Welcome to An Unmarked Place
The need to completely clone an Oracle instance is something that arises fairly frequently, for example, to create a Production instance from a Development instance or vice versa. There have traditionally been a number of ways to achieve this: creating an additional instance and migrating data using imp/exp or manually manipulating data files and control files. In particular, the imp/exp route requires sufficient disk space for the intermediate files, and requires the destination database to be manually configured to be equivalent to the source database as far as control files, instance parameters and tablespaces are concerned. What if there was a way to completely clone a database, including tablespaces, data file layouts, server parameters and data? In Oracle 11g, RMAN allows you to do just this.
A busker in Murray St Mall, Perth, performs on a drum kit cobbled together from upturned bins, buckets and a cable spool. He’d drawn quite a crowd and I decided to capture the moment.
I shot this on my Sony V1P using a Sony Mid-Side electret condenser microphone (which I really like for capturing location sound on-camera). Editing was performed in Sony Vegas Movie Studio with some additional colour grading.
Coming from a DSLR background, auto focus on the Sony V1P seemed to be an enigma. With no defined “focus points” like a DSLR, it was never clear to me what subject in a frame the V1P would try to focus on. The manual didn’t help, and trial and error seemed to show that it would simply try and focus by maximising the contrast of the of entire frame – in other words, the highest-contrast subject would be what it would try and focus upon.
Read the rest of this entry »
Although there are a few ways to increase the storage capacity on the HP 100LX, 200LX and 1000CX, there are two main ways that are common today. Firstly, upgrading the RAM and secondly, using memory cards. Initially, SRAM cards were the only realistic choice when the 200LX was released around 1994. These card were typically 5MB to 10MB and used Static Ram with a battery backup as a storage device. Flash memory, such as that used in Compact Flash cards, just wasn’t economically viable in those types of sizes.
Today, CompactFlash cards are a commodity and as cheap as chips (excuse the pun), whereas SRAM cards are scarce, difficult to find and expensive. Given that CompactFlash cards are directly compatible with PCMCIA, all that is needed is a CompactFlash to PCMCIA adaptor and you should be away.
Unfortunately, the age of the the HP Palmtops, as well as slight variations of interface in CompactFlash cards, means that some modern sizes may not work natively with the palmtop. ACE Technologies have implemented a driver that allows modern, large CF cards to operate correctly.
The Acecard3 driver intercepts the BIOS 13h Interrupt handlers and allows modern PCMCIA storage devices to operate in the 100LX/200LX, despite the inherent limitations in the device’s 17 year old BIOS
From the enclosed bigcard.txt:
Dear Large Storage Card and HP 100LX Customer, Large PC Cards and other cards sometimes do not work "out of the box" with the LX series because of size and other incompatibities. The enclosed driver from defunct ACE Technologies allows many cards to work. ACECARD3.COM CONFIG.TMP BIGCARD.TXT ACECARD3.COM is the required software driver. For your convenience, we have included CONFIG.TMP, which may be merged with your current CONFIG.SYS file. BIGCARD.TXT contains a copy of this letter. Installation: 1) CONFIG.TMP contains the following line: DEVICE=C:\ACECARD3.COM /SDP Use MEMO to add this line to the active CONFIG.SYS file located on the C drive. Important Notes: a) If you have a 32, 64, or 96 Meg C drive, then add this line to the end of your CONFIG.SYS file on your F drive (not C drive). b) Since your palmtop will not recognize your large memory card until the driver has been installed, you cannot store the CONFIG.SYS and AUTOEXEC.BAT files on your large memory card in the A drive. 2) Copy ACECARD3.COM to your C drive. (32, 64, and 96 Meg C drive users: if you follow the directions above and place the new line at the end of the file, copy ACECARD3.COM to the C drive. If you place the new line before DEVICE=C:\RDSWAP.EXE, then copy ACECARD3.COM to the F drive.)
I have used the driver with success with the majority of
Of course, the problem remains – how does one get the Acecard driver onto the device if the CompactFlash card can’t be used? Fortunately many smaller CompactFlash cards do work natively with the 200LX through the use of a PCMCIA-CompactFlash adaptor. I’ve had success copying the Acecard driver to the 200LX using an old Kodak 8MB CF card, as well as a few random 16MB cards. I haven’t had much success with cards larger than 16MB.
The other alternative is to use a serial cable in conjunction with the embedded version of Laplink – but this requires both a compatible serial cable as well as a host machine running Laplink, which may be difficult to find today. Furthermore, the 1000CX and 100LX do not include Laplink, so this isn’t an option for these models.
If you’ve got any other suggestions on improving the storage capacity of these great little devices, leave some comments below!
I’m very pleased to announce the immediate availability of a driver for the Tascam FW-1884 audio interface and control surface in Studio One.
I picked up one of these recently, and although it’s a few years old now, not much has changed with control surface design over the past ten years, making this a very serviceable unit. However I wasn’t terribly impressed with its support under Studio One using Mackie HUI emulation, and so this driver was born.
Full details are at the FW1884 Driver Page.
I love my Fuji X10. It’s a camera that is full of soul and produces absolutely magical images.
