What about having just python-stix/cybox serve as the reference implementation? I guess I just don’t see why we must have an equivalent implementation for every
mainstream language, especially if we make implementation easier by eliminating some of the existing ambiguities in complexities in the language.
On that note, while all of the talk around JSON is great (and I'm personally for it), it really needs to happen in tandem with reduction in complexity/ambiguity, as otherwise we’re just pushing
around the same flawed data structures but in a different serialization. Accordingly, if can come to an agreement that this (JSON/other serializations + simplification) is one of the CTI’s priorities for the next release of STIX and CybOX, it would likely
go a long way towards alleviating some of the broader community’s concerns around our efforts.
-Ivan
From: <> on behalf of "Bush, Jonathan"
Date: Monday, August 31, 2015 at 9:55 AM
To: 'Jason Keirstead'
Cc: Bret Jordan, Aharon Chernin, Mark Clancy, ""
Subject: RE: [cti] Thoughts on STIX and some of the other threads on this list
Very fair point Jason – Do we have anyone like Mitre contracted who can maintain a set of libraries? That could be a heavy lift.
If we have that, I would suggest we work with them to build out the full functionality we need, not just skeleton libraries like we have now.
If we don’t have that, I would go back to the thought that our scope should be limited to a conceptual model only. We have to make a choice here, and it has big
implications.
From:
[mailto:]
On Behalf Of Jason Keirstead
Sent: Monday, August 31, 2015 9:22 AM
To: Bush, Jonathan
Cc: 'Jordan, Bret'; Aharon Chernin; Mark Clancy;
Subject: RE: [cti] Thoughts on STIX and some of the other threads on this list
/ If we abstract out the complexity what we have to ‘learn’ is a set of API calls. This is how modern software is built – Not on data formats but on API formats. /
This sounds good in principle, but in order for this to work in practice the OASIS CTI would have to be responsible not just for the STIX standard, but also reference bindings and documentation for STIX in several mainstream languages, I would say Python, Java,
and C++ at a minimum. This would be a very large body of work to undertake and maintain... even the current reference Python bindings by MITRE are pretty bare-bones (they don't "make anything simple", it's really just a data binding - not really what is required
for a widely used reference library) and I don't think the Java ones were ever completed. If you don't have an easy to use library set for everyone to use, then the format of the data is very important.
I will give an example to the list from my own experience. I had to add some STIX support to a system in Python that was running Python 2.6, which I do not have any control over, and did not ship with a C++ compiler. As a result, the MITRE reference libraries
have a dependancy chain that ends up with something needing C++ linking to libraries to build - so I could not use them at all. I ended up having to write my own STIX parser in Python from the XML... which was quite eye-opening as to how convoluted STIX can
be to work with, and I had all of Python to help me. I can't even imagine the job of someone writing a STIX XML parser in C++ based on the 1.1 specification alone.
/ I honestly hate them all, even the XML format. /
Well I know few data formats I have any particular "love" for :) My main beef with the XML format has nothing to do with how it looks, it has to do with markup verboseness and efficiency on the wire.
-
Jason Keirstead
Product Architect, Security Intelligence, IBM Security Systems
www.ibm.com/security |
www.securityintelligence.com
Without data, all you are is just another person with an opinion - Unknown
"Bush,
Jonathan" ---2015/08/28 09:06:30 PM---Bret - I think my point still remains - Why should I have to learn ANY specific implementation forma
From: "Bush, Jonathan" <>
To: "'Jordan, Bret'" <>, Aharon Chernin <>
Cc: Mark Clancy <>, "" <>
Date: 2015/08/28 09:06 PM
Subject: RE: [cti] Thoughts on STIX and some of the other threads on this list
Sent by: <>
Bret – I think my point still remains – Why should I have to learn ANY specific implementation format? I honestly hate them all, even the XML format.
If we abstract out the complexity what we have to ‘learn’ is a set of API calls. This is how modern software is built – Not on data formats but on API formats.
From: [mailto:]
On Behalf Of Jordan, Bret
Sent: Friday, August 28, 2015 7:38 PM
To: Aharon Chernin
Cc: Mark Clancy;
Subject: Re: [cti] Thoughts on STIX and some of the other threads on this list
Consumers use tools, hopefully they never see the format. Vendors, web developers, app developers, and open source developers write the tools. They are the ones that have to pay the XML tax.
Given the progress that Facebook is making I can begin to see a need for vendors even Soltra Edge to start supporting their threat exchange format.
My question still stands.. Will anyone not use STIX if we stopped doing XML? Follow on, how many more vendors and developers will we gain if we adopted JSON?
Let's just use Intelworks' JSON STIX format and be done with it.
Bret
Sent from my Commodore 64
<trimmed>
DTCC DISCLAIMER: This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error, please notify us immediately and delete the email
and any attachments from your system. The recipient should check this email and any attachments for the presence of viruses. The company accepts no liability for any damage caused by any virus transmitted by this email.
