Stream: Archive Mirror: Isabelle Users Mailing List

Topic: [isabelle] isabelle components


view this post on Zulip Email Gateway (Aug 19 2022 at 11:56):

From: Christian Sternagel <c.sternagel@gmail.com>
Dear Developers,

a tiny remark, in the help-message of the components tool:

$ isabelle components -?

Usage: isabelle components [OPTIONS] [COMPONENTS ...]

Options are:
-I init user settings
-R URL component repository (default
$ISABELLE_COMPONENT_REPOSITORY)
-a all missing components
-l list status

Resolve Isabelle components via download and installation.
COMPONENTS are identified via base name.

ISABELLE_COMPONENT_REPOSITORY="http://isabelle.in.tum.de/components"

"download all missing components" or something similar might be more
descriptive for the "-a" flag.

cheers

chris

view this post on Zulip Email Gateway (Aug 19 2022 at 11:56):

From: Tjark Weber <webertj@in.tum.de>
Another tiny remark about help messages (and I may have raised this
before): I think it is unfortunate that the various Isabelle tools
complain about -? with, e.g.,

.../isabelle/lib/Tools/components: illegal option -- ?

Best,
Tjark

view this post on Zulip Email Gateway (Aug 19 2022 at 11:56):

From: Christian Sternagel <c.sternagel@gmail.com>
You are not the only one ;), see for example (the latter part of) this
thread

https://lists.cam.ac.uk/mailman/htdig/cl-isabelle-users/2011-February/msg00082.html

cheers

chris

view this post on Zulip Email Gateway (Aug 19 2022 at 11:56):

From: Christian Sternagel <c.sternagel@gmail.com>
Since there are some blind alleys when staring from the message in the
above link, here is a more precise location:

https://lists.cam.ac.uk/mailman/htdig/cl-isabelle-users/2011-February/msg00090.html

view this post on Zulip Email Gateway (Aug 19 2022 at 11:56):

From: Makarius <makarius@sketis.net>
I've tuned this a bit in Isabelle/2220f0fb5581.

The changeset also reveals that there is some documentation in the system
manual, even though the tool is mostly for administrative purposes --
actual users of Isabelle get a bundle with all components included.

Makarius

view this post on Zulip Email Gateway (Aug 19 2022 at 11:57):

From: Makarius <makarius@sketis.net>
Can you say what is unfortunate here? By provoking a command line error
the tools print the "usage"; there is no explicit help option. This
should work uniformly for all Isabelle tools without special surprises.

When I worked out that scheme many years ago I was looking around how
Unix-oid tools usually do the help, with the main observation that there
are no standards for that, even until today.

So when I have an unknown Unix tool I feed variations of -? -h -help
--help to it to get some feedback, hoping that none of these options mean
"delete all files in the home directory".

Makarius

view this post on Zulip Email Gateway (Aug 19 2022 at 11:57):

From: Tjark Weber <webertj@in.tum.de>
This is the special surprise. I do expect an explicit help option that
prints the usage message and is not considered an "illegal option."

Best,
Tjark

view this post on Zulip Email Gateway (Aug 19 2022 at 11:57):

From: Makarius <makarius@sketis.net>
Digging up these old threads means I don't have to repeat everything
again. The initial confusion by these guys was caused by using odd
shells. Non-portable shells cause endless problems in the Unix-oid world,
and the "standard" /bin/sh is particularly non-portable. This is why
Isabelle has standardized on GNU bash a long time ago.

I've recently considered one more step in reducing the potential of
confusion: the Prover IDE could include some Isabelle system sub-plugin
for the jEdit Console. That would provide bash and only bash on all
platforms, including Windows. (De-facto there are more problems by users
on Mac OS X and Windows not knowing how to start a terminal, than by Linux
users with odd shells.)

Makarius

view this post on Zulip Email Gateway (Aug 19 2022 at 11:57):

From: Makarius <makarius@sketis.net>
OK, so that is just a canonical case of changing the user expectation.

Makarius


Last updated: Apr 19 2024 at 04:17 UTC