I'd love to use Postgres, but need confidence that it's not got a massive deficiency somewhere.
Take a look at Postgres...it is MUCH more like Oracle in terms of a robust RDBMS. I've heard of projects taking pretty large Oracle installs...and converting over to Postgres with minimal pain.
The main reason MySQL is more popular is that it was smaller and easier to configure...but at the cost of robustness, and initially...data integrity. It was a short cut...much like {gag} MS Access proved often to be.
Postgres takes a bit more planning, and know how to install and use, but then again...so does something like Oracle.
You could probably compare:
MySql == Access
Postgres == Oracle
If you want to make some analogies.
Thank you for your replies.
I wrote the original comment - apologies, but I forgot to login (I don't login here often, I tend to lurk)
The reason I ask is because i've been suspicious of MySQL because of the dual licensing, and also because the (expensive) cluster version needs the indices to be in memory - which requires serious hardware for our setup as our data is 'long and thin'. However there is little experience of postgres here, so we spend the money.
Therefore i've never bothered with postgres, which is stupid, but I know that as many open source projects use mysql, it is the 'go to' database of choice. When speccing a database i've always asked around and people have pointed out the deficiencies of postgres as being the clustering and backup support.
To be specific - how does the clustering (any method) of postgres compare to standard mysql? What is the best way of doing hot backups? Where does the performance fall down?
I'd love to use postgres, but unfortunately i'm too busy doing other tasks to give it a good, proper test. Has anybody been through this already and do they mind sharing?
There are a few other players in the field next to teradata, but when you move to that format there is nothing that would be associated with the word cheap.
However, generally when it gets to that level of field the amount of data in storage usually makes it very obvious.
In some scenarios, we have avoided going to those rather massive solutions by really digging down and seeing if we really needed to store everything.
In a previous job at the start of my career, my company bought a Teradata system which came with the requisite sharp suited consultant, who told us how to lay out the DB schema.
Being Teradata all the hashed indexes were in vogue, so it was lightning fast.
Until the day they realised the users mainly did substring searches, which don't really work on a hashed index. Table scans a plenty = unhappy users.
It doesn't mean a RDBMS is bad, it means that technology misapplied always sucks.
Because Geode was crap! In a previous job we used embedded Geode boards in shops. They were useless - the hardware was some virtualised nonsense, the watchdogs didn't work, throughput was awful and documentation non-existent. Definitely the worst boards i've ever used.
I was sorting out the network of a businessman friend of mine. He'd recently spent £4000 on 4 IP cameras and a PC to monitor it all.
I've never seen such suck. The interface only worked with IE7, nothing newer, and had to install an ActiveX control plus three other InstallShield installers to get any sort of picture. Most of the screen was overlaid with cheesy graphics, and some of the tooltips were in Chinese.
The suppliers didn't have a clue about it, all they do is buy this shite by the containerfull from China. The chance of a software fix is nil as the guy who wrote it probably committed suicide after seeing how bad it was.
Anyways I whipped up a Ubuntu Server install with Zoneminder, and it's like night and day.
DVR kit is in one of those vertical markets where appalling quality is the standard, it seems.
But it is an *mp3* listening test, the idea is to see the state of the art with mp3 and to check whether 128kbps is a reasonable bitrate for everyday use.
Also, this is a double blind test - maybe you should have read the link? HydrogenAudio are well known for their scientific approach to audio codec testing.
Maybe a little less of the arrogance next time, eh?
Round here (Gloucestershire, England) they're worse than the boy racers. They don't indicate, take ridiculous changes-of-heart in betweeen manoeuvres, drive at 20mph and piss me off. Get a taxi.
VIA would have had a lot of business from me had their drivers either worked, or their communication been good, or they'd opensourced all this stuff earlier.
In my last job we produced kiosk and signage displays, as well as multiple display units for bookmakers shops. Each shop ran up to 32 displays, and some customers had 4-5000 shops. That's a lot of hardware.
