Enhance documents on project main page

[loveyoupeng]

Could you improvement the doc on the project main page, there are some typo and ambiguous descriptions, which is hard to understand. And could you provide some design descriptions to help people to evaluate if it is a fit.

[peter-lawrey]

I have asked Catherine to review.
Can you clarify which design points you are looking for?

Wire is a wrapper for Bytes.
It add a format e.g. JSON, YAML, Binary to Bytes which is an enhancement of ByteBuffer.
When you call a method on Wire, it calls the appropriate method on Bytes which does this for the format chosen.

[loveyoupeng]

Thanks for the attention.
Yes as you mentioned that it is a wrapper for bytes. Probably you could list some statement regarding to best use cases such as easy API to handle bytebuff. And the pros cons as well.
One more suggestions is could you also share the high level relationships between all the pieces of the projects under openhft umbrella.
Some of them are in different development cycles, if I cherry pick some of them. What's the strategy you suggest to maintain version upgrade

[peter-lawrey]

Ok. We can try to make this clearer

…

On 15 Feb 2017 11:14, "loveyoupeng" ***@***.***> wrote: Thanks for the attention. Yes as you mentioned that it is a wrapper for bytes. Probably you could list some statement regarding to best use cases such as easy API to handle bytebuff. And the pros cons as well. One more suggestions is could you also share the high level relationships between all the pieces of the projects under openhft umbrella. Some of them are in different development cycles, if I cherry pick some of them. What's the strategy you suggest to maintain version upgrade — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#42 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABBU8bR1Dbv4EYRWmYPFWW8Vty1I4Zfsks5rct4LgaJpZM4MBXMX> .

[peter-lawrey]

We have added a table of contents. We have details comparison with other serialisation libraries. We support ByteBuffer interop.
Chronicle Wire uses Chronicle Bytes for bytes manipulation and Chronicle Core for low level JVM access.

Wire handles

• adding and removing fields.
• changing the order of fields.
• changing the type of a field.
• aliases for types so you can change them.

[loveyoupeng]

Great thanks a lot Thanks a lot Yunpeng Li

…

On Feb 25, 2017 8:51 PM, "Peter Lawrey" ***@***.***> wrote: We have added a table of contents. We have details comparison with other serialisation libraries. We support ByteBuffer interop. Chronicle Wire uses Chronicle Bytes for bytes manipulation and Chronicle Core for low level JVM access. Wire handles - adding and removing fields. - changing the order of fields. - changing the type of a field. - aliases for types so you can change them. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#42 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAoScwlRnzsV1n_W_s3ghJCsa3rBOvg6ks5rgCPggaJpZM4MBXMX> .

[peter-lawrey]

Are there products you are comparing which had documentation you think we could emulate?

[loveyoupeng]

actually I'm comparing binary protocols such as avro protobuff, currently we highly relay on json/xml based http protocol between internal services, which is a waste of resources given the message is complex object tree.
in terms of the doc, could you adding some introduction section to the concepts/metaphor used in Chronicle-Wire, such as what is DocumentContext, WireIn ValueIn and their relationships. seems Chronicle contains a few terms and metaphor from finance industry, it is a little hard to understand vanilla means normal/default for people not from the industry :P

[peter-lawrey]

To start with, I suggest you start with the generic use Marshallable and MarshallableIn/Out interface. We have a lower level interface where you control exactly how serialization / deserialization works. A lower level interface where you can work at the byte/primitive level and lastly an interface where you can get a raw native memory address and do what you want. Eg exchange data direct with a GPU. You shouldn't need to understand the whole library to use it or we have done something wrong.

…

On 3 Apr 2017 23:40, "loveyoupeng" ***@***.***> wrote: actually I'm comparing binary protocols such as avro protobuff, currently we highly relay on json/xml based http protocol between internal services, which is a waste of resources given the message is complex object tree. in terms of the doc, could you adding some introduction section to the concepts/metaphor used in Chronicle-Wire, such as what is DocumentContext, WireIn ValueIn and their relationships. seems Chronicle contains a few terms and metaphor from finance industry, it is a little hard to understand vanilla means normal/default for people not from the industry :P — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#42 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABBU8SzmVnUwj7xRSF9d2BKfRv5L694Lks5rsPaqgaJpZM4MBXMX> .

[benwimpory]

I found these 2015 blog posts and the wire unit tests a good introduction.

http://www.rationaljava.com/2015/10/chronicle-wire-tutorial-next-level-in.html?m=1

[loveyoupeng]

Great, thanks for the advices Thanks a lot Yunpeng Li On Apr 4, 2017 4:21 PM, "benwimpory" <notifications@github.com> wrote: I found these 2015 blog posts and the wire unit tests a good introduction. http://www.rationaljava.com/2015/10/chronicle-wire- tutorial-next-level-in.html?m=1 — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#42 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAoSc6erB464p8o3UGCykZ2AEs3p7cbSks5rsf2OgaJpZM4MBXMX> .

[hft-team-city]

Released in Chronicle-Wire-2.20.101, BOM-2.20.134