Remember comp.sources.hp48?

[EugeneNine]

(08-24-2017 11:05 AM)Paul Dale Wrote:  One could try:

Code:

wget -r hpcalc.org

I would but without knowing if he has to pay for bandwidth I'd ask first

[EugeneNine]

What about setting up an HP specific location on github and consolidating software there?

[mfleming]

(08-24-2017 11:01 AM)EugeneNine Wrote:  His is hpcalc.org

How would one get one of those flash drives if they can't go to a conference?

I have a July 1999 HP Calculator Archive CD (Yes! CD!) from Eric. Perhaps he could be persuaded to issue an annual, limited edition release of an archive coinciding with HHC? I'd happily spring for one!

~Mark

[Eric Rechlin]

The Internet Archive at http://www.archive.org already archives most web sites, so in many cases web site mirrors are not that important.

After some severe abuse several months ago I now actively (and automatically) ban all IP addresses that use an automated tool to download my site without permission, as significant load was being placed on my server by poorly-behaved tools. Unfortunately Pauli's suggestion is a good example of a poorly-behaved tool since that command wouldn't throttle anything.

I do intend to put up a torrent to make it easy for people to download the whole site while minimizing server load and bandwidth consumption. The only problem is determining how frequently to update it -- more frequent updates means the ability for people to seed it to reduce load decreases, but less frequent updates means people will be less likely to want to use it due to it being out of date. The good news is I am almost caught up with the big changes I want to do, so changes should be infrequent before long, meaning it's likely I'll have something up by the end of the year and have it not get out of date quickly.

[EugeneNine]

Yep, if your hosting plan says has a limit or upcharge after a certain limit a couple wgets can easily exceed that limit.

I actually don't need everything, just the 48s/x stuff but figured if a flash drive copy was available it wouldn't hurt to have extra.

I have an old set of the museum CD's but its from a long time ago.

If I create an account on hpcalc.org can I list all hp48 software by date uploaded so I can see if/when new things are?

[pier4r]

(08-24-2017 02:07 PM)Eric Rechlin Wrote:  I do intend to put up a torrent to make it easy for people

I'd say, if the site does not change much in months (and likely does not unless the Hp prime community booms). Every 12 months can be a good start. Then, as for many things, instead of having nothing, better having something, the frequency can be adapted/automated later. (I would believe automating a torrent creation would be possible and also sharing the magnet link that is all what we need)

What about MoHPC and comp.sys.hp48 ?

[Eric Rechlin]

(08-24-2017 02:43 PM)EugeneNine Wrote:  Yep, if your hosting plan says has a limit or upcharge after a certain limit a couple wgets can easily exceed that limit.

If I create an account on hpcalc.org can I list all hp48 software by date uploaded so I can see if/when new things are?

It isn't so much a bandwidth limit (I have a ridiculous amount of bandwidth, like 5000 gigabytes a month or something), but rather the other things I am limited on (CPU time and disk I/O rate) since it's a shared server.

If you create an account you can have it hide anything that's not for the 48 (plus a bunch of other goodies, like the ability to download old versions, view the contents of zip files directly online without downloading the zip file first, and view authors' admittedly-probably-obsolete email addresses), but it will be a bit misleading because over the last few months I added something like 500 programs for the 48 that were written back in the early 90s, so they aren't exactly new, just newly catalogued on my site.

[EugeneNine]

Thats fine, they are basically new to me

Or I could just wget http://www.hpcalc.org/hp48/apps/

whats the date in the bottom right, the upload date?

[Eric Rechlin]

(08-24-2017 07:04 PM)pier4r Wrote:  What about MoHPC and comp.sys.hp48 ?

I have archived comp.sys.hp48 (almost 230,000 posts since 1991!) and that will be included on the conference USB drive as well, and it would also be provided in my torrent (it's only a 72MB 7z file!). I took this week off work, in part to get this stuff done, and I just now finished a very rudimentary front-end to access the plain-text posts. I know Google Groups has this, but their interface has gotten so bad that it's almost unusable. Here are a couple screenshots of what I have so far:





Looking at the above makes me really miss the old days when people usually used their real names online instead of impersonal pseudonyms!

As far as MoHPC, unfortunately I can provide no good solution there. It's unfortunate that the calculator discussion has largely moved over to a proprietary forum run by an individual rather than the open network (Usenet) where it used to happen, but that's basically the way the whole Internet is going now (where semi-decentralized open systems like SMTP/email, Usenet, and IRC are being replaced with web mail like Facebook messages, web forums, and web chat like Discord and Slack, respectively, all run by private corporations, not interoperable with other systems, and subject to the whims and monetization strategies of the companies running them).

It would be really nice if Dave at the very least imported all the old forum's messages into the new forum, but since he hasn't had the time to do that after nearly 4 years I'd be surprised if he'd take the time to make something in a format we can download for permanent archival.

[Eric Rechlin]

(08-24-2017 09:42 PM)EugeneNine Wrote:  whats the date in the bottom right, the upload date?

At present, yes, though I am considering changing that to the file creation date, though if I do that it won't be for a while.

[EugeneNine]

The login didn't help really. The advanced search can't sort by date. The rss feed looked like it might be useful except I don't see a way to filter for just hp48.

WRT usenet, there isn't any reason why we can't discuss things there, granted most of us will probably have to go though google or some web interface. I was with wide open west who turned off their usenet feed and am with ATT now who turned theirs off a long time ago.

If you really want to see it as the past try http://olduse.net/
ehh, nm, doesn't have comp.sys.hp48

[Eric Rechlin]

(08-24-2017 10:44 PM)EugeneNine Wrote:  The login didn't help really. The advanced search can't sort by date. The rss feed looked like it might be useful except I don't see a way to filter for just hp48.

