Search Results

Search found 1657 results on 67 pages for 'writes on'.

Page 2/67 | < Previous Page | 1 2 3 4 5 6 7 8 9 10 11 12  | Next Page >

• Linux buffer cache effect on IO writes?

  - by Patrick LeBoutillier

  Hi, I'm copying large files (3 x 30G) between 2 filesystems on a Linux server (kernel 2.6.37, 16 cores, 32G RAM) and I'm getting poor performance. I suspect that the usage of the buffer cache is killing the I/O performance. To try and narrow down the problem I used fio directly on the SAS disk to monitor the performance. Here is the output of 2 fio runs (the first with direct=1, the second one direct=0): Config: [test] rw=write blocksize=32k size=20G filename=/dev/sda # direct=1 Run 1: test: (g=0): rw=write, bs=32K-32K/32K-32K, ioengine=sync, iodepth=1 Starting 1 process Jobs: 1 (f=1): [W] [100.0% done] [0K/205M /s] [0/6K iops] [eta 00m:00s] test: (groupid=0, jobs=1): err= 0: pid=4667 write: io=20,480MB, bw=199MB/s, iops=6,381, runt=102698msec clat (usec): min=104, max=13,388, avg=152.06, stdev=72.43 bw (KB/s) : min=192448, max=213824, per=100.01%, avg=204232.82, stdev=4084.67 cpu : usr=3.37%, sys=16.55%, ctx=655410, majf=0, minf=29 IO depths : 1=100.0%, 2=0.0%, 4=0.0%, 8=0.0%, 16=0.0%, 32=0.0%, >=64=0.0% submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0% complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0% issued r/w: total=0/655360, short=0/0 lat (usec): 250=99.50%, 500=0.45%, 750=0.01%, 1000=0.01% lat (msec): 2=0.01%, 4=0.02%, 10=0.01%, 20=0.01% Run status group 0 (all jobs): WRITE: io=20,480MB, aggrb=199MB/s, minb=204MB/s, maxb=204MB/s, mint=102698msec, maxt=102698msec Disk stats (read/write): sda: ios=0/655238, merge=0/0, ticks=0/79552, in_queue=78640, util=76.55% Run 2: test: (g=0): rw=write, bs=32K-32K/32K-32K, ioengine=sync, iodepth=1 Starting 1 process Jobs: 1 (f=1): [W] [100.0% done] [0K/0K /s] [0/0 iops] [eta 00m:00s] test: (groupid=0, jobs=1): err= 0: pid=4733 write: io=20,480MB, bw=91,265KB/s, iops=2,852, runt=229786msec clat (usec): min=16, max=127K, avg=349.53, stdev=4694.98 bw (KB/s) : min=56013, max=1390016, per=101.47%, avg=92607.31, stdev=167453.17 cpu : usr=0.41%, sys=6.93%, ctx=21128, majf=0, minf=33 IO depths : 1=100.0%, 2=0.0%, 4=0.0%, 8=0.0%, 16=0.0%, 32=0.0%, >=64=0.0% submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0% complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0% issued r/w: total=0/655360, short=0/0 lat (usec): 20=5.53%, 50=93.89%, 100=0.02%, 250=0.01%, 500=0.01% lat (msec): 2=0.01%, 4=0.01%, 10=0.01%, 20=0.01%, 50=0.12% lat (msec): 100=0.38%, 250=0.04% Run status group 0 (all jobs): WRITE: io=20,480MB, aggrb=91,265KB/s, minb=93,455KB/s, maxb=93,455KB/s, mint=229786msec, maxt=229786msec Disk stats (read/write): sda: ios=8/79811, merge=7/7721388, ticks=9/32418456, in_queue=32471983, util=98.98% I'm not knowledgeable enough with fio to interpret the results, but I don't expect the overall performance using the buffer cache to be 50% less than with O_DIRECT. Can someone help me interpret the fio output? Are there any kernel tunings that could fix/minimize the problem? Thanks a lot,

  Read the article

• Access writes issue in MAC

  - by user594738

  Hi All, I have a MAC machine running MAC OS Version 10.6.6. I have added my mac machine to a domain controller (which runs Windows version of server) and allow users from domain controller to login. Because of our server migration, we removed the mac machine from old domain controller and added it to the new domain controller. Users in the old domain controller are copied to new domain controller. When I login into mac (which is added to new domain controller) with the domain controller user credentials, I was not able to access my desktop folders and any other folders in my users directory. It says "The folder Desktop can't be opened because you don't have permission its contents" Any idea why I was not able to access my home directory and how can I resolve this issue...!

