Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.arch.embedded > #30606 > unrolled thread

(Semi-) formal methods

Back to article view | Back to comp.arch.embedded

Contents

  (Semi-) formal methods Don Y <blockedofcourse@foo.invalid> - 2021-05-11 18:25 -0700
    Re: (Semi-) formal methods Theo <theom+news@chiark.greenend.org.uk> - 2021-05-13 15:25 +0100
      Re: (Semi-) formal methods Don Y <blockedofcourse@foo.invalid> - 2021-05-13 11:57 -0700
        Re: (Semi-) formal methods Theo <theom+news@chiark.greenend.org.uk> - 2021-05-14 10:43 +0100
          Re: (Semi-) formal methods Don Y <blockedofcourse@foo.invalid> - 2021-05-14 12:37 -0700
            Re: (Semi-) formal methods HT-Lab <hans64@htminuslab.com> - 2021-05-15 08:44 +0100
              Re: (Semi-) formal methods Don Y <blockedofcourse@foo.invalid> - 2021-05-15 02:48 -0700
            Re: (Semi-) formal methods Theo <theom+news@chiark.greenend.org.uk> - 2021-05-15 12:02 +0100
              Re: (Semi-) formal methods Don Y <blockedofcourse@foo.invalid> - 2021-05-15 15:07 -0700
    Re: (Semi-) formal methods Dave Nadler <drn@nadler.com> - 2021-05-13 19:32 -0400
      Re: (Semi-) formal methods Don Y <blockedofcourse@foo.invalid> - 2021-05-13 16:57 -0700
      Re: (Semi-) formal methods Paul Rubin <no.email@nospam.invalid> - 2021-05-14 12:22 -0700
    Re: (Semi-) formal methods Don Y <blockedofcourse@foo.invalid> - 2021-05-14 21:55 -0700
    Re: (Semi-) formal methods Don Y <blockedofcourse@foo.invalid> - 2021-05-14 21:52 -0700
      Re: (Semi-) formal methods Clifford Heath <no.spam@please.net> - 2021-05-15 18:14 +1000
        Re: (Semi-) formal methods Don Y <blockedofcourse@foo.invalid> - 2021-05-15 02:10 -0700
          Re: (Semi-) formal methods Clifford Heath <no.spam@please.net> - 2021-05-15 20:54 +1000
            Re: (Semi-) formal methods Don Y <blockedofcourse@foo.invalid> - 2021-05-15 12:26 -0700
              Re: (Semi-) formal methods Clifford Heath <no.spam@please.net> - 2021-05-16 11:41 +1000
                Re: (Semi-) formal methods Don Y <blockedofcourse@foo.invalid> - 2021-05-16 00:35 -0700
                  Re: (Semi-) formal methods Clifford Heath <no.spam@please.net> - 2021-06-02 12:03 +1000
                    Re: (Semi-) formal methods Don Y <blockedofcourse@foo.invalid> - 2021-06-01 22:23 -0700
                      Re: (Semi-) formal methods Clifford Heath <no.spam@please.net> - 2021-06-02 22:39 +1000
                        Re: (Semi-) formal methods Don Y <blockedofcourse@foo.invalid> - 2021-06-02 09:01 -0700
                          Re: (Semi-) formal methods Clifford Heath <no.spam@please.net> - 2021-06-03 11:54 +1000
                            Re: (Semi-) formal methods Don Y <blockedofcourse@foo.invalid> - 2021-06-03 01:56 -0700
    Re: (Semi-) formal methods Clifford Heath <no.spam@please.net> - 2021-05-15 14:15 +1000

Page 1 of 2  [1] 2  Next page →

#30606 — (Semi-) formal methods

How prevalent are (semi-) formal design methods employed?
Which?

[I don't have first-hand knowledge of *anyone* using them]

[toc] | [next] | [standalone]

#30612

Don Y <blockedofcourse@foo.invalid> wrote:
> How prevalent are (semi-) formal design methods employed?
> Which?

We use theorem provers to find bugs in ISA specification:
https://www.cl.cam.ac.uk/~pes20/sail/

They're quite handy for finding bugs before they hit silicon...

Theo

[toc] | [prev] | [next] | [standalone]

#30615

On 5/13/2021 7:25 AM, Theo wrote:
> Don Y <blockedofcourse@foo.invalid> wrote:
>> How prevalent are (semi-) formal design methods employed?
>> Which?
> 
> We use theorem provers to find bugs in ISA specification:
> https://www.cl.cam.ac.uk/~pes20/sail/
> 
> They're quite handy for finding bugs before they hit silicon...

But, presumably, only of value if you're a SoC integrator?

I.e., given COTS devices, what might they reveal to users of
said devices?

[toc] | [prev] | [next] | [standalone]

#30621

Don Y <blockedofcourse@foo.invalid> wrote:
> On 5/13/2021 7:25 AM, Theo wrote:
> > Don Y <blockedofcourse@foo.invalid> wrote:
> >> How prevalent are (semi-) formal design methods employed?
> >> Which?
> > 
> > We use theorem provers to find bugs in ISA specification:
> > https://www.cl.cam.ac.uk/~pes20/sail/
> > 
> > They're quite handy for finding bugs before they hit silicon...
> 
> But, presumably, only of value if you're a SoC integrator?
> 
> I.e., given COTS devices, what might they reveal to users of
> said devices?

They can reveal bugs in existing implementations - where they don't meet the
spec and bad behaviour can result.

However CPU and FPGA design is what we do so that's where we focus our
efforts.  Depends whether FPGA counts as COTS or not...

