Need unique number for some objects ...

> Dennis,
>
> Use the #identityHash. If you need it unique between object spaces then bit-join the value with a unique image identifier integer and keep it as an inst var. The first x bits would represent one part and the last x bits would represent the last part. You'd override #identityHash to lazy-initialize the cached value. Once assigned, it would be unique and enduring.
>
> Paul Baumann
>
>
>
> -----Original Message-----
> From: [hidden email] [mailto:[hidden email]] On Behalf Of Dennis Smith
> Sent: Friday, February 04, 2011 08:44
> To: VWNC,
> Subject: [vwnc] Need unique number for some objects ...
>
> We normally run against gemstone and we use the gemstone "oop" as
> a unique number of objects.  This is fine.
>
> However, during development we don't use gemstone but work with just
> a memory-based system.
>
> We need a "unique number" for our high-level objects (Customer, Order,
> Person, ...).
>
> Anyone have thoughts on how to do this easily?  We current keep a weak array
> and use the index, but we have issues as to when to erase it and start
> over, so before
> I go "fix that" -- I thought I would ask if anyone else has had a
> similar need and how the
> solved it??
>
> --
> Dennis Smith                                     +1 416.798.7948
> Cherniak Software Development Corporation   Fax: +1 416.798.0948
> 509-2001 Sheppard Avenue East        [hidden email]
> Toronto, ON M2J 4Z8              sip:[hidden email]
> Canada                           http://www.CherniakSoftware.com
> Entrance off Yorkland Blvd south of Sheppard Ave east of the DVP
>
> _______________________________________________
> vwnc mailing list
> [hidden email]
> http://lists.cs.uiuc.edu/mailman/listinfo/vwnc
>
>
> This message may contain confidential information and is intended for specific recipients unless explicitly noted otherwise. If you have reason to believe you are not an intended recipient of this message, please delete it and notify the sender. This message may not represent the opinion of IntercontinentalExchange, Inc. (ICE), its subsidiaries or affiliates, and does not constitute a contract or guarantee. Unencrypted electronic mail is not secure and the recipient of this message is expected to provide safeguards from viruses and pursue alternate means of communication where privacy or a binding message is desired.
>

> We do use the identityHash for non-critical itentification (we modified
> Trippy to show the hash
> as part of every object so we can tell if two trippy windows are showing
> the same object or one that just looks the same).
>
> But I thought that identityHash was not unique??
>
> On 04/02/2011 9:17 AM, Paul Baumann wrote:
>> Dennis,
>>
>> Use the #identityHash. If you need it unique between object spaces then bit-join the value with a unique image identifier integer and keep it as an inst var. The first x bits would represent one part and the last x bits would represent the last part. You'd override #identityHash to lazy-initialize the cached value. Once assigned, it would be unique and enduring.
>>
>> Paul Baumann
>>
>>
>>
>> -----Original Message-----
>> From: [hidden email] [mailto:[hidden email]] On Behalf Of Dennis Smith
>> Sent: Friday, February 04, 2011 08:44
>> To: VWNC,
>> Subject: [vwnc] Need unique number for some objects ...
>>
>> We normally run against gemstone and we use the gemstone "oop" as
>> a unique number of objects.  This is fine.
>>
>> However, during development we don't use gemstone but work with just
>> a memory-based system.
>>
>> We need a "unique number" for our high-level objects (Customer, Order,
>> Person, ...).
>>
>> Anyone have thoughts on how to do this easily?  We current keep a weak array
>> and use the index, but we have issues as to when to erase it and start
>> over, so before
>> I go "fix that" -- I thought I would ask if anyone else has had a
>> similar need and how the
>> solved it??
>>
>> --
>> Dennis Smith                                     +1 416.798.7948
>> Cherniak Software Development Corporation   Fax: +1 416.798.0948
>> 509-2001 Sheppard Avenue East        [hidden email]
>> Toronto, ON M2J 4Z8              sip:[hidden email]
>> Canada                           http://www.CherniakSoftware.com
>> Entrance off Yorkland Blvd south of Sheppard Ave east of the DVP
>>
>> _______________________________________________
>> vwnc mailing list
>> [hidden email]
>> http://lists.cs.uiuc.edu/mailman/listinfo/vwnc
>>
>>
>> This message may contain confidential information and is intended for specific recipients unless explicitly noted otherwise. If you have reason to believe you are not an intended recipient of this message, please delete it and notify the sender. This message may not represent the opinion of IntercontinentalExchange, Inc. (ICE), its subsidiaries or affiliates, and does not constitute a contract or guarantee. Unencrypted electronic mail is not secure and the recipient of this message is expected to provide safeguards from viruses and pursue alternate means of communication where privacy or a binding message is desired.
>>

