For a few hundred kilobyte file sure, the difference is like pocket change. For a larger one you’d choose the right tool for the job though, especially for things like a split archive or a database.
Username checks out! Also you’re absolutely right, just last month I was looking for the best compression algorithm/packages to archive a 70gb DB
What did you find?
I ended up with xz. According to this page it’s the one with the best compression ratio. It’s also the slowest but since it was one off I didn’t mind about it.
Nowadays it matters if you use a compression algorithm that can utilize multiple cores for packing/unpacking larger data. For a multiple GB archive that can be the difference between “I’ll grab a coffee until this is ready” or “I’ll go for lunch and hope it is done when I come back”
In that case, which file type would you recommend?
I personally prefer bzip2 - but it needs to be packed with pbzip, not the regular bzip to generate archives that can be extracted on multiple cores. Not a good option if you have to think about Windows users, though.
Ah I have to use Windows for work and that’s the source of most of my compression needs. Thanks for the info though, I’ll look into this
In before the .tar.gz/.tar.bz2 gang…
POSIX is on .pax.gz and .ustar.gz now, what are you doing?
hell yea
Why isn’t everyone using .7z ?
Because gzip and bz2 exists. 7z is almost always a plugin or addon, or extra application. While the first two work out of the box pretty much everywhere. It also depends on frequency of access, frequency of addendum, size, type of data, etc. If you have an archive that you have to add new files frequently, 7z is gonna start grating on you with the compression times. But it is Ok if you are going to extract very frequently from an archive that will never change. While gz and bz2 are overall the “good enough at every use case” format.
7z can be at least decompressed in macOS & FreeBSD out of the box.
On windows tar.bz/gz/xz unpacks to tar and then to actual files. As tar is a separate archive format
Windows having tar.gz support is great.
I have scripts for generating log bundles on user computers and sending to a share. tar.gz is great for compressing ~2.5GB text to send over VPN, and then I can open the .tar.gz direct from the network drive with minimal additional delay opening a 500MB text file inside.
I still prefer 7z for compression
For archiving/backupping *NIX files, tar.whatever still wins as it preserves permissions while 7z, zip and rar don’t
Oh, and while 7z is FOSS and supported out of the box on most Linux desktop OSes and on macOS, Windows users will complain they need to install stuff to open your zip. Somehow, tar.gz is supported out of the box on Linux, macOS, and yes Windows 10 and 11!
Because .7z was a pain in the ass back in the day while .rar just worked.
Xmodem, Ymodem, Zmodem.
Kermit
AT&FM1S11=35
Those were the days. For anyone under 40 see this for what we dealt with. https://support.usr.com/support/s-modem/s-modem-docs/usrv90.pdf The plug and play section is especially amusing these days.
V.42bis was the shit
I tried to look it up but google senseless and bing seems to stall when you type it. If I remember, AT&FM1 is return to factory settings option 1. What does the rest or all of it mean? Is S11 the dial speed
Yeah &F is factory default, M1 is speaker on only until connect, S11=35 is the dial speed (although we later learned that 50 ms is the minimum). Dial speed was important because we’d have Telemate on constant redial trying to get into the BBSes that were popular but were busy because they only had one or two phone lines.
Nice. Thank you. I’m proud I almost remembered that. I never used telemate but I did partner in hosting a small bbs with a friend. FarpointBBS. We had tradewars and the main dude even got it hooked up to the internet in the late days. As you know, way before the www took off.
Sooo many good memories.
M1 is mute true. Now it’s coming back!
punter
No love for hs/link?
I’ve got a 9600 baud modem! It’s impossible for POTS to ever be faster!!
Now we have so much bandwidth it doesn’t matter
Squints eyes
Now we just don’t care about even the slightest modicum of efficiency
In the early days of the internet, WinZip was a must have tool. My college had a fast internet connection. I say fast but I bet it was less than 1Mb shared between everyone. Way faster than the 33k modem I had at home.
I used my college connection to download so much and then took it home on floppy disks. For files larger than 1MB I’d use WinZip to split files up.
And then you get it all off the floppy’s only to realize 10% of it is corrupted
Oh yes. CRC Error was it?
And that’s when you learned about parity files…
When par2 came out that was huge for me. Didn’t use floppies anymore, but the ability to only download the required amount of parity blocks was great.
How about when peoples websites would put the sizes of linked images and files so you could estimate how long it would take to download a given image and such? Basically anything 30KB and above would have a size warning attached.
I used to use Opera with image loading disabled
On my phone I still block images over 64KB when on data
I’m old, aren’t I?
… yeah lol
still using 7z. less space, and easier to browse, since the operating system doesnt have to deal with all the files, easier for the cloud to tag. not caring about space makes the storage more expensive, even games are bigger now with little to none content.
pkz204g.exe will always hold a place in my heart
Just reserve your dislike for the ones still doing .bin, .img, and .cue.
Depends on what you’re doing. Dumps of multitrack CD media should always be bin+cue or a compressed version thereof, such as chd. DVDs and Blu-rays can dump as iso. There are also some extremely niche cases such as specific copy protection that require mdf+mds for a proper dump, but that won’t be something the average user ever encounters. Basically, those formats exist and are still used for a reason, whether you understand them or not.
I do reserve some hatred for people who dump PS1 games as iso, or who use ccd+img+sub for things where the subchannels have no valid usage.
1:1 copies of the bits on the disc is a valid option that some people prefer. Especially if you want to make your own physical disc or make compressed files encoded in a very specific way. It’s also the most reliable way to archive a disc for long-term storage.
sad .ace noises
That was the shit on the FXP scene
I still used uuencode/uudecode to transfert some file between terminal, a few weeks ago
I’m many cases there is some network level compression going on, particularly on higher speed LANs
.sit stuffit files