The DSM-320 is a so-called streaming media-player. It contains no local storage, so the media must be stored somewhere else. You could compare it with a DVD player. A DVD player does not contain any local storage (the storage is the DVD you insert), all it can do is navigate the media that sits on the DVD, it can decode the media you select and send sound and picture through its plugs to a TV and/or an amplifier. It is the size of a thin DVD player, in fact in my home it sits underneath the DVD player. I have had a love/hate relationship with it. I love the idea of a small set-top box with decoding built-in and the rest of the functionality located elsewhere, the actual device has however had a number of irritating bugs which D-Link hasn’t seemed very committed to fix.

Basic description of how it works

The idea behind the DSM-320 is that you have all your media located on a computer somewhere on you network. This media can be music, video and photos. On this computer you run a media-server (a uPnP server). The media-player has cables connecting it to your TV, your stereo and of course your network. When the media-player is turned on, it connects to the media-server over the network, and through the user interface of the media-player you select the media you want. The media-server sends that media to the player, the player decodes it and your TV shows the decoded data (if it was video or photos) and/or your stereo plays the sounds.

Using the DSM-320 media-player

Music

Firing it up (using the TV to show the user interface) I can select whether I want Music, Video or Photos. If I select Music I can select a particular genre, a particular year, an artist, an album or… well, the choices for how I want to find the music I want is huge. The choices I have are actually not dependant on the media-player, but on the media-server running in the background. I’ve got close to 200 CD’s that have all been ripped and MP3 encoded so I have a fair amount to choose from. It is a lot easier to find the music I want through the DSM-320 than hunting for it in the bookshelf (I have tried to institute some kind of system a couple of times and it is always messed up within a month). The data that allows me all this choice in how to select is held in the MP3 tags in the MP3 files, the media-server reads that and organises the lists accordingly. Being able to choose music (or audio books) as easily as I can and then play it on the stereo is great

Video

I have also encoded several older VHS tapes (particularly children’s videos) that are now selectable and playable via the DSM-320 without having to find the right tape and (even better) without lots of tapes taking up space in the bookshelf. There is no such thing as an MP3 tag for MPEG2 files, so here I have had to decide on a directory structure on the hard-disk. Video is selectable either by name or by browsing the directory structure. Things have not been good on the video front for the DSM-320. The DSM-320 remote control has a fast-forward and a rewind button. Within the setup for the DSM-320 you decide whether pressing the fast-forward button makes the video play twice as fast or whether it will bring up a question about where (in hour:minutes format) you want it to start playing. Sometimes I would like to have the one, sometimes the other function, but I can’t have both (which is probably understandable). With some video formats, the “play faster” option just doesn’t work, with all video formats, the time-based format doesn’t work past 1 hour. So if I want to “fast-forward” into a film after the first hour and 20 minutes, sorry, you can go to the 59 minutes into the film location and then play normally from there. I have some MPEG4 files that played fine when I first started using the media-player, after applying the last European firmware upgrade it could no longer play those files (Unsupported Media was the message I got).

Photos

Displaying photos is a feature I haven’t found that usefull, Photos just don’t look that good on a normal TV. But we have occasionally sat down to look at holiday snaps on it rather than sit in front of the computer (where the photos look much better). The DSM-320 will read the date and time from the EXIF header of JPEG pictures, but I haven’t found that way of selecting the pictures I want that great. As with the videos, a good directory structure (Year, Month, Occasion) is much better at finding the images I want.

Internet Radio

D-Link have made it possible to subscribe to music from services such as Rhapsody and Napster. I haven’t tried those services, the media-server software from TwonkyVision I used makes it possible to listen to Internet radiostations using Shoutcast.

Media Server Software

The media-server software that came with the DSM-320 was fairly basic, but to be honest, I only ever used it once as an initial test so I can’t really comment on it. There are several media server implementations around, I have primarily used TwonkyMedia from TwonkyVision as their software can run on Linux.