> Ok, so that makes it even more of a pain :-).
>
> Thanks
>
> David Buck
>
> Sent from my BlackBerry device on the Rogers Wireless Network
>
> -----Original Message-----
> From: Travis Griggs<[hidden email]>
> Sender: [hidden email]
> Date: Fri, 4 Feb 2011 06:58:33
> To: VWNC List<[hidden email]>
> Subject: Re: [vwnc] Need unique number for some objects ...
>
>
> On Feb 4, 2011, at 6:08 AM, David Buck wrote:
>
>> <snip>  It's possible to allocate the objects in Fixed Space and use their addresses, but that's a pain and won't work across image saves and re-launching.
>> <snip>
>
> And Fixed Space is only good for byte type objects anyway:
>
> Behavior>>newInFixedSpace:
> "Fail if the class is not bits-indexable...."
>
> --
> Travis Griggs
> Objologist
> Light travels faster than sound. This is why some people appear bright until you hear them speak...
>
>
>
>
>
> _______________________________________________
> vwnc mailing list
> [hidden email]
> http://lists.cs.uiuc.edu/mailman/listinfo/vwnc
>
> _______________________________________________
> vwnc mailing list
> [hidden email]
> http://lists.cs.uiuc.edu/mailman/listinfo/vwnc

> In fact I just proved that identityHash is not unique.
> I took all our business objects (2209 right now) and checked, there were
> 1136 unique identityHash values.
> Oh well!
>
> On 04/02/2011 9:49 AM, Dennis Smith wrote:
>> We do use the identityHash for non-critical itentification (we modified
>> Trippy to show the hash
>> as part of every object so we can tell if two trippy windows are showing
>> the same object or one that just looks the same).
>>
>> But I thought that identityHash was not unique??
>>
>> On 04/02/2011 9:17 AM, Paul Baumann wrote:
>>> Dennis,
>>>
>>> Use the #identityHash. If you need it unique between object spaces then bit-join the value with a unique image identifier integer and keep it as an inst var. The first x bits would represent one part and the last x bits would represent the last part. You'd override #identityHash to lazy-initialize the cached value. Once assigned, it would be unique and enduring.
>>>
>>> Paul Baumann
>>>
>>>
>>>
>>> -----Original Message-----
>>> From: [hidden email] [mailto:[hidden email]] On Behalf Of Dennis Smith
>>> Sent: Friday, February 04, 2011 08:44
>>> To: VWNC,
>>> Subject: [vwnc] Need unique number for some objects ...
>>>
>>> We normally run against gemstone and we use the gemstone "oop" as
>>> a unique number of objects.  This is fine.
>>>
>>> However, during development we don't use gemstone but work with just
>>> a memory-based system.
>>>
>>> We need a "unique number" for our high-level objects (Customer, Order,
>>> Person, ...).
>>>
>>> Anyone have thoughts on how to do this easily?  We current keep a weak array
>>> and use the index, but we have issues as to when to erase it and start
>>> over, so before
>>> I go "fix that" -- I thought I would ask if anyone else has had a
>>> similar need and how the
>>> solved it??
>>>
>>> --
>>> Dennis Smith                                     +1 416.798.7948
>>> Cherniak Software Development Corporation   Fax: +1 416.798.0948
>>> 509-2001 Sheppard Avenue East        [hidden email]
>>> Toronto, ON M2J 4Z8              sip:[hidden email]
>>> Canada                           http://www.CherniakSoftware.com
>>> Entrance off Yorkland Blvd south of Sheppard Ave east of the DVP
>>>
>>> _______________________________________________
>>> vwnc mailing list
>>> [hidden email]
>>> http://lists.cs.uiuc.edu/mailman/listinfo/vwnc
>>>
>>>
>>> This message may contain confidential information and is intended for specific recipients unless explicitly noted otherwise. If you have reason to believe you are not an intended recipient of this message, please delete it and notify the sender. This message may not represent the opinion of IntercontinentalExchange, Inc. (ICE), its subsidiaries or affiliates, and does not constitute a contract or guarantee. Unencrypted electronic mail is not secure and the recipient of this message is expected to provide safeguards from viruses and pursue alternate means of communication where privacy or a binding message is desired.
>>>

