2 new clients, but points?

[MerePeer]

So yesterday afternoon they announced a new client. I downloaded, installed, and ran it -- I confirmed for twice the work it was giving the same # of points as the Mar. 17 client, as seen by examining the blueprint url before & after a gen.

Today they've posted a newer client again. I wonder if the point structure has been adjusted or if we're supposed to enjoy the extra work in the name of science -- while cows find greener pastures!:confused:

[magicfan241]

I'm waiting until they suspend the old client, so that I get the best number of points.

Or I have to leave and am too lazy to install a new client.

Take your pick...

magicfan241

[FoBoT]

wtf is going on? :spank:


2004/03/26
Please download slower client package
The client posted yesterday was mistakenly linked to an older version. This has now been corrected. To clarify, this is not an update of any sort, so there will be no auto-download. You have to manually download the package and install it, or at least use the new executable from the package. The more people running the new client, the better the results we get.

Until the Free-DC forum is back online, please check here frequently for news and updates.

2004/03/25
New slower client available; FTP problems resolved
There is a complete set of new packages available on the download page with a much slower client. None of the code is changed, and it is still the same protein, but there are now 200 structures per generation. The new client should half the frequency of uploads to the server, as well as the amount of buffered data. Since this is a faster protein, we can do more sampling per generation, which should lead to even better RMSDs. Furthermore, this will space out uploads and relieve the load on our servers. Please make sure to download the new package to slow down the rate of uploads and to generate better data.

Note that if you had problems accessing the FTP site, this was due to unannounced network changes in the hospital, and has now been resolved.


have they gone mad? :bonk:

they should have thought of that BEFORE they changed to a fast protein, not as an afterthought :rolleyes:

[Anteraan]

Exactly, FoBoT. Things like this are the reason some people (self included) are a bit annoyed with the project right now. It will pass, I know, but it certainly plants some seeds of doubt.

[rofn]

i dont think they would suspend the old 100 structure per gen client...i think its just a temporary solution to take off the high load from the servers....and when they do points will get the same for all teams & members

personally i won't upgrade (as long they dont suspend the 100er) when points won't change...but who will do anyway?
i don't think that there are so many people crunching who are only after a good scientific result...even when you got the best rmsd you don't get more points for it...

when you look at the big teams @ dc projects its more a pointage hunt nowadays than a scientific crunching...sad but true...

and yes i hunt points, too

anyway...

i'm sure annoyed too with the current state of the project (client, pointage, servers, blabla)...especially when they come up with things like such a alternative client...and want everybody to jump without giving more points...
sometimes i wonder if they just don't want free-dc on #1 :rotfl:

perhaps we should handle out a one day/week no crunching strike with ars & the cows to show our protest :confused:

[Robert Radford]

"personally i won't upgrade (as long they dont suspend the 100er) when points won't change...but who will do anyway?
i don't think that there are so many people crunching who are only after a good scientific result...even when you got the best rmsd you don't get more points for it..."

I've upgraded both my machines. Science rules.

[rshepard]

Likewise-- I changed all 7 of mine over Thursday and Friday.

[alpha]

I haven't upgraded to the newer (slower) client yet.

I just noticed that the work I've been uploading over the past couple of days, hasn't been increasing my stats. Is credit no longer given for work completed by the "old" client?

Has anybody else noticed this happening to them?

[rofn]

not really...

the only thing i notice is that the stats systems are a bit strange the last days...
sometimes they just update after 4 hours...but the points look ok

at the moment i even cant get macnn's stats page loaded and the ars stats are also a bit strange (always having a black frame) :confused:

:notworthy :notworthy holy Dyyryath :notworthy :notworthy
please get our stats back...or the new one ready...
amen

rofn

[alpha]

My points have been stuck on 1,151,392 for a while. I'm uploading as I type this, and uploaded a bunch of gens yesterday evening, but it doesn't seem to be increasing.

[the-mk]

I didn't change to the "new" slower client either...