10 things I hate about you

Well maybe not 10, I am well pleased with the overall functionality, but the DSM-320 is not particularly well thought out and it is buggy, in no particular order:

• The remote was obviously not designed by someone who never used the device. Look at this picture of the remote control. This is how you would use it to play some music: First, turn the DSM-320 on (red button at the very top), then press the “Music” button (4′th row of buttons from the bottom), select the artist or album you want to listen to (use the navigation arrow keys, this actually works fine), now press “Play” (2′nd row of buttons from the bottom). If you want to skip to the next song, you do that with one of the buttons on the very bottom of the control. It is not a one-handed affair, you need one hand to hold the remote and the other hand to press the right buttons. On the remote for my TV and my DVD player, the buttons I use the most are located in such a way as to make it possible to operate with one hand. Ok, so maybe the usability of the remote control is not the end of the world, but it sure is the most irritating remote control I have ever used.
• On the European version, the plugs were placed by someone who never actually tried plugging it in. The SCART connector is too close to the network connector. A SCART cable sits at an angle to the plug, and on the DSM-320 this means that the network cable is bent close to the connector (of course this only happens if the device is cable connected, if the wireless option is used this is not an issue).
• Connect the device wirelessly and the only security option is WEP. This may be something they can’t do anything about, but if it is something a firmware update could “fix” I sure would like WPA. Until recently it hasn’t been an issue as it was connected by cable.
• The bugs, none of them cause the product to be dysfunctional, but there are several in different parts. Which leads us to…
• Firmware updates are very easy to perform. The DSM-320 will itself check for them over the internet (if it can reach it over your network) and if there is an update will suggest that it applies it. I think there have been 2 firmware updates in the 2 years I’ve had it which considering the known bugs is not enough. It is not very difficult to find irritated owners who on the internet have documented reporting a bug to D-Link and then nothing happening. In summary, the firmware updates have been few and far between and when they have happened they haven’t fixed all the known issues and they have usually introduced some new ones.

Summary

Overall I have liked the DSM-320. The bugs it has I have been able to overlook. Some of the known bugs in the DSM-320 media-player also exist in newer products such as the DSM-520. That does give one the impression that it is a product-range D-Link don’t intend to focus much on. It does make one wonder when one reads that D-Link claim to have a strategy for the digital home and to be the leader of the digital home arena. I’m still using my DSM-320, but I am looking for something to replace it.

Other DSM-320 reviews:

• NoDRM.com review
• PC Magazine
• Geekzone
• OzCableguy

So wi-fi was an occasional access method and an excellent one at that. Now I have two PCs in different parts of the house that are permanently connected via wi-fi and I plan on connecting other devices as well. It has been a lot fiddlier than I thought it would be. After some initial problems with getting a good enough signal to both PCs, it is working alright. But wi-fi is no match for a good old cabled connection.

I’m currently renting the house we live in so I can’t go around drilling holes in the walls, but next time I buy a house again I will have ethernet cables installed in all the rooms. Maybe newer WLAN standards will get rid of the problems with WLANs but here are some of the major issues with using it as the primary data transfer method:

Bandwidth

All my wi-fi devices are configured for 802.11g, in other words, a theoretical bandwidth of 54 Mbit/sec, but in practice somewhere closer to half of that. The difference in speed is so obvious when doing anything other than ordinary web-browsing where the limiting factor is the internet connection. I’ve got my photo albums stored on the server and there is a noticeable time lag between viewing photos on one of the wirelessly connected PCs and a cable connected PC using fast Ethernet (100 Mbit/sec). I have in the last couple of years occasionally digitised old VHS videotapes. I shudder at the thought of transferring a video by wi-fi to the server.

Shared media

related to the Bandwidth issue. With wi-fi you are using a shared media, and if you are not careful about which wi-fi channels you are using, you may even be sharing the media with your neighbours. Last time I ran Netstumbler on the laptop I saw (including my own) 16 different WLANs. 10 of them use channel 11, 3 use channel 1, 2 use channel 6 and 1 uses channel 12. With 802.11b and 802.11g there is an overlap in the frequency range used by the 14 (Europe) different channels. That frequency overlap means that WLANs that unless channels are chosen carefully, transmissions on one WLAN can slow down the performance of the other WLANs. To ensure there is no frequency overlap, the “safe” channels are 1, 6 and 11. Most resources I have seen on the internet refer to channel 6 being the factory-default channel on most wi-fi devices. My own network scanning around the world shows channel 11 to be the default. So my Access Point is set to channel 1. The other networks in my neighbourhood also using channel 1 are not very powerful. For more information on the frequency overlap of wi-fi channels I would recommend these articles:

• IEEE 802.11
• Channel Deployment Issues for 2.4-GHz 802.11 WLANs
• How To: When Wireless LANs Collide!
• Video Streaming Need To Know: Part 2 – The Real World

Wireless phones and wi-fi

One related aside to shared media; many articles and manuals mention that microwave ovens and wireless phones use the same frequency band used by 802.11b and 802.11g WLANs. I think all newish wireless phones in Europe use the DECT standard, at least mine do. I did a quick internet search and DECT doesn’t use the same frequency band as wi-fi, so this must be a US only problem.

