PKZIP

ZXNet echo conference «code.zx»

From Stanislav Udin To All 19 October 2001

Hello everyone! And for what reasons is the subject still missing on Spec? The reason is that there is no one to write to or is it because the subject on Spec is not real? Or maybe there are others reasons? It seems that the sources of the written subject are in free circulation... Stanislav

From Kirill Frolov To Stanislav Udin 22 October 2001

Press RESET immediately, Stanislav! 21 Oct 01 20:57, Stanislav Udin wrote to WoWan Nickolaev: SU>>> And for what reasons is the subject still missing on Spec? No brains. In both senses. More literally, of course. :-( SU>>> is the subject not real on Spec? Really on Real... Only now you will have to write the subject practically from zero, that's why no one does it. Now, if only you could just take the source and fix under spec... But on the PC they are used to wasting memory - a megabyte here, a megabyte here... And the spec is only 64kb (without cans)! SU> What incentive is needed? Isn't it an incentive to write something that doesn't exist yet SU> for which people will say thank you. What incentive did the authors of the anzip have? Well, let’s say that Kondratiev is a natural cooler and he didn’t need any incentive. And the other authors were only engaged in fine-tuning the finished assembly code -- There may be more work, but overall everything is much simpler.

From Stanislav Yudin To All 28 August 2002

Hello everyone! I wanted to do a good deed, write PKZIP for Spectrum, but I failed :( A the idea was good... To begin with, so as not to bother myself with writing an interface, I thought of doing Packing a TR-DOS disk into a zipped TRD-pack is also a necessary program it would work out. With difficulty I found an official detailed description on the Internet PKZIP format, although it is available on www.pkware.com. At first I was happy - so there everything is simple. However, further acquaintance with the document plunged me into sadness - there The compression algorithm itself is not described and on this occasion it is said that this issue goes beyond the scope of the document and those who wish can use the literature indicated at the end of the document :( Who knows how much time I spent on the Internet with the purpose of obtaining a description of the algorithm used in PKZIP. In vain. Or rather There are descriptions of all the necessary algorithms, but this information is only suitable in order to understand what the algorithm is based on and this information is enough to unpack the archive, but not vice versa, or I'm so stupid that I can't move in. Apparently this is why the Spectrum PKUNZIP was born... Actually, PKZIP uses several methods. See what method The first file in the ZIP archive has been reaped by looking at the 9th byte of the file (counting from the first from the beginning of the file). If there is 0 there, then the file is either not compressed or the first filearchive is the header of the directory in which all files and other directories are located archive. I analyzed the results of the following programs: 1. PKZIP for DOS. 2.WinRAR 3. WinACE 4.Windows Commander Compressed files at all possible compression levels for the specified programs. It turned out that the output is files that in all cases use the same compression method - method No. 8. It turns out that the other methods Compressions are not interesting and can be ignored when writing PKZIP? Method number 8 is called Deflating and is a modified algorithm Lempel-Ziv (LZ-compression) with the size of a sliding dictionary (sliding dictionary) up to 32 KB and Huffman/Shannon-Fano encoding. I tried to search for PKZIP sources from other platforms. Yes, they exist, but for me There's zero benefit from them - the devil will break his legs, especially since there's nothing except ASMA Z80 I don't really know. You can easily write a PKZIP that will use the null Store method, then There is a simple way to put uncompressed files into an archive. But who needs it? By the way, it would be easy to pack the archives taking into account the directories DirSys. I don’t know about is-DOS, since I don’t know how everything is stored there, but I suspect that it would be possible too. In general, it’s somewhat strange that it’s still not available on your favorite platform such a useful program... P.S. Before I asked this question, I had no ideaabout the principles of data packaging. Stanislav

From Valerij Kozhevnikoff To Stanislav Yudin 3 September 2002

Hello, Stanislav! 28 Aug 02 23:53, Stanislav Yudin -> All: SY> I wanted to do a good deed, write PKZIP for the Spectrum, but I failed :( A SY> the idea was good... Hmm. It's a good thing. But maybe LZH under Isdos is better? Under SR/M, after all, there is. Somewhere there were even sources. WBR, Jason. /*e-mail: jason2000(scary dog)yandex.ru ICQ: 62235830*/ /np:/ *silence*