  Read the article

• Server nearly unusable when doing disk writes

  - by Wikser

  My question closely relates to my last question here on serverfault. I was copying about 5GB from a 10 year old desktop computer to the server. The copy was done in Windows Explorer. In this situation I would assume the server to be bored by the dataflow. But as usual with this server, it really slowed down. At least I could work with the remote session, even there was some serious latency. The copy took its time (20min?). In this time I went to a colleague and he tried to log in in the same server via remote desktop (for some other reason). It took about a minute to get to the login screen, a minute to open the control panel, a minute to open the performance monitor, ... Icons were loading maybe one per second. We saw the following (from memory): CPU: 2% Avg. Queue Length: 50 Pages/sec: 115 (?) There was no other considerable activity on the server. The server seldom serves some ASP.NET pages, which became also very slow in this time. The relevant configuration is as follows: Windows 2003 SEAGATE ST3500631NS (7200 rpm, 500 GB) LSI MegaRAID based RAID 5 4 disks, 1 hot spare Write Through No read-ahead Direct Cache Mode Harddisk-Cache-Mode: off Is this normal behaviour for such a configuration? What measurements could give further clues? Is it reasonable to reduce the priority of such copy I/O and favour other processes like the remote desktop? How would you do that? Many thanks!

  Read the article

• DFS-R (2008 and R2) 2 node server cluster, all file writes end in conflictAndDeleted

  - by Andrew Gauger

  Both servers in a 2 server cluster are reporting event 4412 20,000 times per day. If I sit in the conflictAndDetected folder I can observe files appearing and disappearing. Users report that their files saved by peers at the same location are overriding each other. The configuration began with a single server, then DFS-R was set up using the 2008 R2 wizard that set up the share on the second server. DFSN was set up independently. Windows users have drives mapped using domain based namespace (\domain.com\share). Mac users are pointed directly to the new server share created by DFS-R. It is PC users indicating most of the lost files, but there has been 2 reports by Mac users about files reverting.

  Read the article

• How cpu writes data to dynamic RAM

  - by Krit

  Hello, I would like to know what kind of electrical signals does a cpu send to a dynamic RAM when it wants to write one bit (a 1 or 0). Is it simply that cpu sends just a single electric pulse, and if that electric pulse's voltage is higher than a certain level, it charges the capacitor to a voltage level that is "1" and if it is at a lower voltage band, it charges capacitor to level that it is "0"?

  Read the article

• slow disk writes between host and guest

  - by Jure1873

  I've got a ubuntu (server kernel) on a amd x4, 4gb ram, 2x seagate sata 1 tb disks for testing virtual machines and the write performance is very slow. The two disks are in a software raid1 array, one small boot ext3 partition, 10gb system partition and the rest is a xfs partition (about 980) gb for data (virtual machines). If I'm copying files from the virtual machine to the host with rsync or scp the copy frequently stalls or goes at about 1mb/s. What's wrong? I've tried disabling barriers on xfs, increased logbufs, allocsize, but it seems nothing helps. The strange thing is that await (for example during copying) for sda is usually under 100, while for sdb is around 400. Any ideas on what could be wrong / what could I do to improve this setup?

  Read the article

• What can i use to journal writes to file system

  - by Dmitry

  Hello, all I need to track all writes to files in order to have synchronized version of files on different place (server or just other directory, not considerable). Let it: all files located in same directory feel free to create some system files (e.g. SomeFileName.Ext~temp-data) no one have concurrent access to synced directory; nobody spoil ours meta-files or change real-files before we do postponed writes (like a commits) do not to care recovering "local" changes in case of crash; system can just rolled back to state of "server" by simple copy from it significant to have it transparent to use (so programmer must just call ordinary fopen(), read(), write()) It must be guaranteed that copy of files which "server" have is consistent. That is whole files scope existed in some moment of time. They may be sufficiently outdated but it must be fair snapshot of all files at some time. As i understand i should overload writing logic to collect data in order sent changes to "server". For example writing to temporary File~tmp. And so i have to overload reads in order program could read actual data of file. It would be great if you suggest some existing library (java or c++, it is unimportant) or solution (VCS customizing?). Or give hints how should i write it by myself. edit: After some reading i have more precision requirements: I need COW (Copy-on-write) wrapper for fopen(),fwrite(),.. or interceptor (hook) WriteFile() and other FS api system calls. Log-structured file system in userspace would be a alternative too.

