Talk:Grub

[edit] Server side source code

What license is the server side source code under, will it be open sourced or kept proprietary???? - Ché [1]

Server code will be open source as well, soon soon I hope! Jer 04:43, 7 August 2007 (UTC)

[edit] Loss of Momentum?

No apparent progress in grub? End of May 2008. Their appears to be a complete loss of momentum.

[edit] Possible Feature

How do I search the grub database? Wouldn't that be a useful feature? User: Brianherman July 30 2007

You can search it using Wisenut [2] or you can access the Web Services yourself by writing a client.

Is there a feature planned like the proxy's for distributed.net? I mean that anyone can be a server to supply another clients with work? I think your server are very busy at this time. Best regards --Cappa 19:49, 2 August 2007 (UTC)

[edit] Source Code

Where can I get the source code for GRUB so I can work on porting it over to MacOS X?

Should be here very soon.----Jimbo Wales 00:50, 29 July 2007 (UTC)

[edit] Download

Why is the linux grub client link broken? Will it be back up soon?

the correct link is http://www.grub.org/ftp/pub/client/grub-client-1.0.5.tar.gz --stephantom

The Windows Grub link is broken too. http://www.grub.org/ftp/pub/client/2.6.0/grub-client-2.6.0.exe Something's wrong with your counter. Agreed - downloading from the link above works perfectly.

[edit] Data Download

When will it be possible to download the complete data and index files (using e.g. bittorrent)? This would be a real incentive to help with the Grub indexing process for developers/companies.

Yep, the plan is definitely to make it openly available. Jer 04:46, 7 August 2007 (UTC)

[edit] Windows Version doesn't work

Is the windows version not working at all? My client is not able to connect. Is this being fixed?--Hdurina 12:34, 29 July 2007 (UTC)

Windows Client Can not connect to grub servers

When was this program re-activated? The copy I downloaded and installed doesn't exactly match the screenshots given on the setup instructions webpage. I don't have a "run" menu option. The program apparently can't connect to any websites (something about a lookup key not being found in any active activation context; ShellExecuteEx command failed and returned 14007). I'll wait for a new version to come out, I guess. Ctaylor503 18:44, 29 July 2007 (UTC)

my win32 client can't connect...I did run grub way back when...great acquisition. the server source would be damn nice hopefully that will add some gas to the fire and make it like you are serious and not afraid of being truly open and above any forking that will go on anyways....tried to build the linux client on ubuntu...looked good after ./configure but bombed out on make /usr/lib/gcc/i486-linux-gnu/4.0.3/../../../../include/c++/4.0.3/backward/backwar d_warning.h:32:2: warning: #warning This file includes at least one deprecated o r antiquated header. Please consider using one of the 32 headers found in sectio n 17.4.1.2 of the C++ standard. Examples include substituting the <X> header for the <X.h> header for C++ includes, or <iostream> instead of the deprecated head er <iostream.h>. To disable this warning use -Wno-deprecated. Crawler.cpp: In member function ‘void Crawler::initCurlOpts(CURL*, MemoryStruct* , MemoryStruct*)’: Crawler.cpp:1047: error: invalid conversion from ‘int’ to ‘CURLoption’ Crawler.cpp:1047: error: initializing argument 2 of ‘CURLcode curl_easy_setopt (CURL*, CURLoption, ...)’ make[1]: *** [Crawler.o] Error 1 make[1]: Leaving directory `/media/30gig/torrents/grub-client-1.0.5/crawler' make: *** [install-recursive] Error 1 root@haxo4s:/media/30gig/torrents/grub-client-1.0.5# pls feel free to add this page somewhere else in the wiki. thx

Yeh I can't get my client to connect either. I believe this is because Grub is in "test" mode at the moment. Matt

I agree. The windows version is not working. I turned mine off until further notice. G.

