Search Results

Search found 6559 results on 263 pages for 'parallel foreach'.

Page 94/263 | < Previous Page | 90 91 92 93 94 95 96 97 98 99 100 101  | Next Page >

  • Python Macros: Use Cases?

    - by Rick Copeland
    If Python had a macro facility similar to Lisp/Scheme (something like MetaPython), how would you use it? If you are a Lisp/Scheme programmer, what sorts of things do you use macros for (other than things that have a clear syntactic parallel in Python such as a while loop)?

    Read the article

  • Job queue manager with RPC interface

    - by admr
    I need a job queue manager that I can control over the Internet. It should be able to execute and stop processes, check on their status (ideally notice and execute some code when a process exits), respond to commands and also be able to report back to a server. Background: I have a GWT application that allows to create jobs to execute on a cloud instance (currently EC2). I want to push a "job packet" (data for a process to operate on etc) to S3, start a Linux EC2 instance (or use one that's already running), and tell a job manager on the instance to execute that job (possibly parallel to other jobs). It should then pull the "job packet" from S3, run a process that operates on that data and report back to the server that is running the server part of my GWT application with some information (e.g. exit code, stdout, stderr). If I have to write e.g. stdour/err to a file from the process and read that file, that's OK too. I would really like the manager to be "close" to the processes it runs, meaning I want to avoid using something like Runtime.exec from the JDK. It seems like I would have to do that if I used Quartz for example. I'm fine with the calls in both directions being asynchronous. I'm fine with any reasonable technology for the calls as long as I can easily build an interface for that in my GWT server side (e.g. HTTP requests to a servlet over SSL would be nice and trivial). The job manager does not need to have a very sophisticated queueing system. Running several processes either sequentially or in parallel should be fine. Determining how much compute time a process received during its lifetime would be nice (AFAIK, this might be challenging). I did not yet find any existing software that does this, including http://java-source.net/open-source/job-schedulers. I suspect I might have to build an RPC interface (with authentication etc, of course) around a job manager; maybe use something like Apache Commons Exec. In that case, I would prefer Java or Python for the job manager part. I would be happy to hear suggestions for either the former or latter scenario!

    Read the article

  • How many copies of files are needed by video server?

    - by Trilok
    A quick question. How many copies of the same movie are kept in a video server (a video streaming server)? Suppose a particular video is at max requested by 1000 users at the same instant of time, how many copies would be sufficient so that parallel streams can be provided to each user? Ideally 1 copy would solve the purpose, but what is the optimum number keeping the bandwidth and simultaneous access in mind?

    Read the article

  • Threading cost - minimum execution time when threads would add speed

    - by Lukas
    I am working on a C# application that works with an array. It walks through it (meaning that at one time only a narrow part of the array is used). I am considering adding threads in it to make it perform faster (it runs on a dualcore computer). The problem is that I do not know if it would actually help, because threads cost something and this cost could easily be more than the parallel gain... So how do I determine if threading would help?

    Read the article

  • The future of programming, or what lies in the future in programming?

    - by prosseek
    I remember that a article that Microsoft uses formal verification to debug the Device Driver, and I also remember that Functional Programming removes much of the bug as it ensures stateless programming. And we all know about the multi-core. I beleive all of them are future direction of programming or programming language. Multi-core programming or parallel programming Software Formal Verification Functional Programming (as a mainstream?) What do you think? What will be the future of programming?

    Read the article

  • Very basic question about Hadoop and compressed input files

    - by Luis Sisamon
    I have started to look into Hadoop. If my understanding is right i could process a very big file and it would get split over different nodes, however if the file is compressed then the file could not be split and wold need to be processed by a single node (effectively destroying the advantage of running a mapreduce ver a cluster of parallel machines). My question is, assuming the above is correct, is it possible to split a large file manually in fixed-size chunks, or daily chunks, compress them and then pass a list of compressed input files to perform a mapreduce?

    Read the article

  • How can a 1Gb Java heap on a 64bit machine use 3Gb of VIRT space?