  Read the article

• Dynamic Text limit when users writes in a textbox (HTML)

  - by DomingoSL

  Hello, I need to implement in Html/javascript (or php is possible) a script which works like twitter, when the user writes text in a field a label change showing the characters left. Can you help me with a simple and easy script example?

  Read the article

• Dinamic Text limit when users writes in a textobox (HTML)

  - by DomingoSL

  Hello i need to implement in Html/javascript (or php is possible) a script who works like twitter, when the user writes text in a field a label change showing the caracters left. Can you help me with a simple and easy script example?

  Read the article

• atomic writes to ehcache

  - by Jacques René Mesrine

  Context I am storing a java.util.List inside ehcache. Key(String) --> List<UserDetail> The ordered List contains a Top 10 ranking of my most active users. Problem Concurrent 3rd party clients might be requesting for this list. I have a requirement to be as current as possible with regards to the ranking. Thus if the ranking is changed due the activities of users, the ordered List in the cache must not be left stale for very long. Once I've recalculated a new List, I want to replace the one in cache immediately. Consider a busy scenario whereby multiple concurrent clients are requesting for the ranking; how can I replace the cache item in an fashion such that: Clients can continue to pull a possibly stale snapshot. They should never get a null value. There will only be 1 server thread that writes to the cache.

  Read the article

• Regasm writes mscoree.dll into Registry key InprocServer32

  - by Stiefel

  When I register my .NET Assembly with regasm.exe the registry key HKEY_CLASSES_ROOT\CLSID{111E32AD-4BF8-495F-AB4D-6C61BD463EA4}\InprocServer32 is set to "mscoree.dll". However, I am trying to mimic an existing COM-Server that was written in C. When registering this old COM-server the InprocServer32 is set to the full path to this component. Unfortunately the existing system (a plugin host that I can not change) reads and use this value - an is confused by the "mscoree.dll" value. My solution might be to patch this registry entry manually - but I would like to understand why regasm writes "mscoree.dll" into InprocServer32 .

  Read the article

• program that writes the even and odd numbers

  - by user292489

  enter code herei was writting a program that can read a set of numbers file called dog.txt; and also writes to two file separating odd and even. i was able to compile my program however, the output expected is not the same which was supposed to be even numbers in one file called EVEN, odd numbers in file odd. #include <stdio.h> #include <stdlib.h> int main(int argc, char *argv[]) { int i; int even,odd; int num; if (argc != 4) { printf("Usage: executable in_file output_file\n"); exit(0); } FILE *dog = fopen(argv[1], "r"); FILE *feven= fopen(argv[2], "w"); FILE *fodd= fopen (argv[3], "w"); while (fscanf(dog, "%d", &num) != EOF) { if (0==i%2){ i++; printf("even= %d\n", num); } else if(i!=0){ i++; printf("odd= %d\n", num); } } fclose(feven); fclose(fodd); fclose(dog); return 0; } output: even= 1 odd= 2 even= 34 odd= 44 even= 66 odd= 78 even= 94 odd= 21 even= 23 odd= 54 even= 44 odd= 65 even= 78 odd= 68 even= 92

  Read the article

• Atomic int writes on file

  - by Waneck

  Hello! I'm writing an application that will have to be able to handle many concurrent accesses to it, either by threads as by processes. So no mutex'es or locks should be applied to this. To make the use of locks go down to a minimum, I'm designing for the file to be "append-only", so all data is first appended to disk, and then the address pointing to the info it has updated, is changed to refer to the new one. So I will need to implement a small lock system only to change this one int so it refers to the new address. How is the best way to do it? I was thinking about maybe putting a flag before the address, that when it's set, the readers will use a spin lock until it's released. But I'm afraid that it isn't at all atomic, is it? e.g. a reader reads the flag, and it is unset on the same time, a writer writes the flag and changes the value of the int the reader may read an inconsistent value! I'm looking for locking techniques but all I find is either for thread locking techniques, or to lock an entire file, not fields. Is it not possible to do this? How do append-only databases handle this? Thanks! Cauê

