3 Elmstone Close
ME16 8DQ
United Kingdom

Don't forget to
Postage and VAT.

26th of March, 2004

My Netgear PS110 Printer Server turned up today, where at least this package unlike the one received yesterday was very will packed. As it arrived in a box, each item was wrapped multiple times in bubble wrap, where the empty space was filled with rolled up newspaper.

Sure enough I soon had it connected up and powered on, where after the one reset my server's DHCP system allocated this Printer Server an IP address, where this Printer Switch called itself PRINTER. That name I still have to change, when after all it is not exactly a printer, but I will have that sorted out soon enough.

My first measure of success was to ping this new device, which indicated that it was connected up fine.

Only problem was to now communicate with my HP Deskjet 870cxi Printer through this Printer Server, where since I never used one before, then I traweled through most of the manual on the software CDN before giving it a try.

Before I got this device I did some research on how to best connect it up, where the general advice was to not bother with the default software and to just connect it up through Windows. Since I was unaware on how to exactly do that, then I decided to install the default software and to see what it could do.

Sure enough when running this wizard for the first time it managed to find this device on my network, where it returned the correct IP address, the allocated name of this device and even the model. That is the point that I encounter the first real problem, when after selecting this listed device for confirguation, then returned was an error message saying that it could not talk to it.

Strange, when if it really could not talk to it, then why was it listed in the first place?

Anyway, after lots of hassle I managed to discover the Windows method for controlling this device. To begin with I selected the wrong "Network Printer" configuation option, which I now see wanted to connect to an established printer on another computer. Since that is not what I wanted to do, then so was the only other option to set up a local printer.

The plug and play detection routine for a few second worried me that this would not work it, but sure enough this detection failure made it default to a screen where you could set up or connect to an established port. As the Netgear PS110 had a port already established here, added by the setup routine no doubt, then so did I give that a try first of all.

Sure enough my HP Deskjet 870cxi model printer drivers could be attached to this new port, which meant that we was all ready to print. Only problem here was that the Windows test page failed to print, where after powering on and off the devices a few times, then the only option was to delete this new printer option and start again.

Following the same choices again this time I selected to establish a TCP/IP connection, where I could selected my Netgear print server from a list, where once again I attached my HP printer to it.

As once again I got printing failure, then just in case I decided to borrow the Lazer printer's cable. After all that one I know for sure works, where this other cable has not been used for a year, where this printer also had previous communication problems.

Since this cable swap helped not one bit, then I gave this problem some thought. As the configuration programs could not talk with this device, even if they could see it. And also my printing attempts failed, even if communication with the Printer Switch and the HP Printer should be fine.

My answer to these series of facts is that this Printer Server wanted some network communication protocol that was not already installed, where after a scan of what was not already installed on network card already, then I decided to add both Client fot Netware Networks and the Appletalk protocol.

After a reset of my Server to complete installation, then before booting had even completed, then the Windows Test Page from the previous session popped out of my HP Deskjet. Solved I thought, even if I have to now figure out just what it needs.

I gave the configuation program another shot, where this time it got past where it normally gets up to. I gave try changing the name a shot, but communation failure kicked in. Going back and trying to do it again also failed, where going back another page stopped it going past the point it usually gets stuck at.

Since another Windows test page also failed to print, then I suspected that maybe this installed software is responsible for all this. And so after uninstalling it a second Windows test page printed out just fine. Guess I will try uninstalling those additional protocols now and see if it was the installed software all along.

Anyway, the first few test print outs came out quite well for a printer that has not been used in a year. Unfortunately, after further use the black ink dried up and the colour one was prone to banding.

Suspecting that these ink cartridges needed cleaning after all, then I took out the died up black one first of all. This one I saw was nearly full of ink already, where it was my latest reused cartridge.

My print cartridge cleaning method has been subject to the years worth of experimenation, when this old printer has been used to print out many thousands of pages in the past. That is why I know that the best cleaning method is just to turn the cartridge up-side-down, where a few drops od cartridge flush should be applied directly to the print head.

This soon broke down the dried up ink on the print head, but the reason why this cleaning method is much better than other methods, is that this cartridge flush gets sucked down by gravity into the print head, which cleans out the inside as well. That is no problem, when this cartridge flush comes back out when you do the first print.

Anyway, do that once, gently wipe it clean, do it again, where as my next test print proved this black cartridge was working like it was brand new.

I gave that a try with the colour cartridge as well, but in this case the light banding remained. Not too bad, but I will soon gave this cartridge my second cleaning method, which is just to stand this print head in a small bowl of cartridge flush and break down whatever comes out.

Connecting back up my red printer cable showed that indeed this cable worked just fine, which just as well as this is the type that I sell and this one is the very one in the photo.

However, next to this red printer cable is a green printer cable, where this one connects to my Canon Multipass C20 device. This device is more commonly known as my fax machine, even if it is capable of a whole load more.

What I mean is that apart from being a fax machine this is also a copier, scanner and printer, which explains the green printer cable.

The only problem with this green cable is that this Canon Multipass C20 multi-function device is the only machine I have ever purchased that became obsolete even before I plugged it in. As during the years that I have owned this device, then never once have I ever printed to it or done a scan using it.

Now, the reason why this happens to be is that this device I purchased shortly after Windows XP was released, where at the time I was indeed using Windows XP and NT Server 4.0 on the Server.

The problem here is that Canon has only ever made Windows 95/98 and ME drivers for this device, which means that even though it was still being sold through the early XP era, then Windows 2000, XP and 2003 support is and will always be zero.

Sure enough I just had another check today, where anyone owning one of these earlier Multipass models won't be using it with these modern operating systems. As Canon's support sucks, which is why I won't buy another one of their obsolete right out the box devices.

And so this model is and will always be my fax machine and copier. However, I was thinking today that even though there is no drivers for this model, then maybe I could still use it with a ye olde printer driver.

What I mean by this is that back during the DOS days the subject of printer drivers was a more rare subject, when if you wanted to printer, then you just did a raw dump to ther printer port. And I bet that if I connect this up to my LPT1 port, then I could run the old COPY hello.txt LPT1 command and have basic text spewing out.

Since that is better than nothing, then I had a look to see if any Windows default driver could do that, where their standard text only driver has so far failed to work. Still, I will connect it up to LPT1 and so some testing to see just what is compatible with this printer.

Plain text is good enough, where graphics would be great, where so far I have nothing.

The other device I wished would work with these latest O/Ss and does not is my multi CD changer. As this great device back in the Win98 days could hold six CDs at once, where switching between them only took a couple of seconds. This model was not really suitable for sharing, but having six software or game CDs on-line at once was a great feature.

Windows XP killed every such device when their drive interface system only allowed one driver letter per device. Sure I can use it as a single CD drive and manually change disks, but that defeats the point of it.

