[issue46833] Installer Wizard is unclear and has redundant settings

2022-02-23 Thread Christian Buhtz


New submission from Christian Buhtz :

Hello together,
this is is about the installer of Python 3.9.10 on Windows 10 64bit.

I have problems to interpret the installer wizard/dialog. And my argument is 
that no matter if there are good reasons for the current options some of the 
users are confused by it.

The goal should be to make the installer more clear about what this options do.

Lets see the "Install for all users" option:
 This appears on all three pages.
 I am not sure but would say that the first two options are related to the 
py-launcher not the the python interpreter itself. OK, but why two options?
 The third option is for the interpreter?
 And I do not see an advantage in making a difference between launcher and 
interpreter for that option.

Lets see about PATH/environment variables:
 This appears on the first page ("Add Python 3.9 to PATH") and on the third 
page ("Add Python to environment variables").
 I do not understand why.

And all this options are not synchronized. It means when I Enable "Add Python 
3.9 to Path" on the first page the "Add Python to environment variables" on the 
third page is not updated (enabled) also.

Again:
I am sure there are very good reasons for this separated options. But the 
wizard should make this reason clear to the user (or her/his admins) so that 
she/he can make an well informed decision.

--
components: Installation, Windows
files: python3_9_10_install_wizard_page1-3.png
messages: 413777
nosy: buhtz, paul.moore, steve.dower, tim.golden, zach.ware
priority: normal
severity: normal
status: open
title: Installer Wizard is unclear and has redundant settings
versions: Python 3.9
Added file: 
https://bugs.python.org/file50638/python3_9_10_install_wizard_page1-3.png

___
Python tracker 
<https://bugs.python.org/issue46833>
___
___
Python-bugs-list mailing list
Unsubscribe: 
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com



[issue46833] Installer Wizard is unclear and has redundant settings

2022-02-23 Thread Christian Buhtz


Christian Buhtz  added the comment:

Thank you very much for your quick replay and for taking my problems and 
thoughts into account.

I have to dive deeper into the topic but still have some ideas how to re-design 
the wizard.

IMHO the primary problem is that on the first page the decision between 
"normal" and "advanced" is mixed with the settings for "normal".

--

___
Python tracker 
<https://bugs.python.org/issue46833>
___
___
Python-bugs-list mailing list
Unsubscribe: 
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com



[issue46833] Installer Wizard is unclear and has redundant settings

2022-02-23 Thread Christian Buhtz

Christian Buhtz  added the comment:

In the attachment you will find a PDF with variants A to D on each side.

I tried to think into the design decisions made by the team who created the 
current installer. I am not sure of course but I tried to take this (assumed) 
decisions into account.

Variant A:
This is nearest to the current release version.
 - The settings on page 1 are moved onto that page directly to the bottom of 
"Install Now" and before(!) "Customize installation".
 - The "all users" option is now available for the interpreter and launcher.
Modifying the position of the two checkboxes make‘s it more clear to which 
decision way (simple or customized/advanced installation) they belong.

Variant B:
The same as A but Page 2 and 3 (from A) are joined together. If there is enough 
room on the wizard dialog this would be OK. But very important is to visually 
separate the two sections "Interpreter" and "Launcher" on that one dialog. You 
could do that with bigger bold text like headings or you could use a horizontal 
bar.

Variant C (would be my second favourite):
Page 1 is more minimal. The user only have to make a decision between simple 
installation and advanced/customize installation.

Variant D (my favourite):
Page one offers the simple options about "PATH" and "all users" for interpreter 
and launcher. This should be separated in a visual way of course.
btw: From a technical point of view I do not see an advantage of separating the 
decision about "PATH" and "all users" between python and py. I would assume if 
py should goes to PATH and installed for "all users" the interpreter should 
treated the same.
More important on D is that the way to the "advanced" (currently named 
"customize") installation way is "hidden" behind a simple GUI button. A lot of 
other installers doing it the same way. It is just a simple button. Not big, no 
special colours or something like that. The page 1 of the current release 
version of the installer is to much bling-bling. ;)

Some more Notes and Thoughts

„Customize“ is not a good term, because it is still possible to „customize“ the 
installation on that first page (the two check boxes on the bottom) without 
clicking „Customize installation“ and
When clicking on „Customize installation“ the next (2nd) page is named 
„Optional Features“ which is different from „Customization“. I would suggest 
"Advanced" or "Expert".
It is similar with “Advanced” on page 3. What is the difference between 
“Advanced” and “Customize”?

Add a „What is the py launcher for“ link to the wizard.

Add a „What is pip launcher for“ link to the wizard.

In the What-for-pages: Do not describe what py/pip can do but describe what the 
user can do with it. Modify the perspectives/view points! I would help you to 
review this texts.