Hi i download grub yesterday and i works. At the beginning i has problems with the connection too but after a few minutes it connect and crawl the web. But the stats on website don't work. noxin

It's also working for me now too, glad to see it starting to gain steam! :)

Yeah, windows version is working great. But there a much time wasted to connect the server. Some hours it couldn't connect and i get a timeout. But at this seconds, it work perfectly. --Cappa 18:46, 2 August 2007 (UTC)Cappa

The links to the grub website should probably be removed until the site is working again. it's down hard right now. 206.176.249.130 18:53, 3 August 2007 (UTC)

I agree. I clicked everything that looked like a link on the Windows side of the page and nothing meaningful happened. Annoying. Surely ineffectual links can be commented out until ready, so that visitors don't get turned away? Robin Patterson 02:19, 5 August 2007 (UTC) (former Zealot)

I have mine working regularly - but it crashes all the time and routinely stops processing the queue of new work units. Mjesales 23:46, 8 January 2008 (UTC)

[edit] Grub on Linux

After patching up some outdated parts of the source I was able to compile this thing against the latest versions of all libs under Ubuntu Gutsy (unstable). However, the script that should provide the settings (http://brainbug.grub.org/settings/grub-client-1.0.5.php?clientid=$clientid&password=$password) seems to be missing on the server. I will post a patch to get this working in Gutsy shortly. Maybe even a *.deb... -- stephantom I have a patch ready at this location. Apply with patch -p 1 -i $filename to make it work it's magic. My problem when I compile on Kubuntu Feisty: http://pastebin.com/m3fd55063

I found that exact same problem. I tried to look in the source code and fix it I couldn't. Magikid 00:48, 31 July 2007 (UTC)

I can not seem to get the source to compile on linux at the time. I applied the patch and still get errors with the Crawler.cpp file...
Crawler.cpp: In member function 'void Crawler::initCurlOpts(CURL*, MemoryStruct*, MemoryStruct*)':
Crawler.cpp:1047: error: invalid conversion from 'int' to 'CURLoption'
Crawler.cpp:1047: error: initializing argument 2 of 'CURLcode curl_easy_setopt(CURL*, CURLoption, ...)'
make[1]: *** [Crawler.o] Error 1
I am not much of a coder so I don't know exactly what that means but if someone can find the fix that would be wonderful.
Thanks, G.

Just started working for me (July 31st, 2007)

Still having the same issue. I know its a problem with the source and the latest versions of cURL (or rather I assume). Any assistance with patching this would be appreciated. I would like to get all my Linux servers working on this. They have the bandwidth and the processing power.

Thanks, G.

The latest version of the Linux client is not able to connect to the current version of the server. It uses an old incompatible protocol. So, even if you get a perfectly operational Linux client, you cannot run anything with it. It is possible to recreate the old server that will work with it (after the server source code is released to the public). However, this task would not be useful, as that old code is deprecated. I don't recommend that you do this. What would be more of use is work on porting the existing client to Linux. This means there is a lot of work to be done before the client is available. The Linux-based Grubsters will regrettably need to wait for the time being. --Igor Stojanovski 08:39, 8 August 2007 (UTC)

[edit] Bootloader?

"Grub" is a bootloader: http://www.gnu.org/software/grub/. that might be very confusing.

Maybe the Semantic Lab will help out with this. In fact, that is exactly the thing where google is not working well and where we could beat google. ;-) — MovGP0 13:07, 30 July 2007 (UTC)

Maybe the problem is not in search results on "grub" but in the fact, that something, which is wanted to be well known, is a synonym for something already well known. — me, 20:37, 30 July 2007 (UTC)

[edit] Prelim Source Tarball (windows client)

We need to get everything hosted properly in a source repository somewhere yet, but lots of people are eager to see something so please start digging. It's not even been stamped with an open source license yet, so this is pretty raw:

http://grub.org/client/grub_client2.tgz

The linux client isn't compatible with the SOAP-based API anymore and probably needs to start from scratch, but we'll get it posted as well in case someone just wants to swap out the server-access parts. The server sources will also be showing up for anyone that wants to run their own instance.

[edit] Grubclient status

I realize this project is still a beta phase, but my Grubclient says it has processed 424 MB, and 2432 work units. Am I helping somebody? Ctaylor503 07:06, 4 August 2007 (UTC)

I would like to see some updates on the homepage. I have no clue what is going on with this project. I too am processing some work units.

Well, my Grubclient stopped working yesterday, or the day before (I didn't exactly notice). That is, it seems to not be able to connect to any servers. Ctaylor503 06:39, 7 August 2007 (UTC)

And now it works again, yea! Ctaylor503 23:25, 8 August 2007 (UTC)

[edit] grub.org status

grub.org doesn't seem to be transmitting any html data at the moment. Ctaylor503 23:29, 8 August 2007 (UTC)

Looks like it was a temporary problem... http://grub.org/ is working for me -- sannse

(talk) 19:13, 9 August 2007 (UTC)

[edit] GRUB client 2.6.0 (windows) working on CrossOver Linux

For anyone interested, the GRUB client 2.6.0 (windows) installs and runs perfectly with CodeWeavers CrossOver Linux 6.1, although not with stock WINE due to MSI/WindowsInstaller issues (note, I only tried through wine-doors, and haven't yet fiddled with manual installations of dependencies). dhart 05:57, 10 August 2007 (UTC)

[edit] When might a BOINC version be available?

