I don't see why this is so hard. Let them charge for allowing companies to bypass the spam filters, fine no big deal... IF they also allow users to opt out of any such paid email.. This would still require that the senders have the user's permission to send, but prevent the sender from getting trapped by a stupid filter.
So am I the only one to figure this out? This pay for spam has the makings of a really great idea, but it is there yet..
"Who you know" calling bullshit on that. It is a hell of a lot how you present yourself. I have responded to job listings on the Internet with an email and gotten the job soon after, knowing that many other people responded, but they didn't seem to understand what the customer was looking for. I did and got it. Same thing goes for the numbers game. Sure we hire a programmer and get tons of resumes, more than 99% are total crap. So it you are actualy *good* at what you do, you will win:)
Wish I had mod points for you. There is always a clue to such posts as your parent, things like, complaining about sending emails. So Net::SMTP was really hard to find? You have to know the modules and tools people write in Perl just like with anything else. But with Perl it is quite easy to fix the damn thing yourself. Recently we were building a home spun(for very good reason) webmail app. The use of Email::StripMIME save massive amounts of time and worked very well. When we found a bug, we worked around it and sent an email to the author. We are now handling hundreds of mail accounts with hundreds of thousdands of emails and doing quite well. The sad part is that not many people seem to understand what you and I do about Perl. But a few do, one great example of a LARGE app that I know of is Bricolage.. Which is getting cooler all of the time, now support PHP templates.
Unless you consider what Sun has done with Zones in Solaris 10.. Now we can divide up that 8 core chip into 8 virutal machines or more. That combined with a decent RAID 10 setup would be a pretty beautifl thing, and as easy or easier to maintain as 8-10 real machines. And guess what about price?
Yeah right, IMAP will not come close to dealing with the kind of volume that a large company would need. It has to be an SQL db, but one that can also be accessed with IMAP clients. In other words DBmail (dbmail.org)
Your system sounds pretty good, I love rsync as well. But if you are dealing with very large amounts of data that 100M network is a bad bottleneck. So I would go with Gigabit, just so your backups don't start taking half a day. You also should have the fastest CPU you can get for the machine that is actualy running rsync.
Just try to configure user privileges with MS Active Directory. It is the best example I know about of a GUI that went horribly wrong that would be a lot easier to understand with a config file.
Another good example is just about any DNS interface. There is some safty in a GUI for DNS, but there are also lots of problems that can be delt with a lot faster in DNS if you can just directly edit the zones if that be by hand or script..
This is a fundamental flaw in PHP, that Perl and Java and I would imagine other lang don't suffer. There is no attempt to do cross database coding with the defaults, and any kind of database support is a compile time option, like SO many other things in PHP. That esp I think is insane. Why in the hell should I have to recompile PHP to support postgress? I should be able to make the change as easily as you suggest, change the DBI drvier, or the JDBC driver. Which amounts to having the driver and a one line code change.
This is why I use Perl.
http://cpan.uwinnipeg.ca/dist/GD-Barcode
But it looks like it is not that hard to find PHP stuff either..
http://www.hotscripts.com/PHP/Scripts_and_Programs/E-Commerce/Barcode_Generation/
So I guess you didn't search very hard. This took me a solid 5 minutes. Barcodes are blindly simple, but if they are messed up, you can't do much about it with a closed source app, send it back, get a refund. Yeah, great. With these simple examples you can have full control and not have to be the victim of a bug you can't fix yourself.
So harder must mean you can't use google and never heard of CPAN. And of course you can use Perl from PHP and as a result you have an outstanding collection of tools free and at your fingertips.
Thanks for the post, but no. Only for two at a time, "Currently pgpool supports up to 2 PostgreSQL servers. " That is the problem, everywhere I look there are serious hard limitations like this.
So for $1000 I am still having to deal with all of these restrictions, NONE of which exist with mysql replication. And I still have not even gotten to my issues with mysql replication.
What Mamoth can't do that mysql can:
The table to be replicated must contain a Primary Key. Multi Column keys are o.k.
The table must not inherit columns from other tables.
