[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Starting at 4.2 to end of 4.2 (4.3 follows in separate post)
Greetings! Comments on the latest draft, starting at 4.2. I won't repeat my comments for the figures which we discussed at the last meeting. I will paste parts of 4.2 as quotes and my suggestions follow: > A versionable artifact is a representative copy, a specific versioned > copy, or a private working copy of an artifact version series. A versionable artifact is: 1) a representative copy 2) a specific versioned copy 3) a private working copy "of an artifact version series." ? Question: In the next sentence, a "representative copy" is said to be the only version of a version series. Mixing apples and oranges aren't we? If a copy isn't under version control, there is no version series. Yes? > When a versionable artifact is not under version control, a > representative copy of a versionable artifact is the only version of a > version series and represents the versionable artifact itself, i.e. > has the same objectId. Suggest: "When a versionable artifact is not under version control, it has no version series but has the objectId of the artifact." The next paragraph is why I saved this section for later. ;-) The first list item: > * a representative copy of a versionable artifact is itself a > versionable artifact, i.e. it has its own objectId which is > different from the objectId of any versioned copy or private > working copy of the versionable artifact. Appears to be saying that there can be versioning of a representative copy of something subject to versioning. OK, but we haven't dealt with versions of the original artifact yet. Seems premature to jump to versioning of a copy of what is subject to version control. Or did we mean something else? (I would re-word in any event, depending upon what was meant.) > * a representative copy of a versionable artifact is a copy of > the latest versioned copy or private working copy of a version > series, depending on the check out state of the version series > and the user loading the artifact. Out of order to say the least. I have no idea at this point what a "versioned copy" might be. Or a "private working copy." > A specific versioned copy of a versionable artifact is an explicit > ”deep” copy of the content and state of a versionable artifact, > preserving its content and state at a certain point in time. Each > versioned copy of a versionable artifact is itself a versionable > artifact, i.e. it has its own objectId. A versioned copy has a version > number, label, and check in comment. Hmmm, let's try: A *versioned copy* of a versionable artifact is a copy of the content and state of a versionable artifact that has: *unique objectId *version number *label *check in comment BTW, "state of a versionable artifact" seems vague to me. Can we list the "state" items explicitly? > A private working copy of a versionable artifact is a versionable > artifact created by an explicit checkout operation on a versionable > artifact under version control. The properties for a private working > copy SHOULD be identical to the properties of a versioned copy of a > versionable artifact on which a checkout operation was performed. > Certain properties such as objectId and creationDate SHALL be > different from a versioned copy. The content of a private working copy > MAY be identical to the content of a versioned copy. A private working > copy need not be visible to users who have permissions to view other > versions in a version series. Until it is checked in using an explicit > checkin operation, a private working copy MUST NOT be considered the > LatestMajorVersion in a version series. If we are going to say this (I think we should) it probably should be part of a restatement of a versioning model. Sticking it all in this paragraph is just confusing but I will try to sort it out. A private working copy is created by an explicit checkout operation on a versionable artifact. (saying properties are the same and then saying some properties are different isn't helpful) Upon checkout, the content of the private copy is identical to the versioned artifact subject to the checkout operation. Upon checkout, the objectId and creationDate properties of the private copy SHALL be different from the versioned artifact. Delete - first is obvious and unnecessary. Second is a question of privilege and not appropriate here. > The content of a private working copy MAY be identical to the content > of a versioned copy. A private working copy need not be visible to > users who have permissions to view other versions in a version series. **** Delete: Until it is checked it, it has no status at all with any known versioning system. Why prohibit the impossible? > Until it is checked in using an explicit checkin operation, a private > working copy MUST NOT be considered the LatestMajorVersion in a > version series. Comment: I don't know why we are saying the first sentence in this note. Second sentence is false. I could be an actor with no privileges to see earlier versions. > Note: The container of a versionable artifact can hold a > representative copy of an artifact version series. Starting from a > representative copy in a container, an actor can traverse a version > series to retrieve any versioned copy or private working copy. Is this part of the earlier note? > ICOM version control model is based on the CMIS version control model > specified in Section 2.1.9 of Content Management Interoperability > Services Version 1.0 [CMIS]. Comment: We need to review CMIS on this point. Ah, you are presuming the private working copy is still subject to the overall application. Is that true? That is how you get into what status the private working copy has in the system. Had to read CMIS to get that. OK, so what does: > icom_doc:Versionable Mean? I chased around the document in a circle try to find it. Yes, we say later in that section: > The VersionControlMetadata class MAY include additional property > definitions which are implementation-defined. But we need to define properties that are going to make ICOM based applications interchangeable. One of the reasons for standards. I will try to find the time to rip out all the line numbering, graphics, etc. and create a concordance based on what is left. That should make spotting used by not defined terms easier. Maybe I can have a version of that for everyone before or by the next ICOM meeting. BTW, icom_doc:Versionable is all over icom_doc:versionHistory, icom_doc:versionableHistory, (btw, bad naming practice to make names so similar), icom_doc:latestVersionedCopy, icom_doc:privateWorkingCopy. icom_doc:totalSize - "The total size of all versioned copies of a versionable artifact in a version series." That seems like an odd property to standardize. Rationale? 4.2.4.1 Description > A version object is a version control metadata of a versioned copy or > a private working copy of a versionable artifact. This one caught my eye, need to check the others: A version object *contains* version control metadata of a versioned copy or a private working copy of a versionable artifact. The use of "is a" is incorrect. 4.2.4.3 Property Definitions > The Version class MUST have the property definitions; definitions; -> definitions: 4.2.5 VersionType > There are three version types defined by ICOM: > > * icom_doc:RepresentativeCopy: to express that a versionable > artifact is a representative copy of an artifact version > series. > > * icom_doc:VersionedCopy: to express that a versionable artifact > is a versioned copy of an artifact version series. > > * icom_doc:PrivateWorkingCopy: to express that a versionable > artifact is a private working copy of an artifact version > series. Are these value types? BTW, are they all for version series? (which isn't clear in any event) 4.2.6.1 Description Not sure this is helpful or distinguishes documents from other artifacts. Not sure what to suggest as a replacement but surely we can find something. Let me end here and start a separate post on 4.3 Message Module. Hope everyone is having a great week! Patrick -- Patrick Durusau patrick@durusau.net Chair, V1 - US TAG to JTC 1/SC 34 Convener, JTC 1/SC 34/WG 3 (Topic Maps) Editor, OpenDocument Format TC (OASIS), Project Editor ISO/IEC 26300 Co-Editor, ISO/IEC 13250-1, 13250-5 (Topic Maps) Another Word For It (blog): http://tm.durusau.net Homepage: http://www.durusau.net Twitter: patrickDurusau Newcomb Number: 1
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]