> Guess I'm back to my weakArray -- I will find and fix my bug :(
>
> On 04/02/2011 10:19 AM, David Buck wrote:
>> Ok, so that makes it even more of a pain :-).
>>
>> Thanks
>>
>> David Buck
>>
>> Sent from my BlackBerry device on the Rogers Wireless Network
>>
>> -----Original Message-----
>> From: Travis Griggs<[hidden email]>
>> Sender: [hidden email]
>> Date: Fri, 4 Feb 2011 06:58:33
>> To: VWNC List<[hidden email]>
>> Subject: Re: [vwnc] Need unique number for some objects ...
>>
>>
>> On Feb 4, 2011, at 6:08 AM, David Buck wrote:
>>
>>> <snip>  It's possible to allocate the objects in Fixed Space and  
>>> use their addresses, but that's a pain and won't work across image  
>>> saves and re-launching.
>>> <snip>
>>
>> And Fixed Space is only good for byte type objects anyway:
>>
>> Behavior>>newInFixedSpace:
>> "Fail if the class is not bits-indexable...."
>>
>> --
>> Travis Griggs
>> Objologist
>> Light travels faster than sound. This is why some people appear  
>> bright until you hear them speak...
>>
>>
>>
>>
>>
>> _______________________________________________
>> vwnc mailing list
>> [hidden email]
>> http://lists.cs.uiuc.edu/mailman/listinfo/vwnc
>>
>> _______________________________________________
>> vwnc mailing list
>> [hidden email]
>> http://lists.cs.uiuc.edu/mailman/listinfo/vwnc
>
> --
> Dennis Smith                         +1 416.798.7948
> Cherniak Software Development Corporation   Fax: +1 416.798.0948
> 509-2001 Sheppard Avenue East        [hidden email]
> Toronto, ON M2J 4Z8              sip:[hidden email]
> Canada         http://www.CherniakSoftware.com
> Entrance off Yorkland Blvd south of Sheppard Ave east of the DVP
>
> _______________________________________________
> vwnc mailing list
> [hidden email]
> http://lists.cs.uiuc.edu/mailman/listinfo/vwnc

> We normally run against gemstone and we use the gemstone "oop" as
> a unique number of objects.  This is fine.
>
> However, during development we don't use gemstone but work with just
> a memory-based system.
>
> We need a "unique number" for our high-level objects (Customer, Order,
> Person, ...).
>
> Anyone have thoughts on how to do this easily?  We current keep a weak array
> and use the index, but we have issues as to when to erase it and start
> over, so before
> I go "fix that" -- I thought I would ask if anyone else has had a
> similar need and how the
> solved it??
>

