The problem here is that changes to one subsystem often require changes to another subsystem. Very few channges happen totally in isolation.
Another issue to consider is the constant advancement in hardware. One reason for changes at the memory management level was the the existing system was inefficient when run on large memory machines. Your mechanism will lead to stagnation in the development tree. Commercial distributions will ignore this, releasing patched kernels with source that hasn't been fully audited, leading to stability problems.
I had a collegue who was fired the day he returned from his honeymoon.
The manner of his firing disgusted me. He was talking to his wife on his honeymoon flight about the inadequacies of the customer site he was working at, and was overheard by one of their own staff. They complained and managed to get him dismissed.
The fact that the statments he made about this third-party were 100% true had nothing to do with it.
I can comment on this. My livelihood was dedicated to a particular operating system, and I was one of the UK's top troubleshooters. Unfortunatly the manufacturer got bought into by Microsoft, and the product was slowly strangled until the customer base dropped off dramtically. I'm sure that serious filesystem corruption bugs were introduced and deliberatly not fixed for months.
Whilst all this was going on, my then employer was reluctant to invest serious money in either quality training or equipment - even as late as the end of 1998 test equipment was still 486 machines with 16MB of RAM. I was a resource too good to lose, yet I wasn't given the opportunity to learn new skills except in my own time outside work, which I dedicated to Linux, of course. When the inevitable collapse of our supplier occurred, I had no recognised other skills to fall back on.
Unfortunatly Linux take up in the UK is still pretty slim, except in the ISP market. The PHBs are driven too much by the power of marketing, and not by the substance of the product. (The same is also affecting our political system). I've decided to get out of IT altogether, and have some other positions in the pipeline.
Where are the version numbers, particularly of Nessus. How are we to know that they tested the current version, or a version bundled with a 6-month old Linux distribution?
One other valid point missed in the review is the frequency of product updates. What mechanisms exist to check for newly discovered vunerabilities? Nessus can be made to automatically install updated scripts, but I have absolutely no idea of the other products reviewed. An out of date security tool can be worse than no security tool at all, as it installs a false sense of confidence. Would you put all your trust in a 1-year-old security scanner - I wouldn't.
...that if this does go ahead, Railtrack have absolutely nothing to do with this.
For those of you outside the UK, Railtrack are the company that took over the running of the railway tracks during privatisation. There was a fatal accident a few months ago, then it was realsised that Railtrack had not been performing enough preventative maintenance over the years. Consequentially we had 20 mile-an-hour limits placed on many sections of track. I recently took 4 hours to complete a journey that usually takes less than two.
I've a journey to make this weekend, from Leeds to Leicester and back. Railtrack are rebuilding Leeds station, and the work was supposedly to have been completed over the holiday period. It hasn't, resulting in trains being replaced by buses. I've no idea if my train will be running this weekend.
The swastika was also a military emblem for Finland and Latvia. The Finns used the swastika until the end of WWII, using a light blue on on a blue background for aircraft identification.
A couple of years ago I saw a model kit of Japanese origin of a Hawker Hurricane in Finnish markings. On the box art all swastikas had been airbrushed out. The decal set did, however, contain the correct decals.
I've worked with Domino/Notes since version 2, and it is the most overated pile of junk I've ever had the displeasure of working with. I have not touched version 5, though.
Notes/Domino may have potential, but it has been implemented badly. The mail client under version 4 was horrible, full of user-interface inconsistancies and lacking some common features (such as reply all). The web server is the slowest I've ever seen.
IMHO Notes/Domino should be wiped from the planet.
Re:They've missed something.
on
A Year of Linux
·
· Score: 1
No. Virgin.net, a smaller UK national ISP who are practically NTL under another name.
I went with them as they were one of the first in the UK to support USR 56K modems. I need to reflash my modem, but I can't find an upgrade for my model, which is one of the UK specific ones. We have problems getting modems approved, and we have to use UK specific models (basically we connect to the phone line by induction instead of directly).
No mention of XFree 4.0.2, which includes anti-alias support. This is going to have a significant impact on UI readability in the future. QT has already been modified, and relevant updates have gone into Debian unstable.
