Search Results

Search found 19525 results on 781 pages for 'say'.

Page 326/781 | < Previous Page | 322 323 324 325 326 327 328 329 330 331 332 333  | Next Page >

• Remote Desktop Services In A Virtual VMWare Environment

  - by Christopher W. Szabo

  I have a quick question regarding Microsoft Remote Desktop Services in a virtualized environment using VMWare. This environment will actually be hosted in a large data center with in a cloud that is offered. This particular data center has the ability to establish high speed point to point connections with customers via metro-ethernet who are hosted in the cloud. The result is that customers can actually host their corporate domain in the data center's cloud. Put the merits of such a configuration aside for the time being. Believe me when I say that the cloud is stable and had enough hardware behind it to rival a dedicated cabinet. My question has to do with RDS in a virtual environment, which would amount to virtual desktops hosted on a virtual server. I've read that this works without issue using Hyper-V and VMWare. But before I take the plunge I wanted to get some feedback from the community.

  Read the article

• How to indicate reliability when reporting availability of competencies

  - by Jan Doggen

  We have employees with competencies: Pete Welder Carpenter Melissa Carpenter Assume they both work 40 hours/week, and have not yet been assigned work. We need to report the availability of these competencies, expressed in hours. As far as I can see now, we can report this in two ways: Method A. When someone has multiple competencies, count them both. Welder 40 hours Carpenter 80 hours Method B. When someone has multiple competencies, count an equal division of hours for each Welder 20 hours Carpenter 60 hours Method A has our preference: - A good planner will know to plan the least available competency first. If 30 hours of welding is planned, we will be left with 10 welder, 50 carpenter. - Method B has the disadvantage that the planner thinks he cannot plan the job when 30 hours of welding is required. However, if we report this we would like to give an estimate of the reliability of the numbers for each competency, i.e. how much are these over-reported? In my example A, would I say that carpenter is 100% over-reported, or 50%, or maybe another number? How would I calculate this for large numbers of competencies? I'm sure we are not the first ones dealing with this, is there a 'usual' way of doing this in planning? Additionally: - Would there be an even better method than A or B? - Optionally, we also have an preference order of competencies (like: use him/her in this order), Pete could be 1. welder 2. carpenter. Does this introduce new options?

  Read the article

• More Tables or More Databases?

  - by BuckWoody

  I got an e-mail from someone that has an interesting situation. He has 15,000 customers, and he asks if he should have a database for their data per customer. Without a LOT more data it’s impossible to say, of course, but there are some general concepts to keep in mind. Whenever you’re segmenting data, it’s all about boundary choices. You have not only boundaries around how big the data will get, but things like how many objects (tables, stored procedures and so on) that will be involved, if there are any cross-sections of data (do they share location or product information) and – very important – what are the security requirements? From the answer to these types of questions, you now have the choice of making multiple tables in a single database, or using multiple databases. A database carries some overhead – it needs a certain amount of memory for locking and so on. But it has a very clean boundary – everything from objects to security can be kept apart. Having multiple users in the same database is possible as well, using things like a Schema. But keeping 15,000 schemas can be challenging as well. My recommendation in complex situations like this is similar to a post on decisions that I did earlier – I lay out the choices on a spreadsheet in rows, and then my requirements at the top in the columns. I  give each choice a number based on how well it meets each requirement. At the end, the highest number wins. And many times it’s a mix – perhaps this person could segment customers into larger regions or districts or products, in a database. Within that database might be multiple schemas for the customers. Of course, he needs to query across all customers, that becomes another requirement. Share this post: email it! | bookmark it! | digg it! | reddit! | kick it! | live it!

  Read the article

• What is the ideal length of a method?

  - by iPhoneDeveloper

  In object-oriented programming, there is no exact rule on the maximum length of a method , but I still found these two qutes somewhat contradicting each other, so I would like to hear what you think. In Clean Code: A Handbook of Agile Software Craftsmanship, Robert Martin says: The first rule of functions is that they should be small. The second rule of functions is that they should be smaller than that. Functions should not be 100 lines long. Functions should hardly ever be 20 lines long. and he gives an example from Java code he sees from Kent Beck: Every function in his program was just two, or three, or four lines long. Each was transparently obvious. Each told a story. And each led you to the next in a compelling order. That’s how short your functions should be! This sounds great, but on the other hand, in Code Complete, Steve McConnell says something very different: The routine should be allowed to grow organically up to 100-200 lines, decades of evidence say that routines of such length no more error prone then shorter routines. And he gives a reference to a study that says routines 65 lines or long are cheaper to develop. So while there are diverging opinions about the matter, is there a functional best-practice towards determining the ideal length of a method for you?