I have always wondered if Microsoft did that on purpose to allow manufacturers to sell more CD drivers in enviroments that need lots of CDs on-line, but clearly they have no interesting in fixing their interface driver, when they could have easily done so long ago.

Anyway, I had better get back to getting this Printer Server fully installed and making sure that their are no issues with it, but it has clearly allowed me to get my HP Deskjet 870cxi working again. Now I only have to solve the paper feed problem on this model, when all those thousands of pages have worn down the roller.

I will have that fixed soon enough now that I have reason to.

As to this monitor problem, then the one that turned up yesterday imploded. What I mean by this is that due to this serious damage to this monitor, then the screen was already weakened.

Since they have these tubes at low pressure, then about four hours after delivery there was a loud pop of heavy glass breaking. And sure enough this monitor now has a big crack right from the top to the lower right corner of the screen.

The person who sold me this monitor mentioned that Parcel Force approved of him using bubble wrap and no box, where the distinctive shape of this monitor would mean that they would be careful with it. Nice idea, but this monitor got destroyed for a very good reason.

It seems that I will get my money back just as soon as the inspector from either Parcel Force or their insurance company come to take a look at it, which hopefully won't be long, when I had one too many broken monitors here before this one turned up. And I am rather eager to dispose of it and other broken hardware with it.

As that won't happen until they have checked it out, then I will be pressing for urgency in that matter.

Anyway, I will be buying me yet another monitor sometime this weekend, where I have a feeling that a certain 19" model will be mine soon enough. Then of course hopefully next week I will get my money back on this broken one.

25th of March, 2004

It seems that it is one of "those days", when nothing can go right. Where the two computer items that I expected to arrive safely today, for one reason or another, did not do so, which I will now explain.

To begin with, then I heard late yesterday that my Netgear PS110 Print Server would not be with me today, when this expected dispatch did not happen for unimportant reasons. Hopefully, he manages to dispatch it today, when I dislike having to wait over a weekend for a delivery.

However, today was the day of the big monitor (well 17") delivery, which would get Workstation 1 back in action. That occurred at about 12:30 today, which is common for my area and these services delivery schedule.

The problem here is that this monitor has been destroyed by Parcel Force, which is not unsurprising considering the packing. As sure enough at some point this monitor was dropped, which resulted in the multiple fracture of the tube and the wires hanging out the side.

As it seems that this seller had the idea to wrap this monitor in a combination of brown paper and bubble wrap. And if you imagine a sledge hammer coming into contact with such a package, then that gives you an idea of what this monitor now looks like.

Clearly a thick walled box and lots of packing material would have been ideal, but this poses an interesting question of who pays for it. As this delivery was insured, where Parcel Force destroyed it due to unsuitable packing.

Well I will leave that one for them to sort out, but clearly I will be needing to find me another monitor. And this time I will make extra sure that it is sent in a suitably packed box, when obviously one seller has not dealt with Parcel Force before.

As of course back in January 2001 during the mad ELVIS rush it was them who did the impossible and managed to lose a fridge/freezer sized full collection of boxes containing these ELVIS programmers. Since this is hard enough to pick up let alone lose, then even now you have to wonder just how they managed this.

The insurance company must have passed an elephant that day considering the cost of this loss, where of course back then it took nearly a month before this manufacturer got compensation and I got my replacement order.

I no longer use Parcel Force for that reason, when they can lose huge packages and not have a clue where they got to. This latest breakage is not really their fault, but I don't have these problems with other delivery companies.

Today I have been buying my aquarium fish some freeze dried food, when they must be bored with the usual flakes by now, where obviously very soon I will be needing to buy me another monitor as well.

24th of March, 2004

The new fans for my Server did arrive today, where I can now report that these two fans are now mounted on this Server's 1 GHz Athlon XP slot CPU and everything is working just fine.

For a quick reminder; then this CPU has a large heatsink and two cheap fans, when the 1 GHz model is the hottest and fastest in this CPU range. And since one of these fans failed months ago, then this CPU has been running on one fan for a long time.

Unfortunately the second fan has recently started to fail, which made a good time to replace them both.

This second fan shortly before I replaced it was getting progressively worse, where these fans normally run at around 4000 rpm. Today it was running at an average of 2000 rpm, which was constantly moving between 3000 rpm and close to zero. And of course my Server's alarm system will activate if the speed drops below 1500 rpm, which is why it has been going on and off all this morning.

Sure I could have switched the alarm off, just like I have already done with the first of these failed fans, but as I am unhappy running this CPU on no fans, then I left it on so that it did not fail for two long.

Due to this second fan running slowly, then the temperature has been pushing 60 degrees C recently, where this fan when working fine could keep the temperate in the low 50s. And when both fans were working, then the temperature was down in the mid to high 40s.

Getting this slot CPU out was easier than I was expecting, when I just undid the latches and pulled it right out. Unscrewing the fans were also easier than expected, which was mostly due to the fact that the last time I tried it the Server was running and I was trying to unscrew it sideways.

So the fans were now off, but as these new fans had labels on both sides, then the first problem was that I did not know what way they had to go around, when of course you need to blow air through the heatsink. And so I wired one up, powered on this CPU-less computer and noted down what way the air came out of this fan.

Before I mounted the new fan, then I cleaned all the dust out of the heatsink. It was not too bad compared to what is has been in the past, but the five fans on this case sure like to move the dust about.

Also since I saw how the heatsink could be removed, then so did I have the idea of applying some of my Arctic Silver 5 paste. This I doubt will help this CPU too much, but each one of these tubes can do like 16 CPUs and my use so far has been one. Maybe I should add some to this 2.6 GHz laptop, when running at full speed this CPU tends to get a little too hot.

Anyway, I quickly cleaned the former blue mess off this CPU and heatsink and added this silver based paste. It was when I was putting this CPU and heatsink back together did I notice that this heatsink had been mounted up-side-down previously.

Sure enough I connected it back up the correct way, where only later on did I discover that it was mounted that wrong way around for a very good reason, which is so that the heatsink is lifted away from the motherboard components so that it could actually fit. And so I had to take back out the CPU, heatsink and fans to turn it back up the wrong way before it could all connect back together.

However, before all that I had to mount the new fans, which is where my second main problem crept up. As I was already aware from my careful measurements these new fans were thicker than the two that used to be on it. That I thought was fine to provide extra cooling, but I was simply hoping that the removed screws would be long enough to mount these new fans.

Given my luck, then that answer was "no", when they were too short. And so I had to hunt around for screws in the same design, but just slightly longer.

Naturally I did not find any, but I did find some metal screws twice as long and slightly thicker in the same thread design. Then I also found some wood screws in the right length, but the wrong thickness and thread type.

