I just realized that the news report could be more useful if it cites
the information source. Here we go:
debgpt -Hx ldo:debian-devel/2024/10 -a 'write a news report based on the
provided information. Cover as many topics as possible. You may expand a
little bit on important matter. include links to the report.' --no-render
BTW, which mailing lists should I cover with those montly reports?
I currently tried on debian-ai, debian-devel, and debian-science.
------------------------------------------------------------------
THE FOLLOWING CONTENTS ARE GENERATED BY AI
------------------------------------------------------------------
### Recent Developments in the Debian Community: A Snapshot
**Date:** October 31, 2024
The Debian development community has seen a flurry of activity in recent
weeks, marked by new package proposals, discussions concerning
architecture support, and enhancements in package management practices.
Here’s a comprehensive overview of the latest developments.
#### New Package Proposals for Tryton
Developer Mathias Behrle has put forth several promising new modules for
the Tryton application platform, which is gaining traction in business
and healthcare sectors. The proposed modules include:
- **tryton-modules-account-invoice-watermark**: This module introduces
the ability to add watermarks to invoices, enhancing document authenticity.
- **tryton-modules-marketing-campaign**: Aimed at improving marketing
efforts, this module facilitates the collection and analysis of data
related to campaigns.
- **tryton-modules-account-stock-shipment-cost**: This module allows for
the allocation of shipment costs based on invoices, streamlining
financial processes.
- **tryton-modules-document-incoming-invoice**: This innovative module
automates the creation of supplier invoices from incoming documents,
improving efficiency.
These proposals signify a strong commitment to bolstering the
functionalities of the Tryton platform. For more details, visit [Tryton
Downloads](https://downloads.tryton.org/7.0).
#### Lomiri Applications Packaging
In a parallel effort, Mike Gabriel has commenced the packaging of
several applications designed for the Lomiri operating environment,
which is essential to the Ubuntu Touch mobile OS. Key applications being
packaged include:
- **Lomiri Address Book App**: A user-friendly application for managing
contacts.
- **Lomiri Dialer App**: An application providing essential telephony
features.
- **Lomiri Messaging App**: A messaging platform facilitating seamless
communication.
These applications aim to significantly enhance the user experience
within the Lomiri environment. Further information can be found in the
[Debian Development Mailing
List](https://lists.debian.org/debian-devel/2024/10/msg00368.html).
#### Discussions on s390x Architecture Support
A significant topic of discussion has been the future of the s390x
architecture in Debian. Developer Chris Hofstaedtler has raised concerns
regarding the dwindling number of active maintainers and a growing
number of packages that do not support s390x. The community is currently
evaluating the viability of continuing support for this architecture,
especially in light of IBM's involvement and commitment to maintaining
s390x. For more insights, see the ongoing discussion
[here](https://lists.debian.org/debian-devel/2024/10/msg00295.html).
#### Challenges with Lintian and Package Uploads
The community continues to address challenges associated with Lintian,
the Debian package checker. Recent issues concerning license compliance
in the `warzone2100` package have sparked discussions on how to manage
Lintian overrides more effectively. Developers are actively exploring
solutions to streamline the package upload process and minimize
friction. More information is available
[here](https://lists.debian.org/debian-devel/2024/10/msg00203.html).
#### Upcoming Changes and Improvements
In a bid to enhance package management practices, the Debian community
is focusing on refining the Non-Maintainer Uploads (NMUs) process.
Developers are advocating for standardized workflows that would improve
collaboration and efficiency, ultimately reducing confusion and workload
for maintainers. For further details on NMUs, check out the discussion
[here](https://lists.debian.org/debian-devel/2024/10/msg00276.html).
#### Technical Updates and Backlogs
Colin Watson has confirmed that the Debian upload server is currently
facing delays due to a backlog caused by a mass rebuild for Pointer
Authentication (PAC) and Branch Target Identification (BTI) support.
Developers are advised to anticipate longer processing times for uploads
during this period. For updates on the upload delays, refer to the
discussion
[here](https://lists.debian.org/debian-devel/2024/10/msg00365.html).
### Conclusion
These developments highlight the vibrant and dynamic nature of the
Debian community, characterized by ongoing efforts to enhance software
offerings, improve user experiences, and address architectural support
challenges. The collaborative spirit remains robust as developers work
together to navigate these changes and continue to uphold the principles
of software freedom.
For more detailed reports and discussions, please refer to the [Debian
Development mailing list archives](https://lists.debian.org/debian-devel/).
On 11/9/24 07:51, Mo Zhou wrote:
The LLM I used to produce that exact news report was gpt-4o-mini,
from openai. ChatGPT is the name of openai's LLM web interface and
its underlying LLM model name could change. It took roughly 3
minutes to perform the bulk API calls.
That said, I basically implemented support for all commonly seen
LLM inference services:
(4 commercial ones)
openai, anthropic, google, xai,
(4 self-hosted)
llamafile, ollama, vllm, zmq (built-in but kind of outdated.)
Other services missing from the list are also supported as long
as it has compatibility mode to the openai api.
For the particular use case like summarizing a mailing list, self-hosted
one will be much slower to respond to the bulk API call unless it is
hosted on a GPU cluster :-)
Small LLMs are not necessarily smart enough. The open llm leaderboard[3]
is a good reference for figuring out the best open-access llm for
self-hosting.
In terms of "Debian hosted computer with AMD GPU for LLM inference" --
that is exactly one of the long term goals of debian deep learning
team (debian-ai@l.d.o). Team members are working to prepare the ROCm
packages and the ROCm version of pytorch.
I find ollama[1] and llamafile[2] quite handy to use locally if do not
mind using software from outside of debian archive, with a spare GPU.
[1] https://github.com/ollama/ollama
[2] https://github.com/Mozilla-Ocho/llamafile
[3]
https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard
On 11/9/24 05:19, PICCA Frederic-Emmanuel wrote:
is it via ChatGPT or an llm self hosted ?
Can we imagine having a Debian hosted computer with and AMD GPU
dedicated to this use case ?
Se should provide these summaries letter for most of our mailing list :)
cheers
Fred
----- Le 9 Nov 24, à 14:09, Hector Oron zu...@debian.org a écrit :
Hello Lumin,
El sáb, 9 nov 2024 a las 10:27, DebGPT (<lu...@debian.org>) escribió:
This is an experiment, by letting LLM go through all 369 emails from
debian-devel on Oct. The command for producing the news report
is included below. Use debgpt's git HEAD if you want to try.
First time I see this kind of email, I thought time ago that'd be a
really cool use of AI, to produce a summary of mailing lists - since I
struggle to read everything.
I just want to thank you for putting this together and, at least from
my side, this is very much appreciated.
Regards
--
Héctor Orón -.. . -... .. .- -. -.. . ...- . .-.. --- .--. . .-.
