1. Forum
  2. Usenet
  3. UK.D-I-Y

  • Re: Laptop replacement

    From Chris J Dixon@21:1/5 to Theo on Wed Mar 26 20:06:53 2025
    Theo wrote:

    FYI there was a typo in the uk.d-i-y in the newsgroup line of your post:

    Thanks Theo. Fat fingers and ageing eyes. :-(

    Chris
    --
    Chris J Dixon Nottingham UK
    chris@cdixon.me.uk @ChrisJDixon1

    Plant amazing Acers.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From mm0fmf@21:1/5 to Chris J Dixon on Thu Mar 27 17:15:12 2025
    On 26/03/2025 20:06, Chris J Dixon wrote:
    Theo wrote:

    FYI there was a typo in the uk.d-i-y in the newsgroup line of your post:

    Thanks Theo. Fat fingers and ageing eyes. :-(

    Chris
    I've had a number of Dell laptops for my job and one HP laptop. My son
    has had a number of Dell laptops and he never managed to destroy any of
    them as a teenager.

    I'm quite happy with them and currently have an i5 13th gen running
    Win11 for work but it's used just as a terminal to access the compute
    farms in our data centre. I've no idea what it's like as a computer.

    Before that I had a Dell 7490 laptop. i7 8th gen 16GB / 256Gb, thin
    light, long battery life. It was a great workhorse and was used for lots
    of software development. Ran Win10Pro just fine. It was replaced by the
    i5 which I don't feel is as well made.

    I so liked my 7490 that I bought one for myself from eBay refurbished.
    It was £189 for a 16GB/256GB i7 8th gen. This one was from 2019 and
    still the battery lasts a good 6hrs+ when browsing/emailing. It came
    with Win11 Pro installed which was quickly replaced with Debian. But
    there will be a Win11 licence in the BIOS if the one you buy doesn't
    have it installed.

    It maybe 6 years old and *ONLY* an 8th gen i7 but laptops got more power
    and resources than most people need sometime around 2015. All they do
    now is wait faster for the user to do something. Maybe it's not quick
    enough for games. Buy a console for them!

    There are plenty on eBay, just make sure you buy one without cracks etc.
    in the plastic and not too much wear on the keyboard. 32GB/1TB would be
    nice but if this is not to be a main machine, 16GB/256GB does me fine.
    Though I do run Debian Linux not Win11.

    My wife has a Lenovo T480 with issues (keyboard legend wearing off keys, intermittent battery fault)... I think I'll pick up another 7490 for her
    as the prices are not much as they are perceived as "a bit old".

    Of course, you mileage may vary.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From David Wade@21:1/5 to Bernard Peek on Thu Mar 27 18:02:50 2025
    On 27/03/2025 16:38, Bernard Peek wrote:
    ["Followup-To:" header set to uk.d-i-y.]
    On 2025-03-27, Jeff Gaines <jgnewsid@outlook.com> wrote:
    On 27/03/2025 in message <vs2ptj$3jchh$2@dont-email.me> Lawrence
    D'Oliveiro wrote:

    On Wed, 26 Mar 2025 19:02:39 +0000, The Natural Philosopher wrote:

    Well I have had pretty good results for MY needs with HP laptops BUT
    anything to run Windows 11 seems very expensive. Like £400

    Whereas a Windows ten capable refurbished is under £100

    Would you entrust mission-critical business operations to obsolete,
    unsupported software?

    I use Windows 8.1 on all my machines that will run it because it allows me >> to download updates and install them when convenient to me. Why this
    facility doesn't exit on Win 10 goodness knows.


    Because the vast majority didn't install updates resulting in huge
    bot-nets. As 8.1 is unsupported Microsoft no longer confirms or
    acknowledge security bugs. As it shares code with 10 and 11 it probably contains many of the same vulnerabilities.

    The scripts the script kiddies use target vulnerabilities not Windows
    versions so any use of an old OS is risky.

    I see you've been suckered by MSFT marketing bullshit :-)

    Yes and no. As an IT Manager I described myself as custodian of the corporate paranoia. One of the options I had was to check that any machine connected had the right OS and had all of the patches I considered
    essential. If I had been managing financial or other critical data the patches might have been added to the critical list within a few days of release.

    It takes about three days for miscreants to download a patch and reverse-engineer an exploit. So exploits by skript-kiddies are most likely three days after a patch. They get progressively less likely after that.


    I had the same approach. At one point we deployed fixes to test machines
    for a week to check their impact on production machines. After a couple
    of incidents we swapped to deploying as soon as possible as we decided
    the impact of a bad patch was less than that of a security breach...

    Dave

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Nick Finnigan@21:1/5 to All on Thu Mar 27 23:43:35 2025
    On 27/03/2025 17:15, mm0fmf wrote:

    It maybe 6 years old and *ONLY* an 8th gen i7 but laptops got more power
    and resources than most people need sometime around 2015. All they do now
    is wait faster for the user to do something.

    ... or wait with the clock running very slowly for all the cores.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Paul@21:1/5 to Nick Finnigan on Thu Mar 27 22:26:13 2025
    On Thu, 3/27/2025 7:43 PM, Nick Finnigan wrote:
    On 27/03/2025 17:15, mm0fmf wrote:

    It maybe 6 years old and *ONLY* an 8th gen i7 but laptops got more power and resources than most people need sometime around 2015. All they do now is wait faster for the user to do something.

     ... or wait with the clock running very slowly for all the cores.


    Core state transitions are ultra-fast now.

    Previously, the F-V curve (as depicted at the URL below),
    required making small steps of voltage and frequency, and it might take
    a millisecond doing it that way, to go from
    the lowest P-state to the highest P-state.

    But I recollect a diagram being drawn, showing a
    core jumping pretty well directly from one frequency
    to another. And the time scale was at the microsecond
    level.

    https://www.anandtech.com/show/14873/reaching-for-turbo-aligning-perception-with-amds-frequency-metrics-/7

    The staff at Anandtech were laid off, and there is no one
    to write articles like that any more. But I understand
    the latest 14900K or the like, there is a capability
    at the microcode level, to modulate the clock at the
    instruction level (I know this sounds ridiculous on the
    face of it). Such a thing might be proposed as
    a response to the "system shock" of execution of
    an AVX512 instruction. There had to be some microcode
    patches, to correct errors in what was going on there.
    I only mention that, to give some idea just how
    small the timescales for this sort of shenanigan,
    have become.

    Paul

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)