MUD-Dev
mailing list archive
[ Other Periods
| Other mailing lists
| Search
]
Date:
[ Previous
| Next
]
Thread:
[ Previous
| Next
]
Index:
[ Author
| Date
| Thread
]
Re: Unique id's
- To: "Multiple Recipients of MUD Design Mailing List" <mud-dev#null,net>
- Subject: Re: Unique id's
- From: coder#ibm,net
- Date: Sun, 13 Apr 97 09:16:01 -0700
On 12/04/97 at 12:55 PM, "Jon A. Lambert" <jlsysinc#ix,netcom.com> said:
>> From: coder#ibm,net
>> On 12/04/97 at 12:51 AM, "Jon A. Lambert" <jlsysinc#ix,netcom.com> said:
>> There's no problem if two ObjectID's contain the same timestamp. It
>> doesn't matter at all. What has to be unique is the combination of
>> record# and timestamp -- and that combo must be unique for the entire life
>> of the database. Happily this is not a problem until April 2038, longer
>> if you chose a different timestamp format.
>>
>Ahh your ID has two parts. I use something called a CriticalSection
>object to block the ' id++; ' function. Its very similar to a mutex,
>except the reset occurs automatically upon leaving the code section.
>Its supposed be slightly more efficient, but I believe its really a mutex
>under the covers. I use it alot, all over the place. :-)
Check into the base code for your CriticalSection. I've seen far to many
such blobs whose definition of a critical section is that it is the ONLY
thread in the application allowed to execute for the duration. <shudder>
(This how OS/2 and Windows handle their "critical section" API's, and I've
seen something similar in a library that shall remain nameless for SunOS).
>> I'm rather pleased actually -- the last time I had the server in a shape
>> it would actually run, I had it creating and destroying about 100,000
>> objects per hour while mobiles wandered about the game going in and out of
>> spoof/watcher areas, and breeding like crazy.
>>
>> Figure, that's what, about 40+ million objects being created and
>> destroyed... Certainly not enough to cause a 32bit wrap, but a good
>> enough test to demonstrate that the idea works.
>Whoa Nelly! I don't expect to reach anything approaching this rate of
>creation and destruction.
Neither do I for actual production databases. It was a contrived stress
test of the area. I tend to do a lot of those sorts of things.
Okay, we'll create two rooms linked by a single door.
Now lets create a simple mobile that just bounces back and
forth between those rooms thru that door.
Now lets clone that puppy a few thousand times.
Turn on the box and see what happens!
>Have you checked your machine for viruses lately? ;-)
Heck. I am the virus.
--
J C Lawrence Internet: claw#null,net
----------(*) Internet: coder#ibm,net
...Honourary Member of Clan McFud -- Teamer's Avenging Monolith...
- Thread context:
- Re: Unique id's, (continued)
- Re: Unique id's,
clawrenc clawrenc#cup,hp.com, Sat 12 Apr 1997, 06:03 GMT
- Re: Unique id's,
Jon A. Lambert jlsysinc#ix,netcom.com, Sat 12 Apr 1997, 11:49 GMT
- Re: Unique id's,
coder coder#ibm,net, Sat 12 Apr 1997, 13:55 GMT
- Re: Unique id's,
Jon A. Lambert jlsysinc#ix,netcom.com, Sat 12 Apr 1997, 23:53 GMT
- Re: Unique id's,
coder coder#ibm,net, Mon 14 Apr 1997, 00:01 GMT
- Re: Unique id's,
Jeff Kesselman jeffk#tenetwork,com, Mon 14 Apr 1997, 02:40 GMT
- Re: Unique id's,
Jon A. Lambert jlsysinc#ix,netcom.com, Mon 14 Apr 1997, 04:31 GMT
- Re: Unique id's,
clawrenc clawrenc#cup,hp.com, Mon 14 Apr 1997, 23:24 GMT
- self-intro: Cynbe,
Cynbe ru Taren cynbe#laurel,actlab.utexas.edu, Wed 09 Apr 1997, 10:54 GMT
[ Other Periods
| Other mailing lists
| Search
]