I am already running 4 or 5 projects and would be more then happy to run one more :-)

http://boinc.berkeley.edu/

+1 for BOINC dhart 02:59, 11 August 2007 (UTC)

[edit] Want GUI Source?

Want the windows source?

There seems to be quite a bit of stuff at a certain URL.

The entire layout of the website, login scripts, database handling. (Someone should fix this)

Looks like the source for the gui is in vb and is there.

Might try to build my own java client at some point. Deathbob 23:44, 15 August 2007 (UTC)

Yeah i discovered this a few weeks back (ref: website layout, login scripts, db, etc) and notified one of the Wikia staff. --Joejaxx 01:53, 16 August 2007 (UTC)

So did I, still waiting for response Deathbob 02:07, 16 August 2007 (UTC)

It's down now, but that also kills the client so I don't know if this was intentional. Deathbob 14:15, 16 August 2007 (UTC)

I'll see about building a java client, I 'm developing a project to compete with BOINC. Deathbob 14:20, 16 August 2007 (UTC)

[edit] SOAP Protocol

I'm definitely interested in a Java client -- I can't seem to get the SOAP url to work without basic AUTH, and my registered credentials from grub.org doesn't seem to work. The URL for the SOAP protocol *is* soap.grub.org/soap, right? --Mbs3 02:33, 17 August 2007 (UTC)

Seems to be but I haven't had the server stable long enough to figure out how it's authenticating because it sure isn't just your user-name and password We'll I can't say that, it may just be redirecting if proper headers aren't sent. I'll break out wireshark tommarow. Deathbob 04:55, 17 August 2007 (UTC)

Maybe we should go ahead and make some other Grub pages to document the protocol? I've tried with Apache Axis, generating client classes and stubs from the WSDL file I found in the released source, and I've also resorted to Apache Jakarta Commons's HttpClient project just to see if I could fight with authentication. I ended up using Wireshark the other day, but that was just an attempt to find the URL to post to. I'll also run back through it again now that I know what IP/Host the SOAP requests and responses use. I'm wondering how much internal documentation existed at Grub, and if we could get our hands on that and open source it :). --Mbs3 13:07, 17 August 2007 (UTC)

Grub Protocol Deathbob 13:53, 17 August 2007 (UTC)

[edit] Local URLs

How can I add local URLs so that grub crawls my website?

Login to grub.org and there should be a local url subpage. It's buggy and I couldn't get it to work.Deathbob 21:27, 16 August 2007 (UTC)

Local URLs still broken. I have crawled over a million URLs but want to crawl my own sites instead of countless of 'religious' sites grub feeds me. FIX IT BEFORE WE GET TIRED OF WAITING!

grrrrrrrrrrrr It's 2008 and still NOT WORKING

J. Lunau 12 January 2008 (UTC) I wonder, if I am doing something wrong? I also can't add any of my local URLs! Does anybody know how to add local URLs?

GRUB HAS A BUG SINCE MONTHS AND THEY NEVER FIXED IT. SO PLEASE HURRY GUYS TO FIX THIS ISSUE. THANKS.

[edit] Server overloaded?

I'm not receiving work units form the server anymore. Connecting to it just fine, but not receiving a thing from it. Sean Fennel 17:29, 17 August 2007 (UTC)

Ya, it's been up and down. The stats make this last week look REALLY bad. Just wait you'll get a few units here and there. The database server is probably being DDOS attacked from the clients. Deathbob 17:41, 17 August 2007 (UTC)

Servers are still down for me on Aug 20th. It seems the servers have now lost the user DB?

