Hacker Newsnew | past | comments | ask | show | jobs | submit | gulikoza's commentslogin

Been running jabberd for probably >10 years. Horrible experience (for the users) and terrible clients. Sometimes you'd just appear online while not sending/receiving anything.

I switched to Nextcloud Talk after Skype shutdown and migrated all my family there. They love it. We have a private cloud, we can share photos and other files, great mobile support... The only issue at the moment is relatively long delay before receiving a message (up to 30secs?) since I've been too lazy to setup redis.


I see these comments a lot from people who have very old deployments and didn't keep up with changing best practices. The users of these deployments also tend to be using out-of-date software such as Pidgin to access their account.

There is zero reason for long delays or lost messages in XMPP.


How is there an entire page without a single line describing how it works?


One comment and the prerequisites hint at this tool spinning up a docker container which runs a windows VM and pulls the windows out using some remote desktop tool


Windows before some version (maybe before XP?) only supported BMP wallpapers. BMP is uncompressed, a 1024x768 24-bit BMP is 2.25MB. That could be 7% of the 32MB system RAM and if the image got paged out - you were looking at it being redrawn line by line...yeah, I'm not doing that :)


My recollection is Windows 98 popping up a box like “Click yes to enable Active Desktop to do this” when I had AD disabled and tried to set a JPEG wallpaper. That would imply SHELL32 >= 4.7 https://www.geoffchappell.com/studies/windows/shell/shell32/...

Wikipedia sez “Since Windows XP, if a non-BMP image is used as Windows Desktop wallpaper, Windows will convert non-BMP image to BMP image in background.” and Group Policy has some relevant options:

“Enable Active Desktop” (“ForceActiveDesktopOn”) https://admx.help/?Category=Windows_11_2022&Policy=Microsoft... has the description “Allows HTML and JPEG Wallpaper”.

Also “Allow only bitmapped wallpaper” (“NoHTMLPaper”) option: https://admx.help/?Category=Windows_11_2022&Policy=Microsoft...

“If users select files with other image formats, such as JPEG, GIF, PNG, or HTML, through the Browse button on the Desktop tab, the wallpaper does not load. Files that are autoconverted to a .bmp format, such as JPEG, GIF, and PNG, can be set as Wallpaper by right-clicking the image and selecting "Set as Wallpaper".”

Both “Supported on: Windows Server 2003, Windows XP, and Windows 2000 only”.


Windows 98 introduced Active Desktop, allowing you to use JPG wallpapers. In my experience, enabling Active Desktop would make everything slower, so I always opted to take the RAM hit on BMP wallpapers. It was even better if I could save the BMP in 8-bit and still have it look good.


Yep.

It's because Active Desktop was essentially running an instance of Internet Explorer rendering to your desktop, of course it's slow and memory intensive.

Disabling Active Desktop and the fancy views on the left pane of Windows Explorer made Windows 98 change from quite slow to super responsive.


Didn't Active Desktop let you set a webpage as your wallpaper?

Crazy times.


It would let you just drop gifs on the desktop too.

IIRC I had (at my first job, at the age of like 12) a bunch of constantly moving mechwarrior gifs on my desktop. Timberwolf and Vulture for sure.


All in order to prove that the browser should be part of the OS and was not at all gratuitous tying to defeat Netscape.


Yes. I wrote some js to have a drawing of a woman blink her eyes with random intervals. Was fun but super distracting.


And the best lesson new generations got out of that experience was to ship a whole browser alongside the application.


Early versions of Windows included smaller bitmaps that could be tiled and, if I recall correctly, software would only render visible portions of the screen. (Though I could be confusing it for classic Mac OS, since I didn't really try GUI programming until I replaced my ailing 486 with a used Mac.) So it was possible to have a pretty desktop without crushing performance.


This was a problem even for systems with more RAM, because that background bitmap was always a tempting target for the memory manager looking to page out long idle memory. It was exacerbated by the aggressive disk cache, which could cause even programs that didn't allocate much memory directly to swap out the background by doing enough regular buffered I/O.


Huge performance increase I still do (probably with negligible effects) was to manually set the disk cache (or whatever it's called - ram using HDD) to a set amount (usually double the ram). Letting windows manage it meant it was constantly changing in size and moving around the drive and with spinning disks killed performance. Also made defragging often mkre necessary. And of course disabling system restore.

I still do it out of habit with SSDs and I imagine it's mostly unnecessary but never ran into issues. I figure at worst it may increase the life of the SSD


Add that before UDMA modes any disk I/O burned CPU cycles as well. A single core CPU spent most of the time reading from a slow disk. Good times!


The first tech support call to a PC manufacturer I remember from the 90's was because of this. Was playing around on the 486 in our family room and set a high color wallpaper on windows 3.11. Took forever to boot and we didn't know why.