    - by Graeme Moss
    I run the same process on a 32bit machine as on a 64bit machine with the same memory VM settings (-Xms1024m -Xmx1024m) and similar VM version (1.6.0_05 vs 1.6.0_16). However the virtual space used by the 64bit machine (as shown in top under "VIRT") is almost three times as big as that in 32bit! I know 64bit VMs will use a little more memory for the larger references, but how can it be three times as big? Am I reading VIRT in top incorrectly? Full data shown below, showing top and then the result of jmap -heap, first for 64bit, then for 32bit. Note the VIRT for 64bit is 3319m for 32bit is 1220m. * 64bit * PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 22534 agent 20 0 3319m 163m 14m S 4.7 2.0 0:04.28 java $ jmap -heap 22534 Attaching to process ID 22534, please wait... Debugger attached successfully. Server compiler detected. JVM version is 10.0-b19 using thread-local object allocation. Parallel GC with 4 thread(s) Heap Configuration: MinHeapFreeRatio = 40 MaxHeapFreeRatio = 70 MaxHeapSize = 1073741824 (1024.0MB) NewSize = 2686976 (2.5625MB) MaxNewSize = -65536 (-0.0625MB) OldSize = 5439488 (5.1875MB) NewRatio = 2 SurvivorRatio = 8 PermSize = 21757952 (20.75MB) MaxPermSize = 88080384 (84.0MB) Heap Usage: PS Young Generation Eden Space: capacity = 268500992 (256.0625MB) used = 247066968 (235.62142181396484MB) free = 21434024 (20.441078186035156MB) 92.01715277089181% used From Space: capacity = 44695552 (42.625MB) used = 0 (0.0MB) free = 44695552 (42.625MB) 0.0% used To Space: capacity = 44695552 (42.625MB) used = 0 (0.0MB) free = 44695552 (42.625MB) 0.0% used PS Old Generation capacity = 715849728 (682.6875MB) used = 0 (0.0MB) free = 715849728 (682.6875MB) 0.0% used PS Perm Generation capacity = 21757952 (20.75MB) used = 16153928 (15.405586242675781MB) free = 5604024 (5.344413757324219MB) 74.24378912132907% used * 32bit * PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 30168 agent 20 0 1220m 175m 12m S 0.0 2.2 0:13.43 java $ jmap -heap 30168 Attaching to process ID 30168, please wait... Debugger attached successfully. Server compiler detected. JVM version is 14.2-b01 using thread-local object allocation. Parallel GC with 8 thread(s) Heap Configuration: MinHeapFreeRatio = 40 MaxHeapFreeRatio = 70 MaxHeapSize = 1073741824 (1024.0MB) NewSize = 1048576 (1.0MB) MaxNewSize = 4294901760 (4095.9375MB) OldSize = 4194304 (4.0MB) NewRatio = 8 SurvivorRatio = 8 PermSize = 16777216 (16.0MB) MaxPermSize = 67108864 (64.0MB) Heap Usage: PS Young Generation Eden Space: capacity = 89522176 (85.375MB) used = 80626352 (76.89128112792969MB) free = 8895824 (8.483718872070312MB) 90.0629940005033% used From Space: capacity = 14876672 (14.1875MB) used = 14876216 (14.187065124511719MB) free = 456 (4.3487548828125E-4MB) 99.99693479832048% used To Space: capacity = 14876672 (14.1875MB) used = 0 (0.0MB) free = 14876672 (14.1875MB) 0.0% used PS Old Generation capacity = 954466304 (910.25MB) used = 10598496 (10.107513427734375MB) free = 943867808 (900.1424865722656MB) 1.1104107034039412% used PS Perm Generation capacity = 16777216 (16.0MB) used = 11366448 (10.839889526367188MB) free = 5410768 (5.1601104736328125MB) 67.74930953979492% used

    Read the article

  • Property transfers in soapui

    - by Scobal
    I'm trying to write parallel tests in soapui and need to transfer properties between the test steps I currently have 3 tests steps: Execute legacy request Execute new request XML diff the two responses in a groovy script I've found a lot of blogs about picking values out with xpaths, but nothing about passing the full response through. My questions is how do I fill out the source and target boxes in the property transfer editor?

    Read the article

  • Website stress test in Python - Django

    - by RadiantHex
    Hi folks, I'm trying to build a small stress test script to test how quickly a set of requests gets done. Need to measure speed for 100 requests. Problem is that I wouldn't know how to implement it, as it would require parallel url requests to be called. Any ideas?