> We do use the identityHash for non-critical itentification (we modified
> Trippy to show the hash
> as part of every object so we can tell if two trippy windows are showing
> the same object or one that just looks the same).
>
> But I thought that identityHash was not unique??
>
> On 04/02/2011 9:17 AM, Paul Baumann wrote:
>> Dennis,
>>
>> Use the #identityHash. If you need it unique between object spaces then bit-join the value with a unique image identifier integer and keep it as an inst var. The first x bits would represent one part and the last x bits would represent the last part. You'd override #identityHash to lazy-initialize the cached value. Once assigned, it would be unique and enduring.
>>
>> Paul Baumann
>>
>>
>>
>> -----Original Message-----
>> From: [hidden email] [mailto:[hidden email]] On Behalf Of Dennis Smith
>> Sent: Friday, February 04, 2011 08:44
>> To: VWNC,
>> Subject: [vwnc] Need unique number for some objects ...
>>
>> We normally run against gemstone and we use the gemstone "oop" as
>> a unique number of objects.  This is fine.
>>
>> However, during development we don't use gemstone but work with just
>> a memory-based system.
>>
>> We need a "unique number" for our high-level objects (Customer, Order,
>> Person, ...).
>>
>> Anyone have thoughts on how to do this easily?  We current keep a weak array
>> and use the index, but we have issues as to when to erase it and start
>> over, so before
>> I go "fix that" -- I thought I would ask if anyone else has had a
>> similar need and how the
>> solved it??
>>
>> --
>> Dennis Smith                                     +1 416.798.7948
>> Cherniak Software Development Corporation   Fax: +1 416.798.0948
>> 509-2001 Sheppard Avenue East        [hidden email]
>> Toronto, ON M2J 4Z8              sip:[hidden email]
>> Canada                           http://www.CherniakSoftware.com
>> Entrance off Yorkland Blvd south of Sheppard Ave east of the DVP
>>
>> _______________________________________________
>> vwnc mailing list
>> [hidden email]
>> http://lists.cs.uiuc.edu/mailman/listinfo/vwnc
>>
>>
>> This message may contain confidential information and is intended for specific recipients unless explicitly noted otherwise. If you have reason to believe you are not an intended recipient of this message, please delete it and notify the sender. This message may not represent the opinion of IntercontinentalExchange, Inc. (ICE), its subsidiaries or affiliates, and does not constitute a contract or guarantee. Unencrypted electronic mail is not secure and the recipient of this message is expected to provide safeguards from viruses and pursue alternate means of communication where privacy or a binding message is desired.
>>

> We do use the identityHash for non-critical itentification (we modified
> Trippy to show the hash
> as part of every object so we can tell if two trippy windows are showing
> the same object or one that just looks the same).
>
> But I thought that identityHash was not unique??
>
> On 04/02/2011 9:17 AM, Paul Baumann wrote:
>> Dennis,
>>
>> Use the #identityHash. If you need it unique between object spaces then bit-join the value with a unique image identifier integer and keep it as an inst var. The first x bits would represent one part and the last x bits would represent the last part. You'd override #identityHash to lazy-initialize the cached value. Once assigned, it would be unique and enduring.
>>
>> Paul Baumann
>>
>>
>>
>> -----Original Message-----
>> From: [hidden email] [mailto:[hidden email]] On Behalf Of Dennis Smith
>> Sent: Friday, February 04, 2011 08:44
>> To: VWNC,
>> Subject: [vwnc] Need unique number for some objects ...
>>
>> We normally run against gemstone and we use the gemstone "oop" as
>> a unique number of objects.  This is fine.
>>
>> However, during development we don't use gemstone but work with just
>> a memory-based system.
>>
>> We need a "unique number" for our high-level objects (Customer, Order,
>> Person, ...).
>>
>> Anyone have thoughts on how to do this easily?  We current keep a weak array
>> and use the index, but we have issues as to when to erase it and start
>> over, so before
>> I go "fix that" -- I thought I would ask if anyone else has had a
>> similar need and how the
>> solved it??
>>
>> --
>> Dennis Smith                                     +1 416.798.7948
>> Cherniak Software Development Corporation   Fax: +1 416.798.0948
>> 509-2001 Sheppard Avenue East        [hidden email]
>> Toronto, ON M2J 4Z8              sip:[hidden email]
>> Canada                           http://www.CherniakSoftware.com
>> Entrance off Yorkland Blvd south of Sheppard Ave east of the DVP
>>
>> _______________________________________________
>> vwnc mailing list
>> [hidden email]
>> http://lists.cs.uiuc.edu/mailman/listinfo/vwnc
>>
>>
>> This message may contain confidential information and is intended for specific recipients unless explicitly noted otherwise. If you have reason to believe you are not an intended recipient of this message, please delete it and notify the sender. This message may not represent the opinion of IntercontinentalExchange, Inc. (ICE), its subsidiaries or affiliates, and does not constitute a contract or guarantee. Unencrypted electronic mail is not secure and the recipient of this message is expected to provide safeguards from viruses and pursue alternate means of communication where privacy or a binding message is desired.
>>