Theo

[toc] | [prev] | [next] | [standalone]

#30625

On 5/14/2021 2:43 AM, Theo wrote:
> Don Y <blockedofcourse@foo.invalid> wrote:
>> On 5/13/2021 7:25 AM, Theo wrote:
>>> Don Y <blockedofcourse@foo.invalid> wrote:
>>>> How prevalent are (semi-) formal design methods employed?
>>>> Which?
>>>
>>> We use theorem provers to find bugs in ISA specification:
>>> https://www.cl.cam.ac.uk/~pes20/sail/
>>>
>>> They're quite handy for finding bugs before they hit silicon...
>>
>> But, presumably, only of value if you're a SoC integrator?
>>
>> I.e., given COTS devices, what might they reveal to users of
>> said devices?
> 
> They can reveal bugs in existing implementations - where they don't meet the
> spec and bad behaviour can result.
> 
> However CPU and FPGA design is what we do so that's where we focus our
> efforts.  Depends whether FPGA counts as COTS or not...

Understood.  Tools fit the application domains for which they were designed.

How did *adoption* of the tool come to pass?  Was it "mandated" by corporate
policy?  Something <someone> stumbled on, played with and then "pitched"
to management/peers?  Mandated by your industry?  etc.

[Just because a tool "makes sense" -- logically or economically -- doesn't
mean it will be adopted, much less *embraced*!]

[toc] | [prev] | [next] | [standalone]

#30629

On 14/05/2021 20:37, Don Y wrote:
> On 5/14/2021 2:43 AM, Theo wrote:
>> Don Y <blockedofcourse@foo.invalid> wrote:
>>> On 5/13/2021 7:25 AM, Theo wrote:
>>>> Don Y <blockedofcourse@foo.invalid> wrote:
>>>>> How prevalent are (semi-) formal design methods employed?
>>>>> Which?
>>>>
>>>> We use theorem provers to find bugs in ISA specification:
>>>> https://www.cl.cam.ac.uk/~pes20/sail/
>>>>
>>>> They're quite handy for finding bugs before they hit silicon...
>>>
>>> But, presumably, only of value if you're a SoC integrator?
>>>
>>> I.e., given COTS devices, what might they reveal to users of
>>> said devices?
>>
>> They can reveal bugs in existing implementations - where they don't 
>> meet the
>> spec and bad behaviour can result.
>>
>> However CPU and FPGA design is what we do so that's where we focus our
>> efforts.  Depends whether FPGA counts as COTS or not...
> 
> Understood.  Tools fit the application domains for which they were 
> designed.
> 
> How did *adoption* of the tool come to pass?  Was it "mandated" by 
> corporate
> policy?  Something <someone> stumbled on, played with and then "pitched"
> to management/peers?  Mandated by your industry?  etc.
> 

It became a must have tool about 2-3 decades ago for the safety 
critical/ avionics/medical industries. Design where becoming so complex 
that simulation could no longer answers questions like, is dead-lock or 
life-lock possible on our statemachine, can you buffers overflow, do you 
have arithmetic overflow, deadcode, race condition etc. The tools are 
now well established and most of the above questions can be answered 
(with some user constraints) by a simple push button tool. They are 
still expensive (you won't get much change from 20K UK pounds) but most 
high-end FPGA/ASIC companies use them. They are not a replacement for 
simulation but one of the tools you need to complete your verification.

Regards,
Hans
www.ht-lab.com


> [Just because a tool "makes sense" -- logically or economically -- doesn't
> mean it will be adopted, much less *embraced*!]

[toc] | [prev] | [next] | [standalone]

#30632

On 5/15/2021 12:44 AM, HT-Lab wrote:
> On 14/05/2021 20:37, Don Y wrote:
>> On 5/14/2021 2:43 AM, Theo wrote:
>>> Don Y <blockedofcourse@foo.invalid> wrote:
>>>> On 5/13/2021 7:25 AM, Theo wrote:
>>>>> Don Y <blockedofcourse@foo.invalid> wrote:
>>>>>> How prevalent are (semi-) formal design methods employed?
>>>>>> Which?
>>>>>
>>>>> We use theorem provers to find bugs in ISA specification:
>>>>> https://www.cl.cam.ac.uk/~pes20/sail/
>>>>>
>>>>> They're quite handy for finding bugs before they hit silicon...
>>>>
>>>> But, presumably, only of value if you're a SoC integrator?
>>>>
>>>> I.e., given COTS devices, what might they reveal to users of
>>>> said devices?
>>>
>>> They can reveal bugs in existing implementations - where they don't meet the
>>> spec and bad behaviour can result.
>>>
>>> However CPU and FPGA design is what we do so that's where we focus our
>>> efforts.  Depends whether FPGA counts as COTS or not...
>>
>> Understood.  Tools fit the application domains for which they were designed.
>>
>> How did *adoption* of the tool come to pass?  Was it "mandated" by corporate
>> policy?  Something <someone> stumbled on, played with and then "pitched"
>> to management/peers?  Mandated by your industry?  etc.
>>
> 
> It became a must have tool about 2-3 decades ago for the safety critical/ 
> avionics/medical industries.

But these are industries with inherently high levels of overhead -- possibly
suggesting (market) "efficiency" (in other industries) as a reason that
discourages adoption.

So, if it/they have value *there*, why aren't it/they embraced EVERYWHERE?
Obviously, other product offerings in other industries face similar design
problems...

> Design where becoming so complex that simulation 
> could no longer answers questions like, is dead-lock or life-lock possible on 
> our statemachine, can you buffers overflow, do you have arithmetic overflow, 
> deadcode, race condition etc. The tools are now well established and most of 
> the above questions can be answered (with some user constraints) by a simple 
> push button tool. They are still expensive (you won't get much change from 20K 
> UK pounds) but most high-end FPGA/ASIC companies use them. They are not a 
> replacement for simulation but one of the tools you need to complete your 
> verification.
> 
> Regards,
> Hans
> www.ht-lab.com
> 
> 
>> [Just because a tool "makes sense" -- logically or economically -- doesn't
>> mean it will be adopted, much less *embraced*!]
> 

[toc] | [prev] | [next] | [standalone]

#30634

Don Y <blockedofcourse@foo.invalid> wrote:
> Understood.  Tools fit the application domains for which they were designed.
> 
> How did *adoption* of the tool come to pass?  Was it "mandated" by corporate
> policy?  Something <someone> stumbled on, played with and then "pitched"
> to management/peers?  Mandated by your industry?  etc.
> 
> [Just because a tool "makes sense" -- logically or economically -- doesn't
> mean it will be adopted, much less *embraced*!]