I have yet to try this myself, though. I run Debian woody, but I want to dist-upgrade to sid. I only have a 56K modem, and my ISP has swapped out their modem banks, and my modem doesn't work at full speed with their new modems until I flash it to V90. I can't find any V90 flash files for my modem. It looks like I'm going to be spending this holiday period looking for flash upgrades.
I bet that when these come out, most of the systems running this will not be running the correct device drivers.
I've lost count of the number of UDMA capable machines I've seen running in PIO mode under both Linux and NT. Not only are data-transfer speeds lousy, but CPU utilisation during data-transfer is obscene. Then they wonder why their applications are running slowly.
The Atari Casino cartridge included a game called Poker Solitare. You were dealt 25 cards from the deck and had to arrange them on a 5x5 grid. You effectivly had 12 hands, 5 verticle, 5 horizontal, and 2 diagonal and received points for each hand.
There was a maximum possible score hinted in the manual, but it seemed impossible to get the correct 25 cards to acheive this. I worked out the highest scoring layout - you needed to obtain 4 Royal flushes, as well as a straight flush involving a 9.
Did anyone manage this, or was the card sequence impossible to obtain?
Be careful. Even if the OS itself is in ROM, and you manage to get the correct filesystem installed, you could still have problems.
I worked with some since dead specialised hardware which required the disks to contain special signatures written outside of the filesystem, otherwise the filesystem wouldn't get mounted. They may have used special disk firmware as well. You had to use an undocumented mechanism to format a drive sucessfully, only known to the manufacturers own tech-support engineers.
Last year I read a review in one of the better UK magazines. They were doing a comparative test of a number of SCSI and EIDE add-on boards, including a number that did RAID.
The benchmark results varied significantly, particularly for one RAID conroller which gave utterly appalling results. It then dawned on me why this was the case. The magazine had performed all their benchmarks under Windows95, yet the RAID controller had no Windows95 drivers available. For this controller they had run their benchmarks using the 16-bit emulation mode.
This is not the first time I've seen such benchmarks run. I've lost count of the number of IDE based systems I've seen not running in optimum UDMA mode.
But 10 years ago you never needed to change your BIOS, something which is commonplace nowadays. I only swapped the BIOS once during those days, and that was to install a special version with no POST to minimalise boot times - counting 16MB of memory took a long time!
The OP was pointing out that quality control has slipped dramatically over the years. Today's hardware has become overtly complicated, leading to numerous teething problems, just to squeeze an extra 5% performance from the hardware. Most of this complexity would not be required if we didn't have the need to run excessivly complex software instigated by the latest programming techniques. OOP with late linking and bytecode interpreters have a lot to answer for.
I've attended a number of interviews, but I've never been given a practical test.
Practical, under pressure tests are the only real way of weeding out the capable from the incapable. A really good test would be to give someone a broken system using an OS variant or piece of software they'd never seen before, and ask them to fix it. Give them a PC with web. Most will panic, others will chose options at random, hoping things will work, but some will tackle it laterally, using whatever resources are available, performing web-searches and.
I've had to do things like this under pressure. I once was sent to a customer to fix a number of things, one of which was some obscure NT backup software which wouldn't backup. I didn't even have any practical experience of NT at the time. By a process of elmination I discovered the the cause of the fault and fixed it in around 10 minutes - the customer had had the problem for months.
I also attended a training course where we were given servers that had been broken in about 5 different ways (missing executables, corrupted data files etc) and was expected to get everything working ASAP, without resorting to a full system restore from OS media and data backup. A really nasty test, but I was the only one to pass it with flying colours.
This situation is very simillar to the position prior to deregualtion, where you were only allowed to use BT telephones, rented from BT at an astronomical rate.
The television license (around 120 GBP per year) pays for almost all BBC output, plus a bit going to maintain the shared terrestial transmitters.
As the BBC has 2 main terrestial TV channels, a few more digital ones, 5 national radio stations and 30+ regional radio stations. And no advertising on any of those, except for other BBC programmes.
IMHO the TV license is a bargain alone just for BBC radio.
If I ever came across such stupidity as NDA reporting of a problem I'd let everyone know, and screw the NDA. It's probably illegal under some EU regulations anyway to restrict infomation in such a way.