In the end I used the wood screws, but as I soon found out the screw type did not matter much anyway. As these holes on the heatsink were not real holes and were just screwed in between the slots on the heatsink. As I found out only the outer ones were suitable for screwing the fans into, but surre enough these wood screws went in just fine.

I just made sure that they provided a tight fit, when these fans should not be allowed to move.

And so about an hour ago I wired it all up and powered back on the Server. For a couple of seconds I feared that I had killed it, but that was just the damned AGP graphics card coming loose in the slot once more.

So it tuned on fine, I re-enabled this formerly failed fan alarm in the CMOS setup, where the Server booted back up to the Win2k3 Server O/S just fine. Once I remembered to plug back in the two removed USB cables, then so was this upgrade complete.

Since you will desire the results, then I must say that they are very pleasing. These new slightly thicker and seemingly better fans run at a faster rate of slightly over 5500 rpms. If they will stay that high after months of use remains to be seen, but for now it is certainly helping to keep the temperate down.

And so after cleaning out the dust, replacing both fans, then quickly adding some Arctic Silver 5 thermal paste, then after an hour of use I can report that the CPU's temperature has just reached 40 degrees C after running for a long time in the high 30s.

That is the lowest that it has ever been, but it could still drop slightly, when this Arctic Silver takes time to set in. Anyway, I am going to get the Server to run a number crunching program in order to see how hot this CPU can get maximum, which will also help to break it in.

My new Print Switch should hopefully arrive tomorrow, which is a less important network addition. As it is this new 17" monitor that is the important one, where I have heard nothing from this seller yet, concerning an urgent delivery and mention of when it is dispatched and when it is due here.

Actually, scrap that, when I have now found this e-mail. As this monitor was dispatched to me yesterday, where it is due here tomorrow. I did not notice this e-mail to begin with, when it was sent to my PayPal payment address instead of my usual EBay address. And that PayPal address is of course serviced by Workstation 1, which was not being used like normal yesterday due to the Gigabit link testing.

Still, I have found a better way to use Workstation 1 than through my portable TV these days, when remote desktop connection allows me to use this important computer though one of my other computers. That proved useful in my network speed testing yesterday, when I was changing about the network card settings on both the Server and Workstation 1 from Laptop 1.

Well I was annoyed at keep having to switch between those computers with logging in each time, which is where I remembered that I could do it this other way instead.

Anyway, my Server's CPU cooling system only this morning was days away from total failure, where these fans arrived just in time to make this CPU cooling better than ever before. And these new fans should certainly last long enough until the day when this system gets replaced, which won't be any time soon.

Earlier News...

As my auction won Print Server and CPU fans did not turn up yesterday, as I hoped they would, then hopefully the two CPU fans at least turn up today. After all the single fan on my Server's CPU stopped twice yesterday, which is not at all helpful. Fortunately, it started back up within seconds each time, but it won't be long now before only a push start gets it moving.

The Print Server I heard from this seller would be dispatched to me today, which means that it should ideally be here tomorrow.

Also I won my new, or more correctly second hand, monitor yesterday, to replace the one that went up in smoke, which I can report is a Lyama 17" Vision Master Pro 400 model. Since this is two inches smaller than the one that just died, then I will have to reduce my desktop from the 1600x1200 mode down to the 1280x1024 mode.

Sure this functional monitor can do 1600x1200 mode at a high refresh rate, but there is something about seeing the text, when even the 1600x1200 mode on my 19" monitor was pushing it a bit. I certainly could have got another 19" monitor had I desired, but this one is a nice model and I can see that 17" has some advantages over a 19" one.

By the way the price is quite reasonable as well, when it cost me less than the delivery cost. And yes I could have waited around and got one for 1p had I desired, but the better models tend to go for a bit more and well I am in a rush.

I did consider getting a tape backup drive yesterday, but the cheap yet usually expensive model that I had my eye on sure enough went back up to a higher price that made it all not worth it.

My issue here of course is that I would desire to backup the less important data on my Server's RAID drive, but tape backup systems are either too small, or are so expensive that a cheaper and better option is to buy another RAID drive to back it up to. The other option of a DVD writer are still too small at 4.7gb a throw, when that makes for a whole load of DVDs in order backup up about 300 GB.

So my favoured option at the moment is simply to wait for the new generation of DVD writers to come about, when if I recall correctly this new DVD media will be able to handle about 20 GB. And my RAID system should last just fine until then, excluding any major disaster, when after all 360 GB out of this 720 GB space is used to back-up the other 360 GB.

Yesterday I tried sorting out the speed of my new Gigabit Ethernet link, when 12.1 MB/s one way and 14.2 MB/s the other way was just not what it should be. And I am pleased to say that I made large gains, even if things are still not what they should be.

First of all with the understandable section, then file transfers between my Server and Workstation 1 now goes at 22.61 MB/s. This 8 MB/s increase is mostly due to "Jumbo Frames", where 9k and 16k frames produce about the same results.

For those that do not know, then moving the speed up to Gigabit Ethernet presents one problem. As suitable hardware can work at Gigabit speed, where it is just that you need really powerful CPUs to handle all the 8,000 odd interrupts a second for these data frame from/to the network.

And so to remove a lot of this burden from the processor, then that is where these Jumbo Frames come in. As by increasing the amount of data that each frame handles results in there being less frames to deal with. Increasing the number of buffers also helped, but it is Jumbo Frames that give my Server's CPU more space to send and receive data. As even the data transfer from Workstation 1 to the Server increased from 12.1 MB/s to the new speed of 13.74 MB/S.

However, it is the 22.61 MB/s transfer the other way that is understandable, when that is very close to the write speed limit of my fake RAID system in Workstation 1. The reason why I call it a fake RAID system is that the claimed RAID 1.5 system containing striping and mirroring is just not worth the effort.

Let me explain, when those who remember my RAID addition to my Server last year will know that after days of testing these were my final scores...

RAID Configuration: 1+0 (Mirroring + Stripping)
Block Size: 16k
Used Drives: 4
Available Space: 360 GB
Random Access Time: 10.8 ms - 4.2 ms (latency) = 6.6 ms
Average Transfer Rate (read): 65140 KB/S
Average Transfer Rate (write): 23782 KB/S

In fact while I am on this subject, then while finding out why my Gigabit connection was not working as expected, then I came across a surprise. As I ran this speed test once again, where this time I came up with an Average Transfer Rate (read) of 75.9 MB/s.

I have no idea how that has happened, but multiple testing has highlighted that my Server's RAID system has increased by over 10 MB/s. Maybe it was moving the RAID card down a slot and reinstalling the drivers that did it, but that improvement was certainly pleasing to see. Now, if only I could get that 75.9 MB/s down the Gigabit Ethernet connection, but that is the easier of the two problems.