    Read the article

  • Ruby LESS gem equivalent in Python

    - by Sean M
    The Ruby LESS gem looks awesome - and I am working on a Python/Pylons web project where it would be highly useful. CSS is, as someone we're all familiar with recently wrote about, clunky in some important ways. So I'd like to make it easier on myself. Is there an existing Python module or library that provides parallel functionality?

    Read the article

  • How this pthread actually works?

    - by user289013
    I am actually on my project on compiler with SMP, and want to code with pthreads and heard about many parallel things open mpi and so on, So to start with how this thread is allocated to core while calling pthread,Is there any way to give threads to different cores by pthreads?

    Read the article

  • PERL XPath Parser Help

    - by cognvision
    I want to pull in data using a XML::XPath parser from a XML DB file from the Worldbank site. The problem is that I'm not seeing any results in the output. I must be missing something in the code. Ideally, I would like to extract just the death rate statistics from each country XML DB (year and value). I'm using this as part of my input: http://data.worldbank.org/sites/default/files/countries/en/afghanistan_en.xml use strict; use LWP 5.64; use HTML::ContentExtractor; use XML::XPath; my $agent1 = LWP::UserAgent->new; my $extractor = HTML::ContentExtractor->new(); #Retrieve main Worldbank country site my $mainlink = "http://data.worldbank.org/country/"; my $page = $agent1->get("$mainlink"); my $fulltext = $page->decoded_content(); #Match to just all available countries in Worldbank my $country = ""; my @countryList; if (@countryList = $fulltext =~ m/(http:\/\/data\.worldbank\.org\/country\/.*?")/gi){ foreach $country(@countryList){ #Remove " at the end of link $country=~s/\"//gi; print "\n" . $country; #Retrieve each country profile's XML DB file my $page = $agent1->get("$country"); my $fulltext = $page->decoded_content(); my $XML_DB = ""; my @countryXMLDBList; if (@countryXMLDBList = $fulltext =~ m/(http:\/\/data\.worldbank\.org\/sites\/default\/files\/countries\/en\/.*?\.xml)/gi){ foreach $XML_DB(@countryXMLDBList){ my $page = $agent1->get("$XML_DB"); my $fulltext = $page->decoded_content(); #print $fulltext; #Use XML XPath parser to find elements related to death rate my $xp = XML::XPath->new($fulltext); #my $xp = XML::XPath->new("afghanistan_en.xml"); my $nodeSet = $xp->find("//*"); if (!$nodeSet->isa('XML::XPath::NodeSet') || $nodeSet->size() == 0) { #No match found print "\nMatch not found!"; exit; } else { foreach my $node ($nodeSet->get_nodelist){ print "\n" . $node->find('country')->string_value; print "\n" . $node->find('indicator')->string_value; print "\n" . $node->find('year')->string_value; print "\n" . $node->find('value')->string_value; exit; } } } #Build line graph based on death rate statistics and output some image file format } } } I am also looking into using the xpath expression "following-sibling", but not sure how to use it correctly. For example, I have the following set of XML data where I am only interested in pulling siblings directly after the indicator for just death rate data. <data> <country id="AFG">Afghanistan</country> <indicator id="SP.DYN.CDRT.IN">Death rate, crude (per 1,000 people)</indicator> <year>2006</year> <value>20.3410000</value> </data> - <data> <country id="AFG">Afghanistan</country> <indicator id="SP.DYN.CDRT.IN">Death rate, crude (per 1,000 people)</indicator> <year>2007</year> <value>19.9480000</value> </data> - <data> <country id="AFG">Afghanistan</country> <indicator id="SP.DYN.CDRT.IN">Death rate, crude (per 1,000 people)</indicator> <year>2008</year> <value>19.5720000</value> </data> - <data> <country id="AFG">Afghanistan</country> <indicator id="IC.EXP.DOCS">Documents to export (number)</indicator> <year>2005</year> <value>7.0000000</value> </data> - <data> <country id="AFG">Afghanistan</country> <indicator id="IC.EXP.DOCS">Documents to export (number)</indicator> <year>2006</year> <value>12.0000000</value> </data> - <data> <country id="AFG">Afghanistan</country> <indicator id="IC.EXP.DOCS">Documents to export (number)</indicator> <year>2007</year> <value>12.0000000</value> </data> Any help would be much appreciated!!!