We first looked at VIA miniitx boards as they had everything we needed, technically speaking. But the drivers sucked, bore no resemblance to any documentation, were incapable of being accelerated, and couldn't be rotated (for portrait displays) no matter how much they were supposed to. If they supported opengl properly it wouldn't have been an issue. This is without going into the problems of the thousands of different revisions of VIA hardware, all of which failed in subtly different ways.
I left the company but they're now using NVidia boards which are more capable, but not as suitable for an embedded solution. Like me the rest of the techies got sick of fighting VIA stupidity.
If this new open methodology gives us proper opengl support and video decode in a mainline kernel they will win a lot of customers. As it is, it's still some way from that. A real shame.
Wrong. The whole point of the hoopla over this version is that it removes the last bit of protection from the image. FFMPEG is starting to implement playback, so it's entirely possible that i'll be able to play backups I make now in the future under Linux.
I'd imagine port knocking is easier to implement, as you don't have to keep changing the port an app is listening on. You could cryptographically change the knocking sequence, which would be far easier than the solution described here.
Slashdot Mirror
I used to work at a nuclear fuel processing plant here called Springfields:
http://www.nuclearsites.co.uk/site.php?LocationID=2
Certainly in the UK, where I live, they're not. I don't think the poster stated where they live.
http://www.national-lottery.co.uk/player/p/help/playinginstore/faqs.ftl
Out of interest, what's the problem with master/slave under postgres?
Take a look at Postgres...it is MUCH more like Oracle in terms of a robust RDBMS. I've heard of projects taking pretty large Oracle installs...and converting over to Postgres with minimal pain.
The main reason MySQL is more popular is that it was smaller and easier to configure...but at the cost of robustness, and initially...data integrity. It was a short cut...much like {gag} MS Access proved often to be.
Postgres takes a bit more planning, and know how to install and use, but then again...so does something like Oracle.
You could probably compare:
MySql == Access
Postgres == Oracle
If you want to make some analogies.
Thank you for your replies.
I wrote the original comment - apologies, but I forgot to login (I don't login here often, I tend to lurk)
The reason I ask is because i've been suspicious of MySQL because of the dual licensing, and also because the (expensive) cluster version needs the indices to be in memory - which requires serious hardware for our setup as our data is 'long and thin'. However there is little experience of postgres here, so we spend the money.
Therefore i've never bothered with postgres, which is stupid, but I know that as many open source projects use mysql, it is the 'go to' database of choice. When speccing a database i've always asked around and people have pointed out the deficiencies of postgres as being the clustering and backup support.
To be specific - how does the clustering (any method) of postgres compare to standard mysql? What is the best way of doing hot backups? Where does the performance fall down?
I'd love to use postgres, but unfortunately i'm too busy doing other tasks to give it a good, proper test. Has anybody been through this already and do they mind sharing?
Ta,
Sean
Maybe it's time to push atheism as a 'religion'? Maybe we can succeed where Dawkins can't?
There are a few other players in the field next to teradata, but when you move to that format there is nothing that would be associated with the word cheap.
However, generally when it gets to that level of field the amount of data in storage usually makes it very obvious.
In some scenarios, we have avoided going to those rather massive solutions by really digging down and seeing if we really needed to store everything.
In a previous job at the start of my career, my company bought a Teradata system which came with the requisite sharp suited consultant, who told us how to lay out the DB schema.
Being Teradata all the hashed indexes were in vogue, so it was lightning fast.
Until the day they realised the users mainly did substring searches, which don't really work on a hashed index. Table scans a plenty = unhappy users.
It doesn't mean a RDBMS is bad, it means that technology misapplied always sucks.
Because Geode was crap! In a previous job we used embedded Geode boards in shops. They were useless - the hardware was some virtualised nonsense, the watchdogs didn't work, throughput was awful and documentation non-existent. Definitely the worst boards i've ever used.
Very true.
I was sorting out the network of a businessman friend of mine. He'd recently spent £4000 on 4 IP cameras and a PC to monitor it all.