This year at CES, Fujifilm have seen fit to improve on a good thing, announcing the Fujifilm X20.
The X20 is the successor to the X10, and whilst little has been done to alter its gorgeous external appearance, its the changes on this inside that count.
Whilst the X20 keeps what appears to be the same lens and chassis, internally it is sporting a new 2/3″ X-Trans sensor – the same style of sensor as found on the Xpro1. The advantage of this sensor design is that its unique 6×6 Bayer colour array means that aliasing is reduced substantially, removing the need for a low-pass (anti-aliasing) filter. This should result in images that maintain higher levels of details, and a camera that performs better in low-light compared to its predecessors.
The other massive advantage of the new sensor is the inclusion of phase-detect autofocus, which Fuji is claiming will give the X20 the fastest focus ability of any camera in its class (fixed-lens compacts with a non APS-C sensor). It’s pretty unusual for Fuji to be ahead of the pack in terms of focus speed, so time will tell if the camera lives up to its marketing hype.
The new sensor means that the old 2/3″ EXR sensor featured in the X10 is gone. Whilst that might be a good thing in terms of the “white orb issue” experienced with the X10, it’s a bit of a sad loss for a very unique and powerful sensor. I will eagerly await to see how the X-Trans sensor stacks up against the EXR sensor, but one thing I will not miss is having to decide which mode to shoot in for a given scene to get the best results!
A few other nice features of the X20 include an upgraded EXR-II processing engine, which apparently Fuji claim will also help reduce noise and improve dynamic range etc. To be clear, the X10 was no slouch in this area, having some of the best out-of-camera JPEGs I have seen from any camera. In fact, I generally find myself shooting (and using) JPEG over RAW on the X10.
One other big upgrade is the inclusion of a data display in the optical viewfinder, which will hopefully make it a little bit more usable. Shame it’s still only 85% coverage however. But the inclusion of such a feature I think will help cement this camera with enthusiasts.
Finally, the X20 will be able to shoot 1080p at 60 fps, which is nice – but the problem is, Fuji’s cameras generally suck badly at video, with horrible compressors and a complete lack of manual control, so it’s probably not something to get excited about.
No word on pricing or availability yet, but hopefully it will be around or less than the AU$600 mark as the X10 was. Will I upgrade? Probably not, at least for a while. For me, I’m still becoming intimate with the X10, with much more to learn in order to get the best out of it. The X20 is a nice upgrade, but doesn’t introduce enough to justify the cost of an upgrade in my opinion. But hopefully it will attract new purchasers, and help cement Fujifilm as a deserved player in the high-end compacts market.
Sometimes it’s amazing what machines will tell you if you just randomly mash the buttons. I was unable to repeat it, but somehow I was able to coax a City of Perth parking machine to print some diagnostic information.
Not that there’s much of interest to report, but the information may be of interest to someone.
Coin payment appears to be vastly more popular at this machine than card payment, with the last card payment being almost 6 days before the last coin payment. Also, the machine was last opened 16 days prior, which may indicate that maintenance and cash collections don’t occur terribly frequently. There’s a GPRS link, so it’s quite possible that cash is only collected on an “as-needed” basis, so overall there’s a possibility that this machine isn’t terribly popular.
I present a program for determining the primality of a number on the HP 35s Programmable Calculator. Whilst this is not a particularly interesting algorithm, it should be noted that this implementation only uses the stack – no registers or variables are used whatsoever. The program uses a total of 4 entries in the stack, of which typically only one or two are directly accessible at any point in time.
The program is also intended to be relatively short. Whilst faster or shorter implementations are certainly feasible, the objective with this implementation was to use the stack exclusively, whilst not being completely naive.
The algorithm is as follows:
- Copy the candidate prime (x) to all levels of the stack
- Check if x is even – if it is, the number is not prime (An additional check should be performed to check if the candidate is 2, which is prime).
- Calculate the square root of x. Take the integer portion only.
- If the integer portion result is not odd, round to the next highest odd number. Let this be (y)
- If the remainder of x divided by y is 0, the number is not prime. Display the canidate x and the found factor. This will be the highest factor the candidate is divisible by. End.
- Decrease y by 2.
- If y is greater than 1, repeat step 5.
- The number is prime, no factors could be found. Restore x to the first position on the stack.
Few things to note:
- Complexity of this algorithm is O(√x/2)
- Despite the limited size of the stack, the expensive task of calculating the composite ceiling is only performed once.
- The largest factor of the number, as well as the original number to be tested, are preserved in the stack upon conclusion of the program.
A001 LBL A A002 ENTER A003 ENTER A004 ENTER A005 2 A006 RMDR A007 x=0? A008 GTO A037 A009 R↓ A010 √x A011 IP A012 ENTER A013 ENTER A014 2 A015 RMDR A016 1 A017 + A018 + A019 1 A020 x=y? A021 GTO A032 A022 R↓ A023 RMDR A024 x=0? A025 GTO A034 A026 LASTx A027 2 A028 - A029 x<>y A030 R↓ A031 GTO A019 A032 PRIME A033 RTN A034 LASTx A035 x<>y A036 R↓ A037 NOT PRIME A038 RTN