Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [jnosql-dev] NoSQL Endgame - Werner KeilThodorisBais-PerconaLive ONLINE 2020 - NoSQL Tutorial

There is no double Standard, but the CORE modules are 16+all Tinkerpop supports (because JNoSQL includes Tinkerpop while Spring so far doesn’t) while those core modules under the Spring artifactID which are truly NoSQL are 6 at most. Less for other Systems like GORM.

 

The number of third Party extensions created by someone in the community can vary, there could be others for GORM or Micronaut, but that is out of scope here as simple as that.

 

So there are no false claims you see we apply the same criteria to each of them.

Instead of trying to change the criteria or discourage those participating in JNoSQL in their free time why don’t you try convince your colleagues to support Apache Tinkerpop, we were asked this question at least once in the session and I don’t think the answer "No" was incorrect. So why not address that in a future version, then it might change the numbers, but again those 16 that are supported by "org.eclipse.jnosql“ are already more than the 6 under the Spring artifactId or 3 supported by GORM.

 

If you try to deny that simple math you’d hurt your own reputation and seem like those weird guys who still insist that Trump won the US Presidential election, I do’t think anybody would want that, do you?

 

Werner

 

Von: Oliver Drotbohm
Gesendet: Mittwoch, 30. Dezember 2020 01:07
An: jnosql developer discussions
Betreff: Re: [jnosql-dev] NoSQL Endgame - Werner KeilThodorisBais-PerconaLive ONLINE 2020 - NoSQL Tutorial

 

Allow me a final note as you keep on misrepresenting this in this thread here: even if you leave LDAP out of the picture, we’re down to 16 supported modules. I don’t quite get why you insist to ignore the Spring Data Community modules. They‘re effectively at the level of support *all* JNoSQL ones live at *by your very own standards*: maintained by the community. 

 

Again, please refrain from imposing double standards and making false claims when publicly speaking about projects you’re not involved with. It will only hurt your personal reputation in the long term. As you’re affiliated with JNoSQL, it also doesn’t cast a positive light on the project itself as there’s enough interesting and positive to talk about in it in the first place and none of those moves I criticize here are needed to show that it’s a cool project.

--

Sent while on the run…



Am 30.12.2020 um 00:37 schrieb Werner Keil <werner.keil@xxxxxxx>:



Yes let’s leave it.

Feel free to think of LDAP as the missing one in "half a dozen" since the genuine NoSQL DBMS are just 5 ;-)

Sure you may also store all sorts of stuff in an LDAP system but ist main purpose is access control and there is already a Jakarta spec and API for that (Security) so if someone in the community wold like to write a driver for it, sure, they are free do do that and Maybe it even exists out there, but we don’t see it as a priority to access LDAP Systems they can already access via Soteria.

 

16>6>3 (the NoSQL Systems GORM supports out of the box according to http://gorm.grails.org/latest/)

 

Whether there are 14 or 20 community drivers for JNoSQL in addition to the , feel free to check it out yourself, we don’t have time for that Research right now, but https://tinkerpop.apache.org/providers.html may also give you a hint. Even if you exclude duplicates already supported directly in the list of 16 that adds about 10 more.

 

I don’t know what’s the reason Spring Data has not supported Tinkerpop yet, but maybe it’ll do some day, then the list of supported DBs would grow roughly comparable to JNoSQL.

 

Cheers,

Werner

 

Gesendet von Mail für Windows 10

 

Von: Oliver Drotbohm
Gesendet: Mittwoch, 30. Dezember 2020 00:21
An: jnosql developer discussions
Betreff: Re: [jnosql-dev] NoSQL Endgame - Werner Keil ThodorisBais-PerconaLive ONLINE 2020 - NoSQL Tutorial

 

> Am 30.12.2020 um 00:02 schrieb Werner Keil <werner.keil@xxxxxxx>:

>

> Again Forget About LDAP we list a framework like Tinkerpop, but LDAP is not really a DB.

> The others compared to e.g. JNoSQL seem acceptable, maybe the best room for improvement would be to list the community Based ones also in the top list.

 

Well, I didn't start this thread to discuss what you personally consider a DB or not. The structure of the Spring Data website is also not in the scope of this discussion. The point of the discussion is that you pull numbers out of thin air into a presentation. I think this email thread is just acknowledging my point.

 

> You’ll find 16 here https://github.com/eclipse/jnosql-communication-driver alone, the others are also community-based, so you may need to do a little digging or analyzing the dependencies on GitHub or MavenCentral if you are curious.