Many a time I've come across serious problems in commercial software or hardware which the manufacturers have known about, yet not bothered to fix.
I once worked on a project run by the biggest fool I've ever worked for. Testing ran to a schedule as follows:
Meeting at the beginning of the week to determine which new features/bug fixes would get implemented that week.
Three days of programming, followed by a cut to CD-ROM at the end of the third day. This cut could be at any time, once it occurred after midnight.
On Friday test everything to make sure it worked, including 100% components that had not even been touched in the previous month. Ship to customer no matter what was found, but report bugs at the following Monday meeting.
No wonder the software never worked properly. Problems were being introduced due to tired developers. Not enough time was spent squashing some really evil bugs. And one developer was so utterly hopeless he should have been fired.
If you think the Spice Girls are bad, you should listen to some of the other junk we have to put up with. S-Club7, Westlife, Steps, Robbie Williams. Somedays I wish I was deaf!
Slashdot Mirror
Another issue to consider is the constant advancement in hardware. One reason for changes at the memory management level was the the existing system was inefficient when run on large memory machines. Your mechanism will lead to stagnation in the development tree. Commercial distributions will ignore this, releasing patched kernels with source that hasn't been fully audited, leading to stability problems.
The manner of his firing disgusted me. He was talking to his wife on his honeymoon flight about the inadequacies of the customer site he was working at, and was overheard by one of their own staff. They complained and managed to get him dismissed.
The fact that the statments he made about this third-party were 100% true had nothing to do with it.
Whilst all this was going on, my then employer was reluctant to invest serious money in either quality training or equipment - even as late as the end of 1998 test equipment was still 486 machines with 16MB of RAM. I was a resource too good to lose, yet I wasn't given the opportunity to learn new skills except in my own time outside work, which I dedicated to Linux, of course. When the inevitable collapse of our supplier occurred, I had no recognised other skills to fall back on.
Unfortunatly Linux take up in the UK is still pretty slim, except in the ISP market. The PHBs are driven too much by the power of marketing, and not by the substance of the product. (The same is also affecting our political system). I've decided to get out of IT altogether, and have some other positions in the pipeline.
One other valid point missed in the review is the frequency of product updates. What mechanisms exist to check for newly discovered vunerabilities? Nessus can be made to automatically install updated scripts, but I have absolutely no idea of the other products reviewed. An out of date security tool can be worse than no security tool at all, as it installs a false sense of confidence. Would you put all your trust in a 1-year-old security scanner - I wouldn't.
For those of you outside the UK, Railtrack are the company that took over the running of the railway tracks during privatisation. There was a fatal accident a few months ago, then it was realsised that Railtrack had not been performing enough preventative maintenance over the years. Consequentially we had 20 mile-an-hour limits placed on many sections of track. I recently took 4 hours to complete a journey that usually takes less than two.
I've a journey to make this weekend, from Leeds to Leicester and back. Railtrack are rebuilding Leeds station, and the work was supposedly to have been completed over the holiday period. It hasn't, resulting in trains being replaced by buses. I've no idea if my train will be running this weekend.
A couple of years ago I saw a model kit of Japanese origin of a Hawker Hurricane in Finnish markings. On the box art all swastikas had been airbrushed out. The decal set did, however, contain the correct decals.
Notes/Domino may have potential, but it has been implemented badly. The mail client under version 4 was horrible, full of user-interface inconsistancies and lacking some common features (such as reply all). The web server is the slowest I've ever seen.
IMHO Notes/Domino should be wiped from the planet.
I went with them as they were one of the first in the UK to support USR 56K modems. I need to reflash my modem, but I can't find an upgrade for my model, which is one of the UK specific ones. We have problems getting modems approved, and we have to use UK specific models (basically we connect to the phone line by induction instead of directly).
I have yet to try this myself, though. I run Debian woody, but I want to dist-upgrade to sid. I only have a 56K modem, and my ISP has swapped out their modem banks, and my modem doesn't work at full speed with their new modems until I flash it to V90. I can't find any V90 flash files for my modem. It looks like I'm going to be spending this holiday period looking for flash upgrades.
I've lost count of the number of UDMA capable machines I've seen running in PIO mode under both Linux and NT. Not only are data-transfer speeds lousy, but CPU utilisation during data-transfer is obscene. Then they wonder why their applications are running slowly.
Clueless MCSEs appear to do this all the time!
There was a maximum possible score hinted in the manual, but it seemed impossible to get the correct 25 cards to acheive this. I worked out the highest scoring layout - you needed to obtain 4 Royal flushes, as well as a straight flush involving a 9.
Did anyone manage this, or was the card sequence impossible to obtain?
I worked with some since dead specialised hardware which required the disks to contain special signatures written outside of the filesystem, otherwise the filesystem wouldn't get mounted. They may have used special disk firmware as well. You had to use an undocumented mechanism to format a drive sucessfully, only known to the manufacturers own tech-support engineers.
The benchmark results varied significantly, particularly for one RAID conroller which gave utterly appalling results. It then dawned on me why this was the case. The magazine had performed all their benchmarks under Windows95, yet the RAID controller had no Windows95 drivers available. For this controller they had run their benchmarks using the 16-bit emulation mode.
This is not the first time I've seen such benchmarks run. I've lost count of the number of IDE based systems I've seen not running in optimum UDMA mode.
The OP was pointing out that quality control has slipped dramatically over the years. Today's hardware has become overtly complicated, leading to numerous teething problems, just to squeeze an extra 5% performance from the hardware. Most of this complexity would not be required if we didn't have the need to run excessivly complex software instigated by the latest programming techniques. OOP with late linking and bytecode interpreters have a lot to answer for.
Practical, under pressure tests are the only real way of weeding out the capable from the incapable. A really good test would be to give someone a broken system using an OS variant or piece of software they'd never seen before, and ask them to fix it. Give them a PC with web. Most will panic, others will chose options at random, hoping things will work, but some will tackle it laterally, using whatever resources are available, performing web-searches and .
I've had to do things like this under pressure. I once was sent to a customer to fix a number of things, one of which was some obscure NT backup software which wouldn't backup. I didn't even have any practical experience of NT at the time. By a process of elmination I discovered the the cause of the fault and fixed it in around 10 minutes - the customer had had the problem for months.
I also attended a training course where we were given servers that had been broken in about 5 different ways (missing executables, corrupted data files etc) and was expected to get everything working ASAP, without resorting to a full system restore from OS media and data backup. A really nasty test, but I was the only one to pass it with flying colours.
I believe the world would be a better place without lawyers and marketing types.
Can't we make up a story about a ravenous star-goat and ship them to some distant planet?
Oh! By the way this book's copyright has expired. You can do absolutely anything with it you like.
This situation is very simillar to the position prior to deregualtion, where you were only allowed to use BT telephones, rented from BT at an astronomical rate.
As the BBC has 2 main terrestial TV channels, a few more digital ones, 5 national radio stations and 30+ regional radio stations. And no advertising on any of those, except for other BBC programmes.
IMHO the TV license is a bargain alone just for BBC radio.
If I ever came across such stupidity as NDA reporting of a problem I'd let everyone know, and screw the NDA. It's probably illegal under some EU regulations anyway to restrict infomation in such a way.
Many a time I've come across serious problems in commercial software or hardware which the manufacturers have known about, yet not bothered to fix.
And miss the football?
Meeting at the beginning of the week to determine which new features/bug fixes would get implemented that week.
Three days of programming, followed by a cut to CD-ROM at the end of the third day. This cut could be at any time, once it occurred after midnight.
On Friday test everything to make sure it worked, including 100% components that had not even been touched in the previous month. Ship to customer no matter what was found, but report bugs at the following Monday meeting.
No wonder the software never worked properly. Problems were being introduced due to tired developers. Not enough time was spent squashing some really evil bugs. And one developer was so utterly hopeless he should have been fired.
Oh look! A flock of pigs has just gone past my window!
If you think the Spice Girls are bad, you should listen to some of the other junk we have to put up with. S-Club7, Westlife, Steps, Robbie Williams. Somedays I wish I was deaf!
I live in a town with a population of 10,000 in Yorkshire. We've as much chance of getting cable as Ian Paisley has of becoming the next pope!
There was the period under Oliver Cromwell, but that ended around 1660 when Charles II came back.