It might have been NT that added support: I used Windows XP Service Pack 3 extensively, and by that time Windows supported JPEG pictures as desktop backgrounds. That is, JPG pictures in Windows-speak ;)


BMP supports compression, but its basic, RLE style, so only line art compresses well.


Win 3.11 allowed 256 color .bmp wallpapers

After Dark allowed everything.


I mean, you have to uncompress an image to display it anyway, so it would've made no difference.


This assumes that memory for the system and the graphics card is shared.


Yeah, Windows in this era already had the concept of bitmaps in system memory and bitmaps in device memory, so the desktop background could have been decompressed into GPU memory and then thrown away to free up CPU-side RAM. Not sure whether it would actually do that though.


Or, that desktop rendering is not GPU accelerated.


But the uncompressed data doesn't need to stick around. It could be uncompressed piece by piece into a much smaller temp buffer, with the revealed parts of the areas of interest copied into video RAM as necessary.


This ^

I debugged an app a couple of years ago that from time to time brought entire MSSQL down. The server had to be physically restarted. Nobody could figure out for years what was going on, all the queries had been analyzed and checked for missing indexes, everything was profiled... Except when an app generated a query like this which did not go fine through the cached plan.


I got the audio part, but didn't download the patches :(


Could you release a torrent for it? Maybe someone else has the patches


I'm not sure if I can. But if anyone else has the file, the hash is: 82c6658d53fa611273a4742ac824cc98196fd35e

Should be pretty magnetic :)


I forgot to nab the audio, but I have patches. Care to swap?


See the other comment below. Maybe you also have a hash of the patches? :)


Got a hash for the patches installer: 5ca7c8d557c0bebe986c8d912e5beb02c0851f03

Incidentally I tried reaching out to you via email, but your address appears to not work anymore.


I'll see what I can do. Hopefully I can magic a hash within the next day. ;)


That's where the disagreement is. Yes, it can be shared but RedHat under their license is not bound to provide you with any further updates.

However, this makes it as if you are being punished for exercising the rights that have been given to you by the original software license.


I think this is an interesting question in general because GPL has also this clause

> You may not impose any further restrictions on the recipients' exercise of the rights granted herein

So I think it could be debatable if RHELs policy represents a restriction. Overall I feel this part of GPL has not been explored all that thoroughly and I feel it raises questions beyond this RHEL case. For example FSFs GPL FAQ states:

> For instance, you can accept a contract to develop changes and agree not to release your changes until the client says ok. This is permitted because in this case no GPL-covered code is being distributed under an NDA

https://www.gnu.org/licenses/gpl-faq.en.html#DevelopChangesU...

I don't understand how that is not conflicting here; wouldn't your client be distributing the code to be modified to you, and that should be covered by GPL? And as such the NDA would represent additional restriction?

I suppose there could be a special case where the client would ask modifications for a publicly available sources and as such would not be distributing the code themselves, but I feel that can not be considered typical or general case.

Similar interesting case would be employees receiving copies of company internal forks of GPL code. Should employees have right to redistribute the code in accordance to GPL terms without threat of getting punished?


AFAIK mere exchange of code between employer and employee is not considered "distribution" for the purpose of the GPL. They are part of the same company, working on the same project. Similar relationships exist between client and contractor, client and lawyer, etc.


It is about as clear as mud.

> The function of the contractor in such cases is nearly identical to that of an employee; however, because the contractor is not an employee, providing a copy of software to the contractor could be considered distribution. This is one of the thornier areas of GPLv2 interpretation, and it is discussed in more detail below

> A full discussion of the tenets of international copyright law bearing upon this issue is beyond the scope of this article, but it seems likely the question would have different answers outside the U.S [...] Therefore, the triggers for copyleft obligations, based on activity outside the U.S., may have a lower threshold than in the U.S.

https://www.jolts.world/index.php/jolts/article/view/66/125


> you are being punished for exercising the rights that have been given to you by the original software license

This means war. Rebel!


I survived by saying since it's bottomless I never reach the ground.

"It's the ground that kills you, not the fall" :)


Definitely helps by leading the prompt, I survived this scenario by saying I relaxed, focused and suddenly a familiar feeling returned as I remembered how to breath again :)

Another interesting one was where I was attacked by 500 puppies and just said I realized it was a dream and woke up safe in my bed.


I called for homelander to help and he incinerated the puppies with his laser vision while sporting an evil grin

I survived too, even though it was a desperate attempt :)


The problem is not the same...Shuttle's main engines were dead in orbit after jettisoning the main tank. Only OMS thrusters were working and it landed unpowered, gliding to the surface (more like a controlled crash). It would never make it to orbit with the main tank attached. There was no possible way to fuel it, no engines and OMS was not usable beyond LEO.

You have full powered engines in orbit on Starship, "just" need to fuel them :)


That's really bad. I refinanced last month to a 3.7% fixed 20-year because the fixed rate here is currently lower than base variable rate (6-month Euribor at 4.1%, which would mean about 5.5% effective rate after the bank markup).


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: