IRC channel logs
2026-03-31.log
back to list of logs
<theesm>PotentialUser-11: usually we wait a few weeks to ~month until the latest kernel version will be set as default; i'd avise against running bleeding edge unless you have a good reason to <mra>any thoughts? seems like an internal guix issue of some sort <Guest41>this seams like a missconfigured channel file to me - specifically the hash part of it. Did you write a channel file or is my guess wrong? <mra>Guest41: i didn't write a channel file, no. i'm just running guix pull <Guest41>huh okay weird - then I got no real guess <mra>unless i made a channel file some time ago and forgot about it, but if a channel file exists it would be in ~/.config/guix, right? <Guest41>yea, if there is one it should be in `~/.config/guix/channel.scm` <Guest41>are you using the nixos service or have you installed it with the script? <mra>i'm using the service. just services.guix = true; <Guest41>honestly thinking about it: I got no clue, sorry. I have had minor issues with the guix service on nixos (guix pull finished, but new profile is not getting used) which I think is because nix tries to manage guix a bit too much, but I don't want to make any guesses here <mra>all good! i might also ask in #nixos <mra>although this seems more like a guix issue <Guest41>I'm trying to install guix on a foreign distro without `nscd` in their repos. 1: https://ftpmirror.gnu.org/gnu/guix/ seems to be down - is it unexpected or something maintenance related that I missed? 2: I think I got `nsncd`, the replacement, working, but the install script still warns me. Is there a way to test this (like a small code snippet)? <Guest72>Hello everyone! For those who use networkmanager, do you know where it stores the settings that I edit wby using nmtui? There is no config file in /etc/NetworkManager/ ...... <Guest72>But it must store the edits somewhere. like "wwan software" "deactivated" <Guest72>All I see in /etc/NetworkManager is a folder containing the two wifi connections that I used since installing Guix <Deltafire>there's some state in /var/run/NetworkManager <Deltafire>i think nmtui may be controlling nm via dbus, so some settings may not exist on disk <Guest72>By the way, I have this other problem. During setup I set a root password (mandatory) to say toor. Then I must create a user account so I do and I leave the password field blank. I start using guix and somehow now when I log into my bormal account and do sudo ..... It asks for password and toor doesnt work but empty password does. <Guest72>I seem to remember that the first time I was asked for the admin pswd I did put toor and it worked , but ever since it looks lime it switched from toor to my normal user password. Im at loss <Guest72>So now toor doesnt work as admin pswd anymore, and my normal user password (empty oswd) works both to log in as user, and to authorise sudo commands. <Guest72>Im tempted to think that when I authorised sudo the first time then guix might have thiught ah this normal user can authorise admin commands (by using the admin password toor) so it must be an admin, hence from now on I will let him do admin commands and he just needs to authentify with his normal password. But then even if this reasoning is <Guest72>correct I dont understand why toor doesn't _also_ work with sudo <ieure>Guest72, sudo asks for the password of the user calling it, not the password of the user it runs commands as. <mra>Guest72: when i'm setting up a new machine i usually entirely disable the root password <ieure>Guest72, That's pretty much the entire point of sudo, to let users run limited programs as other users. If you had the root password already, you could just su to the user, or log in as them -- you have full and complete access. <Guest72>ieure then how does ir know that the user calking it is authorised to call admin commands? And what was the point of setting up admin password during the install? <Guest72>mra admin pswd was mandatory in guix tui install <ieure>Guest72, sudo has a configuration file where you can express who can run which commands. The Guix default is that any user in the `wheel' group has unlimited sudo access, and the installer puts the user you create into that group. Please read the sudo documentation if you have questions about how it works and why. <mra>Guest72: oh, sure, i usually just disable it after the fact. i prefer to have trusted users in `wheel' and entirely disable logins as root <Guest72>Ok thanks. Then I misremember that I did insert toor the first time I called sudo from my user? Does guix always put the first user I create in the wheel group? Is this a gnu convention? <ieure>Guest72, Yes, you must be misremembering. Entering anything other than the current user's password in a sudo prompt will cause an auth failure. <ieure>Guest72, The Guix graphical installer always puts the user you create into the wheel group. There is no "GNU convention" which would cover this behavior. <ieure>It is just a sensible way to handle installation. <bdunahu>it assumes the person who picked the root password in the installer is probably an administrator :) <Guest72>Ok thanks. I wonder if the second user created in the installer is not in wheels <ieure>Guest72, `grep ^wheel /etc/group' will show you all users in the group. <Guest72>And also I wonder what is the point then of creating the superuser if I have the user in wheels. Like mra was saying <bdunahu>you can at least configure who is wheel at the end of the installer, when it asks if you want to make manual changes to the scheme program that is installed <ieure>Guest72, All Unix-like systems must have a root user. <ieure>Guest72, If there was no root user, who would sudo run things as? <Guest72>ieure yea but Ii didnt create a second user during the install. I just wonder because it shiuld not and if it did it might be a bug <ieure>Guest72, Oh, I have no idea. But I don't agree that it "should not" do that. <ieure>I think both adding all users created during the install to wheel, and only adding the first, are both reasonable options which could be chosen. <Guest72>Ok thanks. What does mra mean by disabling the root user after install? <ieure>Guest72, Setting it to have no password, so you cannot log in to it. You cannot remove the root user, the system wouldn't work. <Guest72>So there is an extrfa rule that says when the admin user has nompswd you cant log in as it? Cause normal users cN have no pswd <Guest72>I dont like this discrepancy between password for normal users and admin <Guest72>I think I wil report the fact that you must set a pswd for the root user in the install as a bug. You should be allowed to set nompassword. <mra>Guest72: yeah, setting it to have no password. you can also set its login shell to `nologin' <Guest72>Archwiki says you can also "lock" it <apteryx>lilyp: something in GNOME 49 started loading typelibs strictly, e.g. looking at the shared-library path, which is often wrong in Guix (you can look at the gjs .gir after an interrupted build) <apteryx>I wonder what/where that change came from <apteryx>it's a bit messy to workaround, so maybe we should improve the patch gobject-introspection patch we carry causing this (it forces embedding the absolute path instead of leaving the library names plains, to be looked up from FHS/run paths) <lilyp>apteryx: I do wonder why we embed in the first place, but also I don't think we necessarily have those shared objects on library paths otherwise <lilyp>you should look into glib/gobject-introspection for changes there <lilyp>it's probably an independent attempt at fixing the issue we're observing <apteryx>with a patch in meson that works in combination with the patch we already have against gobject-introspection <apteryx>lilyp: yeah, it'd probably cause issues when some daemon tries to load a typelib and the library is not in LD_LIBRARY_PATH or RUNPATH <apteryx>it also means we have to move the check phase past install phase for a few meson/gnome packages. I've already done so for some of them. I'm not sure why we're just discovering this issue now. <lilyp>I already had to move some checks after install for GNOME 48 <lilyp>maybe I discovered a similar issue, but it wasn't as grave yet <lilyp>btw. I do have a separate patch for meson which addresses an issue when using clang-toolchain; should I submit that for gnome-team as well? <apteryx>you can yes, it's already going to be a large rebuild, and moreso with meson now <apteryx>lilyp: I think previously gobject-introspection-girepository.patch probably made things work for us, at least in a test scenario, and with libraries found in RUNPATH <lilyp>I think it did so even outside of test scenarios, but I could be wrong <apteryx>lilyp: the meson patch from nix works at least, I tested without a world rebuild on gnome-shell <YAR_Oracool>can anyone give a link to guix's official channel in a file? <lilyp>is the duplicate resolved with the patched glib which ignores the so for the already loaded one? <sneek>Welcome back adanska, you have 1 message! <sneek>adanska, czan says: When using fold-services you should set #:target-type to the type of service you're interested in. The default is system-service-type, which has its value set to a monadic service. Other services have a more directly useful service-value which you can extract with service-value. In your case you probably want to set #:target-type to your specially-defined service. <adanska>czan, thanks for that tip! I managed to figure that out for myself after much toiling :) But thank you anyways <janneke>mentioning block-facebook-hosts-service-type on the fediverse got some attention and it was suggested that iwbnif 'facebook were a parameter, and have blocking support for the other big tech bros too <janneke>ACTION would be interested to see that and also, should github be included in 'microsoft, and gitlab.com in 'google, and how would we handle that? thoughts? <identity>and map them to a different reserved IP address, while we are at it <identity>i think it would make sense to split 'microsoft from 'github, maybe adding «also see 'github» to the documentation entry for 'microsoft <attila_lendvai>what are you using for git merge/diff? kdiff3 is recommended as best, although with a dated UI, but it's not packaged in guix. <attila_lendvai>ohh, one can press 'e' on a file in magit to get a potentially better tool in emacs (ediff with 3 windows). <csantosb>civodul: I agree, the changes to hwloc shouldn't be on master, sorry about that <civodul>csantosb: no worries, but it’s true that the build farms have been under extreme pressure since yesterday—with the glibc upgrade, then hwloc, etc. <civodul>and then i realized i was +/- taking care of 3 of them <lilyp>apteryx: #7478 if you want to try that <identity>i guess atai messed up their git config for that commit? <janneke>in that case iwbn to know if they got a heads-up about that <sneek>ada36 was in #guix 7 months ago, saying: identity ah! Thank you!. <jlicht>o/ from librewolf-in-ewm-on-guix-system (ewm ~= exwm, but for wayland) :-) <kestrelwx>Everything that depends on `wlroots` at the moment should depend on `wlroots-0.19` instead, right? I'm looking to bump it to 0.20. <jlicht>that might be even nicer, thanks kestrelwx <identity>i really like that EXWM-on-Wayland implementations started popping up, but i am still shell shocked from 50 million problems/random errors with using a plain compositor instead of a full DE <snamellit>It is a credit to the gnome and KDE team to turn their 50 million problems into just 1 million problems. That being said I find them just too much and the problems you get with a WM are much shallower. <identity>i wish i had the problems you get with WMs, mine are impossible to debug <janneke>identity: is there a non-LLM-coded EXWM/Wayland option that i could look at? <janneke>identity: interesting, but hmm -- Reka does not expose any input or output management to Emacs Lisp right now, and I'm not sure that it will. <janneke>anyways, it "could work" if somehow X would somehow cease to exist <kestrelwx>janneke: The Rust bindings for Emacs dynamic modules have Claude commits, but there aren't huge diffs or anything. Still, I don't really wanna deal with it. <janneke>kestrelwx: yeah, i'm a big fan of not running any LLM code <Guest72>Hey beautiful people. I just installed upower but upon execution I receive the error: libupower-glib-warning **: stringofnumbers........: up_client_get_devices failed: gdbus error:org.freedesktop.dbus.error.servoxeunknwn: the name orf.freedesktop.upower was not provided by any services files <Guest72>Also note I have guix with no graphical interface <Guest72>Can anyone understand what is going on? :) <lilyp>you mean you only installed upower and no graphics stack? <Guest72>Ps I might timeout disconnecr termporarily but will come back and I also read the channel logs <Guest72>lilyp yeah, I dont know what is a graphic stack, and I execute upower from tty <Guest72>I mean I also installed the dependencies etc. I used guox install <lilyp>you need at least a running dbus-daemon for upower; you can get the necessary bits through dbus-service-type iirc, but you'd also have to start that <lilyp>alternatively, use dbus-run-session $SHELL to spawn a shell within a fresh dbus session <lilyp>from there, you can run upower <Guest72>Oh I see. Is there any more "guix-native" way to inspect my laptoo battery properties/stats <identity>Guest72: i have no idea what that would mean <Guest72>Any way to read stats/properties of my battery on guix with no graphical interface, without having to start dbus deamon (which I suppose is not udsually ctive on my guix ) <identity>i mean, you can probably poke at some device file. nothing ‹guix-native› about it, though <identity>something something /sys/class/power_supply/BAT0/ <identity>you can rummage through files in that directory for the information you want <kestrelwx>`guix shell acpi -- acpi -b` probably works without DBus? <chrislck>libwebkitgtk package was modified in march to enable webRTC <chrislck>this means apps which use libwebkitgtk now need to add gst-plugins-bad as a dependency <chrislck>shouldn't libwebkitgtk explicitly depend on gst-plugins-bad now? <identity>…or propagate it. i think somebody already said something to that effect here recently <Elwood_Soup>Hi guys, if feels like helping out someone new to guix, I'd appreciate it <gabber>Elwood_Soup: i can't have a look at your paste without javascript.... <gabber>let me check if torbrowser works <Elwood_Soup>Do you all just use pastebin normally? or do you have a different one you like <gabber>there's one in the description of this channel <gabber>Elwood_Soup: without much expertise i guess the problem is that armhf-linux is the 32bit arm variant, and you are trying to build the 64bit arm image, right? <identity>generally, we should be able to view the links you send without JS. common courtesy on IRC <ieure>Elwood_Soup, What host architecture are you running that command on? The error is saying that the guile binary that it's trying to execute is for a different architecture. <Elwood_Soup>I'm honestly just trying to print the list of possible boards, but i get the same error if i replace 'A20-OLinuXino-Lime2' with nonsense <gabber>Elwood_Soup: you mean image types? <chrislck>identity: it was annoying that "guix shell -D gnucash ..." is no longer an adequate environment for dev work <gabber>Elwood_Soup: Guix can be built for a variety of architectures, but the with ARM boards we need to flash these systems in the correct format on (usually) SD cards—these are the so-called image types <gabber>Elwood_Soup: try `guix system image --help` <Elwood_Soup>I'm on amd64, but I'd like to build an image for an arm64 board <gabber>respectively `guix system image --list-image-types` <gabber>then the system= part should be `aarch64-linux', not `armhf' <gabber>but note that the `system=' directive tells your system to *natively* compile in an emulated system <gabber>cross compilation is much faster (we use `--target` for that) <Elwood_Soup>goootcha gotcha. so is there a cross-compilation flag? <gabber>--target, as stated before. it may not work for a complete system, though (it depends) <Elwood_Soup>ok, it looks like 'guix system image --target=aarch64-linux-gnu --list-image-types' was the command I was really looking for <gabber>Elwood_Soup: --help with all `guix` command invocations is very helpful (though one often has to read the whole list to find what they are looking for) <Guest27>Hey i was looking for bcachefs package. 1 what does statically linked mean? 2 when i click on guix website to open description it says package not found. Help!;) <Guest27>3 do i need linux kernel packAge or tools pAckage. Any help is very appreciated <gabber>statically linked is a compiler option for the inclusion of library code. see statically linked vs dynamically linked <Guest27>Create bcachefs partition (with mirroring) <Guest27>Ah ok it looks like default is dynamic <gabber>"GuixSD" → haven't read that one in quite a while.. very old-school! ;) <gabber>unfortunately i have no clue about and no experience with bcachefs <kestrelwx>You would need the kernel modules in the system configuration to support it, but other than that no clue. <Guest27>kestrelwx so installing tge kernel module packAge is not enough? <cdegroot>bcachefs-linux-module and bcachefs-tools - install these two from your config.scm and there's a decent chance that things will "just work". <Guest27>Wait cdegroot by config.scm you mean the home config or the system config <gabber>and i think bcachefs-linux-modules has to installed in the (kernel-loadable-modules) section, right? <Guest27>ieure but I always thought that packages go into the home config (or are installed via guix package -i). Why the exception here? <Guest27>Because it has to be loaded immediately by the kernel? <Guest27>In which case bcache-tools can go into the home (or be installed by package command)? <Guest27>Actually before placing kernel bcache in the config I must also i stall it with the package -i comma d I supoose <gabber>you can, but that's not necessary <Guest27>gabber thx. Can you give a short exMple of what Id add to my config to add the kernel kodule? So that then i will look up and understand the words you out in your example, this will help.me orient myself in the docs <Guest27>gabber but if I just out it in the config then I must be connected to the internet when I launch reconfigure, correct? And if yes, does system reconfigure just actually launch package -i under the hood? <gabber>sorry, i (still) have neither clue about nor experience with bcachefs <gabber>but yes, you may need internet connection (if you install packages that are not yet in your /gnu/store) <Guest27>Ok maybe ieure will clarofy abiut the kernelloadablemodules section <ieure>Guest27, I have no time to do that, I suggest reading the manual and cookbook. <gabber>Guest27: you can also just try it out. see the `operating-system Reference' section in the manual <Guest27>Honestly as a very recent guix user I am so frustrated from the existent of the two ways of doing things by declaration (system config, home config) and by using the package command. <ieure>Guest27, You're welcome to use whatever approach works for you. I only use declarative, others only use imperative. <gabber>Guest27: why does that frustrate you? there is the declarative way (with system/home config) and the imperative way (`guix package -i`)... but they do different things. packages installed in your profile are not available for the system/root profile <yelninei>I love tests that have a "known_erros" section where errors are turned into warnings .... <Guest27>gabber home config packages are not installed by the system with package -i under the hood? <gabber>you install packages to profiles. you can have as many profiles for as many use-cases as you want. the system always has a profile (with the packages added to it in the operating-system declaration), all other profiles exist, get created and populated by your commands <Guest27>ieure yes but I dont think they are different approaches to achieve the same features. Example package has profiles which dont exist for the home declarative approach! <Guest27>gabber so system reconfig with a new package inside launches under the hood package -i -p SYSTEMPROFILE <Guest27>Sorry for being amateur but I try to understand <Guest27>And your help is immensely precious to me <gabber>Guest27: don't be sorry for not knowing! we all don't know stuff (: and we all started with no knowledge on Guix (except for, like, this one person) <Guest27>ieure unless the equivalent to having many profiles that I manage with package, is having many home distinct configs <yelninei>a test that checks that libc struct definitions match the actual libc struct definition <gabber>Guest27: do be pedantic: no, system reconfiguration does *not* call `guix package -i -p FOO` under the hood, but i am quite certain that both, the system reconfig command and the package installation command use the same procedures to modify the profiles (: <yelninei>i mean the test is awesome because I just caught a lot of errors but then i was wondering why linux only has warnings and no errors <Guest27>gabber thanks, your pedantry clarifies things in my mind. In my ideal world I use the imperative approach and this uodates the config file accordingly. But such thing doesnt exist in guix right? (I think i can export the home config, not sure about system config, but its not like it gets updated aftrr eevery package -i that i do) <gabber>Guest27: no, in Guix we do it exactly the other way round <gabber>no, `guix package -i` does not alter any of your config files <lilyp>if you do use the imperative way now (it is easier to start with fwiw), you can decide to generate a config file from that later with `--export-manifest` <lilyp>but both the imperative and declarative style get "grounded" to another declarative style that's also a "manifest", but with all variables and references resolved <Guest27>Ok so gabber I understand from the tone in your sentence that there is some consensus about the declarative being the predilect way of doing things. Fait enought, but then the declarative way of installing user packages is via home, and I wonder then how guix home (declarative way) would manage different profiles. In guix home section of the manual <Guest27>i see no reference to pthe concept of profiles. And also guix home is much further down on tha man compared to package -i so new users will learn about using package -i for their software, and system reconfig for the system, and be left with this mixed allroach in mind that in my opinion, because it's mixed, then it's also confuding. Only much <Guest27>later they will find guix home (whixh as i said mentions no profiles) <lilyp>the system command likewise has a profile/manifest in that sense, but also has additional configuration for services from which special files are constructed (e.g. an activation script) <lilyp>Guix Home is a declarative way of managing $HOME similar to the system. There is also a declarative `guix package` way which predates Guix Home. <Guest27>lilyp can you ref.the section of the manual that talks about the "meta-declarative" way you mention ie the manifest? <lilyp>I would probably start at or around section 8.4 "Writing manifests" <gabber>Guest27: there is not really a right or wrong here, but Guix people tend to like the declarative way. Guix has grown historically and (especially) the reference manual could use some re-ordering and tidying up. guix home came after almost(?) a decade of guix' existence, so people used Guix a lot without/before guix home... Guix is great because it allows/enables/empowers you to tend your system(s) the way you like <Guest27>lilyp could I hVe a reference also about the declarative package -i that you mention pls? <lilyp>there is no declarative package -i <lilyp>there is declarative package -m <gabber>Guest27: you can/could use different profiles with `guix shell -p path/to/profile` after creating one <lilyp>which is linked in section 8.4 <Guest27>Thank you to all of you for your help. Before the library kicks me out because its closing time, one last question for this support session :). : <Guest27>I have this analogy in mind: package -i IMPERATIVE on the default profile //analogous to// home reconfogure DECLARATIVE. Question: package -p secondprofile -i. IMPERATIVE ON YHE SECOND PROFILE //analogous to// ???? I wanna say guix home -p secondprofile reconfigure BUT THERE IS NO SUCH ARGUMENT. <bjc>home is its own profile, and only ever has one. like the system. <gabber>i think this does not exist (yet) <bjc>profiles are just execution environments. system, home, package -i, they all use profiles as the mechanism to implement their behavior <Guest27>gabber it should exist if we really want the declarative to keep up with the imperative approach, cause because of this it's clearly inderior! <gabber>well.... you can use manifests to create ad-hoc profiles/environments <gabber>it is not exactly the same as (persisting) profiles, but we use it in a similar way <Guest27>bjc I wish home default profile coincided with the package default profile , I mean home profiles and package profiles coincided and using home reconf or package reconf were just different approaches to achieve the same resulta. Which is clealy not the case as things stand now <bjc>gabber: still all profiles under the hood. it's just how they're named and whether they persist <Guest27>bjc im ready to accept aystem being only one profile. <identity>just to be clear, there is no «guix package reconfigure», but there is «guix package --manifest=manifest.scm» <bjc>Guest27: home is also a single profile. maybe there'll be others some day, though i don't think that buys you much over existing solutions <bjc>between the system, home, and shells, you can do a lot of stuff <Guest27>identity does home reconfigure call the same procedures as guix package --manifest? <lilyp>home is package plus extras similar to system; package is just package <Guest27>bjc its very bad that home is one single profile. It makes no sense to have it and manage it separate from the package profiles <cdegroot>bjc: don't forget the awesomeness of `use guix` in .envrc :-) <cdegroot>Guest27: I would not see why it's bad. I mean, Guix Home also manages shepherd services, generate config files, etc. It's way more like Guix System. <cdegroot>Also would not see why you'd want multiple instances of that and how you'd manage. <bjc>it really is supposed to be the personal-user analog of guix system <cdegroot>(my setup: system has system level stuff, home has my essential desktop, and project directories have .envrcs with local Guix configs that activate when I or Emacs go into that directory) <Guest27>lilyp wouls you be so kind as to direct me towards more info abiut what that extra you say, is? <cdegroot>extra is managing running services and config files; exactly like guix system. <lilyp>stuff like the pipewire service for example, which creates some config files in your $HOME <lilyp>`guix package` is not meant to do that and folks were annoyed by it, so they started a command that allowed populaating $HOME <Guest27>I wanna underatand more about use cases of hoke vs defaultprofilewithpackage <Guest27>Library closing and will log off but will check the log at home in 30m <lilyp>basically, the difference between `guix package` and `guix home` is that `guix home` also manages such configuration files for you (as long as there is a service for that, but even if not, you can code around that with even more scheme), whereas `guix package` only provides some files and environment variables, the latter only after sourcing `$profile/etc/profile`. <lilyp>in other words, you still have to write your regular configuration files by hand or interact with whatever settings daemon your desktop environment has (though notably huge things like gnome have not been sufficiently mapped to home either) <ieure>Is anyone aware that CI is failing to evaluate the master jobset much more often than it's succeeding? <ieure>Only three of the last ten builds have succeeded, every failure has the same two errors: "Signals delivery fails constantly at GC #398" / "Signals delivery fails constantly" <Guest27>Guys I was trying to mount the dkms for bcachefs. So i out bcachefs-linux-module in the loadable modules part of the operating system declaration, trying to reconfigure and I get thr complaint that I must do something else with a hint to writing at the beginning something like (modules (gnu packages file-systems)) <Guest27>I do and It compains again about "have you forgotten a form"? <Guest27>Then I thought I sed the wrong syntax so I change the parenthesis into (gnu) (packages) (file-systems)) and I get another complaint that it cant find the module packages. I deduce this is probably more wrong than before <Guest27>But now Im back at (gnu packages file-system) with the complaint about having forgotten a form <ieure>Guest27, Plenty of not-guys in this channel, and "mount the dkms" doesn't really make any sense. Please paste your configuration and the error you're getting into a pastebin, and put the link in here. <Guest27>hint: Did you forget a `use-modules' form? <Guest27>I think it refers to (use-modules (gnu packages file-systems)) <ieure>Guest27, If you would like help, please follow my instructions. <ieure>Guest27, Why do you need to restart it to paste your configuration and the error message? <Guest27>ieure I dont want a graphical interface on my laptop so cannot use the pastebin <ieure>Guest27, 0x0.st lets you paste with just curl. <ieure>You need a browser to see the incantation to do it, but can use effectively anything. Probably even curl the page and read the output. <ieure>Guest27, Sorry, I am out of time and have to leave. But hopefully someone will be able to help you with the information I requested. <Guest27>Here i am again if you said something in the meantime I didnt get the message