bwwilson@lion.waterloo.edu (Bruce Wilson) (03/29/89)
Hi, There have been a lot of people comparing the various compression programs, quoting the speed of *compression*. But what about the speed of un-compression? It seems to me that this is a more common operation for most people. I know that probably 90% of my time spent with archives is looking at them to see if they are useful, or not. I don't really care if the archive took the author 2 minutes or 20 minutes to create. Does anyone know if Zoo is just as fast as Zip or Arc in this process? (I think they are about the same but I haven't done any timing.) I understand Zip has various levels of compression that take more time to compress but do they also take more time to uncompress? (My guess is that they don't.) In trying to sort out the Arc-Zoo-Zip debate I've realized that there are four main factors (there are many more but I think these are the main ones) [quotes are from memory]: 1. COMPRESSION: "I just converted all my Arc's to Zip and I saved 2 meg. and that's nothing to scoff at..." 2. SPEED: "Zoo takes 2.5 minutes to archive these files whereas Zip only took 2 minutes..." 3. PORTABILITY: "Zoo works on all three machines I work with but Zip doesn't.." 4. DISTRIBUTION/FEES: "We need a FREELY DISTRIBUTABLE file archiving system..." I'm not sure it's possible to find a file archiving that will be the best of all the above so the decision should be based on a weighted combination. In other words we have to decide what the important features are before choosing an archiving system. I've read a lot of messages by various people supporting the various programs (all with good, valid points) but they don't get us any closer to a decision because we all have different priorities. Personally, I think (4) is the most important because of the nature of the Net and noone has can say they don't want to use an arc program because it costs. This is followed very closely by (2) because I would guess that most people on the net are using more than one system (I'm reading this group on a Unix machine but downloading for my PC). The next would be (1) because this saves space for archives sites (and me) making room for more stuff. The last is (2), not because it isn't important but rather I think it's the *least* important. If you accept my priority list (I don't expect you to) then it's fairly clear that Zoo wins. Maybe the algorithm for LHarc could be 'plugged-into' Zoo (and change the extension) to create a new program to satisfies the compression fanatics (then they could only gripe about speed). While I'm posting, I would definitely like to see something in the first article indicating a bit about the licensing of the program. IE- "this binary is public domain|free but copyrighted|requests a $10 donation|requires a $10 payment". Then *I* can decide if I will go to the effort of trying a Rolodex program that wants $100. Also, Rahul, you are doing a great job and I thank you for your efforts. Thank you for your time, bruce bruce Wilson | "what you don't spend you don't have to earn..." bwwilson@lion.waterloo.edu | from Blake (a film by Bill Mason)
Chris.Maidt@p8.f30.n147.z1.FIDONET.ORG (Chris Maidt) (04/01/89)
BW> Hi, BW> There have been a lot of people comparing the various compression BW> programs, quoting the speed of *compression*. But what about the BW> speed of un-compression? It seems to me that this is a more BW> common operation for most people. I know that probably 90% of my BW> time spent with archives is looking at them to see if they are BW> useful, or not. I don't really care if the archive took the author BW> 2 minutes or 20 minutes to create. Besides this being a test to see if this inter-domain stuff really works, I want to comment that for ME, speed and size are the most important parts of compression. I write my own stuff, and have to constantly update it and send it off again ... I want it arced NOW and I want to be able to send it in nothing flat. De-compression, on the otherhand ... I like care how long it takes to uncrunch ... although SEA's 6.01 takes the cake here unarcing whole tree structures in 10-15 minutes. -- Chris Maidt - via FidoNet node 1:147/10 UUCP: ...!uokmax!metnet!30.8!Chris.Maidt INTERNET: Chris.Maidt@p8.f30.n147.z1.FIDONET.ORG