Anyway, back to the fake RAID system in Workstation 1, when I have promised the RAID 1.5 results for a long time. And so when I first connected up this two disk RAID system, then I decided to give RAID 0 a shot to see the maximum capabilities of this hardware. Where I can report that RAID 0 with a 16k block size gave me an Average Transfer Rate (read) of 75.6 MB/s.

Very nice, but then I tried RAID 1 to show the capabilities of reading from a single HDD. And testing here gave an Average Transfer Rate (read) of 43.9 MB/s, which is in line with what is to be expected.

However, since RAID 1.5 promises stripping, then so was I expecting improvements on this. And testing on RAID 1.5 highlights an Average Transfer Rate (read) of a minor 44.1 MB/s. In other words no real gain over a RAID 1 system, or single HDD for that matter, where the only reason that I use it is because the access time drops by almost 2 ms over the normal rate.

Further testing using different configurations was not possible, when since RAID 1.5 does not support real stripping, then so could not the ideal 16k block size be specified. And that is what is really holding RAID 1.5 back, when instead it uses a system where each disk can be separated and used on it's own like a normal HDD.

Thank you for that idea, but I would prefer the faster RAID 0 like speed obtained by using the 16k block size, even if I had to replace a broken HDD to get the system working again.

So since as the RAID system on Workstation 1 has been next to useless, then I have been thinking about changing it. The easiest option here is to add another two 120 GB HDD on the slave positions on these two controllers and making another RAID 1+0 array.

Sure using the slave positions on these controllers would not be using these IDE channels at optimum single drive speed, but at minimum this non-ideal RAID 1+0 system should be faster than what I have at the moment. And of course adding another RAID controller card to one of these spare PCI slots could have them all working on a master only channel, which would make a system faster than the Server.

The problem with my 22.61 MB/s limit from the Gigabit Ethernet link is that this is now very close to the HDD writing speed limit on these HDDs, where the bus on this board would not really delay things from the mirrored write any.

I can only imagine that this speed would increase if I moved to a RAID 1+0 array, when the real write 16k block stripping could double the write speed. As of course the data write would be divided between all four drive, where the mirrored write would also make all four drives write at once.

The 200 MHz bus on my 800 MHz quad piped Pentium 4 Hyper-Threading board would be able to handle this a lot better than the 100 MHz bus on the Server can, where to improve the Server speed further I would really need to replace the motherboard (and everything with it).

So we can now see why my Server's 75.9 MB/s sustained data read is limited to this 22.61 MB/s, when that is simply how fast this other computer can write it. And so better use of this Gigabit link can be had by other Gigabit supporting computers pushing the Server further.

I did hope to give this a test soon with my 2.6 Ghz laptop, but alas I have the 5620 model with Fast Ethernet built in, where Gigabit Ethernet support was only added two generations later with the very latest 5800 model. Still, I am not sure what is the point of Gigabit support on a Laptop, when it's slower than normal HDD system is not even close to RAID speed.

Since the 22.61 MB/s transfer rate is understandable, and greatly faster than the previous Fast Ethernet link, then I am fully happy with that, even if my future Gigabit Switch will also need to support Jumbo Frames, when not all of them do. As where the area I have yet to understand comes in is with the opposite transfer of 13.74 MB/s, when this RAID system can easily handle a HDD write of double that.

My Server transfer systems have always been biased like that, which makes me think that maybe the O/S thinks "this is a file server and we are here to serve files and not to receive them". However, my biggest clue to what is going on is with my latest HDD testing, when with the HDD system on Workstation 1 the CPU Utilization was a low 3.7%.

I am still wondering how it does that, when both my RAID controllers are made by the same company, but since this same value on my Server was a very high 82.8%, then I believe I see what is going on. As with the combined CPU handling of the RAID Array and some major data streaming down the PCI bus, then the CPU on the Server has simply been overloaded.

Combine that with all the frame handling of this Gigabit NIC, then that could well explain why the write to the Server is no faster. I will have to do some local file copying on the Server to find out for sure, but if so that means that the Server would either need a faster CPU or a faster bus.

Either way that means replacing the motherboard and everything with it, which is very undesirable. Still, if I do ever go that route, then an on-board Gigabit NIC and RAID controller is a must have, just like with the one already in Workstation 1. The other option of course is to offload this processing to a real hardware based RAID card, but you still have the problem that when you fill the bus that fills the CPU as well.

Something to think about none the less, but for now I will see if I can inch any more out of it. And hopefully one or two of the three items that I am waiting for will arrive today, which will help restore my computer systems to a fully working state.

23rd of March, 2004

Well since the three new computer items that I expected to arrive yesterday did so, then so have I now completed the first upgrade to my Server.

This upgrade basically involved removing one Fast Ethernet NIC and replacing this with a Gigabit Ethernet NIC instead. Then the other change was to remove the Twin Printer Port PCI card and replace this with my new Five Port USB 2.0 PCI card. Two of those USB ports are on the back and the other three are on a tray that slots into a 3.5" slot on the front.

Everything went well enough, even if I had to open the case back up a few times, when I first forgot to screw in this 3.5" tray. Then my AGP graphics card had come loose in the slot and needed to be pushed back in, where the third problem was that my RAID card that I had moved down a slot had one of the cables trapped under the end, causing the card to not connect correctly, where the RAID card had to be removed and then inserted again in order to move this cable out of the way.

Since I have yet to obtain a Gigabit speed Switch, then this new Gigabit NIC is connected directly to the Gigabit NIC built into the motherboard on my Workstation 1 computer. And since I have gave it a quick speed test in both directions, then a large file write to the Server goes at 10.2 mb/s, while the read from the Server went at 14.1 mb/s.

Those are certainly new speed records for my network, but since both my HDDs and PCI bus on these two computers are capable of higher performance than this, then I do not yet believe that things are working as they should. After all the RAID 1+0 system on my Server has already been clocked at 23.2 mb/s sustained write (or 46.4 mb/s if you don't wait for the mirror write to complete) and 63.6 mb/s sustained read.

I would have expected at least double these values, which means that I will have to look into it further. Maybe adding a Gigabit Switch will help, when on my former Fast Ethernet link, then things were going slowly until a Switch was added taking the speed up to a value that I now forget.

My current Fast Ethernet Switch I have now moved to running on the second Fast Ethernet NIC in the Server, when putting it on this new main Gigabit Ethernet NIC will only slow down this new link. And sure enough things are working fine with my Belkin WAP (Wireless Access Point) plugged into it, where the new Print Server will be added to this soon.