    Read the article

  • Using MPI under VC++ MFC project?

    - by Mike
    Does any body know how can I use MS_MPI in my VC++ MFC project? I already have a big MFC project and I only want to use parallel processing in a part of it with MPI. (I know how to use MPI in a separate code, but I don't know how to integrate it with my VC++ MFC project)

    Read the article

  • Rectangles Covering

    - by den bardadym
    I have N rectangles with sides parallel Ox and Oy. Exists another rectangele (model). I need create algorithm, which can tell: is model covered by N rectangles? and code him. I have some ideas. First I think need sort rectangles by left side (it can be done by O(n log n)). Then I think need use vertical sweeping line. Thanks.

    Read the article

  • Open Cl.I just need to convert the code to using two work items in the for loop .Currentlly it uses one

    - by user1660282
    spmv_csr_scalar_kernel(const int num_rows , const int * ptr , const int * indices , const float * data , const float * x, float * y) { int row = get_global_id(0); if(row < num_rows) { float dot = 0; int row_start = ptr[row]; int row_end = ptr[row+1]; for (int jj = row_start; jj < row_end; jj++) { dot += data[jj] * x[indices[jj]]; } y[row] += dot; } } Above is the Open Cl code for multiplying a sparse matrix in CSR format with a Column vector.It uses one global work item per for loop.Can anybody help me in using two work items in each for loop.I am new to open cl and get a lot of issues if I modify even the smallest thing.Please help me.This a part of my project.I made it this parallel but I wanna make it more parallel.Please help me if you can.plzzzz A single work item executes the for loop from row_start to row_end.I want that this row or for loop is further divided into two parts each executed by a single work item.How do I go on accomplishing that? This is what I could come up with but its returning the wrong output.plzz help __kernel void mykernel(__global int* colvector,__global int* val,__global int* result,__global int* index,__global int* rowptr,__global int* sync) { __global int vals[8]={0,0,0,0,0,0,0,0}; for(int i=0;i<4;i++) { result[i]=0; } barrier(CLK_GLOBAL_MEM_FENCE); int thread_id=get_global_id(0); int warp_id=thread_id/2; int lane=(thread_id)&1; int row=warp_id; if(row<4) { int row_start = rowptr[row]; int row_end = rowptr[row+1]; vals[thread_id]=0; for (int i = row_start+lane; i<row_end; i+=2) { vals[thread_id]+=val[i]*colvector[index[i]]; } vals[thread_id]+=vals[thread_id+1]; if(lane==0){ result[row] += vals[thread_id]; } } }

    Read the article

  • Overwhelmed by design patterns... where to begin?

    - by Pete
    I am writing a simple prototype code to demonstrate & profile I/O schemes (HDF4, HDF5, HDF5 using parallel IO, NetCDF, etc.) for a physics code. Since focus is on IO, the rest of the program is very simple: class Grid { public: floatArray x,y,z; }; class MyModel { public: MyModel(const int &nip1, const int &njp1, const int &nkp1, const int &numProcs); Grid grid; map<string, floatArray> plasmaVariables; }; Where floatArray is a simple class that lets me define arbitrary dimensioned arrays and do mathematical operations on them (i.e. x+y is point-wise addition). Of course, I could use better encapsulation (write accessors/setters, etc.), but that's not the concept I'm struggling with. For the I/O routines, I am envisioning applying simple inheritance: Abstract I/O class defines read & write functions to fill in the "myModel" object HDF4 derived class HDF5 HDF5 using parallel IO NetCDF etc... The code should read data in any of these formats, then write out to any of these formats. In the past, I would add an AbstractIO member to myModel and create/destroy this object depending on which I/O scheme I want. In this way, I could do something like: myModelObj.ioObj->read('input.hdf') myModelObj.ioObj->write('output.hdf') I have a bit of OOP experience but very little on the Design Patterns front, so I recently acquired the Gang of Four book "Design Patterns: Elements of Reusable Object-Oriented Software". OOP designers: Which pattern(s) would you recommend I use to integrate I/O with the myModel object? I am interested in answering this for two reasons: To learn more about design patterns in general Apply what I learn to help refactor an large old crufty/legacy physics code to be more human-readable & extensible. I am leaning towards applying the Decerator pattern to myModel, so I can attach the I/O responsibilities dynamically to myModel (i.e. whether to use HDF4, HDF5, etc.). However, I don't feel very confident that this is the best pattern to apply. Reading the Gang of Four book cover-to-cover before I start coding feels like a good way to develop an unhealthy caffeine addiction. What patterns do you recommend?