  Read the article

• How soon does nginx's token bucket replenish when limiting at requests per minute?

  - by Michael Gorsuch

  Hi all. We've decided that we want to experiment and limit requests per minute instead of requests per second on our sites. However, I am confused by the burst parameter in this context. I am under the impression that when you use the 'nodelay' flag, the rate limiting facility acts like a token bucket instead of a leaky bucket. That being the case, the bucket size is equal to the burst parameter, and every time that you violate the policy (say 1 req/s), you have to put a token in the bucket. Once the bucket is full (being equal to the burst setting), you are given a 503 error page. I am also under the impression that once a violator stops going against the policy, a token is removed from the bucket at a rate of 1 token/s allowing him to regain access to the site. Assuming that I have the above correct, my question is what happens when I start regulating access per minute? If we chose 60 requests per minute, at what rate does the token bucket replenish?

  Read the article

• How to manage security of these self hosted web apis, to ensure that the request coming for accessing data is authenticated?

  - by Husrat Mehmood

  Let's pretend I am going to work on an enterprise application. Say I have 11 modules in the application and I would have to develop Dashboards for every role in the organization for whom I are going to develop application. We Decided to use Asp.Net Web Api and return json data from our apis. We are going to include 11 Self hosted web apis projects in our application (one self hosted web api) for every module. All 11 modules are connected to one Sql server 2012 Database. Then once api is ready we would have to create Business Dashboards (Based upon roles in Organization). So Now my web api client is Asp.Net Mvc application.Asp.Net mvc will consume those web apis. Here is the part for whom all explanation is done. How should I manage Security of all 11 self hosted web apis? How should I only authenticated request is coming? If I authenticate user by login and password and then redirect user to appropriate Dashboard designed for the role that user have and load data by consuming web apis. How should I ensure that the request coming for accessing data is authenticated?

  Read the article

• How to view status of software RAID 1 resynching?

  - by tputkonen

  I have two 500 GB disks and yesterday I mirrored first drive to the second one using software RAID 1. PC has now been on for 30 hours. Both disks say "Resynching", but there is no progress indicator. In addition, there is a small yellow exclamation mark on both disks. My questions are: How long could the synch take for 500 GB drive with about 150 GB of data? PC has 4 GBs of RAM and AMD dual core 4000+ Is there a way to monitor status of the synching? How can I check what the exclamation mark means?

  Read the article

• Understanding top output in Linux

  - by Rayne

  Hi, I'm trying to determine the CPU usage of a program by looking at the output from Top in Linux. I understand that %us means userspace and %sy means system/kernel etc. But say I see 100%us. Does this mean that the CPU is really only doing useful work? What if a CPU is tied up waiting for resources that are not avaliable, or cache misses, would it also show up in the %us column, or any other column? Thank you.

  Read the article

• How can I tag sequences inside videos?

  - by Antoine

  For example, let's say I find a sequence in a WWDC session (Apple videos) that explains a technical point very well. I would like to be able to store the exact range in the video, tag it with keywords, and manage all that information. And when I want to share this information, I could simply export to a plain English sentence: "Have a look at the video named …". It would be great to be able to launch a video player for local videos and go straight to the beginning of the sequence.

  Read the article

• How can I slightly delay the pop up of the task bar?

  - by Xavierjazz

  Windows XP SP3 Many times as I head to the bottom of my desktop, I will slightly overshoot the target at the bottom and the task bar will pop up, hiding the target. I then have to move the cursor out of it so it retreats and then again try to access the target. Is there a way to add an interval, say, 1 second, to the task bar popping up so I can adjust to the target before the task bar covers it? EDIT: as per my answer below, "What I ended up doing is just docking it on the LH side of the screen. There is no change in response but I don't go to that location so often so it's much better.".

  Read the article

• How can I unit test rendering output?