The value of the primary key must be of less than 8k bytes in actual length. For example if you have a primary key value of 10000, the value is only 5 bytes.
It currently works with 7.3+ & 7.4+ of the Open Source or Mammoth PostgreSQL.
You must be running RedHat(7.3,8,9,AS,ES,Fedora)/Suse Linux (8.2,9), or Solaris 9, (Mac OS X coming soon).
The database name on the Master & Slaves must be the same.
This leads me to believe this is not based on logs like you suggested at first.
Again, I wish I could find this great replication system for postgress, we would be moving to it from mysql in a second if it did exist.
Did you not read what I wrote. Structure changes to dbs don't get replicated with Slony, HOW THE HELL can anyone use that even as a backup?? You didn't look close enough at mysql's replication. And get my tone, I am saying this even though I am not happy with mysql replication right now because it still doesn't do enough. But from what is documented on the Slony site, mysql's does a lot more and is much more useable. I WISH that were not the case and wonder what the hell is taking so long for Postgress that is so far ahead in just about every other way to catch up???
>Now with Slony-I, PostgreSQL has a decent, robust, >and open source replication solution, I will expect >continued interest in this area.
I have been wishing for something like this for 3 years now! But Slony is still not nearly up to speed with mysql replication, just looking at the amount of work you have to do to set it up, but more than that the very idea that schma changes don't replicate? (I take that from the current limitations on their website) That makes it a pile of garbage:( And I am sad to say that, not happy. The whole reason we are using mysql is because of its replication feature, yet as good as that is this is still a long way from what we really need which is more like the Oracle multi master. But that is way out there for us price wise when that is the only feature we really need.
Thanks,
Eric
Slashdot Mirror
I don't see why this is so hard. Let them charge for allowing companies to bypass the spam filters, fine no big deal... IF they also allow users to opt out of any such paid email.. This would still require that the senders have the user's permission to send, but prevent the sender from getting trapped by a stupid filter. So am I the only one to figure this out? This pay for spam has the makings of a really great idea, but it is there yet..
"Who you know" calling bullshit on that. It is a hell of a lot how you present yourself. I have responded to job listings on the Internet with an email and gotten the job soon after, knowing that many other people responded, but they didn't seem to understand what the customer was looking for. I did and got it. Same thing goes for the numbers game. Sure we hire a programmer and get tons of resumes, more than 99% are total crap. So it you are actualy *good* at what you do, you will win :)
Wish I had mod points for you. There is always a clue to such posts as your parent, things like, complaining about sending emails. So Net::SMTP was really hard to find? You have to know the modules and tools people write in Perl just like with anything else. But with Perl it is quite easy to fix the damn thing yourself. Recently we were building a home spun(for very good reason) webmail app. The use of Email::StripMIME save massive amounts of time and worked very well. When we found a bug, we worked around it and sent an email to the author. We are now handling hundreds of mail accounts with hundreds of thousdands of emails and doing quite well. The sad part is that not many people seem to understand what you and I do about Perl. But a few do, one great example of a LARGE app that I know of is Bricolage.. Which is getting cooler all of the time, now support PHP templates.
I hear that happens when you get older.
How about China? Everywhere? Or just the west? Think before your post... Yes, I know this is Slashdot...
I would just like to thank you for making me spit up on my keyboard :)
Or just get a girlfriend for god's sake!
Anyone for speeding it up? How many nukes can we toss in?
Unless you consider what Sun has done with Zones in Solaris 10.. Now we can divide up that 8 core chip into 8 virutal machines or more. That combined with a decent RAID 10 setup would be a pretty beautifl thing, and as easy or easier to maintain as 8-10 real machines. And guess what about price?
This is slashdot, it is pronouced Wessel..
There is nothing, nothing as lame as quoting yourself.
Yeah right, IMAP will not come close to dealing with the kind of volume that a large company would need. It has to be an SQL db, but one that can also be accessed with IMAP clients. In other words DBmail (dbmail.org)
Your system sounds pretty good, I love rsync as well. But if you are dealing with very large amounts of data that 100M network is a bad bottleneck. So I would go with Gigabit, just so your backups don't start taking half a day. You also should have the fastest CPU you can get for the machine that is actualy running rsync.