> Oh, I see what you are saying. Yes, there are only 16383 unique identityHash values and obviously more than 16K objects in your image. Sorry. Silly mistake. Use an application incremented value and combine that with an image-specific value.
>
> Paul Baumann
>
> -----Original Message-----
> From: Paul Baumann
> Sent: Friday, February 04, 2011 15:19
> To: 'Dennis Smith'; [hidden email]
> Subject: RE: [vwnc] Need unique number for some objects ...
>
> Dennis,
>
> The identityHash would be reused from a GC event like the number of scavenges. Include something like numScavenges too:
>
> identityHash
>          uniqueHash isNil ifTrue: [
>                  uniqueHash := (Processor activeProcess hash bitShift: 20)
>                  + ((ObjectMemory current numScavenges bitAnd: 2r111111) bitShift: 14)
>                  + super identityHash
>          ].
>          ^uniqueHash
>
> The hash of the active process is a simple (yet not perfect) alternative to an image identifier. One form of image identifier can come from an image start time. Another would be an image identifier that you assign. You decide how many bits for each hash component according to your needs.
>
> Keep in mind that operations like #become: will preserve the hash:
>
> | a aHash b bHash |
> aHash := (a := Array new: 1) identityHash.
> a at: 1 put: aHash.
> bHash := (b := Array new: 1) identityHash.
> b at: 1 put: bHash.
> (a identityHash = aHash and: [b identityHash = bHash]) ifFalse: [self unexpected].
> (a identityHash = a first and: [b identityHash = b first]) ifFalse: [self unexpected].
> a become: b.
> (a identityHash = aHash and: [b identityHash = bHash]) ifFalse: [self unexpected].
> (a identityHash ~= a first and: [b identityHash ~= b first]) ifFalse: [self unexpected].
>
> Your objects would also need to override #become: to exchange the hash values you've assigned.
>
> Paul Baumann
>
>
>
> -----Original Message-----
> From: [hidden email] [mailto:[hidden email]] On Behalf Of Dennis Smith
> Sent: Friday, February 04, 2011 10:07
> To: [hidden email]
> Subject: Re: [vwnc] Need unique number for some objects ...
>
> In fact I just proved that identityHash is not unique.
> I took all our business objects (2209 right now) and checked, there were
> 1136 unique identityHash values.
> Oh well!
>
> On 04/02/2011 9:49 AM, Dennis Smith wrote:
>> We do use the identityHash for non-critical itentification (we modified
>> Trippy to show the hash
>> as part of every object so we can tell if two trippy windows are showing
>> the same object or one that just looks the same).
>>
>> But I thought that identityHash was not unique??
>>
>> On 04/02/2011 9:17 AM, Paul Baumann wrote:
>>> Dennis,
>>>
>>> Use the #identityHash. If you need it unique between object spaces then bit-join the value with a unique image identifier integer and keep it as an inst var. The first x bits would represent one part and the last x bits would represent the last part. You'd override #identityHash to lazy-initialize the cached value. Once assigned, it would be unique and enduring.
>>>
>>> Paul Baumann
>>>
>>>
>>>
>>> -----Original Message-----
>>> From: [hidden email] [mailto:[hidden email]] On Behalf Of Dennis Smith
>>> Sent: Friday, February 04, 2011 08:44
>>> To: VWNC,
>>> Subject: [vwnc] Need unique number for some objects ...
>>>
>>> We normally run against gemstone and we use the gemstone "oop" as
>>> a unique number of objects.  This is fine.
>>>
>>> However, during development we don't use gemstone but work with just
>>> a memory-based system.
>>>
>>> We need a "unique number" for our high-level objects (Customer, Order,
>>> Person, ...).
>>>
>>> Anyone have thoughts on how to do this easily?  We current keep a weak array
>>> and use the index, but we have issues as to when to erase it and start
>>> over, so before
>>> I go "fix that" -- I thought I would ask if anyone else has had a