  - by stephelton

  I've been embracing Test-Driven Development (TDD) recently and it's had wonderful impacts on my development output and the resiliency of my codebase. I would like to extend this approach to some of the rendering work that I do in OpenGL, but I've been unable to find any good approaches to this. I'll start with a concrete example so we know what kinds of things I want to test; lets say I want to create a unit cube that rotates about some axis, and that I want to ensure that, for some number of frames, each frame is rendered correctly. How can I create an automated test case for this? Preferably, I'd even be able to write a test case before writing any code to render the cube (per usual TDD practices.) Among many other things, I'd want to make sure that the cube's size, location, and orientation are correct in each rendered frame. I may even want to make sure that the lighting equations in my shaders are correct in each frame. The only remotely useful approach to this that I've come across involves comparing rendered output to a reference output, which generally precludes TDD practice, and is very cumbersome. I could go on about other desired requirements, but I'm afraid the ones I've listed already are out of reach.

  Read the article

• Unable to mount external hard drive - Damaged file system and MFT

  - by Khalifa Abbas Lame

  I get the following error when i try to mount my external hard drive. UNABLE TO MOUNT Error mounting /dev/sdc1 at /media/khalibloo/Khalibloo2: Command-line `mount -t "ntfs" -o "uhelper=udisks2,nodev,nosuid,uid=1000,gid=1000,dmask=0077,fmask=0177" "/dev/sdc1" "/media/khalibloo/Khalibloo2"' exited with non-zero exit status 13: ntfs_attr_pread_i: ntfs_pread failed: Input/output error Failed to read of MFT, mft=6 count=1 br=-1: Input/output error Failed to open inode FILE_Bitmap: Input/output error Failed to mount '/dev/sdc1': Input/output error NTFS is either inconsistent, or there is a hardware fault, or it's a SoftRAID/FakeRAID hardware. In the first case run chkdsk /f on Windows then reboot into Windows twice. The usage of the /f parameter is very important! If the device is a SoftRAID/FakeRAID then first activate it and mount a different device under the /dev/mapper/ directory, (e.g. /dev/mapper/nvidia_eahaabcc1). Please see the 'dmraid' documentation for more details. It doesn't mount on windows either: "I/O Device error" it's an ntfs hard drive with a single partition Of course, i tried chkdsk /f. it reported several file segments as unreadable, but didn't say whether it fixed them or not (apparently not). also tried with the /b flag. ntfsfix reported the volume as corrupt. TestDisk was able to fix a small error with the partition table by adding the "80" flag for the active (only) partition. TestDisk also confirmed that the boot sector was fine and it matched the backup. However, when attempting to repair the MFT, it couldn't read the MFT. It also couldn't list the files on the hard drive. It says file system may be damaged. Active@ also shows that MFT is missing or corrupt. So how do i fix the file system? or the MFT?

  Read the article

• The Endeca UI Design Pattern Library Returns

  - by Joe Lamantia

  I'm happy to announce that the Endeca UI Design Pattern Library - now titled the Endeca Discovery Pattern Library - is once again providing guidance and good practices on the design of discovery experiences.  Launched publicly in 2010 following several years of internal development and usage, the Endeca Pattern Library is a unique and valued source of industry-leading perspective on discovery - something I've come to appreciate directly through  fielding the consistent stream of inquiries about the library's status, and requests for its rapid return to public availability. Restoring the library as a public resource is only the first step!  For the next stage of the library's evolution, we plan to increase the scope of the guidance it offers beyond user interface design to the broader topic of discovery.  This could include patterns for architecture at the systems, user experience, and business levels; information and process models; analytical method and activity patterns for conducting discovery; and organizational and resource patterns for provisioning discovery capability in different settings.  We'd like guidance from the community on the kinds of patterns that are most valuable - so make sure to let us know. And we're also considering ways to increase the number of patterns the library offers, possibly by expanding the set of contributors and the authoring mechanisms. If you'd like to contribute, please get in touch. Here's the new address of the library: http://www.oracle.com/goto/EndecaDiscoveryPatterns And I should say 'Many thanks' to the UXDirect team and all the others within the Oracle family who helped - literally - keep the library alive, and restore it as a public resource.