This new Print Server to connect between this Switch and my colour Inkjet Printer could turn up later today, when this "like new" model was won in an EBay auction this weekend just gone. This Netgear PS110 model brand new from one of the lowest cost suppliers around costs around £58 (or £68 including VAT), where this one just as good cost me a lesser £31.

There were two auctions on this (UK located) model on EBay at the time, where the second one sold for a higher price of £42. So I believe that I got a bargain by saving £27 or £37 in the process.

Also later today could turn up my two new fans, where this time they are brand new. And if they do indeed turn up, then that will mean that I will have to open the Server's case once again in order to pull out this Athlon XP 1 Ghz slot CPU. Where it is just a case of then unscrewing the failed fan and the second one soon to fail, where these two new fans will have my Twin fan CPU cooling working once more.

Sure I technically need no CPU fans in my five fan case cooling if I was to be silly, when set on the high fan speed that is, but as I have seen this strong suction likes to suck dust through every hole and gap possible.

The other new item that I got in yesterday was one of D-Link's AirPlus XtremeG PCMCIA cards, model number DWL-G650. Where this is not plugged into this laptop at the moment due to "issues".

What I mean is that after discovering where the controls were hidden to set up my Wireless parameters, then I soon had this Wireless card connected to my Belkin WAP at 11 mbps speed. However, since my Belkin WAP has now been identified as the very source of my long term Wireless problems, then this WAP needs to be removed from the system if I am ever to read files from the Server.

And so for the very reason that I obtained my USB 2.0 PCI card, then my USB Wireless NIC is now connected to it (after some hassle with a really tight USB port), where this NIC on the Server is just working fine.

However, this D-Link PCMCIA NIC is failing to achieve an 11 mbps ad-hoc link to it. Under Windows control; first of all it spots this connection, which causes this NIC on the Server to report a connection, but this connection is quickly lost. I have also tried putting it on manual control and specifying the exact parameters, but still no luck.

So for now I am just using my old 11 mbps PCMCIA NIC, which achieves a connection just fine. And with this new USB 2.0 PCI card, then this link to the Server is perfect, without any issues with the Wireless link or USB bus.

I am soon to give this new D-Link NIC another try to see if I can get it working in 802.11b ad-hoc mode, but that at this time does not seem hopeful.

This D-Link NIC also has other issues that I will now explain. As even before I plugged it in did I notice that D-Link's web site did not even mention my new B3 version, which left me wondering if the B1 and B2 drivers are also fine for use with my B3 model. Since the default install CD went on to install the B1 drivers anyway, then I soon had these drivers uninstalled and the very latest drivers with the 108 mbps support installed instead.

The second issue crops up when you try removing this D-Link PCMCIA NIC from this slot, when you get an immediate "blue screen of death" system error message before the computer resets. Since I am just normally a person to take them straight out, then on the second time I thought that I would use Window's "it is now safe to remove your hardware" device removal option.

As soon as I clicked on it to shut down this device for removal did the "blue screen of death" return, which seems to clearly indicate that D-Link has driver issues that are not fully compatible with Windows system. However, I do have one theory on how to remove this card without having to reset the computer, when disabling this NIC in the Network Properties seems to shut it down fine.

Only problem then is that when you plug it back in it will need to be enabled again.

The final issue that may in part go to explain my ad-hoc connection problem is that this NIC is interfering with the touchpad on my Laptop, when only during the ad-hoc channel scanning does this mouse pointer go into really slow staggered movement mode. And this got so bad that my touchpad drivers shut down the advanced features of this touchpad reporting that another pointing device was interfering with it.

And so D-Link's products may be technically advanced, but they still have a lot of teething problems. Still, I am sure things work better at higher speeds, when at minimum D-Link's products should work with D-Link's other products.

Since I have now identified the cause of my Wireless problem, my Belkin  WAP, then maybe I will get me a D-Link WAP soon. Then again, as my Wireless link is now working fine, then there is no rush.

Anyway, my computer system repairs go on, where maybe today I will win me a monitor to replace the one that recently went up in smoke. As all I need to obtain now is this Monitor, a Gigabit speed Switch, then ideally a D-Link WAP.

Further News...

Since I have just gave this new D-Link G650 PCMCIA NIC another try to get it to connect to my 11 mbps USB NIC on the Server, then I am pleased to say that I now have it working just fine.

Trying to get the driver working though, simply did not work out, which is why I decided to uninstall the device and then install the driver again. And sure enough as soon as the driver was activated, then the driver spotted my 11 mbps connection on the Server and wanted me to enter the key, when it obviously knew that the link was encrypted.

Once that was done, then connection was achieve. Only problem was that DHCP allocated the same details that it last gave this NIC when it was connected through the WAP. Since that was on the wrong IP range, then I had to visit the Server to delete this old profile from it. And by disabling and then enabling this D-Link PCMCIA NIC again, then so did DHCP create and allocate a new profile in the correct IP range

Anyway, I hope that finding problem does not return, but for now it is working just fine.

21st of March, 2004

I have now figured out the best way to deal with my 19" CTX PR960F monitor that went up in smoke only a few days ago, when within this day and age it is not even worth bothering to have it repaired.

What I mean is that these days everyone is upgrading their systems to using a flat panel TFT or LCD display, which includes everyone from the single monitor home user to entire business departments with monitors by the dozen or more. This bulk of unwanted monitors is why the bottom has fallen out of the new CRT market, when for new monitors prices are very low.

And the lowest of the low in terms of second hand monitors is to be found on EBay, where the monitors that hardly anyone wants (or at least supply exceeds demand) are certainly being given away to anyone who will pay the postage cost. Monitors of any size from just 1p, up to little more, where the delivery cost is also reasonable, even if some people try to make some profit out of the delivery cost.

People are willing to sell this way simply to give their old perfectly working monitor a good home, where compared to the CRT monitors that exist, and the continuing flat panel upgrade process, then what is on EBay is certainly the tip of the iceberg.

I see a business opportunity for the energetic person exporting to less high tech nations, or at least selling to those who would pay more, but you can see that it is pointless having my monitor repaired, when the true market value of a used CRT monitor these days is 1p. I have considered upgrading to a flat panel model myself, but this technology still has some development to do, CRTs are still better in many ways and I do have the desk space.

So I guess my broken monitor will be about my fourth monitor here that has ended up on the rubbish heap, but then monitor repair is always expensive and as I well know such repair may not solve all monitor issues.

Sure, buying such cheap things from EBay can be a case of "you get what you pay for", but of course sellers have to provide accurate descriptions, where a perfectly monitor has to be just that.

A further development in my Man versus Machine mini-War is that the second fan on the CPU in my Server is showing signs that this could fail as well within the next few weeks. What I mean is that three times within a few minutes of each other did this fan stop, where fortunately it has been working well enough (if slightly slowly) since then.

