Happy to see this forum!!

Linux clients have not released yet......but....fortunately I was told where the source was by Eric Korpela in the dev's mailing list. (Thank you, Eric!)

So I've made optimized clients for Linux compiled with Intel C++ compiler and linked with fftw. I named it OSECL (Optimistic/Optimized Seti Enhanced Client for Linux.)

For details and links, see this post and thread. And use at your own risk ;)

The developer team has done an excellent job...the source is very smart!!

Have fun!! Windows optimized clients will follow....sooner or later...

---

Luckiest in the world. WMD = Weapon of Mass Distraction

Would not an optimized client be contrary to a beta test?

---

I would be happy with just a plain old plain Jane Linux client.

---

I would be happy with just a plain old plain Jane Linux client.

It's important. I made a plainvanilla client for linux. Only mmx is required. Statically linked. I don't know about the speed. If you need more speed, please use the optimized one ;) And as a matter of course, use at your own risk.

seti-enhanced-mmx-linux.tar.gz

At first, attach to the project. You'll get an empty directory for this beta project. Then put app_info.xml and the executable in the project directory. Restart boinc core client and you'll get WUs automatically and processing will begin. In the result, a remark appears in stderr out section so you can distriminate it from the official one and know cvs $id of the source tree. Its version number is set to 4.07 (the default for a custom built client on a unix like system is 4.07), but you don't have to mind it. It's the same as 4.00 on Windows.

---

Luckiest in the world. WMD = Weapon of Mass Distraction

I'll be putting out a linux client for SAH enhanced 4.01 early next week. Don't know if I'll get graphics into it, but I'll try.

I'll be merging Tetsuji's changes into the source tree so people who want to compile with the IPP FFT library will have an easier time of it...

Eric

I'll be putting out a linux client for SAH enhanced 4.01 early next week. Don't know if I'll get graphics into it, but I'll try.

I'll be merging Tetsuji's changes into the source tree so people who want to compile with the IPP FFT library will have an easier time of it...

Eric

Thanks!! IPP is very fast on some of Intel processors....somehow on some it's slower. Anyway the more options, the better. but in my personal opinion, when I pursue speed, I will abandon graphics ;)

BTW somehow Windows version is much slower than Linux version. I found many Windows specific debugging macros and removed all of them (as a matter of course for my own use for security..) but it is still slower than Linux version... On public SETI@home site, they have almost the same speed.

---

Luckiest in the world. WMD = Weapon of Mass Distraction

Would not an optimized client be contrary to a beta test?

No, as optimization is being done to the original source code. It's doing the same science, only quicker. This way, dev can get a lot more WUs to look at from the smaller user base. Net result, app goes from beta to public quicker. :)

I'll be putting out a linux client for SAH enhanced 4.01 early next week. Don't know if I'll get graphics into it, but I'll try.

I'll be merging Tetsuji's changes into the source tree so people who want to compile with the IPP FFT library will have an easier time of it...

Eric

oops....I put 4.07 in my optimized clients...now I renumbered all of my WUs in my cache and my Linux client while the server is down ;) But won't change the version number in my released ones...it would confuse many people.

if there's anything to optimize or something, I'll do it.

---

Luckiest in the world. WMD = Weapon of Mass Distraction

I am still using 4.07 on my linux box. Will my results be invalid until TMR builds the 4.02 optimized client for linux? I just upgraded all my windows boxes to TMR's 4.02 from 4.07. They were showing invalid.

I am still using 4.07 on my linux box. Will my results be invalid until TMR builds the 4.02 optimized client for linux? I just upgraded all my windows boxes to TMR's 4.02 from 4.07. They were showing invalid.

It will give you all invalid results. You need to switch to the latest version because on thie beta site this version upgrade includes a significant change.

---

Luckiest in the world. WMD = Weapon of Mass Distraction

It will give you all invalid results. You need to switch to the latest version because on thie beta site this version upgrade includes a significant change.

Do you have a 4.02 linux client out yet?

It will give you all invalid results. You need to switch to the latest version because on thie beta site this version upgrade includes a significant change.

Do you have a 4.02 linux client out yet?

oops....sorry!! forgot it...Just a moment For the time being please use official one. I'll post them in this message forum (another thread will be fine.)

---

Luckiest in the world. WMD = Weapon of Mass Distraction