  Read the article

• How to limit TCP writes to particular size and then block untlil the data is read

  - by ustulation

  {Qt 4.7.0 , VS 2010} I have a Server written in Qt and a 3rd party client executable. Qt based server uses QTcpServer and QTcpSocket facilities (non-blocking). Going through the articles on TCP I understand the following: the original implementation of TCP mentioned the negotiable window size to be a 16-bit value, thus maximum being 65535 bytes. But implementations often used the RFC window-scale-extension that allows the sliding window size to be scalable by bit-shifting to yield a maximum of 1 gigabyte. This is implementation defined. This could have resulted in majorly different window sizes on receiver and sender end as the server uses Qt facilities without hardcoding any window size limit. Client 1st asks for all information it can based on the previous messages from the server before handling the new (accumulating) incoming messages. So at some point Server receives a lot of messages each asking for data of several MB's. This the server processes and puts it into the sender buffer. Client however is unable to handle the messages at the same pace and it seems that client’s receiver buffer is far smaller (65535 bytes maybe) than sender’s transmit window size. The messages thus get accumulated at sender’s end until the sender’s buffer is full too after which the TCP writes on sender would block. This however does not happen as sender buffer is much larger. Hence this manifests as increase in memory consumption on the sender’s end. To prevent this from happening, I used Qt’s socket’s waitForBytesWritten() with timeout set to -1 for infinite waiting period. This as I see from the behaviour blocks the thread writing TCP data until the data has actually been sensed by the receiver’s window (which will happen when earlier messages have been processed by the client at application level). This has caused memory consumption at Server end to be almost negligible. is there a better alternative to this (in Qt) if i want to restrict the memory consumption at server end to say x MB's? Also please point out if any of my understandings is incorrect.

  Read the article

• Database solution for 200million writes/day, monthly summarization queries

  - by sb

  Hello. I'm looking for help deciding on which database system to use. (I've been googling and reading for the past few hours; it now seems worthwhile to ask for help from someone with firsthand knowledge.) I need to log around 200 million rows (or more) per 8 hour workday to a database, then perform weekly/monthly/yearly summary queries on that data. The summary queries would be for collecting data for things like billing statements, eg. "How many transactions of type A did each user run this month?" (could be more complex, but that's the general idea). I can spread the database amongst several machines, as necessary, but I don't think I can take old data offline. I'll definitely need to be able to query a month's worth of data, maybe a year. These queries would be for my own use, and wouldn't need to be generated in real-time for an end-user (they could run overnight, if needed). Does anyone have any suggestions as to which databases would be a good fit? P.S. Cassandra looks like it would have no problem handling the writes, but what about the huge monthly table scans? Is anyone familiar with Cassandra/Hadoop MapReduce performance?

