The origins of LHA

[Leffmann]

I thought this was interesting, the origins and history of how LHA came to be, by Haruhiko Okumura:

http://oku.edu.mie-u.ac.jp/~okumura/...n/history.html

[KONEY]

utterly interesting, thanks! So it all began in Japan...

[StingRay]

Very interesting indeed, thanks for sharing.

[demoniac]

Insightful read. Didn't realize PKZIP was similar.

[jotd]

I recently rebuilt it froml the source on windows, works but creates files with userid 0, so not readable on Linux without being root... bummer. I'd really like a working windows version.

[Carlos Ace]

Originally Posted by Leffmann

I thought this was interesting, the origins and history of how LHA came to be, by Haruhiko Okumura:

http://oku.edu.mie-u.ac.jp/~okumura/...n/history.html

Cool info.
Thanx.

[kolla]

But how did lha become de-facto standard for Amiga?

[redblade]

Originally Posted by kolla

But how did lha become de-facto standard for Amiga?

Good question. Arc and Zoo were released first I think (I saw it as a option on Diskmaster 1.3) But Lzx (96)was the first archiver I used on the Amiga (pkzip on Dos in 95)

[hth313]

Zoo was quite popular in the early days. I think there was something in between like lharc(?), but then lha came around and it made smaller archives than anything else. I think it was also faster.

I think the reason it became the de-facto standard was that it was better than the alternatives and everyone just switched. This is as far as I remember it some 25 years later.

Whenever something new came around, we Amiga people were not afraid to try it. There was no "we have done it like that for 100 years" thinking in that community.

[Leffmann]

Originally Posted by jotd

I recently rebuilt it froml the source on windows, works but creates files with userid 0, so not readable on Linux without being root... bummer. I'd really like a working windows version.

Wouldn't the change have to be in the Linux version though, so that when it finds an invalid user ID, it replaces it with that of the current user?

Originally Posted by kolla

But how did lha become de-facto standard for Amiga?

Would be interesting to hear Stefan Boberg comment on that.

Originally Posted by hth313

There was no "we have done it like that for 100 years" thinking in that community.

There's a little bit of that these days though. LHA is still overwhelmingly preferred, even with LZX clearly being the better archiver with a wide lead in all the important metrics.

There's something friendly and familiar in the LHA name I guess.

[phx]

Originally Posted by Leffmann

There's a little bit of that these days though. LHA is still overwhelmingly preferred, even with LZX clearly being the better archiver with a wide lead in all the important metrics.

There's something friendly and familiar in the LHA name I guess.

I guess that LHA is just *the* standard archiver on the Amiga. You can expect that every Amiga user has it installed. Just like ZIP is the standard in the Windows world, although there are much better archivers.

So what made it standard? It is older than LZX. And LZX was not free for a long time, which limited its distribution. Also LHA was used on Fish disks and on Aminet.

Today I would still prefer to release software in LHA archives.

[daxb]

The main reason for LHA is IMO that it is used for Aminet archives. I think LHA is still fine today, although it is not the best archiver. Small packed size isn't that important like 25 years ago. Nowadays, in my opinion fast decompress time is the main argument against LHA.

[Leffmann]

Originally Posted by daxb

The main reason for LHA is IMO that it is used for Aminet archives. I think LHA is still fine today, although it is not the best archiver. Small packed size isn't that important like 25 years ago. Nowadays, in my opinion fast decompress time is the main argument against LHA.

Some numbers on this, testing with the contents of the Workbench 3.1 diskette on an A1200 with fastmem:

LZX: 45 seconds to compress to 370 KB, 10 seconds to decompress
LHA: 80 seconds to compress to 435 KB, 16 seconds to decompress

[hth313]

I want to add that back in the days, I think there were some patent or IP restriction that also caused issues using one or more of the alternatives.

Lha was free, did a better job than the other free alternatives and people switched to it and I suppose it just reached critical mass.

[jotd]

Leffman the Amiga version creates correct archives. I had fully automated whdload slave distribution from building to lha packing & sending mail to release, but it used the windows lha and Bert had to "sudo" to be able to extract them (and repack them...)

[daxb]

Originally Posted by Leffmann

Some numbers on this, testing with the contents of the Workbench 3.1 diskette on an A1200 with fastmem:

LZX: 45 seconds to compress to 370 KB, 10 seconds to decompress
LHA: 80 seconds to compress to 435 KB, 16 seconds to decompress

I get the same size here but...

Everything in and from/to RAM: on A1200 040/40 32MB:
LZX: 8.26 seconds to compress to 370 KB, 5.6 seconds to decompress
LHA: 11.32 seconds to compress to 435 KB, 5.8 seconds to decompress

Command lines for compressing:
lzx -bo1024 -Qf -r -R -q -2 af <target> <source>
lha -q -r -Z -Qq a <target> <source>

[demoniac]

I think another reason why LZX wasn't more popular is that the authors only made the compressor on the Amiga side and the DOS version never arrived.

It's interesting to read that MS cabinet files uses LZX algorithm.

[jotd]

The author of LZX got hired by Microsoft at some point.

[demoniac]

Originally Posted by jotd

The author of LZX got hired by Microsoft at some point.

Yes, it seems directly after he finished college.

[Hewitson]

The one thing I always wondered was why the archives were called .lzh on PC and .lha on Amiga.