| Newsgroups |
gmane.linux.lfs.beyond.devel |
| Message-ID |
<[email protected]> |
Am 24.05.25 um 14:13 schrieb Bruce Dubbs ([email protected] via
blfs-dev Mailing List):
> On 5/24/25 3:52 AM, Rainer Fiebig ([email protected] via blfs-dev Mailing
> List) wrote:
>> Am 24.05.25 um 07:59 schrieb Bruce Dubbs ([email protected] via
>> blfs-dev Mailing List):
>>> On 5/24/25 12:04 AM, Xi Ruoyao ([email protected] via blfs-dev Mailing
>>> List) wrote:
>>>> On Sat, 2025-05-24 at 12:15 +0800, Xi Ruoyao wrote:
>>>>> On Fri, 2025-05-23 at 15:21 -0400, Joe Locash wrote:
>>>>>> On 5/14/25 12:21 AM, Xi Ruoyao ([email protected] via blfs-dev
>>>>>> Mailing
>>>>>> List) wrote:
>>>>>>> On Wed, 2025-05-14 at 12:07 +0800, Xi Ruoyao wrote:
>>>>>>>> On Tue, 2025-05-13 at 18:13 -0500, Bruce Dubbs wrote:
>>>>>>>>> On 5/13/25 5:59 PM, Joe Locash ([email protected] via blfs-dev
>>>>>>>>> Mailing List) wrote:
>>>>>>>>>> On 5/13/25 6:53 PM, Bruce Dubbs ([email protected] via
>>>>>>>>>> blfs-dev Mailing List) wrote:
>>>>>>>>>>> On 5/13/25 5:34 PM, Joe Locash ([email protected] via blfs-dev
>>>>>>>>>>> Mailing List) wrote:
>>>>>>>>>>>> On 5/13/25 5:51 PM, Bruce Dubbs ([email protected] via
>>>>>>>>>>>> blfs-dev Mailing List)
>>>>>>>>>>>> wrote:
>>>>>>>>>>>>> I looked at my build log and didn't see where it is using an
>>>>>>>>>>>>> internet connection.
>>>>>>>>>>>>> Can you show where that is happening?
>>>>>>>>>>>>
>>>>>>>>>>>> /usr/bin/python3 -m build zenmap/
>>>>>>>>>>>> * Creating isolated environment: venv+pip...
>>>>>>>>>>>> * Installing packages in isolated environment:
>>>>>>>>>>>> - setuptools
>>>>>>>>>>>> > /usr/bin/python3 -m pip --python
>>>>>>>>>>>> /tmp/build-env-fr8do3me/bin/python install
>>>>>>>>>>>> --use-pep517 --no-warn-script-location --no-compile -r
>>>>>>>>>>>> /tmp/build-reqs-
>>>>>>>>>>>> uok0756y.txt
>>>>>>>>>>>> < WARNING: Retrying (Retry(total=4, connect=None, read=None,
>>>>>>>>>>>> redirect=None,
>>>>>>>>>>>> status=None)) after connection broken by
>>>>>>>>>>>> 'NewConnectionError('<pip._vendor.urllib3.connection.HTTPSConnection
>>>>>>>>>>>>
>>>>>>>>>>>> object
>>>>>>>>>>>> at 0x7f83a5b830e0>: Failed to establish a new connection:
>>>>>>>>>>>> [Errno -3]
>>>>>>>>>>>> Temporary failure in name resolution')':
>>>>>>>>>>>> /simple/setuptools/
>>>>>>>>>>>>
>>>>>>>>>>>> ....
>>>>>>>>>>>>
>>>>>>>>>>>> < ERROR: Could not find a version that satisfies the
>>>>>>>>>>>> requirement setuptools
>>>>>>>>>>>> (from versions: none)
>>>>>>>>>>>> < ERROR: No matching distribution found for setuptools
>>>>>>>>>>> Interesting. I have:
>>>>>>>>>>>
>>>>>>>>>>> /usr/bin/python3 -m build zenmap/
>>>>>>>>>>> /usr/bin/python3 -m build ndiff/
>>>>>>>>>>> * Creating isolated environment: venv+pip...
>>>>>>>>>>> * Installing packages in isolated environment:
>>>>>>>>>>> * Creating isolated environment: venv+pip...
>>>>>>>>>>> * Installing packages in isolated environment:
>>>>>>>>>>> Compiling libnsock
>>>>>>>>>>> ...
>>>>>>>>>>>
>>>>>>>>>>> We are installing setuptools in LFS. Did you not have that
>>>>>>>>>>> installed?
>>>>>>>> It cannot use the setuptools in LFS because they are not creating
>>>>>>>> venv
>>>>>>>> with --system-site-packages (like what we are doing for various
>>>>>>>> Python
>>>>>>>> modules). If you purge ~/.cache/pip and disconnect the network
>>>>>>>> (or use
>>>>>>>> unshare -r -n to create an environment with no network) you can
>>>>>>>> reproduce the issue.
>>>>>>> Fixed with
>>>>>>> https://lists.linuxfromscratch.org/sympa/arc/blfs-book/2025-05/msg00293.html.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>> sed 's/-m pip install/& --no-build-isolation/' -i Makefile.in
>>>>>>
>>>>>> is also needed for the install side of it.
>>>>>
>>>>> This is just strange. We don't have --no-build-isolation for pip3
>>>>> install of anything in
>>>>> https://www.linuxfromscratch.org/blfs/view/systemd/general/python-modules.html.
>>>>>
>>>>>
>>>>>
>>>>> Thus I guess adding --no-build-isolation here is just covering up the
>>>>> real issue. I'll investigate later.
>>>>
>>>> The real issue is "make install" rebuilds the python modules instead of
>>>> installing the wheels produced by "make." This is just stupid IMO.
>>>>
>>>> I'll do something like
>>>>
>>>> sed -r '/pip install/s#(ZENMAP|NDIFF)DIR\)/#&dist/*.whl#' \
>>>> -i Makefile.in
>>>>
>>>> to correct it.
>>>
>>> I like using the -r in the sed. Good example.
>> I suggest to use "-E" instead. "-E" is easier to associate with
>> "extended" than "-r" and thus easier to remember. It's also used with
>> the same meaning in grep where "-r" means "recursive". Or use the long
>> form.
>>
>> Diversity is all very fine but I plead for consistency throughout the
>> books whenever possible. Building LFS/BLFS is hard enough.
>
> Well, we use -r in 6 packages and -E in 2 packages now. Those options
> mean the same thing in sed. I really don't see a problem with using
> both in the book.
There's also the long form, to cover all 3 possibilities. But does that
help the reader/user? That the same option can have 3 different names
is a weakness, not a strength. Such things are rather confusing than
helpful and make learning and using a tool or language harder than
necessary. The epitome of this is - IMO - bash (if, if [], if [[]] etc.).
LFS/BLFS should not try to teach or repeat each and every oddity of
tools or languages but first and foremost try to help the user build and
understand a modern Linux-system. To that end the most contemporary or
meaningful form of syntax should be chosen and used consistently
throughout the books.
Rainer
--
http://lists.linuxfromscratch.org/sympa/info/blfs-dev
Unsubscribe: See the above information page