>>> similar need and how the
>>> solved it??
>>>
>>> --
>>> Dennis Smith                                     +1 416.798.7948
>>> Cherniak Software Development Corporation   Fax: +1 416.798.0948
>>> 509-2001 Sheppard Avenue East        [hidden email]
>>> Toronto, ON M2J 4Z8              sip:[hidden email]
>>> Canada                           http://www.CherniakSoftware.com
>>> Entrance off Yorkland Blvd south of Sheppard Ave east of the DVP
>>>
>>> _______________________________________________
>>> vwnc mailing list
>>> [hidden email]
>>> http://lists.cs.uiuc.edu/mailman/listinfo/vwnc
>>>
>>>
>>> This message may contain confidential information and is intended for specific recipients unless explicitly noted otherwise. If you have reason to believe you are not an intended recipient of this message, please delete it and notify the sender. This message may not represent the opinion of IntercontinentalExchange, Inc. (ICE), its subsidiaries or affiliates, and does not constitute a contract or guarantee. Unencrypted electronic mail is not secure and the recipient of this message is expected to provide safeguards from viruses and pursue alternate means of communication where privacy or a binding message is desired.
>>>
>
> --
> Dennis Smith                                     +1 416.798.7948
> Cherniak Software Development Corporation   Fax: +1 416.798.0948
> 509-2001 Sheppard Avenue East        [hidden email]
> Toronto, ON M2J 4Z8              sip:[hidden email]
> Canada                           http://www.CherniakSoftware.com
> Entrance off Yorkland Blvd south of Sheppard Ave east of the DVP
>
> _______________________________________________
> vwnc mailing list
> [hidden email]
> http://lists.cs.uiuc.edu/mailman/listinfo/vwnc
>
>
> This message may contain confidential information and is intended for specific recipients unless explicitly noted otherwise. If you have reason to believe you are not an intended recipient of this message, please delete it and notify the sender. This message may not represent the opinion of IntercontinentalExchange, Inc. (ICE), its subsidiaries or affiliates, and does not constitute a contract or guarantee. Unencrypted electronic mail is not secure and the recipient of this message is expected to provide safeguards from viruses and pursue alternate means of communication where privacy or a binding message is desired.
>
>
> _______________________________________________
> vwnc mailing list
> [hidden email]
> http://lists.cs.uiuc.edu/mailman/listinfo/vwnc
>

> Try something like :Time microsecondClock on the computer with current
> time.
>
> Dennis Smith wrote:
>> We normally run against gemstone and we use the gemstone "oop" as
>> a unique number of objects.  This is fine.
>>
>> However, during development we don't use gemstone but work with just
>> a memory-based system.
>>
>> We need a "unique number" for our high-level objects (Customer, Order,
>> Person, ...).
>>
>> Anyone have thoughts on how to do this easily?  We current keep a
>> weak array
>> and use the index, but we have issues as to when to erase it and start
>> over, so before
>> I go "fix that" -- I thought I would ask if anyone else has had a
>> similar need and how the
>> solved it??
>>
>

> We normally run against gemstone and we use the gemstone "oop" as
> a unique number of objects.  This is fine.
>
> However, during development we don't use gemstone but work with just
> a memory-based system.
>
> We need a "unique number" for our high-level objects (Customer, Order,
> Person, ...).
>
> Anyone have thoughts on how to do this easily?  We current keep a weak array
> and use the index, but we have issues as to when to erase it and start
> over, so before
> I go "fix that" -- I thought I would ask if anyone else has had a
> similar need and how the
> solved it??
>
>