The techs have been contacted about this problem and action is being taken. Thanks for your interest and feedback. --Markie 21:37, 20 August 2007 (UTC)

[edit] Not duplicating

Apache Lucene[3] already does alot of stuff around search, is duplication between the Lucene project and Grub being kept in mind? - Ché [4]

"Lucene itself is just an indexing and search library and does not contain crawling and HTML parsing functionality."[5] From what I'm reading, it seems the plan is for Grub to include exactly what Lucene says it does not contain, namely distributed crawling and some minimal parsing --Mbs3 13:17, 21 August 2007 (UTC)

Great...so the focus of Grub will be solely the distributed acquisition not necessarily indexing - Ché [6]

[edit] Stats seem off

It seems that, for a lot of people, the number of URLs crawled last month is listed as being greater than the total number of URLs crawled (go into stats, and sort users by "last month"). Is the last month count just an estimate for a full month in the event a user has joined less than a month ago, or is this a glitch? I think I saw this start to happen a few days ago... --HANtwister 20:55, 31 August 2007 (UTC)

hmm yeah it does seem this way. ive opened a ticket here for this.p.s. please sign you posts (use 4 tildes like this ~~~~) thanks --Markie 16:52, 31 August 2007 (UTC)

Sorry about that - I'm not too terribly familiar with editing wikis (for whatever reason I thought my user name might get added automatically :) ). --HANtwister 20:55, 31 August 2007 (UTC)

thats ok, no worries --Markie 22:18, 31 August 2007 (UTC)

Same problem here. Team stats do not properly show all the URLs crawled.

[edit] I've got an error!

My grub client failed to authenticate with the server: I have entered a bad username or password. But I used to be able to connect to the grub server, so I don't know what changed. Ctaylor503 01:15, 6 September 2007 (UTC)

I have exactly the same problem. I don't know how to contact the Grub Administrators to get this problem fixed.

Post a error report at http://dev.grub.org/ Deathbob 19:15, 18 October 2007 (UTC)

I've now done that. I also got the server to e-mail me the correct username and password, and I am using the right ones, but the client still says "bad username or password". I don't know what else to do. The problem must be at the server end.

Hi, I was also having a problem with authentication.. changing the password to more simple one (ascii only) worked for me. Mykhal 12:26, 3 June 2008 (UTC)

Hi, which client you use? One of new?--thindil 13:19, 3 June 2008 (UTC)

[edit] Grub URL status check: MySQL error

Bug: under the page http://www.grub.org/html/tools.php?op=services when I enter an URL in the form to check the status and click submit only a page containing of MySQL errors appear. When will this become fixed?

thanks for reporting this. ive started a new ticket for it here and it should be acted upon soon. thanks --Markie 17:17, 8 September 2007 (UTC)

[edit] robots.txt

Everyone seems to have forgotten about getting permission of Webmasters before using their servers to run tests. Grub-Wikia is not making friends this way.

GRUB should be obeying robots.txt. What do you mean by tests? Deathbob 15:16, 22 September 2007 (UTC)

See the main page, duh: "NOTICE: The Windows client is running in TEST mode right now, watch the Grub wiki for updates and to help us port this thing :)"

Is there something specifically happening to your server? Or is this just a concern about software that is not fully stable. PS: The major reason it is in test mode is because of poor server communication to the GRUB server from the client. Deathbob 17:09, 23 September 2007 (UTC)

[edit] New Linux-Version ?

Will a new Linux-Version be published in near future?--84.137.200.227 23:27, 29 October 2007 (UTC)

[edit] grub.org / grubgui / grubclient not working for a number of days now (empty stats, no work packets any more)

hi there, since a number of days the grubclient (win32) doesnt fetch any workitems any more for crawling. the grubgui display http error 404 (not found) when the grubclient tries do download the work packages.

also the grub.org statistics (team, individual, week....) is empty for the past few days... check for yourselves: http://grub.org/html/users.php?op=member-stats&showTopMembers=today&offset=1&results=100

has somebody broken the grub.org servers? are there gonna be new grub releases in the near future? these clients have been pretty old and buggy, for both win32 and unix/linux too..... any comments of the official developers? grub/looksmart or wikia technical/distributed staff? thanks and cheers.