My 1 Ghz Athlon CPU in this computer has a large heatsink and two fans, where the first one of these failed completely months ago. Since my case has five fans sucking air in and out of it, then as recently proved my Server can run on no CPU fans. Such a thing would kill modern processors in a short space of time, but no fans on this CPU just means that processor intensive applications need to be avoided.

That is why my first action following this fan stopping was to close the number cruncher running on this computer, which had pushed the temperature up to 62 degrees C before I shut it down.

Anyway, since it is time to ensure that this remaining fan does not fail completely, then I am already awaiting delivery of two replacement fans. And as I know from before the first fan failed, then all these fans working at once does nicely in keeping a low temperature.

Fortunately, nothing else has gone up in smoke (yet), but having to reset the USB system on this motherboard frequently is starting to get annoying, which is why it is thankful that my USB 2.0 PCI card turns up tomorrow.

I am going to do further changes to my network system to improve things more, where of course also tomorrow will arrive my new Gigabit NIC for the wired half of my network. Not to forget the PCMCIA card that will one day push my wireless speed up to 108 mbps.

The most interesting thing I am soon to add is a device known as a Print Server, where this device is little more than a couple of Printer Ports with a NIC attached. My server technically has enough Printer Ports already, numbering three, but due to the extra PCI card that is going in my server tomorrow, then I am having to change things around.

What I mean is that my Server has five PCI slots, where three of them are currently used by network cards, where the other two are for my RAID card followed by this twin Printer Port PCI card. This old motherboard also has two ISA slots, which are currently unused due to the Printer Port card taking up both slots.

Not to mention that this eight slot board also has an AGP slot, which is where you will find the graphics card.

I am going to rework these slots so that by removing the currently unused NIC and this Printer Port card, then I would have one PCI slot free, not to mention one ISA slot. Ideally I guess this PCI slot could host either a SCSI card or maybe a second RAID card, when my Server with five HDDs already certainly has space for four or five more.

Until then I may keep the third NIC in there, when it may be useful for some connection. A good example of that would be to host my 10/100 switch for things like the Print Server, while a direct connection between this Server and Workstation 1 can run at Gigabit speed.

I will need to get myself a Gigabit speed switch soon enough, when that will allow more than one computer to use this Gigabit NIC. Not to forget that the slower NIC and Switch would then become redundant. A second switch would also allow me to see if it is the WAP or my existing Switch that is to blame for my now worked around Wireless problem.

Then one day in the future at a suitable time I would add one of D-Link's WAPs to upgrade my Wireless network to 108 mbps speed, when my existing 11 mbps system has always been considered slow.

This new Print Server, when I buy it soon, will also allow me to get my Colour Inkjet Printer working once more, when the Server has always had a problem with this printer. As when you first add it this printer is perfectly fine, but as soon as the Server is reset, then suddenly it can no longer find this printer until you spend quite a while installing all the drivers and applications again.

Since this printer works fine directly connected to other computers, then something is going wrong in this system. And maybe even the second printer port can allow me to one day add a colour lazer printer, which I do not yet have good enough reasons to buy.

My old Inkjet printer currently has a paper pick-up problem after being famed for being able to handle perfect double sided printing without a single feeding fault. I already know what the problem is and how to fix it, when the roller just needs refreshing, when the handling of thousands of A4 pages has just worn the sticky covering to the roller down.

Having colour printing again will be nice, which is one reason why I do not yet need an expensive colour lazer. And of course everything I print out these days, like my sales flyers, comes out of my mono lazer at up to 16 ppm.

Anyway, there will be a lot of hardware upgrade this new week, when all these problems should be resolved once and for all.

Also I did actually manage to print out order forms on Friday, when I remembered that my still new AGP graphics card supports monitor output, DVI output and the all important TV output. And so for this short time I am just using my portable TV, where the resolution sucks and makes things tricky to read, but this is the best solution until I get me one of those give away monitors.

Kind of odd. As when I got this graphics card, then this TV output function was not even something that I thought about, then having no real need for it. So my on-line ordering system is all running on my portable TV at the moment, but that has allowed those one day delayed orders to depart on Friday instead of the expected tomorrow.

Other e-mail is being handled by this Laptop computer, which is on the link that can be taken down by these USB ports. Not really a problem to start this link up again, but today I will try this WAP and Switch on the usually unused NIC to see if it does the same. I expect it will, but this will allow me to set up the required system for when the other NIC is running on Gigabit speed.

No2 who turned up here this weekend thinks that my CAT 5e cable may not be good enough, due to reported problems, where tomorrow I will see just how fast this Gigabit link can go. Where a transfer from the HDDs on Workstation 1 to the RAID 1+0 system on the Server should be quite impressive with both fast read and write.

Then my network would soon achieve the point I have planned for a very long time, which for example is why I went with RAID 1+0 on the Server instead of the slower write RAID 5 system.

Well I hope to order my Printer Server today, where I am now off to enable this third NIC.

18th of March, 2004

Well my past week or so has been quite a battle between myself and my computer hardware to get it working as it should, where at the moment I have made some gains, but yet at this second I am technically losing, even if victory is in sight.

This new battle began due to an aspect that resulted from my previous attempts at fixing my wireless connection to my Laptop2 computer, which for a long time have been subject to a serious problem. This is that any transfers from the wireless network to the wired network went perfectly fine, but any transfers from the wired network to the wireless network was subject to many disconnections.

Since this Laptop2 computer was the one that handled this vast bulk on the normal e-mail addresses, then this is one aspect of why I decided to close them all and only use the one going to my Workstation 1 computer.

Now I had previously spent weeks trying to solve this connection problem, where this new battle began when I discovered that my NIC to my Cable Modem had a lot of services bound to it beyond just TCP/IP. As of course by default only TCP/IP should be attached to Internet connections, when you would not want people on the Internet to login to your system or read your files.

Fortunately I have good passwords protecting my computers, but it was here I made an accident and uninstalled all these services instead of unbinding them. Only minutes later did I discover that it had uninstalled them from my second NIC as well, where before I could sort this out correctly the O/S on my Server required me to reset.

It was then that Win2k3 server fired the first shot in this computer versus human War, when after adding back the services like "Client for Microsoft Networks" and "File and Printer Sharing", then a further couple of resets highlight that things were not working as it should.

More correctly my entire domain was missing, including all the services like user and computer control, my twenty odd shares, and even the ability of all my other computers to login in to the server. And considering the minor thing that I had done, which I had quickly undone, then this was a very serious problem.

During my research into what was going wrong, then I came across a comment that could well solve my Wireless problem, which was that WAPs (Wireless Access Points) can have compatibility problems with Switches. Since I had already tried everything else, then I thought that I would give that a try by removing my WAP from the Switch and plug it directly into a NIC.

