Wednesday, September 03, 2008

DataObjects.Net v4.0 is released!

Product page: http://www.x-tensive.com/Products/DO/

Documentation: DataObjects.Net Wiki and Reference. For now everything here is very crude and quite incomplete, but we'll intensively work on it during the nearest days.

For now there are no samples, just our tests. The sample WindowsForms application we have will be replaced by WPF sample shortly (it was developed few months ago, and actually it is a bit outdated now, so we simply don't want to show not the best practices, moreover, for the outdated technology base).

P.S. You'll obviously find the "release" (it really can't be called a real one) is made up in hurry - that's what we were doing during last 8 months of delay, and we really don't what to delay it any more. It is almost the framework we are going to bring, and even Xtensive.Storage project itself is almost the same as it must finally be.

18 comments:

  1. Congratulations on releasing DO4! I just looked at the licensing/comercial purchasing options and see that ShareIt are listing DO4 at US$4800. Is that correct? That's many time more than the price of DO3 and certainly more than I as a lone developer can justify spending. Can you confirm this pricing please?

    ReplyDelete
  2. Yes - we promised this: "We'll change DataObjects.Net prices shortly after publishing the preview - they'll be noticeably higher. But we're thinking about making Express Edition of v4.0 more featured." - so we did this in extreme fashion.

    Possibly later there will be LGPL \ GPL option (in case when only essential features are used), but for now that's it.

    On the other hand, such pricing policy allows you to start using full-featured product immediately, and pay for it before your own release.

    ReplyDelete
  3. Forgot to add:
    1) It is one-time payment. I.e. you'll get all the further updates for free for it.
    2) Shortly we will add an option allowing to pay this amount in multiple parts (e.g. monthly).
    3) We must spend ~ 1-2 months to show there is really something that worth these money ;) (for now all is really crude).

    ReplyDelete
  4. Something makes me believe that none will pay such enormous high price at least an year until the product will take its final shape. Didn't you think that for products of such kind it will be much better to keep prices as low as possible and to revenue from support instead?

    ReplyDelete
  5. Our vision:
    - Shortly DO will become a unique tool solving some problems others simply can't handle at all (syncronization chains, build-in RDBMS, distributed indexes). So we might be the only solution in a niche.
    - If it will prove it really replaces full featured RDBMS well, its current pricing model shows it's quite attractive to buy it.
    - We understand the price is higher in comparison to other available products. But it is much better and contemporary. We believe it brings much better effect from the point of saved money then it costs, if it is used.
    - We allow you to use it right now, and pay only when your product meets the market. We provide full source code. We show we'll care about it further, and will not leave its development. So it's safe to invest the money to it.
    - The specified amount must be paid just once. There are no more yearly update subscription.

    ReplyDelete
  6. I want to be honest with you because I think you're doing great software so I maybe telling you the things I never would if the money I buy software was money from my own pocket :) But I do care about future of my project which currently connected with future of X-Tensive.
    >The specified amount must be paid just once
    Yes, I understand that and this is exactly the point where people can say to themselvs "Hm... strange, never heard of such pricing model. What these guys want to revenue from to support their company existance? From sales? But they defenetely have no strong marketing department. Soon they can be run out of money due to low level of sales and what will happen then?". I would prefer rather to pay annual fee and be sure the company has constant cashflow to their development department just to be sure for your future, guys. Maybe you should think about two different pricing models? That's imho, would be much more promising and unusual.
    >If it will prove it really replaces full featured RDBMS well...
    And this is 2nd point I am concerned, sorry :)
    I do understand great potential of your product, but I believe, none even could think about using it as a replacment to well known RDBMSes such as Oracle and MS SQL. May be just for projects with low requirements to database part such as Skype, ICQ etc.
    I would suggest you to set-up a poll on pricing models in support forum. Here what other people say. I would be happy find myself mistaken.

    ReplyDelete
  7. Here what other people say.
    Sorry :) a typo - I meant "hear" :)

    ReplyDelete
  8. 3 issues

    1. Licensing. What sort of licensing options do you give to organizations which develop the applications internally, and not for commercial purposes? Is there special licenses for non-profit organisations?

    2. After downloading the Express Edition of DataObjects.NET, i had no clue of how i can piece up an application. I think the usage examples should come in quick.