I've never seen such suck. The interface only worked with IE7, nothing newer, and had to install an ActiveX control plus three other InstallShield installers to get any sort of picture. Most of the screen was overlaid with cheesy graphics, and some of the tooltips were in Chinese.
The suppliers didn't have a clue about it, all they do is buy this shite by the containerfull from China. The chance of a software fix is nil as the guy who wrote it probably committed suicide after seeing how bad it was.
Anyways I whipped up a Ubuntu Server install with Zoneminder, and it's like night and day.
DVR kit is in one of those vertical markets where appalling quality is the standard, it seems.
Our management have been chomping at the bit to get iphones.
Unfortunately they've also mandated we s/mime encrypt all intra-company email, which doesn't work on the thing as you can't install a certificate.
Does anyone with access to the new SDK know if certs have been added to the thing?
Wouldn't it be better to fix GCC so it has the same optimisations?
But it is an *mp3* listening test, the idea is to see the state of the art with mp3 and to check whether 128kbps is a reasonable bitrate for everyday use.
Also, this is a double blind test - maybe you should have read the link? HydrogenAudio are well known for their scientific approach to audio codec testing.
Maybe a little less of the arrogance next time, eh?
... they get a fucking taxi?
Round here (Gloucestershire, England) they're worse than the boy racers. They don't indicate, take ridiculous changes-of-heart in betweeen manoeuvres, drive at 20mph and piss me off. Get a taxi.
VIA would have had a lot of business from me had their drivers either worked, or their communication been good, or they'd opensourced all this stuff earlier.
In my last job we produced kiosk and signage displays, as well as multiple display units for bookmakers shops. Each shop ran up to 32 displays, and some customers had 4-5000 shops. That's a lot of hardware.
We first looked at VIA miniitx boards as they had everything we needed, technically speaking. But the drivers sucked, bore no resemblance to any documentation, were incapable of being accelerated, and couldn't be rotated (for portrait displays) no matter how much they were supposed to. If they supported opengl properly it wouldn't have been an issue. This is without going into the problems of the thousands of different revisions of VIA hardware, all of which failed in subtly different ways.
I left the company but they're now using NVidia boards which are more capable, but not as suitable for an embedded solution. Like me the rest of the techies got sick of fighting VIA stupidity.
If this new open methodology gives us proper opengl support and video decode in a mainline kernel they will win a lot of customers. As it is, it's still some way from that. A real shame.
... would have been a more interesting headline...
Wrong. The whole point of the hoopla over this version is that it removes the last bit of protection from the image. FFMPEG is starting to implement playback, so it's entirely possible that i'll be able to play backups I make now in the future under Linux.
I pay $5 a month, and my IP address is immune from any RIAA lawsuits concerning music torrents.
I would pay that, and so would anyone I know. Somehow however I think their idea won't work like that.
Is there any work being done on hacking these things? As they run Linux and have decent video decode capability it would be a bit of a steal.
My girlfriend looks great in pigtails, it takes 10 years off her age too.
Trouble is she's only 22...
http://it.slashdot.org/comments.pl?sid=426412&cid=22144084
... so that explains why the RDBMS dudes were bitching about mapreduce t'other day:
http://it.slashdot.org/article.pl?sid=08/01/18/1813248
They're taking the piss surely?
OMG! The bloggers are on strike, oh noes!! Where will I get my random crap and aerated opinions from?!!!
It's almost as ridiculous as the 'Students Unions' we have in Universities here...
About time I reckon. VB was genuinely useful for a while, but the industry grew up past it and now it's just an insecure relic.
Corps need to get their bespoke VBA apps onto the web and be done with it.
I'd imagine port knocking is easier to implement, as you don't have to keep changing the port an app is listening on. You could cryptographically change the knocking sequence, which would be far easier than the solution described here.
You're correct, but the writeup uses current tense, thus giving the impression this is still spreading when it isn't.
Facebook have already blocked it, days ago...