Object Spaces Isn't Going To Ship

“ObjectSpaces will be merged with WinFS!

The URL that I was giving was slightly off: go to http://msdn.microsoft.com/data/ to read all about.

Although there were many guesses on what the news would be nobody really hit the spot.

My first impressions on this are twofold:

  1. The API will be tightly integrated into Longhorn and therefore become an integral part of the operating system. This makes it even more important in my view than the previous scenario of an “add-on” to ADO.NET.
  2. It's a delay for even longer than “shortly after Whidbey“. “ [1]

I sure hope this isn't the case. If ObjectSpaces is really going to be delayed till Longhorn, I am going to seriously lose a lot of respect for Microsoft. First, they delay the whole freak'n Visual Studio suite by a year, and now they are saying that even with all they extra time, they can't put ObjectSpaces in? Give me a break! Perhaps those guys at MS that you've been looking up to are all talk. Perhaps they don't code quite as well as they would have you think... I find it quite strange that one guy can put together virtually an entire working ObjectSpaces implementation with a couple months of spare time, and all the ADO.NET geniuses at MS combined can't ship the same thing after 4 years of working on this crap (remember, they demo'd it two PDCs ago and told you it was just around the corner).

[1] http://www.alexthissen.nl/Weblog/PermaLink.aspx?guid=b6e4d556-057c-498b-93c1-0377008440db


  • INteresting.

    Given the resosurce of Microsoft (1+ developer on this) and the timeframe they had (4+ years), I see only three possibilities:

    * They were not happy with the public announcements of acourtt case against them, which may have gone through. At least my lawyer standing on standby said so.

    * Some dude HIGH up has NO CLUE about data access, which would be in line with their ignorance for the whole topic over the last years. I mean, O/R mappers and not exacty something special. I wrote about ADO 2 being a right step in a wrong direction and asked for O/R mappers YEARS ago.

    * They have the most incompetent people on this that I have ever seen trying an O/R mapper, unable to finish a proiduct into a shipping state over many years.

    I start tending it is nr 2.

    Anyhow, this gives thid party tools like my EntityBroker a life again. Guys, if you want an O/R mapper, stop waiting for ObjectSpaces. It will not come.

  • The question is, can that one guy maintain that library interoperating with 30000 other classes over a period of say 25 years?

  • With all respect to Paul Wilson, his O/R mapper is not even close to what Objectspaces can do. It's not a comparison you should make. Paul's tool is somewhat limited but does things it says it does a-ok.

    However, your view is somewhat the same as mine though: in 2 years time, with a team of what, 4 people?, they couldn't produce a fully working O/R mapping layer for ADO.NET which is releasable today.

    I wrote mine in 14 months, alone. If I can put out one alone in that timeframe, why did their team of even higher trained and skilled people fail?

    There must have been something else behind the scenes which made the team to throw away code on a regular basis because of a focus switch.

  • I am still having the impression that Microsoft is not convinced about O/R mapping at all (yet).

    AFAI understood the O/R mapping part in WinFS is also just one of several APIs you can use to access the data so even that the currently plan to release OS with WinFS is not a commitment to O/R mapping.

    We are also happy that we didn't wait for Objectspaces after seeing it first on PDC 2001. We are doing a lot of individual enterprise application development for government and similar organizations and we would have had a serious disadvantage if we would still hand code our DAL. In fall 2003 I was shortly in doubt whether we should migrate our projects from our own product to ObjectSpaces when it comes out (which would have been sad after spending 6 man years in 2 years of development) but after seeing what they came up with on PDC2003 I was confident again.

    So I think that the reason for not coming out with OS is rather political than technical.

    For O/R mapping product vendors like us this is the best situation that could have come: MS was calling .NET developers attention to using O/R mapping and now they are not providing them with an own product.

    People will soon realize that it is more efficent to start developing large enterprise applications RIGHT NOW with any O/R mapping tool they feel that fits their need, no matter if OS is coming out now or not, than to continue doing "Bare Bones SQL the Way their Fathers did" :-). When MS finally will come out with an O/R mapping technology (or any other O/R mapping technology will gain a significant market share) it will be much easier for them migrate their projects and development methodology/blue print architectures as if they would still do handcoded DAL.

  • Thomas,

    * They were not happy with the public announcements of acourtt case against them, which may have gone through. At least my lawyer standing on standby said so.

    And on what were you going to sue them? Microsoft building an O/R mapper?

  • I actually agree with Frans, although its nice to see mentioned. :) I've never tried to claim my O/R Mapper was an ObjectSpaces clone, or that it was all encompassing. Instead, it simply uses a syntax as much like the preview of ObjectSpaces as possible, and it made other similar decisions. I try to market mine as the simple man's O/R Mapper, which suffices for many people, but I also more than acknowledge Thomas' EntityBroker and Fran's LLBLGen Pro as the "better" full-featured products, if you need that instead of simplicity.

  • @Alex Thissen

    ::And on what were you going to sue them? Microsoft

    ::building an O/R mapper?

    Anto competitive laws, like in pülace on every civilized country. You are not allowed to put up a product and give it out below production costs. Incompetitive behavior due to their bundling.


    Go to loaw school.Not more unrealistic than then EU commision court case about the Real player being bundled. And guess what - MS lost this one.

    You are simply not allowed to take unfair advantage, and there is enough logic in there for this to be taken very badly by the courts. THis is called uncaif competition.

  • On the contrary, MS is very committed to O/R problem domain. So much so, that they want to do it right. Because of the overlaps with WinFs, it makes perfect sense to roll this together. We have way too many data access API's already.....ODBC, OleDb, DAO, RDO, ADO, ADO.NET, SQLXML. We don't need another one to be introduced only to be somewhat superseded 2 years later. The Java world has seen this type of churn with O/R mapping and Microsoft hopes to avoid the unnecessary extra churn when it is so forseeable.

    MS could very easily ship ObjectSpaces in the whidbey timeframe - heck they showed it for .NET 1.0. It's not that they can't, but that they won't and it's the right decision IMHO.

  • Barry: If I want to write an application against SqlServer, why do I need an O/R mapping technology that's in the FILESYSTEM? (Unless the filesystem IS hte database, but that would kill Yukon sales, wouldn't it).

    MS is absolutely not committed to the O/R problem domain. If they were, they would have introduced an EJB-CMP equivalent long ago which for example lets you setup safe caches spanning multiple appdomains in a distributed environment, something which is possible in Java but absolutely not in .NET.

  • Thomas,

    You are delusional if you think your threat of a lawsuit had anything to do with Microsoft's decision. I think you've been reading too much of your own press if you think that you have that much stroke in the industry.

  • Yeah, Mark what you said...DITTO!

  • Gratitude is the sign of noble souls.


  • -----------------------------------------------------------
    "Hi there, I observed your blog via Yahoo while seeking for first assist for a heart assault and your publish appears quite interesting for me."

  • -----------------------------------------------------------
    Hi. I read a few of your respective other articles and needed to know if you would be interested in exchanging blogroll links?

  • I would like finding over a broken heart could be so easy as following a couple of steps.!!! but its not…

    Medieval Studies

  • An individual built many advantageous facts presently there. I did looking with regard to inside the challenge and also noticed loads of men and girls could agree within your web page.

Comments have been disabled for this content.