[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Brainstorming ideas for define-configuration
From: |
jbranso |
Subject: |
Re: Brainstorming ideas for define-configuration |
Date: |
Fri, 10 Mar 2023 20:15:52 +0000 |
March 9, 2023 3:25 PM, "Liliana Marie Prikler" <liliana.prikler@gmail.com>
wrote:
> Hi,
>
> Am Donnerstag, dem 09.03.2023 um 02:28 +0000 schrieb Bruno Victal:
>
> I smell bad code ahead.
>
>> We could provide procedures that validate each record type within
>> define-configuration itself instead of validating the value at
>> runtime (i.e. within the body of the service-type).
>>
>> --8<---------------cut here---------------start------------->8---
>> ;; the common case
>> (define-configuration foo-configuration
>> (name
>> string
>> "Lorem ipsum...")
>>
>> ;; ...
>>
>> (validator procname))
>>
>> ;; [bonus] Simpler configurations that only care for mutually-
>> exclusive fields
>> (define-configuration foo-configuration
>> (name
>> string
>> "Lorem ipsum...")
>>
>> (title
>> string
>> "Lorem ipsum..."
>> (conflicts 'name)))
>> --8<---------------cut here---------------end--------------->8---
>
> Instead of providing both a name field and a title field, you might
> provide a field that can either be a name or a title or allow an even
> more powerful value type as long as it makes sense.
While I would agree that a guix service writer should avoid mutually
exclusive fieldnames and instead prefer mutually exclusive records
(and 95% of that time that will work), but may we examine it from a
user's perspective? How does the service writer differentiate from
a string title or string name?
Suppose that you want to respond to a king's rudeness. You can
secretly insult him or obviously insult him:
===Mutually exclusive records===, which are better from a maintainer's
perspective, but perhaps cause the user to write more scheme:
"..your traitor brother. Maybe I’ll feed him to wolves after I’ve
caught him. Did I tell you, I intend to challenge him to single
combat?"
(insult-configuration
(response
(secret-insult-configuration
(secret-insult “I should like to see that, Your Grace.”))))
OR
"You can't insult me."
(insult-configuration
(response
(obvious-insult-configuration
(obvious-insult "We've had vicious kings and we've had idiot kings,
but I don't know if we've ever been cursed with a vicious idiot for
a king!"))))
===Mutually exclusive fieldnames===
"I am the KING!"
(insult-configuration
(secret-insult "Any man who must say, 'I am the king' is no
true king. I'll show you that after I've won your war.")))
OR
"You are Kingsguard!"
(insult-configuration
(obvious-insult "...F*ck the King."))))
These examples are pretty wonky I will admit, but I really like
an option of having mutually exclusive fieldnames. Having said all of this,
I will agree that that mutually exclusive fieldnames are a bit like "goto"
in C. You really should never use them, unless you absolutely have to.
Thanks,
Joshua
P.S. I thought about not sending this email, then realized that someone
might find it funny. Sorry if it wastes your time. :(
- Brainstorming ideas for define-configuration, Bruno Victal, 2023/03/08
- Re: Brainstorming ideas for define-configuration, Attila Lendvai, 2023/03/09
- Re: Brainstorming ideas for define-configuration, Joshua Branson, 2023/03/09
- Re: Brainstorming ideas for define-configuration, Liliana Marie Prikler, 2023/03/09
- Re: Brainstorming ideas for define-configuration, Maxim Cournoyer, 2023/03/10
- Re: Brainstorming ideas for define-configuration,
jbranso <=
- Re: Brainstorming ideas for define-configuration, Ludovic Courtès, 2023/03/15