ygrek / mldonkey

cross-platform multi-network p2p daemon
http://mldonkey.sourceforge.net/
Other
235 stars 43 forks source link

Wiki & forum will shut down on August 21, 2023 #90

Open spiralvoice opened 1 year ago

spiralvoice commented 1 year ago

Hi,

the MLDonkey website https://mldonkey.sourceforge.net/Main_Page, hosted at sourceforge.net and running on php 5.4, will be shut down on August 21, 2023 due to an PHP update: https://sourceforge.net/p/forge/documentation/Project%20Web%20Services/#upgrading-from-php-5-net-subdomains

Due to lack of time and interest I will not update the forum and wiki software. If anyone is interested and capable of doing this job please post in this issue to discuss the transfer of the project. Another idea would be to use Github discussions...

spiralvoice commented 1 year ago

Cross-posted on the MLDonkey forum: https://mldonkey.sourceforge.net/forums/viewtopic.php?t=5843

petchema commented 1 year ago

Hey Spiralvoice! Pango here I received the notification too. I haven't followed PHP progress in years and for the same reasons won't try to keep the forums afloat either. Too bad, lots of good memories attached to this site!

yolanjoy commented 1 year ago

Perhaps move/copy the sites content on here before shut down?

d235j commented 1 year ago

Is Sourceforge forcing this PHP update? The linked web page does not say they are.

spiralvoice commented 1 year ago

Is Sourceforge forcing this PHP update? The linked web page does not say they are.

Yes, quoting the email I received: This notice is to inform you that SourceForge will be upgrading PHP on project websites from PHP 5.4 to PHP 7.4 on August 21, 2023. This is part of an ongoing process to ensure all systems within our infrastructure are on current versions and will be followed with PHP 8 later this year.

carlonluca commented 1 year ago

Can you upload the databases somewhere? This way someone may be able to work on it also after the deadline.

malvinas2 commented 1 year ago

As an alternative, I suggest enabling the discussion here at Github. Since the number of questions and answers in recent years have been manageable anyway, this should be possible without any problems.

carlonluca commented 1 year ago

My proposal was more related to preserving the history than to future discussions. It would be useful to not lose the info currently available in the forums.

petchema commented 1 year ago

That was my idea too. But I looked around in the admin interface I have access to, found nothing that looked like a backup feature. Maybe I looked at the wrong place, maybe I didn't understand what I was looking at, but I saw nothing to extract forums data.

lulol commented 11 months ago

@petchema et all: I'm not familiar with Sourceforge admin tasks, but found this that may be relevant.

While it seems that restoring Sourceforge backups are not supported, creating them definitely are.

Backing up Your Data

Specifically the forum data: Project Data Export

The data export provides a zip file of all your project's metadata, forum, ticket, wiki, and blog content, so you can have a local copy of your own data - for safe keeping, data analysis, custom reporting, etc. The data for each tool is a JSON file, using the same format as our API for individual tickets, wiki pages, etc. You can generate an export from the Admin page on your project. Just choose "Export" from the left menu, check off the tools you want to export, and hit the "Export" button.

You can also request a data export using the Allura API. When complete, the file can be retrieved via scp USERNAME@web.sourceforge.net:/home/project-exports/PROJECT/FILENAME .

Creating and downloading backups of everything only should take a few minutes. Restoring them to Github on the other hand may need a bit of more research/work.

petchema commented 11 months ago

Okay, that requires a Sourceforge project admin account, not a phpBB admin account. From the look of it, project admins are "mlpango, ovidr, spiralvoice" (https://sourceforge.net/projects/mldonkey/) Sadly I have no idea what's mlpango account password, nor how to recover control over that account that I don't have used in years, if ever (I did not participate to the forums setup).

petchema commented 11 months ago

Update: I recovered my admin account, and exported 2 SQL dumps (phpBB and MediaWiki) using the myphpadmin interface, and tarballs of /home/project-web/mldonkey (websites) and /home/frs/project/mldonkey/mldonkey (downloads).

total 947364
-rw-rw-r-- 1 xxxxxxxx xxxxxxxx 744260035 Aug  5 02:57 frs-mldonkey.tgz
-rw-rw-r-- 1 xxxxxxxx xxxxxxxx  51808978 Aug  5 02:09 m156414_phpbb.sql
-rw-rw-r-- 1 xxxxxxxx xxxxxxxx  66825912 Aug  5 02:24 m156414_wiki.sql
-rw-rw-r-- 1 xxxxxxxx xxxxxxxx 107193186 Aug  5 02:44 project-web-mldonkey.tgz
-rw-rw-r-- 1 xxxxxxxx xxxxxxxx       245 Aug  5 03:10 SHA1SUMS
ᐅ cat SHA1SUMS 
5025fc9ff14eea5dc917f777fac882cca2d4fd24  m156414_phpbb.sql
69454aa58f8f74f176cab9563bfe67dd474ed0a7  m156414_wiki.sql
d6c7c67a9f5061639798dc928bc4817b17d0c769  project-web-mldonkey.tgz
8857639d31b856f0326c2f2f49ee04181276a124  frs-mldonkey.tgz

Can you think of anything else that should be saved? Both SQL dumps contain sensitive information, so cannot be made public unexpurged

lulol commented 11 months ago

Awesome!

Since the mailing-list is not hosted at Sourceforge I guess that is all the public data from the Official Project at https://mldonkey.sourceforge.net that I can think of. I hope, I'm not familiar with Sourceforge admin tasks.


The other old related stuff from the DownloadLinks page, hosted at Sourceforge is not directly from MLDonkey but eDonkey2000 and Overnet.

https://ed2k-tools.sourceforge.net/index.shtml http://mldkgui.sourceforge.net/ http://mldiv.sourceforge.net/

Anyway, I guess that the site databases won't be removed right away by Sourceforge after August 21, 2023, even if the sites will be disabled for not being updated.

petchema commented 11 months ago

Tried to scrap the whole website, without logging so it only comprises public pages, that have been public for years now. It's very redundant because the same thread page is being fetched using its thread id (t) but also for each of its post ids (p), the good side is that it's nicely browsable. 7z compresses that very efficiently, sadly github doesn't accept .7z extensions, so I uploaded it elsewhere. Hopefully it'll work for whoever is interested: https://mega.nz/file/miIAzZbD#6hMjnuHNK5nBUnk62CbHpb53hM7nxD0abOODzYNnKSk