The really sad thing is the lowest rated post on this, is the most truthful :)
Just try to configure user privileges with MS Active Directory. It is the best example I know about of a GUI that went horribly wrong that would be a lot easier to understand with a config file. Another good example is just about any DNS interface. There is some safty in a GUI for DNS, but there are also lots of problems that can be delt with a lot faster in DNS if you can just directly edit the zones if that be by hand or script..
You can't learn anything without reading the man pages. Simple as that..
This is a fundamental flaw in PHP, that Perl and Java and I would imagine other lang don't suffer. There is no attempt to do cross database coding with the defaults, and any kind of database support is a compile time option, like SO many other things in PHP. That esp I think is insane. Why in the hell should I have to recompile PHP to support postgress? I should be able to make the change as easily as you suggest, change the DBI drvier, or the JDBC driver. Which amounts to having the driver and a one line code change.
I realized as I posted my last message that I just totaly misunderstood you. My deepest apologies.
This is why I use Perl. http://cpan.uwinnipeg.ca/dist/GD-Barcode But it looks like it is not that hard to find PHP stuff either.. http://www.hotscripts.com/PHP/Scripts_and_Programs /E-Commerce/Barcode_Generation/
So I guess you didn't search very hard. This took me a solid 5 minutes. Barcodes are blindly simple, but if they are messed up, you can't do much about it with a closed source app, send it back, get a refund. Yeah, great. With these simple examples you can have full control and not have to be the victim of a bug you can't fix yourself.
So harder must mean you can't use google and never heard of CPAN. And of course you can use Perl from PHP and as a result you have an outstanding collection of tools free and at your fingertips.
There is NOTHING more usefull that RTFM. That is why it is an acronym..
Thanks for the post, but no. Only for two at a time, "Currently pgpool supports up to 2 PostgreSQL servers. " That is the problem, everywhere I look there are serious hard limitations like this.
So for $1000 I am still having to deal with all of these restrictions, NONE of which exist with mysql replication. And I still have not even gotten to my issues with mysql replication. What Mamoth can't do that mysql can: The table to be replicated must contain a Primary Key. Multi Column keys are o.k. The table must not inherit columns from other tables. The value of the primary key must be of less than 8k bytes in actual length. For example if you have a primary key value of 10000, the value is only 5 bytes. It currently works with 7.3+ & 7.4+ of the Open Source or Mammoth PostgreSQL. You must be running RedHat(7.3,8,9,AS,ES,Fedora)/Suse Linux (8.2,9), or Solaris 9, (Mac OS X coming soon). The database name on the Master & Slaves must be the same. This leads me to believe this is not based on logs like you suggested at first. Again, I wish I could find this great replication system for postgress, we would be moving to it from mysql in a second if it did exist.
Did you not read what I wrote. Structure changes to dbs don't get replicated with Slony, HOW THE HELL can anyone use that even as a backup?? You didn't look close enough at mysql's replication. And get my tone, I am saying this even though I am not happy with mysql replication right now because it still doesn't do enough. But from what is documented on the Slony site, mysql's does a lot more and is much more useable. I WISH that were not the case and wonder what the hell is taking so long for Postgress that is so far ahead in just about every other way to catch up???
Thanks, I almost bought his post, but your points all make instant sense.
>Now with Slony-I, PostgreSQL has a decent, robust, >and open source replication solution, I will expect >continued interest in this area. I have been wishing for something like this for 3 years now! But Slony is still not nearly up to speed with mysql replication, just looking at the amount of work you have to do to set it up, but more than that the very idea that schma changes don't replicate? (I take that from the current limitations on their website) That makes it a pile of garbage :( And I am sad to say that, not happy. The whole reason we are using mysql is because of its replication feature, yet as good as that is this is still a long way from what we really need which is more like the Oracle multi master. But that is way out there for us price wise when that is the only feature we really need.
Thanks,
Eric