We're a university research lab.  We developed the tool in response to two
trends:

- growing complexity of systems and the increasing prevalence of bugs in
implementation (for example in memory coherency subsystems).

- proposing security extensions to architectures and wanting to be able to
show that what we've proposed doesn't have any loopholes in it.  It provides
confidence to us and to people adopting the technology that the technology
is robust, and there won't be problems once silicon is deployed and
impossible to fix later.

(that's not to say there won't be entirely new classes of attacks coming out
of left-field in the way that Spectre surprised a lot of people, but are at
least trying to reason about the attacks we know about)

Because the costs of respinning are so high, chip design is all about
verification.  Doing it in a formal sense is a step up from hand-writing or
randomly-generating tests.  I don't think the industry needs convincing that
it's a good idea in abstract - it's mainly making the benefits outweigh the
costs.  A lot of the work (eg in the RISC-V community) is about bringing
down those costs.

Theo

[toc] | [prev] | [next] | [standalone]

#30636

On 5/15/2021 4:02 AM, Theo wrote:
> Don Y <blockedofcourse@foo.invalid> wrote:
>> Understood.  Tools fit the application domains for which they were designed.
>>
>> How did *adoption* of the tool come to pass?  Was it "mandated" by corporate
>> policy?  Something <someone> stumbled on, played with and then "pitched"
>> to management/peers?  Mandated by your industry?  etc.
>>
>> [Just because a tool "makes sense" -- logically or economically -- doesn't
>> mean it will be adopted, much less *embraced*!]
> 
> We're a university research lab.  We developed the tool in response to two
> trends:
> 
> - growing complexity of systems and the increasing prevalence of bugs in
> implementation (for example in memory coherency subsystems).
> 
> - proposing security extensions to architectures and wanting to be able to
> show that what we've proposed doesn't have any loopholes in it.  It provides
> confidence to us and to people adopting the technology that the technology
> is robust, and there won't be problems once silicon is deployed and
> impossible to fix later.

OK.  Both make sense.  And, both are hard (if not impossible) to "fix"
at layers *above* the hardware.

The next, most practical, question is:  how do you encourage its adoption?
Publishing papers is what professors/grad students "do".  That's
different from actually getting folks to *use* something that you've
developed/written about.

(a casual examination of the amount of "stuff" that has come out of
academia and "gone nowhere" -- despite some value! -- should make
this evident)

> (that's not to say there won't be entirely new classes of attacks coming out
> of left-field in the way that Spectre surprised a lot of people, but are at
> least trying to reason about the attacks we know about)
> 
> Because the costs of respinning are so high, chip design is all about
> verification.  Doing it in a formal sense is a step up from hand-writing or
> randomly-generating tests.  I don't think the industry needs convincing that
> it's a good idea in abstract - it's mainly making the benefits outweigh the
> costs.  A lot of the work (eg in the RISC-V community) is about bringing
> down those costs.

Understood.  In the past, I've had to rely on "good eyes" to spot problems
in designs.  Folks tend to only test how they *think* the design SHOULD
perform.  So, they often omit checking for things that they don't imagine
it (mis?)doing.

<do something; witness crash>
"What did you just do?"
"This..."
"You're not supposed to do that!"
"Then why did you/it LET ME?  Did I break some *law*???"

[Thankfully, I have a nack for identifying MY "assumptions" and the
liabilities that they bring to a design]

A suitably aggressive tool can avoid this bias and just hammer at every
nit it can algorithmically deduce/exploit.

A colleague designed a video scaler for a printer.  When I started poking at
it with real numbers, I discovered cases where the PHYSICAL image size
produced on a B-size page was actually SMALLER than that produced on an
A-size page (presumably, you'd print on larger paper to get a larger image!).

Ooops!

Given the number of variations in how the interface could be configured,
he'd never have been able to exhaustively test all cases.  So, my observation
was just serendipitous.

OTOH, *seeing* what I'd uncovered gave him cause to look more carefully
at his implementation *before* sending it off to fab (which would have cost
money AND calendar time -- leading to a delayed market introduction).

[toc] | [prev] | [next] | [standalone]

#30618

On 5/11/2021 9:25 PM, Don Y wrote:
> How prevalent are (semi-) formal design methods employed?

https://tinyurl.com/thbjw5j4

[toc] | [prev] | [next] | [standalone]

#30620

On 5/13/2021 4:32 PM, Dave Nadler wrote:
> On 5/11/2021 9:25 PM, Don Y wrote:
>> How prevalent are (semi-) formal design methods employed?
> 
> https://tinyurl.com/thbjw5j4

Hardly formal -- no cufflinks, watchfob, cumberbund nor tails!  :>

[toc] | [prev] | [next] | [standalone]

#30624

Dave Nadler <drn@nadler.com> writes:
>> How prevalent are (semi-) formal design methods employed?
> https://tinyurl.com/thbjw5j4

Wrong picture, try the one on this page:

   https://web.stanford.edu/~engler/

Look at the articles linked from there too ;-).

[toc] | [prev] | [next] | [standalone]

#30626

On 5/14/2021 9:52 PM, Don Y wrote:
> Or, said another way, what does a tool/process have to *do* in
> order to overcome this "resistance"?

By "do", I mean in the colloquial sense, not a specific feature
set, etc.