Nice idea, but my WAP will only work in a Switch, when direct connection to a NIC results in a not connected message.

And so I thought that I would try removing the WAP altogether and go back to the ad-hoc system that I used to use before getting a second Laptop and a WAP to control them both. Sure enough as soon as I had connected up one of the Wireless NICs to the server and configured the settings, then a perfect connection and trouble free transfers were clear.

Also during my trying to fix my Wireless problem only days previously I had got my 2.6 Ghz Laptop1 computer working again, after a year of it being non-functional due to the end falling off the power cable. And by fixing it I meant cutting the end completely off of this power cable, cutting the removed end block in half to see how two wires connected to four pins and then to solder these wires to the end section that fell off to create a new power connection cable.

Sure enough Laptop1 lived again, even I have yet to still tape up these exposed wires.

Only problem was that during the initial recharge of these completely flat batteries that it killed one of them, when the recharge was up to 78% before I went to bed. And when I got up recharge was on 100% and it was still trying to charge it, which certainly should not be correct.

After disconnecting this battery and plugging it in again, then the power level was now on 3% and still charging, where by disconnecting the repaired mains cable it was clearly that this battery was providing no power at all. This is an odd thing to happen, where the first battery recharge for over a year killed it, which make me wonder if timed failure is built into it.

Certainly my new power connection was not to blame, when this is a very simple re-connection. And my second (additional purchase) battery charged up just fine all of two times.

So I guess this battle began early, when I fixed Laptop1 only to have the computer kill the battery.

Also my now fixed Wireless link was not to go unnoticed, when the old USB problem returned. As the thing about my server is that the motherboard is an extremely stable model, but it was an early model that had USB 1.0 ports.

And these ports simply do not like modern USB 1.1 hardware, where the mouse alone works fine and even my scanner, but any other hardware like my former USB flash drive, which the Server's USB system killed, results in problems. For my USB Wireless NIC this results in the loss of the entire four port USB system, which only a Server reset or swapping between the two available drivers restarts.

So to make use of my new Wireless link without having visit the Server each 30 minutes or so means that a USB 2.0 PCI card was added to my shopping list. Not to mention a new Gigabit wired NIC for the Server and a D-Link 108 mbit PCMCIA NIC for one of these two laptops, which will only be useful when I have a faster D-Link NIC or WAP on the Server as well.

The next problem was that my former WAP acted like a simple Gateway by having its own IP address within my Wired NICs range, where the Laptops connected to this WAP had their IP address allocated within this same range by DHCP. However, this USB NIC does not work like that, when it's own IP range needs to be allocated within it's own Subnet Mask.

Adding a new scope was easy enough, but then DHCP refused to allocate IP addresses for these Laptops. Sure I could assign these manually, but since DHCP should be doing this automatically, then it was my task to find out what was going wrong.

The answer to that was that DHCP was bound to my Wired NIC, where anything not coming from within that Subnet Mask was being rejected. And the solution to this problem was the little service hidden under Routing and Remote Access called the DHCP Relay, where this will move the DHCP requests from the Wireless NIC to the Wired NIC to allow them to be accepted by DHCP.

And since DHCP can see where they first came from, then so does it allocate IP addresses from my new Wireless scope.

Apart from that during my search of R&RA did I discover the basic firewall on my Cable Modem NIC, which solved another one of my long term problems by allocating a small range to use through this.

And before I forget, then after many days of research I found the cause of my lost Domain, which was that the netlogon service had not been started, where after starting it manually my Domain including all the user and computer control spring to life.

Only problem was that after insuring that this service was set to start automatically, then further Server resets highlighted that Win2k3 server was messing up and was failing to start this service on boot. Since it was time to radical action, then so did I delete all these NICs and services in order to restart them from new.

Sure enough after setting it all up again, including the 20 lost shares, was my system working again. And even after discovery of the NET command did I find a faster way of establishing shares on my Encrypted Business Drive, which needs to be manually mounted an shared after every reboot.

As my new computer hardware arrives Monday to solve these remaining problems, even though I have to find out what was going wrong with my WAP to see it is not repeated on my future Wireless upgrade, then it was clear that my computer systems was soon to lose this War.

It was then that my computer hardware unleashed it's "Weapon of Mass Destruction", when less than 24 hours ago did my 19" CTX PR960F Monitor fail. I mean three and a half years of faithful service and it fails at this point? What are the odds? Beyond conclusive proof that all this time my computer systems have hated me and was engaged in a technological War!

That monitor just happens to be connected to my most important computer (beyond the server maybe), Workstation 1, which I had just spent over a year slowly upgrading to replace everything except the case and the boot HDD. New very stable motherboard, new CPU, new memory, new twin RAID system (one day to move to four drives) and even a new AGP graphics card, when a nearby lightning strike and power surge killed my old one.

Since that was my new most reliable computer, then so has it been handling just about every aspect of my business. Including site support, answering e-mail to most importantly of all the processing of on-line order forms,

As that is a highly complex system, involving complex filtering, then I would not even desire to touch it unless absolutely vital. So all new on-line orders will have to wait until Monday, when Workstation 1 borrows the Server's monitor (a 17" one in a 15" sized case) and this heavy lump of junk is moved out of the way.

This cannot really be done before late Friday, when this is a very heavy monitor and only No2 who is dropping by Friday can just about lift it. I expect that myself and No1 could just about do it, but she is beyond her days of heavy lifting and it would be unfair to ask

I have at least added the vital e-mail addresses to this Laptop1, where NIC-less Laptop2 is currently hibernating. And I did answer just about all e-mail on Workstation1 before failure, but there was three on-line paid for orders noticed just before it died. So they and any other new on-line orders will be going nowhere until Monday, where they should be printed out on Saturday.

As to this monitor, then it is unfortunate that the three year warranty expired six months ago, where this model back on the 27th of September 2000 cost me £317 plus £55.48 in VAT (that I could and did reclaim).

I have been fearing an expensive repair, but if that is the case, then I am better off buying a new model. After all TFT and LCD models are now the choice of the day, where another 19" CRT model, not unlike my one, costs about £130 plus VAT these days. I will explore the repair, exchange and upgrade options, before deciding if buying new is best.

Anyway, unless my computer hardware has any further surprises, then although the latest strike has caused me a large set back, where the USB issue is also very unhelpful, but this War will soon be won and my hardware will achieve a unique state of total obedience.

Well it has certainly taken a lot of hard work (and money), but in all this is a good thing.

9th of March, 2004

As it happens I only uploaded the previous news section just today, when adding that new claim in my Global Domination Plan Special Offer took longer than expected seeing that I was away this weekend.

Not to mention that I have just added a new Special Offer to this site, which you can see for yourself on my Special Offers page.