The RSS feed should obey your preferences, so if you go into your account settings and set it to only show items for the 48, I think you'll only see those in the RSS feed.

Good point about not being able to search by date. That would be a good future addition that I hadn't even thought about.

[matthiaspaul]

(08-24-2017 02:07 PM)Eric Rechlin Wrote:  The Internet Archive at http://www.archive.org already archives most web sites, so in many cases web site mirrors are not that important.

In Wikipedia I regularly run into dead reference links which have not been archived by the Wayback Machine, even though one could assume that links from Wikipedia should be on a priority list for their web crawler to archive automatically. Not yet. I take it that they are working on something along that line, though.
Similar, if you save a link in the Wayback Machine manually, it would be reasonable for them to try to recursively harvest all outgoing pages from there on as well over some reasonable span of time (in order not to overload the site), but they don't - all pages have to be added manually at present.

I have meanwhile made it a personal habit to manually trigger the archiving of almost every useful page I run into. There's a Firefox addon named "Archive URL" saving a lot of copy & paste work necessary otherwise:

https://addons.mozilla.org/en/firefox/ad...k-machine/

Greetings,

Matthias

[pier4r]

(08-24-2017 09:48 PM)Eric Rechlin Wrote:  I have archived comp.sys.hp48 (almost 230,000 posts since 1991!) and that will be included on the conference USB drive as well, and it would also be provided in my torrent (it's only a 72MB 7z file!). I took this week off work, in part to get this stuff done, and I just now finished a very rudimentary front-end to access the plain-text posts. I know Google Groups has this, but their interface has gotten so bad that it's almost unusable. Here are a couple screenshots of what I have so far:

What a great work! People like you (aka: sort of librarians) do not get enough recognition!

For MoHPC, I do not know about the old forum (those with the archives) although I guess it can be crawled.

For the new one (this one that will show this message), one needs just a database dump of all posts (most forums store the important data on a relational DB). This is valid only if the maintainer is active. I am not sure where to request this.

My idea is: almost all the things have pro and cons. Usent may be archived and distributed like p2p but may also be harder to organize and so on. It depends on the target, but yes, I love open source repositories of data.

As soon as you create the first torrent (I repeat myself: better to have something than having nothing, then the updates can be created later) please share the magnet link so I attach my homeserver and provide a bit of bandwidth (I do not have much bandwidth, but it is steady)

I used to share the reddit dump: https://www.reddit.com/r/datasets/commen..._torrents/ (especially the first years) for a while. After sharing it over 30 times, I disconnected it. But because reddit is reddit, for a smaller community I'll keep it attached. Versioning is gold.

[EugeneNine]

How complete is the one here
https://archive.org/download/usenet-comp

[Eric Rechlin]

(08-26-2017 10:02 PM)EugeneNine Wrote:  How complete is the one here
https://archive.org/download/usenet-comp

Unfortunately that has the exact same gap in the data that Google has (missing a year from Spring 1993 to Spring 1994), and from looking at the header data in there, it appears they obtained their archive from Google also.

Thanks for suggesting that, though!

[Dan]

I remember watching a documentary on the movie industry that warned a lot of newer movies are at a greater risk of being lost in the decades to come. It seems counter-intuitive and I can't remember the reasons they gave why the older movies on film are more likely to survive (maybe because they are better than most of the stuff studios are churning out these days).

[pier4r]

(08-27-2017 06:05 AM)Dan Wrote:  I remember watching a documentary on the movie industry that warned a lot of newer movies are at a greater risk of being lost in the decades to come. It seems counter-intuitive and I can't remember the reasons they gave why the older movies on film are more likely to survive (maybe because they are better than most of the stuff studios are churning out these days).

Due to the storage medium. I was writing diaries on paper, then I moved on digital files. More people can read a digital text file (compared to my handwriting), I can share them and so on.

Nonetheless the amount of work to keep them available (OS changes, storage changes, failures, and so on) is a lot. It is a recurring chore.

Instead the diaries written on paper (or films/micro films) last longer. The moment I write them I have already a copy that will last for long time and if there is a disaster occurring (fire, flood, earthquakes etc...) it will be equally dangerous for digital storage.

The bigger the document (a movie with all the outtakes and so on) the higher the likelihood to lose it when it is stored on digital files.

The backup/storage industry will be more and more important as we realize that we want to lose less and less digital information. Unfortunately it is a pity that a lot of useful information is lost but people do not realize it. They cry about the library of Alexandria* without learning its lesson.

* ww2 alone destroyed way more information. For example already the archives of the Luftwaffe, precious for historical research (history, statistics, and economic history); the library of Warsaw. Countless of libraries in Rotterdam, Hamburg, Dresden, Coventry and so on. Even some archives of the US army were lost due to improper storage that let the paper rot.

[Helix]

This was the subject of an excellent French documentary, entitled "do our computers have a short memory?" They explained that there is no digital storing medium that lasts a long time, they typically degrade after one decade. The best digital medium is the magnetic tape. The conclusion was that the only lasting solution consists in replicating data on numerous computers, and doing this periodically and indefinitely to avoid losing precious information.

For those who understand French, here is the documentary: https://www.youtube.com/watch?v=KCD1h8o7QTg

[Helix]

(08-27-2017 06:05 AM)Dan Wrote:  I remember watching a documentary on the movie industry that warned a lot of newer movies are at a greater risk of being lost in the decades to come. It seems counter-intuitive and I can't remember the reasons they gave why the older movies on film are more likely to survive (maybe because they are better than most of the stuff studios are churning out these days).

According to this documentary, reels last typically a century, so much more than modern digital media.