  Read the article

• A script that writes errors and should create a event-error

  - by helmich

  this if it works should check the internet connection if there is a connection it does nothing. if there isn't a connection it should write a error in a txtfile if that happend 5 times it should create a error but it doesn't I will show you the whole code that i have now and the piece of code that i want in a loop. I can't get it in the way i want. I want it to creat 1 Event-error after 5 times writing to the file. this is the whole code i will put the code i want in a loop under it strDirectory = "Z:\text2" strFile = "\foutmelding.txt" strText = "De connectie is verbroken" strWebsite = "www.helmichbeens.com" If PingSite(strWebsite) Then WScript.Quit 'Website is pingable - no further action required Set objFSO = CreateObject("Scripting.FileSystemObject") RecordSingleEvent Dim fout For fout = 1 To 5 : Do If fout = 5 Then Exit Do Set WshShell = WScript.CreateObject("WScript.Shell") Call WshShell.LogEvent(1, "Test Event") Loop While False : next '------------------------------------ 'Record a single event in a text file '------------------------------------ Sub RecordSingleEvent If Not objFSO.FolderExists(strDirectory) Then objFSO.CreateFolder(strDirectory) Set objTextFile = objFSO.OpenTextFile(strDirectory & strFile, 8, True) objTextFile.WriteLine(Now & strText) objTextFile.Close End sub '---------------- 'Ping my web site '---------------- Function PingSite( myWebsite ) Set objHTTP = CreateObject( "WinHttp.WinHttpRequest.5.1" ) objHTTP.Open "GET", "http://" & myWebsite & "/", False objHTTP.SetRequestHeader "User-Agent", "Mozilla/4.0 (compatible; MyApp 1.0; Windows NT 5.1)" On Error Resume Next objHTTP.Send PingSite = (objHTTP.Status = 200) On Error Goto 0 End Function '----------------------------------------------- 'Counts the number of lines inside the text file '----------------------------------------------- Function EventCount(fout) strData = objFSO.OpenTextFile(strDirectory & strFile,ForReading).ReadAll arrLines = Split(strData,vbCrLf) EventCount = UBound(arrLines) End Function This is the whole code, and it doesnt work correctly becaus it creats a event-log rightaway and it should do that after the script has written 5 times to the textfile here is the code that writes to a textfile Sub RecordSingleEvent If Not objFSO.FolderExists(strDirectory) Then objFSO.CreateFolder(strDirectory) Set objTextFile = objFSO.OpenTextFile(strDirectory & strFile, 8, True) objTextFile.WriteLine(Now & strText) objTextFile.Close End sub and here is the code but this part doesnt not work or atleast i think it is this part Dim fout For fout = 1 To 5 : Do If fout = 5 Then Exit Do Set WshShell = WScript.CreateObject("WScript.Shell") Call WshShell.LogEvent(1, "Test Event") Loop While False : next Function EventCount(fout) strData = objFSO.OpenTextFile(strDirectory & strFile,ForReading).ReadAll arrLines = Split(strData,vbCrLf) EventCount = UBound(arrLines) End Function this is the not working part and I don't know what to do anymore so can you please take a look at it tank you very much. btw: this code can be very usefull for a network administrator

  Read the article

• Effect of frequent sdcard writes

  - by Al

  In my chat app, I am adding the ability to log chats. The logs are saved on the sdcard and one BufferedWriter is kept open for every person/channel chat is done with. I'm wondering what effects this might have have on the sdcard and its life. My BufferedWriter's buffer size is to 1024, I'm also wondering if that is too small or too big.

  Read the article

• Where is default location where tracelistener writes txt logs

  - by djerry

  Hey guys, i want to log some traces in my service. When i set initializeData to a location in the d: partition, i can write with no problems. When i set the initializeData to c:\, it doesn't write at all. Now i was wondering 2 things : 1) Does my service not have the rights to write to c:\ partition? 2) if i don't specify the partition, where does it write to? This is the part of my app.config which works: <add initializeData="d:\txtServiceLog.txt" type="MonitoringServerService.FaultTracer, MonitoringServerService" name="txtListener" traceOutputOptions="DateTime, Timestamp, ProcessId, Callstack"> <filter type="" /> </add> When changing to code below, i doesn't write anymore : <add initializeData="c:\txtServiceLog.txt" type="MonitoringServerService.FaultTracer, MonitoringServerService" name="txtListener" traceOutputOptions="DateTime, Timestamp, ProcessId, Callstack"> <filter type="" /> </add> And where should i look if i do this : <add initializeData="txtServiceLog.txt" type="MonitoringServerService.FaultTracer, MonitoringServerService" name="txtListener" traceOutputOptions="DateTime, Timestamp, ProcessId, Callstack"> <filter type="" /> </add> Thanks in advance.

  Read the article

• spring json writes more json then i want

  - by mkoryak

  I am using the json spring view in order to return a simple JSON object from my controller. The problem is that its returning more data then i want. it is returning the validation errors and things inside of my model when all i am doing is this in the controller: Map model = new HashMap() model.put("success", "true"); return new ModelAndView("jsonView", model); If you look at the bottom of this page in the docs it looks like i am getting back data that POST would return. i am not doing a post, i am doing a GET by going directly to the URL with params. How do i get this lib to return just the data in my model?

  Read the article

• Are C++ Reads and Writes of an int atomic

  - by theschmitzer

  I have two threads, one updating an int and one reading it. This value is a statistic where the order of the read and write is irrelevant. My question is, do I need to synchronize access to this multi-byte value anyway? Or, put another way, can part of the write be complete and get interrupted, and then the read happen. For example, think of value = ox0000FFFF increment value to 0x00010000 Is there a time where the value looks like 0x0001FFFF that I should be worried about? Certainly the larger the type, the more possible something like this is I've always synchronized these types of accesses, but was curious what the community thought.