I.e., "It has to make my dinner and wash the dishes in order
for me to consider it worth embracing"  (or, "It has to cut
25% of the development cost from a project")

[toc] | [prev] | [next] | [standalone]

#30627

On 5/14/2021 9:15 PM, Clifford Heath wrote:
> On 12/5/21 11:25 am, Don Y wrote:
>> How prevalent are (semi-) formal design methods employed?
>> Which?
>>
>> [I don't have first-hand knowledge of *anyone* using them]
> 
> Don, IDK if you know about TLA+, but there is a growing community using it. It 
> is specifically good at finding errors in protocol (== API) designs (because 
> "TL" means "Temporal Logic"). I haven't used it so can't really answer many 
> questions, but I have been following the mailing list for some time and greatly 
> admire some of the excellent folk are are using it.

My query was more intended to see how *commonplace* such approaches are.
There are (and have been) many "great ideas" but, from my vantage point,
I don't see much by way of *adoption*.  (Note your own experience with TLA!)

[The "(Semi-)" was an accommodation for *individuals* who may be
using such things even though their work environment doesn't]

So, you either conclude that the methods are all "hype" (not likely),
*or*, there is some inherent resistance to their adoption.  Price?
(Process) overhead?  NIH?  Scale?  Education?  <shrug>

[Note my followup question to Theo as to how *he/they* ended up with
their tool/process]

There seem to be many "lost opportunities" (?) for tools, techniques,
processes, etc.  I'm just curious as to *why* (or, why *not*).
Or, said another way, what does a tool/process have to *do* in
order to overcome this "resistance"?

It's as if a (professional) writer wouldn't avail himself of a
spell-checker...  Or, a layout guy not running DRCs...  (yes,
I realize this to be an oversimplification; the examples I've
given are just mouse clicks!)

[toc] | [prev] | [next] | [standalone]

#30630

On 15/5/21 2:52 pm, Don Y wrote:
> On 5/14/2021 9:15 PM, Clifford Heath wrote:
>> On 12/5/21 11:25 am, Don Y wrote:
>>> How prevalent are (semi-) formal design methods employed?
>>> Which?
>>>
>>> [I don't have first-hand knowledge of *anyone* using them]
>>
>> Don, IDK if you know about TLA+, but there is a growing community 
>> using it. It is specifically good at finding errors in protocol (== 
>> API) designs (because "TL" means "Temporal Logic"). I haven't used it 
>> so can't really answer many questions, but I have been following the 
>> mailing list for some time and greatly admire some of the excellent 
>> folk are are using it.
> 
> My query was more intended to see how *commonplace* such approaches are.
> There are (and have been) many "great ideas" but, from my vantage point,
> I don't see much by way of *adoption*.  (Note your own experience with 
> TLA!)

My own experience is irrelevant, as I was semi-retired when I first came 
across it. On the other hand, the reason I came across it was I received 
a message from Chris Newcombe (admiring my related work), whose success 
in using it to find a potential failure in DynamoDB that could have 
knocked *Amazon* off the air was a stimulus to many *many* folk learning 
TLA+.


> So, you either conclude that the methods are all "hype" (not likely),
> *or*, there is some inherent resistance to their adoption.  Price?
> (Process) overhead?  NIH?  Scale?  Education?  <shrug>

For software folk at least, it requires a very different way of 
thinking. The same problem I had promulgating fact-based modeling: both 
address a massive *blind spot* in developer's consciousness.

Specifically we are unable to consciously detect when there is a failure 
in our logic; because to be conscious of the failure it would have to be 
*not present*. That is, we can only know such things in hindsight, or 
when we deliberately apply specific methods to check our logic. But why 
would we do that when it is "apparent" that our logic is correct?

> It's as if a (professional) writer wouldn't avail himself of a
> spell-checker...  Or, a layout guy not running DRCs...  (yes,
> I realize this to be an oversimplification; the examples I've
> given are just mouse clicks!)

These are not the same at all, because those things have rules. There is 
no rule for correct logic.

Clifford Heath

[toc] | [prev] | [next] | [standalone]

#30631

On 5/15/2021 1:14 AM, Clifford Heath wrote:

>> It's as if a (professional) writer wouldn't avail himself of a
>> spell-checker...  Or, a layout guy not running DRCs...  (yes,
>> I realize this to be an oversimplification; the examples I've
>> given are just mouse clicks!)
> 
> These are not the same at all, because those things have rules. There is no 
> rule for correct logic.

Logic is only "correct" if you are applying a prover.

You can still use formal methods in things like specifications -- where
there is no "proof" implied.  The advantage being that everyone can
unambiguously understand the intent of the specification without lots
of (verbose) "legalese".

E.g., I did my initial design with OCL as the means by which I conveyed
my intent to my colleagues.  It wasn't that much of a "lift" for them
to learn the representation well enough to ask pertinent questions and
"challenge" the implementation.  And, didn't resort to lots of mathematical
abstraction to make those points.

Unfortunately, use in such a document is not suited for "general audiences"
because it lacks rationale for each item in the specification.  (and,
relies on some semi-ubiquitous usage to ensure readers CAN read it)

OTOH, if you're writing the code (or reading it), such documents add
further texture to what you're seeing (in the programming language).
Another set of comments, so to speak.  Or, a roadmap.

The alternative is: an /ad hoc/ specification (with some likely incompletely
specified set of loose rules) *or* an *absent* specification.  Each of these
leave gaping holes in the design that (supposedly) follows.

Again, why the resistance to adopting such a "codified" approach?
There's no capital outlay required to adopt a *methodology* (unless
you want/need tools).  It's as if the effort is seen as an *additional*
effort -- but with no perception of a "return".  Is this because the
"return" doesn't stand out and have flashing lights surrounding it?

[toc] | [prev] | [next] | [standalone]

#30633

On 15/5/21 7:10 pm, Don Y wrote:
> On 5/15/2021 1:14 AM, Clifford Heath wrote:
> 
>>> It's as if a (professional) writer wouldn't avail himself of a
>>> spell-checker...  Or, a layout guy not running DRCs...  (yes,
>>> I realize this to be an oversimplification; the examples I've
>>> given are just mouse clicks!)
>>
>> These are not the same at all, because those things have rules. There 
>> is no rule for correct logic.
> 
> Logic is only "correct" if you are applying a prover.