Security

Security is still an issue. Many people don’t know what they are doing or don’t seem to be thinking it through properly. From where I live I can access 3 open WLANs. I discussed this with an acquaintance who also works in the IT industry. His comment was that his access point also was open as he didn’t have any network traffic worth securing. I then asked him what he would do if somebody used his connection to download paedophile content, it would potentially be traceable to his home. I don’t know about the UK (where I currently live) but according to an article I just read in a German IT magazine, You can (in Germany) be prosecuted if your WLAN is used for illegal activities and you have not done anything to secure it. So security is important. Anybody who as read anything on wi-fi security will know that the first security mechanism available for wi-fi, WEP, is useless. But I’ve got a media-player device that only supports WEP which means I either have to use WEP on all devices or not use that particular device wirelessly. Well, I may have a work-around for that, more on this in a later post. But my issue with WLAN security is that it is too complicated for many people or they don’t take it seriously. The various encryption methods (WEP, WPA with TKIP or AES, WPA2) don’t exactly make it easier. And some older (but not that old) devices in effect don’t support any security.

The centrepiece in my infrastructure is a Linksys WRT54G router/firewall/switch/wifi access-point. Definitely recommended! It has been extremely reliable. For some years now, I’ve been using the DD-WRT third-party firmware with it. Be aware that if bought as new I would recommend buying the WRT54GL. The current “normal” WRT54G does not use Linux but VxWorks does not contain the amount of memory that the GL version has and it is not as simple to apply third-party firmware to it. For more information about the different versions that exist of the WRT54G, see this Wikipedia article.

The router is connected to an ADSL modem (of course it isn’t really a modem but it seems to be the accepted terminology) which in turn connects me to BT (British Telecom). The ADSL modem is an X-Modem M3 from ADSL Nation.

Down and up-stream speeds of the ADSL service are 2Mbit and 256 Kbit/sec respectively. The ISP (BT) and the bandwidth is not my choice as such. I work for IBM and am considered a mobile worker, IBM therefore connects me and pays my ADSL connection.

The server and the laptop I use are connected by ordinary ethernet cables as they are in the same room as the switch. Other systems are connected via PCI WLAN cards with a (theoretic) bandwidth of 54 Mbit/sec.

The Server Hardware

I used to use old PCs that I or extended family members no longer used as workstations, wiped them and “converted” them to server duty. Eventually the services I put on the server(s) became something I relied on so downtime became a real nuisance. The homework of the kids would usually be stored on the server so it usually wasn’t an option to wait for the weekend to arrive before getting the server up and running again if for whatever reason it had gone down. When home servers are used primarily as file servers, the performance is usually not that important, but I increasingly used webserver-based applications and the performance of the machines I used as servers therefore did become an issue. At one stage I had a server setup of two machines where most of the resource intensive programs were running on a laptop which via NFS then used the disk-storage of an old desktop so if, for whatever reason, one machine was down I effectively had no server. I ended up spending too much time keeping these old machines in the air.
Eventually I bought a real server, a Dell PowerEdge 830 equipped with a dual-core 3GHz Pentium 4, 1 GB ECC (Error Correcting Code) RAM and two 300 GB harddisks configured in a RAID-1 array (mirrored disks). This system has performed flawlessly and, other than externally inflicted power cuts or software maintenance (updades necessitating reboots), it has had no downtime.

I would definitely recommend some kind of file-storage in a home network. If you have the skills use a “real” server as adding additional services are possible, otherwise get a NAS (Network Attached Storage) device of some kind. Examples would be the Linksys NSLU2 or the Buffalo TeraStation. I haven’t used any of them, but if I were to go down that route, I would look at getting one of those.

Operating System

On servers I have always used RedHat Linux or a derivative thereof. When Fedora Core came out, I started using that, but I don’t really care to be on the bleeding edge on the server. I want it to be reliable. Fedora Core is updated too often for my server needs. For a server you want something that is security updated for an extended period. a couple of times I found that the version I was using was no longer updat and I didn’t really have the time to perform a major software update of everything. That sort of thing requires backups in case it doesn’t work afterwards, so it is a time-consuming task. Today I use CentOS and I have version 4 installed. The people behind CentOS use the source RPM files that RedHat publish and compile them, ending up with a Linux distribution that for all intents and purposes is identical to RedHat Enterprise Linux, but free.