Use horizontal bars in the GUI to better visualise the separate ways/topics. 
E.g. in Variant B on page 2.

I have some more detailed suggestions about modified wording. But I think at 
this point it is enough. :)

--
Added file: https://bugs.python.org/file50640/py_installer.pdf

___
Python tracker 
<https://bugs.python.org/issue46833>
___
___
Python-bugs-list mailing list
Unsubscribe: 
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com



[issue46833] Windows installer is unclear and has redundant settings

2022-02-26 Thread Christian Buhtz


Christian Buhtz  added the comment:

Dear Steve,
thanks for your feedback. I did not understand all details of your design 
decisions but I am OK with that.

Variant A is fine for me, too. The important to points of A for me are 1) that 
the checkboxes on page 1 are moved up directly under the "Install Now" to make 
clear where they belong to and 2) the (easier to understand) separation between 
interpreter- (page2) and launcher-settings (page3).

My apologize but I am not able to create PRs/patches. I am not familiar with 
your development environment and not able to invest time to learn it because in 
the feature it does not seems like I will create more patches for something 
because I do not have the expertise for a python interpreter.

But I understand that is a question of resources. If there are currently now 
resources I would recommend to keep this ticket open.

--

___
Python tracker 
<https://bugs.python.org/issue46833>
___
___
Python-bugs-list mailing list
Unsubscribe: 
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com



[issue44901] Info about used pickle protocol used by multiprocessing.Queue

2021-08-12 Thread Christian Buhtz


New submission from Christian Buhtz :

I read some of the PEPs about pickeling. But I would not say that I understood 
everything.

Of course I checked the docu about multiprocessing.Queue. Currently it is not 
clear for me which pickle protocol is used by multiprocessing.Queue.
Maybe I missed something in the docu or the docu can be improved?

 - Is there a fixed default - maybe different between the Python versions?
 - Or is the pickle protocol version dynamicly selected depending on the 
kind/type/size of data put() into the Queue?

Is there a way to find out at runtime which protocol version is used for a 
specific Queue instance with a specific piece of data?

Background:
I use Python 3.7 and 3.9 with Pandas 1.3.5.
I parallelize work with hugh(?) pandas.DataFrame objects. I simply cut them 
into pieces (on row axis) which number is limited to the machines CPU cores 
(minus 1). The cutting happens several times in my sripts because
for some things I need the data as one complete DataFrame.
Just for example here is one of such pieces which is given to a worker by 
argument and send back via Queue - 7 workers!


RangeIndex: 226687 entries, 0 to 226686
Data columns (total 38 columns):
 #   Column  Non-Null Count   Dtype
---  --  --   -
 0   HASH_ 
 
 37  NAME_ORG226687 non-null  object
dtypes: datetime64[ns](6), float64(1), int64(1), object(30)
memory usage: 65.7+ MB 

I am a bit "scared" that Python wasting my CPU time and does some compression 
on that data. ;) I just want to get a better idea what is done in the 
background.

--
messages: 399447
nosy: buhtz
priority: normal
severity: normal
status: open
title: Info about used pickle protocol used by multiprocessing.Queue
versions: Python 3.7, Python 3.8, Python 3.9

___
Python tracker 
<https://bugs.python.org/issue44901>
___
___
Python-bugs-list mailing list
Unsubscribe: 
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com



[issue44701] Create a @deprecated decorator (annotation)

2021-08-18 Thread Christian Buhtz


Christian Buhtz  added the comment:

This discussion on python-ideas
https://mail.python.org/archives/list/python-id...@python.org/thread/62CTVNQ2GIS4B6WUBX23K4CCCK5MCGYL/

--
nosy: +buhtz

___
Python tracker 
<https://bugs.python.org/issue44701>
___
___
Python-bugs-list mailing list
Unsubscribe: 
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com



[issue42993] doc xml.etree.ElementTree.ElementTree.write does not mention attribute order

2021-01-21 Thread Christian Buhtz


New submission from Christian Buhtz :

The docs for 'xml.etree.ElementTree.ElementTree.write' in Python 3.7 (and 
possible earlier) does not say a word about the ordering of the attributes.

This makes unittesting hard.

But looking in the code tells me that the attributes are ordered lexically. 
Important to know. So please upgrade the docu about that.

In Python 3.8 the ordering behavior changed again but is mentioned in the docu. 
So no need to change in 3.8 or later.

--
assignee: docs@python
components: Documentation
messages: 385448
nosy: buhtz, docs@python
priority: normal
severity: normal
status: open
title: doc xml.etree.ElementTree.ElementTree.write does not mention attribute 
order
type: enhancement
versions: Python 3.7

___
Python tracker 
<https://bugs.python.org/issue42993>
___
___
Python-bugs-list mailing list
Unsubscribe: 
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com