I was loose with terminology. People tend to think that their 
"reasoning" is correct and doesn't need to be logically analysed or 
proved. They're wrong, but the blind spot is unavoidable.

Of course, there's also the problem that a thing may be "proved" yet be 
(undetectably, until it becomes detectable) not what would have been 
wanted - if it had been possible to foresee the failure mode.


> You can still use formal methods in things like specifications

That's how I used it, for static modelling (modelling all possible 
"states of the world" as they may exist at a point in time). Although 
dynamic modelling is much more exciting, it is rarely difficult once the 
correct static model has been agreed.

<http://dataconstellation.com/ActiveFacts/CQLIntroduction.html>

> Unfortunately, use in such a document is not suited for "general audiences"

The goal of CQL is to make the formal model suitable for (and expressed 
in the language of) anyone generally familiar with the domain being 
modelled.

> because it lacks rationale for each item in the specification.

Actually rationale is seldom needed. What is needed is an example of the 
scenario that is allowed or disallowed by each definition. The example 
is almost always an adequate rationalisation.
> The alternative is: an /ad hoc/ specification (with some likely 
> incompletely
> specified set of loose rules) *or* an *absent* specification.  Each of 
> these leave gaping holes in the design that (supposedly) follows.

That's precisely true. That's why analytic (formal) models are needed.

> Again, why the resistance to adopting such a "codified" approach?

Hubris, mostly. People genuinely don't see the need until enough 
experience has humbled them, and by then, their accumulated caution and 
tentativeness mean their industry sees them as dinosaurs.

Clifford Heath.

[toc] | [prev] | [next] | [standalone]

#30635

On 5/15/2021 3:54 AM, Clifford Heath wrote:
> On 15/5/21 7:10 pm, Don Y wrote:
>> Unfortunately, use in such a document is not suited for "general audiences"
> 
> The goal of CQL is to make the formal model suitable for (and expressed in the 
> language of) anyone generally familiar with the domain being modelled.

I opted for OCL as my system is object-based, UML is relatively easy to
understand and the way things would be expressed closely mimics the
way those constraints would be imposed in the code.  Any "translation"
effort would introduce other opportunities for errors to creep in.

>> because it lacks rationale for each item in the specification.
> 
> Actually rationale is seldom needed. What is needed is an example of the 
> scenario that is allowed or disallowed by each definition. The example is 
> almost always an adequate rationalisation.

I disagree.  I find many cases where I need to resort to prose to
explain why THIS implementation choice is better than THAT.  Just
*stating* (or implying) that it is leaves too much to the Reader,
as an "exercise" ("Now WHY, exactly, would this be a better approach?")

For example, I had to make a decision, early on, as to whether or
not "communications" would be synchronous or asynchronous.  There
are advantages and drawbacks to each approach.  From a performance
standpoint, one can argue that async wins.  But, from a cognitive
standpoint, sync is considerably easier for "users" (developers)
to "get right"; message and reply are temporally bound together
so the user doesn't have to, *later*, sort out which message a
subsequent reply correlates with (icky sentence structure but it's
early in the morning  :< )

The fact that comms then become blocking operations means any
desire for concurrency has to be addressed with other mechanisms
(e.g., set up a thread to do the comms so the "main" thread can
keep working and make the coordination between these two -- which
now appear asynchronous -- to be more visible.)

Or, trying to explain the many ways a particular comm can fail
(e.g., what if the party on the other end never listens?
Or, *can't* listen?  Or, the processor hosting it powers down?
Or...) and how the response/detection of that failure can
vary based on where in the "failure" it occurs.

>> The alternative is: an /ad hoc/ specification (with some likely incompletely
>> specified set of loose rules) *or* an *absent* specification.  Each of these 
>> leave gaping holes in the design that (supposedly) follows.
> 
> That's precisely true. That's why analytic (formal) models are needed.

I'm not sure the models need to be mechanically verifiable.  What *needs*
to happen is they need to be unambiguous and comprehensive.  You should
be able to look at one (before or after the code is written) and convince
yourself that it addresses every contingency -- as well as HOW it does so
(to the appearance of other actors)

>> Again, why the resistance to adopting such a "codified" approach?
> 
> Hubris, mostly. People genuinely don't see the need until enough experience has 
> humbled them, and by then, their accumulated caution and tentativeness mean 
> their industry sees them as dinosaurs.

"Hubris" suggests overconfidence (in their abilities).  I'm not sure
that's the case.

I started looking for a "common ground" in which I could express my current
design when I "discovered" that, not only is there no such thing, but that
there isn't even a common approach (methodology) to design.  It's not
a question of whether method A or method B is the more common (and, thus,
more readily recognizable to adopt in my presentation) but, that NONE OF
THE ABOVE is the clear winner!

[I couldn't even get a concensus on how to *diagram* relationships
between actors/objects just for illustrative purposes!]

This prompted the question, here.

I've been looking back over the past experiences I've had with getting
folks to "move forward" to try to see if I can identify the issue that
leads to this resistance.

Early on, I can argue that a lack of understanding for the (software)
design process led some employers to skimp in ways that they might
not have realized.  E.g., when a development seat cost many thousands of
1975 dollars, you could see how an employer, new to using MPUs *in*
their products, could decide to "share" a single seat among several
developers.

Similarly, when ICEs came on the scene.

And, HLLs.

But, those issues are largely behind us as there are many "free"/cheap
tools that can make these sorts of decisions "no brainers".

I can also see how scheduling pressures could lead to a resistance
to adopt new methods that *claim* to improve productivity.  I've
never been in an environment where time was allotted to explore
and learn new tools and techniques; there's another product that's
anxious to make its way to manufacturing (and YOU don't want to
be the one responsible for stalling the production line!)

