Re: [sqlite] sqlite shared-cache mode usage

2006-08-29 Thread Jay Sprenkle
As you can tell from this forum, locking and synchronization is the area where there is least intuitive understanding among users and is the most consistent source of problems. There must be a deep psychological reason. I don't think it's deep really, just the most complex part to understand.

Re: [sqlite] sqlite shared-cache mode usage

2006-08-29 Thread John Stanton
Jay Sprenkle wrote: On 8/29/06, John Stanton <[EMAIL PROTECTED]> wrote: Thankyou. The Firefox people have merely removed their dependance upon an unreliable resource, cross OS file locking. A prudent design choice. If they come up with an elegant distributed lock protocol it would be worth

Re: [sqlite] sqlite shared-cache mode usage

2006-08-29 Thread Jay Sprenkle
On 8/29/06, John Stanton <[EMAIL PROTECTED]> wrote: Thankyou. The Firefox people have merely removed their dependance upon an unreliable resource, cross OS file locking. A prudent design choice. If they come up with an elegant distributed lock protocol it would be worth propagating

Re: [sqlite] sqlite shared-cache mode usage

2006-08-29 Thread John Stanton
Thankyou. The Firefox people have merely removed their dependance upon an unreliable resource, cross OS file locking. A prudent design choice. If they come up with an elegant distributed lock protocol it would be worth propagating universally in the light of the success of Firefox and its

Re: [sqlite] sqlite shared-cache mode usage

2006-08-29 Thread Martin Jenkins
Ritesh Kapoor wrote: Can you - DRH or someone else provide some more background information on why locking dosen't work on NFS mounted file systems. I just tried to find out what the locking problem was but couldn't find a web page discussing it in any detail. Lots of pages saying there *was*

Re: [sqlite] sqlite shared-cache mode usage

2006-08-29 Thread drh
"Jay Sprenkle" <[EMAIL PROTECTED]> wrote: > > If you run two instances of firefox you trash > your own database. No, you didn't read what I said. Firefox implements their own locking mechanism, so two instances of firefox will play nicely together. The problem is when some other application,

Re: [sqlite] sqlite shared-cache mode usage

2006-08-29 Thread Jay Sprenkle
On 8/29/06, Ritesh Kapoor <[EMAIL PROTECTED]> wrote: Can you - DRH or someone else provide some more background information on why locking dosen't work on NFS mounted file systems. If its a known issue then is there an SQLite compile time option that would remove locking - i couldn't find one.

Re: [sqlite] sqlite shared-cache mode usage

2006-08-29 Thread Jay Sprenkle
On 8/29/06, John Stanton <[EMAIL PROTECTED]> wrote: Jay Sprenkle wrote: > > If it breaks because of something you did, then YOU are the bum. > If it's broken because of the operating system THEY are the bums. > Having someone to blame still doesn't make it work. They are have made the rational

Re: [sqlite] sqlite shared-cache mode usage

2006-08-28 Thread Ritesh Kapoor
Can you - DRH or someone else provide some more background information on why locking dosen't work on NFS mounted file systems. If its a known issue then is there an SQLite compile time option that would remove locking - i couldn't find one. When I got stuck with this problem I had to remove all

Re: [sqlite] sqlite shared-cache mode usage

2006-08-28 Thread John Stanton
Jay Sprenkle wrote: The problem is that they have to work on broken operating systems. I don't know of another way to patch around the problem. Do you? nope. If it breaks because of something you did, then YOU are the bum. If it's broken because of the operating system THEY are the bums.

Re: [sqlite] sqlite shared-cache mode usage

2006-08-28 Thread Jay Sprenkle
The problem is that they have to work on broken operating systems. I don't know of another way to patch around the problem. Do you? nope. If it breaks because of something you did, then YOU are the bum. If it's broken because of the operating system THEY are the bums.

Re: [sqlite] sqlite shared-cache mode usage

2006-08-28 Thread drh
"Jay Sprenkle" <[EMAIL PROTECTED]> wrote: > On 8/28/06, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote: > > Let me emphasize that everything in the previous paragraph is > > hearsay and supposition and could be wildly incorrect. > > Thanks Dr.H. > > I hope it's not true. If it's true It really

Re: [sqlite] sqlite shared-cache mode usage

2006-08-28 Thread drh
"Jay Sprenkle" <[EMAIL PROTECTED]> wrote: > Good morning all, > > I'm in the design stage of a project and had a question about sqlite > shared-cache mode. > The new version of firefox will use mozStorage, which is based on > sqlite using shared-cache mode. I want other programs to be able >

AW: [sqlite] sqlite shared-cache mode usage

2006-08-28 Thread Marc Ruff
in advance. Marc Ruff -Ursprüngliche Nachricht- Von: Jay Sprenkle [mailto:[EMAIL PROTECTED] Gesendet: Sonntag, 27. August 2006 19:07 An: sqlite-users@sqlite.org Betreff: [sqlite] sqlite shared-cache mode usage Good morning all, I'm in the design stage of a project and had a question

[sqlite] sqlite shared-cache mode usage

2006-08-27 Thread Jay Sprenkle
Good morning all, I'm in the design stage of a project and had a question about sqlite shared-cache mode. The new version of firefox will use mozStorage, which is based on sqlite using shared-cache mode. I want other programs to be able read/write to the database but I was told this might be a