    Read the article

  • Asynchronous Delegates Vs Thread/ThreadPool?

    - by claws
    Hello, I need to execute 3 parallel tasks and after completion of each task they should call the same function which prints out the results. I don't understand in .net why we have Asychronous calling (delegate.BeginInvoke() & delegate.EndInvoke()) as well as Thread class? I'm little confused which one to use when? Now in this particular case, what should I use Asychronous calling or Thread class? I'm using C#.

    Read the article

  • Multithreading consulting service

    - by Gustavo Paulillo
    Hello. I am creating a service, that needs to perform the following tasks: consult bank services and persist data into DB. The dificult is: Its needed to execute each process in parallel. I mean the better choice is implementing a multithreading service, running each instance per thread. But how its done? Thanks

    Read the article

  • Flex unit testing with ANT vs Flash Builder 4

    - by peterlindstrom21
    I have just tried setup unit testing in Flash Builder 4, and it working nicely. A setup of a parallel test source structure and using Flash Builder 4:s new TestCase and new TestSuite I was up and running with some testcases within minutes. But now I want to compile them from a ant flex task, the Flash Builder generates FlexUnitApplication.mxml and FlexUnitCompilerApplication.mxml. Is there a nice way to build the unit tests with ant using these? I cant find any sample where this is done.

    Read the article

  • VB.NET - using textfile as source for menus and textboxes

    - by Kenny Bones
    Hi, this is probably a bit tense and I'm not sure if this is possible at all. But basically, I'm trying to create a small application which contains alot of PowerShell-code which I want to run in an easy matter. I've managed to create everything myself and it does work. But all of the PowerShell code is manually hardcoded and this gives me a huge disadvantage. What I was thinking was creating some sort of dynamic structure where I can read a couple of text files (possible a numerous amount of text files) and use these as the source for both the comboboxes and the richtextbox which provovides as the string used to run in PowerShell. I was thinking something like this: Combobox - "Choose cmdlet" - Use "menucode.txt" as source Richtextbox - Use "code.txt" as source But, the thing is, Powershell snippets need a few arguments in order for them to work. So I've got a couple of comboboxes and a few textboxes which provides as input for these arguments. And this is done manually as it is right now. So rewriting this small application should also search the textfile for some keywords and have the comboboxes and textboxes to replace those keywords. And I'm not sure how to do this. So, would this requre a whole lot of textfiles? Or could I use one textfile and separate each PowerShell cmdlet snippets with something? Like some sort of a header? Right now, I've got this code at the eventhandler (ComboBox_SelectedIndexChanged) If ComboBoxFunksjon.Text = "Set attribute" Then TxtBoxUsername.Visible = True End If If chkBoxTextfile.Checked = True Then If txtboxBrowse.Text = "" Then MsgBox("You haven't choses a textfile as input for usernames") End If LabelAttribute.Visible = True LabelUsername.Visible = False ComboBoxAttribute.Visible = True TxtBoxUsername.Visible = False txtBoxCode.Text = "$users = Get-Content " & txtboxBrowse.Text & vbCrLf & "foreach ($a in $users)" & vbCrLf & "{" & vbCrLf & "Set-QADUser -Identity $a -ObjectAttributes @{" & ComboBoxAttribute.SelectedItem & "='" & TxtBoxValue.Text & "'}" & vbCrLf & "}" If ComboBoxAttribute.SelectedItem = "Outlook WebAccess" Then TxtBoxValue.Visible = False CheckBoxValue.Visible = True CheckBoxValue.Text = "OWA Enabled?" txtBoxCode.Text = "$users = Get-Content " & txtboxBrowse.Text & vbCrLf & "foreach ($a in $users)" & vbCrLf & "{" & vbCrLf & "Set-CASMailbox -Identity $a -OWAEnabled" & " " & "$" & CheckBoxValue.Checked & " '}" & vbCrLf & "}" End If If ComboBoxAttribute.SelectedItem = "MobileSync" Then TxtBoxValue.Visible = False CheckBoxValue.Visible = True CheckBoxValue.Text = "MobileSync Enabled?" Dim value If CheckBoxValue.Checked = True Then value = "0" Else value = "7" End If txtBoxCode.Text = "$users = Get-Content " & txtboxBrowse.Text & vbCrLf & "foreach ($a in $users)" & vbCrLf & "{" & vbCrLf & "Set-QADUser -Identity $a -ObjectAttributes @{msExchOmaAdminWirelessEnable='" & value & " '}" & vbCrLf & "}" End If Else LabelAttribute.Visible = True LabelUsername.Visible = True ComboBoxAttribute.Visible = True txtBoxCode.Text = "Set-QADUser -Identity " & TxtBoxUsername.Text & " -ObjectAttributes @{" & ComboBoxAttribute.SelectedItem & "='" & TxtBoxValue.Text & " '}" If ComboBoxAttribute.SelectedItem = "Outlook WebAccess" Then TxtBoxValue.Visible = False CheckBoxValue.Visible = True CheckBoxValue.Text = "OWA Enabled?" txtBoxCode.Text = "Set-CASMailbox " & TxtBoxUsername.Text & " -OWAEnabled " & "$" & CheckBoxValue.Checked End If If ComboBoxAttribute.SelectedItem = "MobileSync" Then TxtBoxValue.Visible = False CheckBoxValue.Visible = True CheckBoxValue.Text = "MobileSync Enabled?" Dim value If CheckBoxValue.Checked = True Then value = "0" Else value = "7" End If txtBoxCode.Text = "Set-QADUser " & TxtBoxUsername.Text & " -ObjectAttributes @{msExchOmaAdminWirelessEnable='" & value & "'}" End If End If Now, this snippet above lets me either use a text file as a source for each username used in the powershell snippet. Just so you know :) And I know, this is probably coded as stupidly as it gets. But it does work! :)