I can see how fear/uncertainty can lead individuals (and organizations
as organizations are, at their heart, just individuals) to resist
change; the process you know (regardless of how bad it might be) is
safer than the process yet to BE known!

For things like specification and modeling, I can see a preception of
it as being a duplication of effort -- esp the more formally those
are expressed.

And, for some, I think an amazing lack of curiosity can explain their
clinging to The Way We Always Did It.  Or, the lack of "compensation"
for the "risk" they may be taking ("Heads, you win; tails, I lose!")

*Most* of these arguments can be rationalized by "employees" -- the
excuse that THEY (personally) don't have any control over THEIR
work environment.  OK, I'll avoid getting into *that* argument...

But, for folks working in more permissive environments (e.g.,
independent contractors), most of the arguments crumble.  YOU can
make the time to learn a technique/tool; YOU can buy the tool;
YOU can gauge its impact on YOUR productivity; etc.

I have a lot of respect for my colleagues -- they've all got
proven track records with significant projects.  Yet, they still
fall into this pattern of behavior -- clinging to past processes
instead of exploring new.  To their credit, they will make an
effort to understand the approaches and technologies that *I*
pursue -- but, that's more the result of a personal relationship
(than a BUSINESS one).  Yet, I never hear any epiphanies where
they exclaim "this is SO COOL"...  is the pull of the familiar SO
overwhelming?

[I left the 9-to-5 world primarily so I could dabble in other
fields, techniques, etc.  I have no desire to make "model 2",
especially if it will follow "process 1"!]

[toc] | [prev] | [next] | [standalone]

#30637

On 16/5/21 5:26 am, Don Y wrote:
> On 5/15/2021 3:54 AM, Clifford Heath wrote:
>> On 15/5/21 7:10 pm, Don Y wrote:
>>> Unfortunately, use in such a document is not suited for "general 
>>> audiences"
>>
>> The goal of CQL is to make the formal model suitable for (and 
>> expressed in the language of) anyone generally familiar with the 
>> domain being modelled.
> 
> I opted for OCL as my system is object-based, UML is relatively easy to
> understand and the way things would be expressed closely mimics the
> way those constraints would be imposed in the code.  Any "translation"
> effort would introduce other opportunities for errors to creep in.

It's very difficult to teach non-programmers to read OCL. But I agree on 
the need for code generation directly from the model, without 
translation. That's what CQL does too.

>> Actually rationale is seldom needed. What is needed is an example of 
>> the scenario that is allowed or disallowed by each definition. The 
>> example is almost always an adequate rationalisation.
> 
> I disagree.  I find many cases where I need to resort to prose to
> explain why THIS implementation choice is better than THAT.

Yes, where that is needed, CQL provides context-notes too. In fact I 
select four specific categories of rationale ("so that", "because", "as 
opposed to" and "to avoid"), with optional annotations saying who agreed 
and when.

> For example, I had to make a decision, early on, as to whether or
> not "communications" would be synchronous or asynchronous.  There
> are advantages and drawbacks to each approach.  From a performance
> standpoint, one can argue that async wins.  But, from a cognitive
> standpoint, sync is considerably easier for "users" (developers)
> to "get right"; message and reply are temporally bound together
> so the user doesn't have to, *later*, sort out which message a
> subsequent reply correlates with (icky sentence structure but it's
> early in the morning  :< )

Yes, I'm well familiar with that problem. Best decade of my working life 
was building a development tool that exclusively used message passing.
For the most part it's amazingly liberating, but sometimes frustrating.

>> That's precisely true. That's why analytic (formal) models are needed.
> 
> I'm not sure the models need to be mechanically verifiable.

Even without verification, the model must only make rules that are *in 
principle* machine-verifiable. If a rule is not verifiable, it's just 
waffle with no single logical meaning. If it has a single logical 
meaning, it is in principle machine-verifiable.

>  What *needs*
> to happen is they need to be unambiguous and comprehensive.

Comprehensive is impossible. There's always a possibility for more 
detail in any real-world system. But unambiguous? Yes, that requires 
that it is a formal part of a formal system which allows its meaning to 
be definitively stated. That is, for any scenario, there exists a 
decision procedure which can determine whether that scenario is allowed 
or is disallowed (this is what is meant when a logical system is termed 
"decidable").

>  You should
> be able to look at one (before or after the code is written) and convince
> yourself that it addresses every contingency

It's a great goal, but it is *in principle* impossible. Correctness of 
every contingency requires that the model matches the "real world", and 
Godel's theorem shows that it's impossible to prove that.


>>> Again, why the resistance to adopting such a "codified" approach?
>>
>> Hubris, mostly. People genuinely don't see the need until enough 
>> experience has humbled them, and by then, their accumulated caution 
>> and tentativeness mean their industry sees them as dinosaurs.
> 
> "Hubris" suggests overconfidence (in their abilities).  I'm not sure
> that's the case.
> 
> I started looking for a "common ground" in which I could express my current
> design

Fact-based modelling uses language as an access point to people's mental 
models, by analysing "plausible utterances" or "speech acts" for their 
logical intent, and building a formal model that captures the domain 
*using their own terms and phrases*.

Poor though it is, there exists no better tool than natural language to 
explore and express common ground. CQL provides a two-way bridge between 
that and formal logic, so that any mathematically formal statement can 
be unambiguously expressed using natural sentences, and every fact in 
the domain can be expressed using at least one agreed sentence using a 
restricted natural grammar that is also mathematically formal (meaning, 
unambiguously parseable to a logical expression).

> when I "discovered" that, not only is there no such thing,

Each model needs to be formulated by agreement in each case. The only 
way to reach agreement is to express scenarios and formalise ways of 
expressing them, so that any acceptable statement can be analysed for 
its logical intent.

