everything wrong with free software

 "obedience breeds foolishness"

### december-update-regarding-mlsd *originally posted:* dec 2021 december was a truly great month for the more libre software directory-- every (initial) link to be added was added, every url is online and you can find the result of that here: => mlsd.html mlsd with that work complete, already the urls are being separated into pages by letter: (there will be a page for 0-9 as well) => mlsd-c.html mlsd-c => mlsd-g.html mlsd-g => mlsd-p.html mlsd-p => mlsd-r.html mlsd-r as urls are moved to pages, they get removed from the main page (which had well over 8,000 entries at its peak) and the main (mlsd) page will ultimately serve the purpose of an index for the directory. you can see how this fits the overall plan on the ewwfs index page: ``` [x] database copied and processed [x] initial github purge [x] all initial urls added... (9,000+ entries) [ processing ] titles derived from projects and split into pages by first letter [ ] additional data ``` to anyone who spots a typo on one of the new pages: dont worry, i know about them. the first priority is to move the urls. two initial quality checks are done, first in the code and then in the output. the first check makes the second check easier, and the second check makes the first check more complete. after that, the pages are left as-is for now, and further fixes (however few will be needed) can be made casually. basically, the pages go up in a "pretty good" state and making each page perfect is put off for later. getting the information from the single (enormous) page into something more manageable is important-- plus, as each page goes up it does not take on every single url for that letter. rather, what letter a url belongs to is based on its WEBSITE TITLE, not (primarily) the url itself. this means that more than 8,000 webpages had to be downloaded just to get the title (theyre cached, they can be used for other purposes as well). ### this is an important part of the forking process, as i didnt want to get any information from the fsd except the urls, which (importantly) are not all used. the only information taken from the fsd is the url to each project-- NOT the name. the name is derived from the website that url goes to. in a situation where that provides no useful information, the title can be derived from the url itself. that has not been done yet, so each page is technically incomplete. but thats alright, since the first phase was to get all the urls into the main "mlsd" index, and all the urls that are NOT separated into pages are there. basically, while information is sorted into new pages-- the public still has access to the unsorted data. as the data is tagged, named and sorted, the unsorted data grows smaller (and loads faster) until hopefully no unsorted data remains. but some urls will simply go nowhere, appending an internet archive link wont even change this sometimes, and those urls will likely get dumped at some point. the goal of the mlsd is not only to fork the fsd, but to add new entries. once the pages are sorted, it will be a more manageable system. everything was designed to streamline the fsd and make it easier for a single person to edit (and to make it easier to fork) and just because a project isnt listed doesnt mean it couldnt be at some point. for example, if it moves from github. getting the name of the project based on the title tag has been very useful-- for example, some of the code.google.com urls actually listed github information instead when loaded, due to redirects. many of these were removed immediately as a result. of course, its not a perfect system either. many urls contain "lint" that could be removed later on. its not a high priority, but it is a goal to remove some of this extra noise (to make the list nicer and easier to look through). because the list is ultimately based on gemtext, it is very easy to parse and clean up (and edit). i love the idea of wikis because html has gotten far too bloated. but these days, even wikis are so bloated that im very happy to edit with gemtext instead. the fact that it is so lightweight on features is its best feature. it makes a project like this far easier to maintain. i was not planning to have this phase online before 2022, so having 4 letters online is really exciting. note that c, g, p, r (the 4 presently online) are also the largest pages (at least by the initial count-- those paying the closest attention may notice that mlsd-r was the first, but is no longer the largest page) so each page that is added is actually less work than the previous one. as already stated, the priority right now is to get the entries named and moved from the main mlsd page to the individual (letter) pages. this makes the data much more manageable, whether to look through it or to edit it, which makes it a good priority. as implied by the status indicators, adding names is merely the second feature of the listing-- the first goal is urls (now complete) and the second is naming, with the third being "additional information". rather than even try to derive this from the fsd, the goal is to get information on this from sources that are more up to date (and closer to the projects themselves). already with the scheme for deriving names, this has provided better, more up-to-date information on where projects are located (if theyre even online) than the fsd has. to me, thats an important upgrade and technically this information could help clue people into editing the fsd to changes they are not yet aware of. the primary goal of the project is still to show people what projects are the least problematic in terms of a corporate overthrow of free software. its not a perfect list, its not the largest free software listing, but it avoids most projects that are relying on the good will of microsoft (very clarice starling of them, though between microsoft and hannibal lecter im pretty sure i prefer dr. lecter) and it gives the backstabbers who joined a years-long coup to overthrow free software on behalf of open source the very justice they demanded-- albeit in the like of a judge who asks a murderer hes about to convict what "the killers" he made up should get as a sentence... and then grants it to the actual culprit. i do not want to (as much as i can possibly avoid it) use software written by people who defrauded the world and tried to kill richard stallman out of petty, narcissistic selfishness. those authors do not fight for freedom, they fight against it, and if i wanted to use software written by the worst scoundrels and frauds i would do everything possible to promote github. ### a "quick" note about that: the grapevine (actually just regular mailing lists, but its more fun to say the grapevine) says that emacs authors want to rely on sourcehut more-- an fsf-hosted sourcehut at that. thats really good if its true, because emacs has drifted too far towards microsoft github, but im sceptical. how roy will probably spin this is by implying that emacs authors care about the whole github thing. some absolutely do, the question is how many. my guess is that emacs (and especially extensions for it) will continue to drift towards redmonds clutches even while a smidgeon of the whole thing moves to sourcehut hypothetically. and i do mean hypothetically. the fsf has made moves like this before, the most prominent example being savannah itself. and thats the point-- savannah, for all its lack of modern features and the sort of hate it attracts from github-huffing types is relatively lightweight and low-bullshit. im definitely not saying its a FANTASTIC design, more like a lesser evil. i dont (yet) have anything against the design of sourcehut (whatever its written in, i probably wouldnt care for too much) though the trend towards bloat is too common, yet hopefully sourcehut will learn from gitlab to avoid becoming too silly in an effort to be the opposite of savannah in every possible way (including the its good points). above all, im not going to get too excited about what i consider a rumour. i know who is working on it, i imagine hes not bullshitting entirely, but it will come down to people actually using it and thats the point where my scepticism comes from. in other words, its not his fault if sourcehut is another flash in the pan upgrade at the fsf. i blame the organisation for its mediocrity and self-defeating priorities. theyre welcome to put the crack down anytime and get on with the real business of fighting for freedom; but until i see it, ive been down that road a few times before-- dont expect me to believe it. the last person (as far as i can tell) to push for an upgrade like this was a stallman backstabber, at least this time around (i checked) it does not seem to be. that part, if nothing else, is likely progress. it does matter a bit who fights for you-- whether theyre actually on your side, or simply looking for a way to sell you out. some people only care about the license. sometimes at least, that license is like a steel door on a bamboo-framed fortress. there are most certainly other matters to contend with. even the fsd considered more criteria than that, the (new, not necessarily better) fsf should as well. since 2011 however, stallman has been at least partially "resigned" about such matters. which would be alright, if other people picked up the slack-- they wont do this at the fsf. instead, they bolstered their defences against whistleblowers. thats more of a nod to their real priorities than anyone is going to admit. => https://wrongwithfreesw.neocities.org