  Read the article

• FULL LIST Difference Edubuntu over Ubuntu

  - by Richard Kemp

  I have always used Edubuntu rather than Ubuntu because I have many children including home educated ones who casually use the education offerings. I appreciate many but not all of the extra offerings. I never got beyond 8.04.4 LTS, so am considering updating. I note now with Edubuntu 12.04 it is 2GB i.e.4 times bigger, even apparently using squashfs. I do now have to consider whether the extra resources are worth it in each computer in my household. I am sick of reading over and over again 'Edubuntu is Unbuntu with extras such as blah blah', and in a different place the same thing but with a different blah blah. I downloaded Edubuntu 12.04 and ran it 'live'. Can't say I got on at all with the Unity Interface (this can be changed when installed), but the apparent education stuff seemed quite small and pathetic, I am obviously not finding it. Is there anywhere a FULL LIST of what Edubuntu has pre-installed has over Ubuntu, so a sensible decision can be made whether to go for it, or just add packages that look relevant to me to Ubuntu. Is there truly no other difference between Unbuntu/Edubuntu but the extra educational packages? (I for example have no school server, but if I understood exactly what is involved, with the number of computers in the house, it may be worth creating one!) Please advise. Thank You.

  Read the article

• A simple Volume Replication Tool for large data set?

  - by Jin

  I'm looking for a solution to the following: Server A (Site A) - Win 2008 R2 - approx 10TB (15TB max) of data - well over 8 million files Server B (Site B) - Win 2008 R2 I want to assynchronously replicate Server A's volume to a volume on Server B for data redundancy. Something that I can say to my users, "go here for data" when/if Server A goes belly up due to machine problems, disaster, etc. Windows 2008 R2 does have DFS, but microsoft does not apparently support this large of a dataset (or more accurately, more than 8 million files - according to the docs I could find). I also looked at Veritas Volume Replication, but this seems almost too much as I would also require Veritas Volume Manager. There are numerous "back-up" software which makes a 1-1 backup, which would be ok, but since it will be transfering over internet, I'd like something that has compression during transfer like DFS has. Does anyone have any suggestions regarding this?

  Read the article

• Is it possible to rate-limit an scp/sftp/rsync transfer from the command-line? ie, manual QoS on a s

  - by warren

  Specifically, I am looking to rate-limit an scp or sftp session in the call itself. For example, let's say I want to copy 100MB to one server, and 1GB to another. I'd like to be able to run both of these at the same time, but maintain a QoS for "normal" computer usage - somewhat similar to how you can rate-limit bittorrent. Is there a way to do this without touching the networking hardware? I'm envisioning something akin to: magic-qos-tool 'scp file user@host:/path/to/file' Or.. scp -rate 40kbps file user@host:/path/to/file

  Read the article

• PHP file_put_contents File Locking

  - by hozza

  The Senario: You have a file with a string (average sentence worth) on each line. For arguments sake lets say this file is 1Mb in size (thousands of lines). You have a script that reads the file, changes some of the strings within the document (not just appending but also removing and modifying some lines) and then overwrites all the data with the new data. The Questions: Does 'the server' PHP, OS or httpd etc. already have systems in place to stop issues like this (reading/writing half way through a write)? i. If it does, please explain how it works and give examples or links to relevant documentation. ii. If not, are there things I can enable or set-up, such as locking a file until a write is completed and making all other reads and/or writes fail until the previous script has finished writing? My Assumptions and Other Information: The server in question is running PHP and Apache or Lighttpd. If the script is called by one user and is halfway through writing to the file and another user reads the file at that exact moment. The user who reads it will not get the full document, as it hasn't been written yet. (If this assumption is wrong please correct me) I'm only concerned with PHP writing and reading to a text file, and in particular, the functions "fopen"/"fwrite" and mainly "file_put_contents". I have looked at the "file_put_contents" documentation but have not found the level of detail or a good explanation of what the "LOCK_EX" flag is or does. The senario is an EXAMPLE of a worst case senario where I would assume these issues are more likely to occur, due to the large size of the file and the way the data is edited. I want to learn more about these issues and don't want or need answers or comments such as "use mysql" or "why are you doing that" because I'm not doing that, I just want to learn about file read/writing with PHP and don't seem to be looking in the right places/documentation and yes I understand PHP is not the perfect language for working with files in this way...