This works because every functioning business already has ways to talk 
about everything that matters to it. Fact-based modeling captures those 
expressions and formalises them, using their own words to express the 
result, so agreement can be reached. It's little use to formalise a rule 
in a way that cannot be verified by the people who proposed it - one 
cannot reach agreement that way. Many MANY development failures fall 
into the trap of "but you said... no but I meant....!", or "I didn't say 
that because it's just common sense! What kind of fool are you?"

> I've been looking back over the past experiences I've had with getting
> folks to "move forward" to try to see if I can identify the issue that
> leads to this resistance.

"Every man's way is right in his own eyes" - Proverbs 21:2

People can't see the flaws in their own logic, because their logic is 
flawed. They resist methodical attempts to correct them, because they're 
already "right".

> I've
> never been in an environment where time was allotted to explore
> and learn new tools and techniques;

I know a number of companies that implement "10% time", for employees to 
explore any technology or personal projects they feel might be relevant 
to the business (or to their ability to contribute to it). I think 
Google is one of these, in fact, though my examples are closer to home.

> For things like specification and modeling, I can see a preception of
> it as being a duplication of effort -- esp the more formally those
> are expressed.

Unfortunately a lot of companies view testing in the same way. As if it 
wasn't possible for them to make a mistake.

> And, for some, I think an amazing lack of curiosity can explain their
> clinging to The Way We Always Did It.  Or, the lack of "compensation"
> for the "risk" they may be taking ("Heads, you win; tails, I lose!")

Folk who have been "lucky" a few times tend to become the "golden child" 
and get promoted. Once in senior positions they're much more likely to 
reject techniques which could discover that "the emperor has no clothes"

> I have a lot of respect for my colleagues -- they've all got
> proven track records with significant projects.  Yet, they still
> fall into this pattern of behavior -- clinging to past processes
> instead of exploring new.

I look forward to hearing of your experiences with TLA+, Alloy, or CQL.
I promise that it will be worth your effort.

Clifford Heath.

[toc] | [prev] | [next] | [standalone]

#30638

On 5/15/2021 6:41 PM, Clifford Heath wrote:
> On 16/5/21 5:26 am, Don Y wrote:
>> On 5/15/2021 3:54 AM, Clifford Heath wrote:
>>> On 15/5/21 7:10 pm, Don Y wrote:
>>>> Unfortunately, use in such a document is not suited for "general audiences"
>>>
>>> The goal of CQL is to make the formal model suitable for (and expressed in 
>>> the language of) anyone generally familiar with the domain being modelled.
>>
>> I opted for OCL as my system is object-based, UML is relatively easy to
>> understand and the way things would be expressed closely mimics the
>> way those constraints would be imposed in the code.  Any "translation"
>> effort would introduce other opportunities for errors to creep in.
> 
> It's very difficult to teach non-programmers to read OCL. But I agree on the 
> need for code generation directly from the model, without translation. That's 
> what CQL does too.

My target audience is programmers.

>>> Actually rationale is seldom needed. What is needed is an example of the 
>>> scenario that is allowed or disallowed by each definition. The example is 
>>> almost always an adequate rationalisation.
>>
>> I disagree.  I find many cases where I need to resort to prose to
>> explain why THIS implementation choice is better than THAT.
> 
> Yes, where that is needed, CQL provides context-notes too. In fact I select 
> four specific categories of rationale ("so that", "because", "as opposed to" 
> and "to avoid"), with optional annotations saying who agreed and when.

I present my descriptions in "multimedia prose" so I can illustrate
(and animate) to better convey the intent.

In some domains, "text" is just a terrible choice for explaining what
you're doing (e.g., speech synthesis, gesture recognition, etc.).
It's considerably easier to emit sounds or animate graphics (in these
mentioned examples) than to try to describe what you're addressing.

>> For example, I had to make a decision, early on, as to whether or
>> not "communications" would be synchronous or asynchronous.  There
>> are advantages and drawbacks to each approach.  From a performance
>> standpoint, one can argue that async wins.  But, from a cognitive
>> standpoint, sync is considerably easier for "users" (developers)
>> to "get right"; message and reply are temporally bound together
>> so the user doesn't have to, *later*, sort out which message a
>> subsequent reply correlates with (icky sentence structure but it's
>> early in the morning  :< )
> 
> Yes, I'm well familiar with that problem. Best decade of my working life was 
> building a development tool that exclusively used message passing.
> For the most part it's amazingly liberating, but sometimes frustrating.

 From the comments I've received from colleagues using my codebase,
the biggest problems seem to be related to (true) concurrency and
the possibility of a remote host failing while the message is in transit.
Folks seem to be used to "local" results in short order (transport delays
not being significant even in an IPC)

>>> That's precisely true. That's why analytic (formal) models are needed.
>>
>> I'm not sure the models need to be mechanically verifiable.
> 
> Even without verification, the model must only make rules that are *in 
> principle* machine-verifiable. If a rule is not verifiable, it's just waffle 
> with no single logical meaning. If it has a single logical meaning, it is in 
> principle machine-verifiable.

Yes.  But *getting* such a tool (or requiring its existence before adopting
a model strategy) can be prohibitive.

>>   What *needs*
>> to happen is they need to be unambiguous and comprehensive.
> 
> Comprehensive is impossible.

"Comprehensive" only has to apply to the world-view seen through the
interface.  If you're dealing with a particular actor/object/server,
all you have to do is be able to define EVERY possible outcome that
a client could experience using that interface.

If the sky turns yellow and every cat on the planet dies... <shrug>
(likely this isn't pertinent to any interface)

OTOH, if a process can die while servicing a request, or run out
of resources, or encounter invalid operands, or... then each of
those possibilities have to be enumerated.

Much of the additional cruft (in my world) comes from the fact
that the request is executing "elsewhere" (even if on the same
host) while YOUR code appears to be running correctly; there's
no guarantee that the request is executing AT ALL.

