fully I'll put the html message that feets etc is looking at. I cant help more now - I need to programme a admin module for anodyne, as a payment for their help with the script.

---

URL is not working... :-(

---

Founder of BOINC GROUP - Objectivists - Philosophically minded rational data crunchers.