I can also say that new stock of FUN4 and FUN5 cards turned up today, where they would have been here last Friday or Saturday had not the Royal Mail gone extra slowly. That puts my stock of these FUN4 cards back up to their normal level, following my Special Offer last month quickly lowered my former large stock.

The FUN5 cards have just been running low for a while, where one customer then turned up and purchased all of my remaining stock, which is why now my new stock won't be at all easy to buy up at once.

Anyway, pending some hard work between today and tomorrow I hope to have support for this CAS Interface 2 sorted out, not to mention that my former SEASON2 details now need to be changed to my new SEASON3 model. So all outstanding orders leave today, where hopefully tomorrow you will be able to see the details of these two new items.

The only other thing is that PayPal has not yet fixed this withdrawal problem, due to their unexpected and flawed change to my account, where their "top priority" fixing of this problem seems to be going rather slowly. Seems that a reminder is in order, when doing this simple name swap is taking far too long.

5th of March, 2004

Anyone reading this News page previously may be aware of my comments about the PayPal on-line payment service made on the 10th of February. And the reason why this news subject continues is because PayPal has been doing recent changes to my account with them, where this time they left me another surprise to discover.

As you may recall I had to contact PayPal last month due to the fact that when I added my Citibank banking details to their service, then it would not allow me to specify the account name. And since my PayPal account was in my business name of Cardman, then so did this bank account carry the wrong name when added to their system.

PayPal's reply to this problem was that my PayPal account could only carry one name at a time, even if my default account page displays both my business / alias name and my real name. And since the PayPal account name for some odd reason was linked to my banking account name, then so could either one or the other of my bank accounts work with their system at any one time.

Since this would change the name that my customer's see, then that is why I have been reluctant to do this so far.

The news from the 10th mentions other factors, but it seems that PayPal has been playing about here, which I will now explains.

As I paid a visit to my on-line PayPal account today intent on withdrawing my latest collection of funds, where like usual I entered the required amount, selected my Cardman business account at Lloyds-TSB and clicked on Proceed. It was then that I noticed their little surprise, when instead of the "Cardman" account name being mentioned, then there in it's rightful place was my real name.

So there I was thinking that PayPal had just killed my desired withdrawal, where during their recent move of my account to being hosted in the UK instead of the US, then they must have got the two names on my PayPal account mixed up.

Since my PayPal / banking name was currently using my real name, then so did I consider withdrawing funds to my Citibank account instead, where I again entered the desired amount, this time selected my Citibank account and clicked on Proceed.

Shock horror, when low and behold attached to this account was the name "Cardman". And in a moment of enlightenment I now understood what PayPal had really done, when they had reconsidered my account name problem and decided to take my recommended action in simply giving these two accounts the correct name.

Even though PayPal has now killed all possible means for me to withdraw my Sterling funds, then in fact this "cock-up" is quite pleasing, when they are now more than half way to giving my bank accounts their correct names. That will certainly help with my future US Dollar account based in the US, whenever I desire to withdraw these funds instead of spending them buying junk on EBay,

Naturally, I have now dispatched PayPal another e-mail, pointing out that this change was a very welcomed idea, but that they had got these two account names mixed up.

In all that is the type of customer support that I would expect from their service, mostly due to the 4.5% charge that I have to pay them. As there they were saying one thing, like the market dominant payment service they are, where all of a sudden they have changed their mind.

That in fact could well go beyond the level that I would expect from them, when they are a big ugly commercial service that in normal times should say "these are our unchanging rules that you have to follow". And so seeing this on my account was in fact a total surprise, when I was planning on a possible work-around system.

In truth I guess that my problem was just one of many faced by all their business customers, where they decided on a more global policy change and that my account was just marked as one needing attention. So with my fingers crossed my bank accounts at PayPal should soon carry the correct names.

What is most pleasing about all this is that all on-line payment service should give their business customers their red carpet attention and go the full distance to make them fully happy. After all for a company like myself I can bring many thousands of new customers to their service, which for my one account has huge effects on their business.

Add all us business customers together and we are the very core of what make their own service successful.

That is no doubt why today I had to give another one of my supported on-line payment services a verbal slap, when without doubt the PayHound service is the worst payment option that I currently offer. And I have directly seen myself that they are so bad that for the first time ever, unless there are some last minute improvements, then this payment option will soon be no more.

Other services like NoChex learned the hard way that you have to treat your business customer's well, when during the time that NoChex did not do so, then they lost a lot of supplier support that they have yet to recover from. NoChex I can say is certainly a good service, where despite their 1% charge they treat me very well.

Even NoChex support recently apologized for having to bother me, when their own financial regulations required them to prove that I really owned my Cardman business account. And it was their hard learned supplier respect that made me play the good little supplier by sending them the required details ASAP.

I have been waiting for PayHound to learn this same lesson and for my account with them to morph into something that so I would term "usable". Instead for months my account with them has been nothing beyond a series of problems, which has certainly been harming my customers using their service.

PayHound when they first started filled a gap left in the on-line payment service market, where to begin with they had considerable promise, due to their good flexible payment and chargeless service. Unfortunately, they pushed things too far, when due to their lack of income and profit their customer support suffered.

When the EU regulations came around, then NoChex did well in tackling this problem, but PayHound fell into this pit and never managed to climb out of it. Even though PayHound are now wise enough to charge nominal fees intent on profit, but it is their poor on-line payment service that is holding them back from turning thing around and instead flooding their support department with customer complaints.

I have certainly waited a very long time for PayHound to turn things around, where I am now feeling that I have waited long enough and that PayHound should go the MinutePay route and be confined to history.

MinutePay by the way lost out due to their desire for a EU wide Euro payment system simply did not match the market reality, when these ancient banking systems have been very hostile towards Euro support across countries. PayHound on their other hand are suffering a management problem, where these people think that they know what is best and obviously do not.

Anyway, PayPal and NoChex are my two most popular on-line payment choices, when they know to treat suppliers well. As it is us suppliers who make their services as popular as they currently are, simply by giving them our support.

The other thing that I had to do today was to upload new versions of my Order Forms, when the popularity of my new Euro based on-line payment options highlighted the fact that I had forget to mention the currency rate. That is of course required when my accounts are in Sterling, when all these EUR and USD payments need this currency rate to be entered as their Sterling value.

I have some other good news today, when as of yesterday my Global Domination Plan (see my special offers page) has had another conquest, when this time it is the Philippines that has been invaded. And if I remember I will add those details when I come to upload this latest news.

The final thing to mention is that adding that CAS Interface 2 should hopefully be complete next week, when testing and adding this new item is taking longer than expected, when I have had other business matters to deal with.

So from the best to the worst of on-line payment services, then that is the end of my news for today.