  Read the article

• NoSQL for concurrent reads/writes

  - by Mickael Marrache

  After getting some performance issues for an application using a MySQL database, I'm thinking of using NoSQL solutions. My architecture is as follows: One application receives messages from the network at a high throughput (i.e. 50000 messages/sec). Each message is stored in the DB, so it's important for the write rate to be as fast as the arrival rate. Then, I also have some PHP pages that accesses the DB to get the data stored by the other application. It's important for me that the retrieved data is as relevant as possible (i.e. not old data, let's say not more than 5 seconds old). Also, the data is not critical, so I don't need any security mechanism to avoid losing the data. I see there are a lot of NoSQL solutions, but I don't know if they are all relevant. Could you please provide me some directions. Thanks

  Read the article

• Java writes bad wave files

  - by Cliff

  I'm writing out wave files in Java using AudioInputStream output = new AudioInputStream(new ByteArrayInputStream(rawPCMSamples), new AudioFormat(22000,16,1,true,false), rawPCMSamples.length) AudioSystem.write(output, AudioFileFormat.Type.WAVE, new FileOutputStream('somefile.wav')) And I get what appears to be corrupt wave files on OSX. They won't play from Finder however using the same code behind a servlet writing directly to the response stream and setting the Content-Type to audio/wave seems to play fine in quicktime. What gives?

  Read the article

• Examples of program that writes programs

  - by Nyan

  Where can I find such programs? I think yacc and lex are this kind of programs? I'd like to see example codes especially in C language.

  Read the article

• Designing a hierarchical structure with lots of reads and writes?

  - by JD01

  I am in the process of working on a video on demand system part of it involves the management of a hierarchical tree structure (think windows explorer) which allows users to upload videos, move folders around, create new folders etc. User action will be allowed depending on which permissions they have. In addition to managing the structure (creating folders and uploading videos), subscribers will be viewing content (read access). The number of reads will be significantly more than the writes. My question (and it is big one) is should I store the data in a database (for the writes) and also have some sort of cache which will be used for the reads? or do I use two databases? or is there a better solution? Also I will have to resolve concurrency issues which I think optimistic locking on the database will resolve. I have a fair bit about CQRS over the last few months but not sure if this is the way to go. Any ideas?

  Read the article

• Who writes the words? A rant with graphs.