BTW: I can't access http://teamstats.macnn.com/dfold/stats.php?TID=5&page=s2 and http://marcc.no-ip.org/tsf/userstats.php?TeamID=5 either :rolleyes:

Do we have any "backup"-stats?


the only thing i notice is that the stats systems are a bit strange the last days...
sometimes they just update after 4 hours...but the points look ok
You know the summer-time, rofn? In the US they update a week later :D And usually a stats-run at http://anteater.blueprint.org/teampages/team5.html takes only some minutes, but with this fast protein it takes 30 to 40 minutes :( I think that's why there is such a big update-hole

[rofn]

Originally posted by the-mk
You know the summer-time, rofn? In the US they update a week later :D And usually a stats-run at http://anteater.blueprint.org/teampages/team5.html takes only some minutes, but with this fast protein it takes 30 to 40 minutes :( I think that's why there is such a big update-hole

This Issue was also there before the changeover here to summertime...
Sometimes there just was no update on the stats sites you mentioned for 4 hours...perhaps its only the scripts there and the official stats (where i usually never take a look at) are updated and due this 30 to 40 minute processing cycle the scripts on the stats sites are getting dumped because they get no data for that period of time...

rofn

[the-mk]

hmm, missunderstood you... now I know what you mean :idea:

There still are questions: :confused:
- what happened to Marcc's stats and MacNN's stats?
- do we have any backup stats?

[pfb]

Originally posted by the-mk
hmm, missunderstood you... now I know what you mean :idea:

There still are questions: :confused:
- what happened to Marcc's stats and MacNN's stats?
- do we have any backup stats?

For backup stats try http://wibble.bounceme.net/Sneakers/TeamStats.php?tid=5 (main page is http://wibble.bounceme.net/Sneakers/) - currently in Beta stage but better than nothing :p

[rofn]

the last one i know is http://statsman.ww-testsites.co.uk/distfoldingstats/html/5.html

its pretty basic...but should do the trick till those 2 are back up...hopefully soon...

[the-mk]

Thanks for the info!

Why didn't I remember the statsman-stats? I just had to look to stats.free-dc.org :bang:

[rofn]

@pfb

nice one...keep it up

nice layout...i like this sets complete feature & the mini-team idea (subteams finally)
never seen that on a stats site before :)

for my taste the graphs could need some rrdtool roundup...but its still beta anyway ;)

[pfb]

Originally posted by rofn
@pfb

nice one...keep it up

nice layout...i like this sets complete feature & the mini-team idea (subteams finally)
never seen that on a stats site before :)

for my taste the graphs could need some rrdtool roundup...but its still beta anyway ;)

cheers - using JPGraph as that seemed the easiest one to use with PHP...might look at rrdtool and see how easy is it to use with PHP under Windows...

[alpha]

Originally posted by alpha
My points have been stuck on 1,151,392 for a while. I'm uploading as I type this, and uploaded a bunch of gens yesterday evening, but it doesn't seem to be increasing.

Sorry to take this thread a step backwards but I figured out my problem. It turned out to be this:

Mon Mar 29 09:55:30 2004 ERROR: [000.000] {foldtrajlite2.c, line 4845} Warning during upload: STATUS 910 MISSING PREVIOUS OR ILLEGAL GENERATION

I applied the fix found on the official website. Not very happy that the client just uploaded the generations without ouputting an error message on-screen. I don't even want to think about how many generations I lost.

[rofn]

Originally posted by pfb
might look at rrdtool and see how easy is it to use with PHP under Windows...

rrdtool is quite windows friendly...
for me it's the best monitoring engine that exists...i use it @ work (isp) as full monitoring engine for all the customers leased lines, servers, routers, firewalls, databases,... i even run it at home for monitoring almost all aspects of my boxen...cpu, mem, hdd and even my hardware weather station which is connected through serial interface to one of my linux box...(temperature, wind-speed, -direction, dew point, humidity, barometer, rainfall)

example pic:
http://members.nextra.at/rofo/weather.png

so as you can see i'm quite addicted to it :D

p.s. sorry for the :|ot|:

rofn

[Cmarc]

My stats should be back online some time tonight. They'll probably be pretty slow while I do some fine tuning but tomorrow everything should be back to normal.
I apologize for the inconvenience.
Cheers,
Marc

[the-mk]

Good to hear that they'll come back again :thumbs: :notworthy

[ECL]

Originally posted by alpha

Mon Mar 29 09:55:30 2004 ERROR: [000.000] {foldtrajlite2.c, line 4845} Warning during upload: STATUS 910 MISSING PREVIOUS OR ILLEGAL GENERATION

I applied the fix found on the official website. Not very happy that the client just uploaded the generations without ouputting an error message on-screen. I don't even want to think about how many generations I lost.

I lost about a thousand generations on three machines this way. I've found that deleting all trace of DF from my machines and moving to less-moronically managed projects works quite well.

[the-mk]

I hate this

Mon Mar 29 09:55:30 2004 ERROR: [000.000] {foldtrajlite2.c, line 4845} Warning during upload:
STATUS 910 MISSING PREVIOUS OR ILLEGAL GENERATION error too :swear:

I'm glad that it only happend at the backup-client which is crunching while the main-client is uploading :D