  Read the article

• App installed in ~/usr launches from terminal but not Applications menu (or why does setting ld_library_path in .profile not work as it should)

  - by levesque

  I have built and installed an application under a directory of my choosing, let's say under /home/jim/usr, so files have been put in three-four folders, all under this $HOME/usr folder (e.g., bin, include, lib, share, etc.). I can launch this application from the command line just fine as I added the proper paths to my environement variables PATH and LD_LIBRARY_PATH in ~/.bashrc. I added the same paths to the ~/.profile file, which, if I'm not mistaken, is supposed to be parsed by Ubuntu. Doesn't work. Nothing. Where can I go from there? EDIT: I logged out/in and restarted my computer. Both didn't change a thing. The problem seems to come from the fact that no matter what I do the LD_LIBRARY_PATH environment variable is not properly passed to Ubuntu. Using log files, I found that the application I'm trying to run in this example doesn't find one it's dependencies located in ~/usr/lib. One solution would be to add the /home/jim/usr/lib folder inside a file located in /etc/ld.so.conf.d/, but I don't have admin rights on this machine. Making a wrapper script like this one works: #!/bin/bash export LD_LIBRARY_PATH=$HLOC/usr/lib application &> $HOME/application_messages.log but that would force me to wrap all my home compiled applications with this script. Any ideas? Why does Ubuntu/Gnome remove the LD_LIBRARY_PATH environment variable from my set variables? Is it because trying to do this is bad practice? UPDATE (and solution): As found by Christopher, there is a bug report about this on launchpad. LD_LIBRARY_PATH is unset after parsing of the ~/.profile file. See the bug report. Seems the only solution for now is to make a wrapper script.

  Read the article

• How best to keep bumbling, non-technical managers at bay and still deliver good work?

  - by Curious

  This question may be considered subjective (I got a warning) and be closed, but I will risk it, as I need some good advice/experience on this. I read the following at the 'About' page of Fog Creek Software, the company that Joel Spolsky founded and is CEO of: Back in the year 2000, the founders of Fog Creek, Joel Spolsky and Michael Pryor, were having trouble finding a place to work where programmers had decent working conditions and got an opportunity to do great work, without bumbling, non-technical managers getting in the way. Every high tech company claimed they wanted great programmers, but they wouldn’t put their money where their mouth was. It started with the physical environment (with dozens of cubicles jammed into a noisy, dark room, where the salespeople shouting on the phone make it impossible for developers to concentrate). But it went much deeper than that. Managers, terrified of change, treated any new idea as a bizarre virus to be quarantined. Napoleon-complex junior managers insisted that things be done exactly their way or you’re fired. Corporate Furniture Police writhed in agony when anyone taped up a movie poster in their cubicle. Disorganization was so rampant that even if the ideas were good, it would have been impossible to make a product out of them. Inexperienced managers practiced hit-and-run management, issuing stern orders on exactly how to do things without sticking around to see the farcical results of their fiats. And worst of all, the MBA-types in charge thought that coding was a support function, basically a fancy form of typing. A blunt truth about most of today's big software companies! Unfortunately not every developer is as gutsy (or lucky, may I say?) as Joel Spolsky! So my question is: How best to work with such managers, keep them at bay and still deliver great work?

  Read the article

• Heartbeat won't start up from a cold boot when a failed node is present

  - by Matthew

  I currently have two ubuntu servers running Heartbeat and DRBD. Let's say one node is down... The servers are directory connected with a 1000Mbps cross over cable on eth1 and have access to a IP camera LAN on eth0 The node that is still functioning won't start up heartbeat and provide access to the drbd resource. I have to manually restart heartbeat by "sudo service heartbeat restart" to get everything up and running. How can I get it to start fine from a cold start? Here is the my ha.cf and some material from the syslog... If I'm missing any information that might be of some help. http://pastebin.com/rGvzVSUq <--- Syslog http://pastebin.com/VqpaPSb5 <--- ha.cf

  Read the article

• NTFS Permission Structure to allow Traversal but no Modification except in Leaf Nodes?

  - by pepoluan

  Assume there's this folder structure: D:\ --+-- Acctg --+-- Payable | +-- Receivable | +-- Fin --+-- Inv | +-- Tax | +-- Treas | +-- Mrktg --+-- Ads +-- Promo Users are not allowed to change the structure, but they are free to create & delete files & folders in the leaf nodes (i.e., the rightmost folders). AGDLP principle said that I should assign permissions on the above folders to DL-Groups. Let's say I have a G-Group of users, G-Accounting-Payable, containing users that have access to the D:\Acctg\Payable folder. The way I see it, I have two strategies: - Strategy 1 Create three DL-Groups and assign them permissions: DL-D-Acctg_T -- allowed traversal of D:\Acctg folder DL-D-Acctg-Pay_LF -- allowed listing of D:\Acctg\Payable folder contents DL-D-Acctg-Pay__RW -- allowed full permissions to the contents of D:\Acctg\Payable folder Add G-Accounting-Payable as member to all the above DL-Groups - Strategy 2 Create just one DL-Group DL-D-Acctg-Pay__RW, and assign it the proper permissions for each level of the folder. Then, add G-Accounting-Payable as member to that DL-Group. - Which strategy is the Recommended Best Practice, and why?