    3. I'm thinking on how you would use DataObjects.NET with ASP.NET 2.0+ membership provider. How would you do that? What security methods have been included in DataObjects.NET.

    I have mission critical project that should start ASAP, and a snappy response will be helpful.

    ReplyDelete
  9. Sorry for delay - I took a short vacation after release.

    1. We didn't thought about this yet. I'm keeping in mind something like: "While the count of employees < 100, a special license for internal usage can be used; otherwise you must buy a commercial license, or use it under GPL".

    2. Yes, we understand this. For now there are only tests. The nearest update making new DO fully usable (including sample(s), etc.) will be published closer to the end this month. It's possible there will be some intermediate ones, but not on this week.

    3. For now nothing is done here. Later (but not ASAP) we'll certainly add support for this.

    "I would prefer rather to pay annual fee and be sure the company has constant cashflow to their development department just to be sure for your future, guys. Maybe you should think about two different pricing models? ... I would suggest you to set-up a poll on pricing models in support forum."

    Let's see - we'll consider this.

    ReplyDelete
  10. Btw, today we've made some performance tests of DO with Memory Provider (in-memory non-transactional index) in comparison to SQL Server 2005. DO beats SQL Server 2005 at least by 7 times on such tests as insertion, ~ 60-70% of time is spent by Xtensive.Storage itself, which isn't optimized almost at all yet (we never profiled it at all - you know, we were in a big hurry. Xtensive.Indexing (eats ~ 30%) and lower parts are optimized well.

    Conclusions:
    - It will be really fast even as OO IMDB engine easily comparable to the fastest competitors.
    - Quite likely our real transactional index implementation will beat SQL Server 2005, since we have about 20 times of "performance reserve" for it! (7/30%). Taking into account the architecture I described here, it should be more then enough.
    - I suspect we can increase the performance of Xtensive.Storage at least by 30-40%, since today's attempt shown there are really many performance-related lacks - in just 5 minutes I've made it 5-10% faster.

    So we're quite optimistic now ;)

    ReplyDelete
  11. Just published:
    http://wiki.dataobjects.net/index.php?title=Roadmap

    ReplyDelete
  12. I tend to think if you start listing at $4800 people will expect a lot from DO4 (software maturity, detailed docs, knowledgebase, impeccable support, and tons of samples and how-to's) and balk at purchasing a license until they see more.

    But if you start at $1500 then charge 50% for annual subscriptions I think you will amass more funds faster allowing you to hire more staff and thus able to mature DO4 much earlier (not to mention you will probably receive more profit per subscriber in 3~4 years versus $4800 one time payment)

    ReplyDelete
  13. Ok, tomorrow:
    - We'll change our licensing policy. Hopefully you'll like a new one :)
    - 30% discount for the first early adopters will be intact till the end of September.
    - Roadmap for the first two weeks of October will be announced.
    - Finally, you'll get current status report.

    And in the end of this week there will be an update of the current DataObjects.Net build (with all the features mentioned in status report).

    ReplyDelete
  14. Quick update:
    - Status report & roadmap update is here: http://wiki.dataobjects.net/index.php?title=Roadmap
    - Licensing policy isn't changed yet. Sorry, but only tomorrow (we need more time to describe the new one).

    ReplyDelete
  15. Done. Please see http://wiki.dataobjects.net/index.php?title=Commercial_licenses

    Licensing policy change will officially happen tomorrow.

    ReplyDelete
  16. Any idea on when we might see a simple example on how to use DO4? I'm sure I'm not alone in being keen to get started trying out DO4, but really need to see an example to get up and running. Thanks.

    ReplyDelete
  17. We'll publish an update today, but there is still no example. We've spent last week on performance related things, and will spend probably 3-4 more days on this. Current result is here.

    What we dislike in it is Bulk Fetch test result - we must spent more time to make our results at least 2-3 times better. Certainly current result is not bad, but on the other hand, we must have a batter one to compete with EF well.

    Concerning the plans: we'll spend this week on optimizations, and the next one - on completing the things we were unable to finish this month (sample and some polishing of what was already done). So the next "big" update is moving to 10...12 October, 2008.

    In "exchange" we'll keep the current discounts till this day.

    ReplyDelete
  18. We've just finished with current benchmarks. Please refer to this article for details and results.

    ReplyDelete