> There's always a possibility for more detail in 
> any real-world system. But unambiguous? Yes, that requires that it is a formal 
> part of a formal system which allows its meaning to be definitively stated. 
> That is, for any scenario, there exists a decision procedure which can 
> determine whether that scenario is allowed or is disallowed (this is what is 
> meant when a logical system is termed "decidable").
> 
>>   You should
>> be able to look at one (before or after the code is written) and convince
>> yourself that it addresses every contingency
> 
> It's a great goal, but it is *in principle* impossible. Correctness of every 
> contingency requires that the model matches the "real world", and Godel's 
> theorem shows that it's impossible to prove that.

Again, it only has to match the portion of the real world visible to
the interface in question.

A filesystem interface doesn't have to address the possibility of
the polar ice caps melting...

>>>> Again, why the resistance to adopting such a "codified" approach?
>>>
>>> Hubris, mostly. People genuinely don't see the need until enough experience 
>>> has humbled them, and by then, their accumulated caution and tentativeness 
>>> mean their industry sees them as dinosaurs.
>>
>> "Hubris" suggests overconfidence (in their abilities).  I'm not sure
>> that's the case.
>>
>> I started looking for a "common ground" in which I could express my current
>> design
>> when I "discovered" that, not only is there no such thing,

You've missed the point.  I was addressing the point that modeling, itself,
is relatively uncommon (at least in "these circles").  So, trying to find
some common subset that everyone (most practitioners) were using was
pointless.

If no one speaks Esperanto, then it is foolish to present your documentation
*in* Esperanto!

>> I've been looking back over the past experiences I've had with getting
>> folks to "move forward" to try to see if I can identify the issue that
>> leads to this resistance.
> 
> "Every man's way is right in his own eyes" - Proverbs 21:2
> 
> People can't see the flaws in their own logic, because their logic is flawed. 
> They resist methodical attempts to correct them, because they're already "right".

That's more pessimistic than I -- a cynic -- would claim.

When I approached my colleagues re: this topic, no one tried to
defend their current practices.  I think they all realize they
*could* be doing things "better".  This led to my contemplation of
why they *aren't* moving in those directions.

It's like the guy who KNOWS he should be backing up his computer...
yet doesn't.   Why????

>> I've
>> never been in an environment where time was allotted to explore
>> and learn new tools and techniques;
> 
> I know a number of companies that implement "10% time", for employees to 
> explore any technology or personal projects they feel might be relevant to the 
> business (or to their ability to contribute to it). I think Google is one of 
> these, in fact, though my examples are closer to home.

Yes, but 10% of 40 hours isn't a helluvalot of time.  I would imagine
other things chip away at that time.

I've seen companies budget for "reading technical journals/rags", etc.
But, unless you're really aggressive with your time management, it's
unlikely that a few hours a week is going to amount to much -- other
than priming you for your after-work pursuit of the problem (on YOUR time)

>> For things like specification and modeling, I can see a preception of
>> it as being a duplication of effort -- esp the more formally those
>> are expressed.
> 
> Unfortunately a lot of companies view testing in the same way. As if it wasn't 
> possible for them to make a mistake.

I think everyone (management) would LOVE to test.  If it didn't take any
TIME!  I don't think the same is true of developers!  Most seem to
consider testing "tedious" -- not very fulfilling.  Hence they only test
for what they KNOW the code/product should do!

[I can't count the number of times I've been present when a developer
witnessed a "can't happen" event.  HE SAW IT WITH HIS OWN EYES -- he's
not relying on a possibly distorted report from the field.  Yet,
unless it jumps up and yells at him -- "The Problem is Right Here!" -- he
will eventually just shrug and move on... as if it had never happened!]

And, tell management that you've baked in X months of testing
into your schedule and they'll cut X months from it -- to get
product out the door sooner.

They *may* pay you lip service:  "You can test while we're
building our first production lots" -- but, "something" will
come up to divert you from even that.

>> And, for some, I think an amazing lack of curiosity can explain their
>> clinging to The Way We Always Did It.  Or, the lack of "compensation"
>> for the "risk" they may be taking ("Heads, you win; tails, I lose!")
> 
> Folk who have been "lucky" a few times tend to become the "golden child" and 
> get promoted. Once in senior positions they're much more likely to reject 
> techniques which could discover that "the emperor has no clothes"

I don't know.

How many ways are there for you to get from your home to your place of
business?  How many have you EXPLORED??  What would it cost for you to try a
different one, each day?  Depart 5 minutes earlier -- just in case there
is some unexpected BIG delay inherent in a particular route.  Chances are,
each route will be roughly comparable in duration...

How many different development styles have you tried, over the years
("tried" meaning "used for an entire project")?

I like to explore different ways of doing things; I already KNOW how
THAT way worked out!  But, my experience suggests this is not a
widespread idea/practice!  Screw up (take too long, etc.) and
it's YOUR fault; discover a better way of doing things -- <yawn>

>> I have a lot of respect for my colleagues -- they've all got
>> proven track records with significant projects.  Yet, they still
>> fall into this pattern of behavior -- clinging to past processes
>> instead of exploring new.
> 
> I look forward to hearing of your experiences with TLA+, Alloy, or CQL.
> I promise that it will be worth your effort.

It's unlikely that I will try any of them!  This is my last "electronics"
project (50 years in a field seems like "enough"; time to pursue OTHER
interests!)  As I don't see any immediate/direct benefit to the folks who
are already using my codebase (nor any overwhelming use of the tools by
others), it would just be an intellectual exercise benefiting only myself
and would likely complicate *their* understanding of what I'm doing!

I'll, instead, put the effort to something that will have more tangible
results.  Maybe some better tools to track the migration of a procedure
through the system in real-time (instead of /ex post factum/)...?

[toc] | [prev] | [next] | [standalone]

Page 1 of 2  [1] 2  Next page →

Back to top | Article view | comp.arch.embedded

csiph-web