 

Again, it's not me who has to make sense of that. If you claim you support 31 database, you should be able to point at that support. Everything else is red herrings and misrepresentation.

 

> Unlike Spring Data or Micronaut there are no permanent full-time employees paid to either maintain the code or just advertise it.

> Yet this community driven effort created support for 3-4x more "main modules " (MongoDB, Redis, Casandra, Geode, GemFire) than Spring Data which is also reflected by " supported out of the box". Therefore I see no reason to Change that, despite doing this in our spare time we did everything correctly. If you want to present more "main modules" in the future, let us know, then we might rephrase something, but right now there is no need for it.

 

So what is it now? Do community modules count as you request for yourself (see above) or don't they as you state here for the Spring Data side of things? Please make up your mind. I don't need "to present" anything to anyone. There's no obligation for me to prove anything. I am just calling out misrepresentation of facts.

 

I am happy to leave it at this as – just as with the comparison in the slide deck – you're constantly moving the goal posts here.

 

All I ask you is to reconsider spending a bit more time investigating a project you're not working on next time you choose to include it in a "comparison". Anyone reasonably familiar with this space will be able to draw their own conclusions anyway.

 

Happy new year and a really heartfelt congrats on what the entire jNoSQL team achieved in 2020!

 

Cheers,

Ollie

 

 

> Von: Oliver Drotbohm

> Gesendet: Dienstag, 29. Dezember 2020 23:22

> An: jnosql developer discussions

> Betreff: Re: [jnosql-dev] NoSQL Endgame - Werner Keil Thodoris Bais-PerconaLive ONLINE 2020 - NoSQL Tutorial

>

> Actually, we're not making them too hard to find [0, 1]. They're listed in the main content part of the page and you're really the first person not finding the info. Please give note, I can screenshot the info for you if needed. We currently list:

>

> Aerospike, ArangoDB, Apache Cassandra, Cloud Datastore, Couchbase, DynamoDB, Elasticsearch, Gemfire, Apache Geode, Hazelcast, Jest, LDAP, MongoDB, Neo4j, Redis, Apache Solr, Vault.

>

> Makes 17 if I count correctly. Am I really wrong to suggest, that if you like to publish a comparison like that, it's your duty to do the due diligence in investigating this and not stop at "Whoa, yeah well, I don't really know better so I just put random stuff on some slides"? In fact, I am happy to consult in advance in case you plan to redo stuff like that.

>

> I am still happy to learn about the discrepancy between the number of logos on the jNoSQL website and the actually available drivers in the repository.

>

> I am not concerned about who "ranks first" in some comparison with completely made up comparison criterias. All I care about is that what is is represented correctly and I see room for improvement in this regard.

>

> Cheers,

> Ollie

>

> [0] https://spring.io/projects/spring-data#main-modules

> [1] https://spring.io/projects/spring-data#community-modules

>

> > Am 29.12.2020 um 20:45 schrieb Werner Keil <werner.keil@xxxxxxx>:

> >

> > Hi,

> >

> > You mean when you pop the list open on the left siide?

> > That list contains all sorts of different connectors, but even if you are as flexible for interpretation as to call Hadoop or Solr a "NoSQL" system, that boils the 17 item list down to around 10. JDBC, JPA, R2DBC or Envers are not NoSQL systems, sorry and I also would not call REST or JPA that way either.

> > So without the likes of Solr or Hadoop (sorry but those are like claiming JNoSQL support TinkerPop ;-) that leaves

> > MongoDB, Redis, Cassandra, Geode, GemFire, Couchbase, Elasticsearch (also highly questionable, but let’s leave it there) and Neo4J.

> >

> > That’s 8, I would say Half a Dozen is Closer to 8 than a Dozen. So unless you come up with a different list, I see no reason to Change the statement, it’s alLittle bizarre since all in all Spring Data currently ends up best, because it is older and more mature while Jakarta NoSQL/JNoSQL or Micronaut Data (that does not support NoSQL at this point, which is why it ranked worse and has even fewer than half a dozen NoSQL DB Systems right now) are boh still moving targets and may only be considered for production under very special circumstances.

> >

> > We also made clear in our talks, whether Otavio, Thodoris myself or a combination of up to 3 co-speakers, that some of the Solutions try to mix "Apples with Oranges" which is why the criteria also cannot be exactly the same for all, they are too different, e.g. some rely on JDBC/JPA only while others are more diverse.

> >

> > So unless that list gets over 10 true NoSQL DB Systems, there is no Need to change the comparison.

> >

> > I hope you understand that and please let us know, when it changes drastically and of course constructive Input on the NoSQL spec and API is also welcome where somebody has suggestions.

> >

> > Cheers,

> > Werner

> >

> > Von: Oliver Drotbohm

> > Gesendet: Dienstag, 29. Dezember 2020 16:50

> > An: jnosql developer discussions

> > Betreff: Re: [jnosql-dev] NoSQL Endgame - Werner Keil Thodoris Bais -PerconaLive ONLINE 2020 - NoSQL Tutorial

> >

> > Hi,

> >

> > > Am 29.12.2020 um 15:17 schrieb Werner Keil <werner.keil@xxxxxxx>:

> > >

> > > Oliver,

> > >

> > > Thanks for the Input. Can you please Point us to the exact list of supported Drivers or DB Systems?

> > > JNoSQL http://www.jnosql.org/ lists around 30 and the list is not even complete because there are a few vendors creating drivers independently, but if Spring Data has a comparable list we’re happy to include it or update the number to „a dozen“ or whatever seems appropriate.

> >

> > The link I had in the original message lists all available ones. The website you point to in turn links to  actual database products and lists 31. The driver implementations' repository [0] lists 16. Care to elaborate on the difference? I am not really inclined to play the numbers game here but there seems to be some serious distortion going on in the external communication.

> >

> > > Please leave this out of the nosql-dev, this is only for the Jakarta NoSQL spec and while Spring Data does not distinguish between API and implementation (that is what Emily and Ed Burns said in their comparison between Spring Boot and Eclipse MicroProfile ;-D) the comparison is based on compatible implementations like JNoSQL.

> >

> > Sorry if I am not aware of the subtleties of all of the different email lists, I just replied to the original email.

> >

> > Cheers,

> > Ollie

> >

> > [0] https://github.com/eclipse/jnosql-communication-driver

> >

> > > Von: Oliver Drotbohm

> > > Gesendet: Dienstag, 29. Dezember 2020 15:10

> > > An: jnosql developer discussions

> > > Cc: nosql developer discussions

> > > Betreff: Re: [jnosql-dev] NoSQL Endgame - Werner Keil Thodoris Bais -Percona Live ONLINE 2020 - NoSQL Tutorial

> > >

> > > Werner, Thodoris,

> > >

> > > would you mind clarifying the last bullet point on slide 53 claiming Spring Data would "only" support "half a dozen" NoSQL stores? In fact, the website [0] lists more than a dozen stores supported and – from the list of competitors – arguably has the broadest list of stores supported.

> > >

> > > In general, when creating such comparison overview slides, it's good practice to not switch the criterias to evaluate each contestant under on the go. Doing so makes slides 50 and following rather meaningless as you move from "huge number" as a plus without actually qualifying this to "only half a dozen" as a negative when it's even factually untrue.

> > >

> > > It's great to see those direct comparisons, but I'd prefer if we stick to the facts and not colorize our evaluation just because we're affiliated with one of the contestants.

> > >

> > > Cheers,

> > > Ollie

> > >

> > > [0] https://spring.io/projects/spring-data

> > >

> > > > Am 29.12.2020 um 13:58 schrieb Otavio Santana <otaviopolianasantana@xxxxxxxxx>:

> > > >

> > > >

> > > > Hello everyone, there is a virtue presentation to compare several Java framework solutions in the Java world by Werner Keil and Thodoris Bais.

> > > >

> > > > https://www.youtube.com/watch?v=5fuWXMHdfoU

> > > >

> > > > https://www.slideshare.net/ThodorisBais/nosql-endgame-devoxxua-conference-2020

> > > >

> > > >

> > > > --

> > > > Otávio Santana

> > > >

> > > >

> > > > twitter: http://twitter.com/otaviojava

> > > > site:     http://about.me/otaviojava

> > > >

> > > > _______________________________________________

> > > > jnosql-dev mailing list

> > > > jnosql-dev@xxxxxxxxxxx

> > > > To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/jnosql-dev

> > >

> > >

> > > _______________________________________________

> > > jnosql-dev mailing list

> > > jnosql-dev@xxxxxxxxxxx

> > > To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/jnosql-dev

> >

> >

> > _______________________________________________

> > jnosql-dev mailing list

> > jnosql-dev@xxxxxxxxxxx

> > To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/jnosql-dev

>

>

> _______________________________________________

> jnosql-dev mailing list

> jnosql-dev@xxxxxxxxxxx

> To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/jnosql-dev

 

 

_______________________________________________
jnosql-dev mailing list
jnosql-dev@xxxxxxxxxxx
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/jnosql-dev

 


Back to the top