-- (09:17 EST, 2008-01-29)

Maybe I'm not official staff, but i can answer ;) Grub are now under development - a new version of protocol, server and client. For now we test few new clients. You can find source code of new clients on http://svn.swlabs.org/grubng/ Thus don't worry,Grub not die ;) --Thindil 20:48, 29 January 2008 (UTC)

hello there, i just tried this new grubng (win32 .exe version) from http://svn.swlabs.org/grubng/trunk/csharp/Releases/Grubng-0.1/grubng-0.1.exe and also the stuff inside http://svn.swlabs.org/grubng/trunk/csharp/Grubng-binaries.zip but they all fail... i get a .net framework crash which wants to report to microsoft, and then telling me in commandline-output/stdout that its missing gtk-sharp version 2.10 or something, although i have installed the go-mono package from ftp://www.go-mono.com/archive/1.2.6/windows-installer/4/mono-1.2.6-gtksharp-2.10.2-win32-4.exe as its stated in the readme.txt file inside the c#-release of grubng for the windows platform....

the exact errors are: C:\Program Files\Grubng>Grubng.exe

Unhandled Exception: System.IO.FileNotFoundException: Could not load file or ass embly 'gtk-sharp, Version=2.10.0.0, Culture=neutral, PublicKeyToken=35e10195dab3 c99f' or one of its dependencies. The system cannot find the file specified. File name: 'gtk-sharp, Version=2.10.0.0, Culture=neutral, PublicKeyToken=35e1019 5dab3c99f' at Grubng.MainClass.Main(String[] args) WRN: Assembly binding logging is turned OFF. To enable assembly bind failure logging, set the registry value [HKLM\Software\M icrosoft\Fusion!EnableLog] (DWORD) to 1. Note: There is some performance penalty associated with assembly bind failure lo gging. To turn this feature off, remove the registry value [HKLM\Software\Microsoft\Fus ion!EnableLog].

and:

C:\Documents and Settings\user\Desktop\New Folder (3)>Grubng.exe

i even tried to install the gtk# stuff from the novell site at http://forge.novell.com/modules/xfmod/project/?gtks-inst4win

the one installer http://forge.novell.com/modules/xfcontent/private.php/gtks-inst4win/Win32%20Runtime%20Installer/v2.8.3/gtksharp-runtime-2.8.3-win32-0.0.exe installs fine, but the other http://forge.novell.com/modules/xfcontent/private.php/gtks-inst4win/Win32%20Installer/v2.8.3/gtksharp-2.8.3-win32-0.0.exe fails telling me that i need .net framework 1.1 or 2.0 installed, which i have installed, to be more precise my .net framework is at this servicepack 1 level for 2.0 and 3.0 also with servicepack 1, and also 1.1 installed just fine.

the system is winxp sp2 win32 latest patches/updates applied, in english

any hint how to get this gtk-sharp with release 2.10 at least, as some gtk site says that the 2.10versions and above have only been released as a source code and no binaries/installers for windows.

how the heck do i get grubng to work with the normal go-mono/gtk# downloads/installers?? thanks for any help.

ok here is the solution to get the gtk-sharp installer with version of at least 2.10upwards on win32: use the binary package of the openvista project on sourceforge for example at: https://sourceforge.net/project/showfiles.php?group_id=74626

download the gtk-sharp installer there from: http://downloads.sourceforge.net/openvista/gtksharp-runtime-2.10.1-1.exe?modtime=1182190610&big_mirror=0

install that on win32 and install the go-mono stuff too from ftp://www.go-mono.com/archive/1.2.6/windows-installer/4/mono-1.2.6-gtksharp-2.10.2-win32-4.exe

that did the trick for me. i dont know why the go-mono package is named gtksharp 2.10 too as it didnt work out for me on my system with only that package installed... :(

-- cheers (2008-01-30, 5:59 AM EST)

On Windows enough is download Mono installer - it has include gtk# (newest). http://svn.swlabs.org/grubng/trunk/csharp/Grubng-binaries.zip - this is always most updated version of C# client, updated near every day. --Thindil 12:22, 31 January 2008 (UTC)

[edit] grubng - bugreports where?

hi there, i have submitted a first bugreport of the new grubng client here Talk:GrubWorkUnit.

-- cheers (2008-01-30, 8:39 AM EST)

Bug reports please send on http://dev.grub.org --Thindil 10:56, 2 February 2008 (UTC)

[edit] Unit size

Hello, why don´t have units 1000 url at least but 10 url only?

Hi, if i good think, you use grub client from www.grub.org - which is obsolete. Try instead clients from source code repository http://svn.swlabs.org/grubng/ --Thindil 17:33, 8 February 2008 (UTC)

[edit] Is Grub Dead Yet?

After the controversy over payment for altering wikipedia entries, has wikia put the can on Grub?

If so why not close it down. 82.26.119.208 15:01, 31 May 2008 (UTC)

Retrieved from "http://search.wikia.com/wiki/Talk:Grub"

This page was last modified 18:45, 3 June 2008. GFDL