  Read the article

• geomipmapping using displacement mapping (and glVertexAttribDivisor)

  - by Will

  I wake up with a clear vision, but sadly my laptop card doesn't do displacement mapping nor glVertexAttribDivisor so I can't test it out; I'm left sharing here: With geomipmapping, the grid at any factor is transposable - if you pass in an offset - say as a uniform - you can reuse the same vertex and index array again and again. If you also pass in the offset into the heightmap as a uniform, the vertex shader can do displacement mapping. If the displacement map is mipmapped, you get the advantages of trilinear filtering for distant maps. And, if the scenery is closer, rather than exposing that the you have a world made out of quads, you can use your transposable grid vertex array and indices to do vertex-shader interpolation (fancy splines) to do super-smooth infinite zoom? So I have some questions: does it work? In theory, in practice? does anyone do it? Does this technique have a name? Papers, demos, anything I can look at? does glVertexAttribDivisor mean that you can have a single glMultiDrawElementsEXT or similar approach to draw all your terrain tiles in one call rather than setting up the uniforms and emitting each tile? Would this offer any noticeable gains? does a heightmap that is GL_LUMINANCE take just one byte per pixel(=vertex)? (On mainstream cards, obviously. Does storage vary in practice?) Does going to the effort of reusing the same vertices and indices mean that you can basically fill the GPU RAM with heightmap and not a lot else, giving you either bigger landscapes or more detailed landscapes/meshes for the same bang? is mipmapping the displacement map going to work? On future cards? Is it going to introduce unsurmountable inaccuracies if it is enabled?

  Read the article

• Is a matching entry in /etc/hosts required for hostname?

  - by JohnWoltman

  I was installing a Tomcat webapp that refused to work until I stumbled on someone else's issue with an unrelated product. The solution was to add the machine's name to /etc/hosts, to match the name returned by hostname. Is this required for general Linux networking to function correctly? My webapp is running in a virtual machine so that I can test the webapp, and I don't normally bother with the /etc/hosts file on VMs. I just shook my fist and cursed Tomcat and webapp's behavior. I read http://serverfault.com/questions/118823, but that doesn't say if it's required or not.

  Read the article

• multi player client server game - polling - wamp

  - by stonebold

  Let's say I make a client-server application. A simple game for example. Where each client polls the server every half a minute. How many clients is it possible to have before it overlaods a wamp server? Basically how robust is Apache for this kind of stuff? Getting a request, aggregating data from mysql server, and then returning the data in an xml format. What solution should I use for my case? Thank you in advance.

  Read the article

• Duplication of code (backend and javascript - knockout)

  - by Michal B.

  We have a new developer in our team. He seems a smart guy (he just came in so I cannot really judge). He started with implementing some small enhancements in the project (MVC3 web application using javascript - jquery and knockout). Let's say we have two values: A - quite complex calculation C - constant B = A + C On the screen there is value B and user can change it (normal texbox). When B changes, A changes as well because C is constant. So there is linear dependency between A and B. Now, all the calculations are done in the backend, but we need to recalculate A as user changes B (in js, I would use knockout). I thought about storing old A and B and when B changes by 10 then we know that new A will be old A + 10. He says this is dirty, because it's duplication of code (we make use of the fact that they are dependent and according to him that should be only in one place in our app). I understand it's not ideal, but making AJAX request after every key press seems a bit too much. It's a really small thing and I would not post if we haven't had long discussion about it. How do you deal with such problems? Also I can imagine that using knockout implies lots of calculations on the client side, which very often leads to duplication of the same calculations from the backend. Does anyone have links to some articles/thoughts on this topic?

  Read the article

< Previous Page | 322 323 324 325 326 327 328 329 330 331 332 333  | Next Page >