  - by Roger Hart

  If you read my rant, you'll know that I'm getting a bit of a bee in my bonnet about user interface text. But rather than just yelling about the way the world should be (short version: no UI text would suck), it seemed prudent to actually gather some data. Rachel Potts has made an excellent first foray, by conducting a series of interviews across organizations about how they write user interface text. You can read Rachel's write up here. She presents the facts as she found them, and doesn't editorialise. The result is insightful, but impartial isn't really my style. So here's a rant with graphs. My method, and how it sucked I sent out a short survey. Survey design is one of my hobby-horses, and since some smartarse in the comments will mention it if I don't, I'll step up and confess: I did not design this one well. It was potentially ambiguous, implicitly excluded people, and since I only really advertised it on Twitter and a couple of mailing lists the sample will be chock full of biases. Regardless, these were the questions: What do you do? Select the option that best describes your role What kind of software does your organization make? (optional) In your organization, who writes the text on your software user interfaces? (for example: button names, static text, tooltips, and so on) Tick all that apply. In your organization who is responsible for user interface text? Who "owns" it? The most glaring issue (apart from question 3 being a bit broken) was that I didn't make it clear that I was asking about applications. Desktop, mobile, or web, I wouldn't have minded. In fact, it might have been interesting to categorize and compare. But a few respondents commented on the seeming lack of relevance, since they didn't really make software. There were some other issues too. It wasn't the best survey. So, you know, pinch of salt time with what follows. Despite this, there were 100 or so respondents. This post covers the overview, and you can look at the raw data in this spreadsheet What did people do? Boring graph number one: I wasn't expecting that. Given I pimped the survey on twitter and a couple of Tech Comms discussion lists, I was more banking on and even Content Strategy/Tech Comms split. What the "Others" specified: Three people chipped in with Technical Writer. Author, apparently, doesn't cut it. There's a "nobody reads the instructions" joke in there somewhere, I'm sure. There were a couple of hybrid roles, including Tech Comms and Testing, which sounds gruelling and thankless. There was also, an Intranet Manager, a Creative Director, a Consultant, a CTO, an Information Architect, and a Translator. That's a pretty healthy slice through the industry. Who wrote UI text? Boring graph number two: Annoyingly, I made this a "tick all that apply" question, so I can't make crude and inflammatory generalizations about percentages. This is more about who gets involved in user interface wording. So don't panic about the number of developers writing UI text. First off, it just means they're involved. Second, they might be good at it. What? It could happen. Ours are involved - they write a placeholder and flag it to me for changes. Sometimes I don't make any. It's also not surprising that there's so much UX in the mix. Some of that will be people taking care, and crafting an understandable interface. Some of it will be whatever text goes on the wireframe making it into production. I'm going to assume that's what happened at eBay, when their iPhone app purportedly shipped with the placeholder text "Some crappy content goes here". Ahem. Listing all 17 "other" responses would make this post lengthy indeed, but you can read them in the raw data spreadsheet. The award for the approach that sounds the most like a good idea yet carries the highest risk of ending badly goes to whoever offered up "External agencies using focus groups". If you're reading this, and that actually works, leave a comment. I'm fascinated. Who owned UI text Stop. Bar chart time: Wow. Let's cut to the chase, and by "chase", I mean those inflammatory generalizations I was talking about: In around 60% of cases the person responsible for user interface text probably lacks the relevant expertise. Even in the categories I count as being likely to have relevant skills (Marketing Copywriters, Content Strategists, Technical Authors, and User Experience Designers) there's a case for each role being unsuited, as you'll see in Rachel's blog post So it's not as simple as my headline. Does that mean that you personally, Mr Developer reading this, write bad button names? Of course not. I know nothing about you. It rather implies that as a category, the majority of people looking after UI text have neither communication nor user experience as their primary skill set, and as such will probably only be good at this by happy accident. I don't have a way of measuring those frequency of those accidents. What the Others specified: I don't know who owns it. I assume the project manager is responsible. "copywriters" when they wish to annoy me. the client's web maintenance person, often PR or MarComm That last one chills me to the bone. Still, at least nobody said "the work experience kid". You can see the rest in the spreadsheet. My overwhelming impression here is of user interface text as an unloved afterthought. There were fewer "nobody" responses than I expected, and a much broader split. But the relative predominance of developers owning and writing UI text suggests to me that organizations don't see it as something worth dedicating attention to. If true, that's bothersome. Because the words on the screen, particularly the names of things, are fundamental to the ability to understand an use software. It's also fascinating that Technical Authors and Content Strategists are neck and neck. For such a nascent discipline, Content Strategy appears to have made a mark on software development. Or my sample is skewed. But it feels like a bit of validation for my rant: Content Strategy is eating Tech Comms' lunch. That's not a bad thing. Well, not if the UI text is getting done well. And that's the caveat to this whole post. I couldn't care less who writes UI text, provided they consider the user and don't suck at it. I care that it may be falling by default to people poorly disposed to doing it right. And I care about that because so much user interface text sucks. The most interesting question Was one I forgot to ask. It's this: Does your organization have technical authors/writers? Like a lot of survey data, that doesn't tell you much on its own. But once we get a bit dimensional, it become more interesting. So taken with the other questions, this would have let me find out what I really want to know: What proportion of organizations have Tech Comms professionals but don't use them for UI text? Who writes UI text in their place? Why this happens? It's possible (feasible is another matter) that hundreds of companies have tech authors who don't work on user interfaces because they've empirically discovered that someone else, say the Marketing Copywriter, is better at it. And once we've all finished laughing, I'll point out that I've met plenty of tech authors who just aren't used to thinking about users at the point of need in the way UI text and embedded user assistance require. If you've got what I regard, perhaps unfairly, as the bad kind of tech author - the old-school kind with the thousand-page pdf and the grammar obsession - if you've got one of those then you probably are better off getting the UX folk or the copywriters to do your UI text. At the very least, they'll derive terminology from user research.

  Read the article

< Previous Page | 1 2 3 4 5 6 7 8 9 10 11 12  | Next Page >