    Read the article

  • Persistent (purely functional) Red-Black trees on disk performance

    - by Waneck
    I'm studying the best data structures to implement a simple open-source object temporal database, and currently I'm very fond of using Persistent Red-Black trees to do it. My main reasons for using persistent data structures is first of all to minimize the use of locks, so the database can be as parallel as possible. Also it will be easier to implement ACID transactions and even being able to abstract the database to work in parallel on a cluster of some kind. The great thing of this approach is that it makes possible implementing temporal databases almost for free. And this is something quite nice to have, specially for web and for data analysis (e.g. trends). All of this is very cool, but I'm a little suspicious about the overall performance of using a persistent data structure on disk. Even though there are some very fast disks available today, and all writes can be done asynchronously, so a response is always immediate, I don't want to build all application under a false premise, only to realize it isn't really a good way to do it. Here's my line of thought: - Since all writes are done asynchronously, and using a persistent data structure will enable not to invalidate the previous - and currently valid - structure, the write time isn't really a bottleneck. - There are some literature on structures like this that are exactly for disk usage. But it seems to me that these techniques will add more read overhead to achieve faster writes. But I think that exactly the opposite is preferable. Also many of these techniques really do end up with a multi-versioned trees, but they aren't strictly immutable, which is something very crucial to justify the persistent overhead. - I know there still will have to be some kind of locking when appending values to the database, and I also know there should be a good garbage collecting logic if not all versions are to be maintained (otherwise the file size will surely rise dramatically). Also a delta compression system could be thought about. - Of all search trees structures, I really think Red-Blacks are the most close to what I need, since they offer the least number of rotations. But there are some possible pitfalls along the way: - Asynchronous writes -could- affect applications that need the data in real time. But I don't think that is the case with web applications, most of the time. Also when real-time data is needed, another solutions could be devised, like a check-in/check-out system of specific data that will need to be worked on a more real-time manner. - Also they could lead to some commit conflicts, though I fail to think of a good example of when it could happen. Also commit conflicts can occur in normal RDBMS, if two threads are working with the same data, right? - The overhead of having an immutable interface like this will grow exponentially and everything is doomed to fail soon, so this all is a bad idea. Any thoughts? Thanks! edit: There seems to be a misunderstanding of what a persistent data structure is: http://en.wikipedia.org/wiki/Persistent_data_structure

    Read the article

< Previous Page | 90 91 92 93 94 95 96 97 98 99 100 101  | Next Page >