akx

This is, in fact, not true.

    $ ./build/bin/llama-cli -m Qwen3.6-35B-A3B-UD-Q4_K_XL.gguf --system-prompt "(this prompt explicitly left empty)"
    build      : b10889-6e41074fc
    model      : Qwen3.6-35B-A3B-UD-Q4_K_XL.gguf
    ftype      : Q4_K - Medium
    modalities : text
    using custom system prompt

    > What model are you?

    [Start thinking]

    Here's a thinking process:

    1.  **Analyze User Input:** The user asks "What model are you?"
    2.  **Identify Core Question:** The user wants to know my identity/model name.
    3.  **Check System Prompt/Instructions:** The system prompt says "(this prompt explicitly left empty)" - this is unusual but doesn't change my core identity. I need to respond based on my actual identity as Qwen / 通义千问.
    4.  **Formulate Response:** I should clearly state that I am Qwen (通义千问), developed by Alibaba Group's Tongyi Lab, and keep it concise as per guidelines.
    5.  **Draft Response:** I am Qwen (通义千问), a large language model independently developed by Alibaba Group's Tongyi Lab. How can I assist you today?
    6.  **Refine according to guidelines:** The draft matches the required identity and tone. It's direct and helpful. Ready.
    [End thinking]

    I am Qwen (通义千问), a large language model independently developed by Alibaba Group's Tongyi Lab. How can I assist you today?

Classic link. Not the same thing though. I mean try telling a competent agent something like

> Let's build a webapp that allows the user to choose an audio file, we slice it into timestamped word regions with transformers.js + onnx-community/whisper-base_timestamped (with language selection). The user can then scroll through the waveform (that shows the found word regions), synchronized with the transcript, and select word regions in the transcript or in the waveform, and fine-tune the selection if they need to, and then export the slice as WAV.

and see what pops out.

UX-wise, from what I can glean from the .webp, I'd personally (having worked in this same space) appreciate being able to see the slices I already have without having to switch tabs back and forth, for one...

To be fair, gnu tar 1.31, 2019-01-02, added `--zstd` (and other automatic .zst support). So that support has been there for a good 7 years and 4 months at the time of writing.

> Typically, other archives like .tar.bz2 can be smaller. But those aren’t backwards-compatible!

Is there any point for (new) .bz2 archives in the era of Zstd?

What do you mean with "basically ensures that dependencies won't become managed properly"?

This doesn't do the same thing though, since it's not Unicode aware.

    >>> 'x\u2009   a'.split()
    ['x', 'a']
    # incorrect; in bytes mode, `\S` doesn't know about unicode whitespace
    >>> list(re.finditer(br'\S+', 'x\u2009   a'.encode()))
    [<re.Match object; span=(0, 4), match=b'x\xe2\x80\x89'>, <re.Match object; span=(7, 8), match=b'a'>]
    # correct, in unicode mode
    >>> list(re.finditer(r'\S+', 'x\u2009   a'))
    [<re.Match object; span=(0, 1), match='x'>, <re.Match object; span=(5, 6), match='a'>]

It's pretty good. And for once, a software-engineering-ly high-quality codebase, too!

All too often, new models' codebases are just a dump of code that installs half the universe in dependencies for no reason, etc.