FrogFind and 68k News
This is a modal window.
The media could not be loaded, either because the server or network failed or because the format is not supported.
Formal Metadata
| Title |
| |
| Subtitle |
| |
| Title of Series | ||
| Number of Parts | 287 | |
| Author | ||
| License | CC Attribution 2.0 Belgium: You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor. | |
| Identifiers | 10.5446/57100 (DOI) | |
| Publisher | ||
| Release Date | ||
| Language |
Content Metadata
| Subject Area | ||
| Genre | ||
| Abstract |
|
FOSDEM 2022236 / 287
2
4
6
8
12
17
21
23
31
35
37
41
44
45
46
47
50
62
65
66
67
68
71
73
81
84
85
86
90
92
94
100
102
105
111
114
115
116
117
118
121
122
124
127
131
133
135
137
139
140
141
142
145
149
150
156
164
165
167
169
170
171
172
174
176
178
180
183
184
189
190
192
194
198
205
206
207
208
210
218
220
224
225
229
230
232
235
236
238
239
240
242
243
244
245
246
249
250
253
260
262
264
267
273
274
277
282
283
287
00:00
Open sourceCodeNeuroinformatikInternetworkingWeb browserWeb service
00:42
Group actionYouTubeNeuroinformatikBitComputer hardwareYouTubeProjective planeCD-ROMFloppy diskComputer animation
01:21
Bookmark (World Wide Web)Denial-of-service attackExtension (kinesiology)Virtual machineFile formatFront and back endsWeb serviceSearch engine (computing)Computer animation
01:55
Software testingCodePoint (geometry)Software developerSoftwareProjective planePlastikkarteWeb serviceSheaf (mathematics)Open sourceCode
02:41
Demo (music)UsabilityWeb pageComputer-generated imageryLink (knot theory)Information securitySoftware testingGoogolBuildingProxy serverContent (media)File formatWeb browserProgrammable read-only memorySkewnessInternetworkingSoftwareMedical imagingWeb browserVirtual machineProcess (computing)InternetworkingReal numberLink (knot theory)Web pageComputer hardwareComputer configurationRevision controlContent (media)Projective planeBitOpen setGoodness of fitFile formatProxy serverWebsiteSoftwareWeb 2.0Inheritance (object-oriented programming)Local ringCategory of beingPoint (geometry)Sheaf (mathematics)Sound effectExtrapolationLevel (video gaming)Electronic visual displayOpen sourceServer (computing)NeuroinformatikQuery languageMetropolitan area networkInformation securityResultantLine (geometry)BuildingBookmark (World Wide Web)View (database)TrailComplexity classDebuggerFront and back endsArithmetic meanQuicksortSpherical capWeb serviceComputer animation
09:24
AlgorithmDirectory serviceWindowComputer configurationComputer fileBookmark (World Wide Web)Inclusion mapCNNHead-mounted displaySoftware engineeringOrbitWeb pageSineQuantum statePeer-to-peerComputerOpen setRankingElectronic visual displayComputerUniform resource locatorComputer simulationGoogolLink (knot theory)Error messageWebsiteNavigationNeuroinformatikInformation securityEncryptionWeb pageElectronic visual displayQuery languageHeegaard splittingSheaf (mathematics)ResultantInternetworkingProxy serverPoint (geometry)Different (Kate Ryan album)Uniform resource locatorCoprocessorBitRight angleFloating-point unitMedical imagingTrailElementary arithmeticPlastikkarteMaxima and minimaReal-time operating systemServer (computing)Source code
13:46
SicCommodore VIC-20PlastikkarteGame controllerStack (abstract data type)Web pageResultantLink (knot theory)Revision controlMedical imagingOperating systemWeb browserFloppy diskSearch engine (computing)Source code
17:14
Software developerOpen sourceControl flowSoftwareBlock (periodic table)Library (computing)GoogolCodeParsingComputerNewton <Computer>InternetworkingLimit (category theory)File formatSoftware developerBuildingNeuroinformatikSoftwareGoogolCross-site scriptingBlock (periodic table)ParsingResultantQuery languageProjective planeLatent heatOpen sourceComputer hardwareBitPersonal digital assistantSource codeLibrary (computing)InformationTrailCodeWeb pagePower (physics)Proxy serverAsynchronous Transfer ModeInformation securityRevision controlQuicksortWeb serviceConnected spaceContent (media)Computer animation
20:34
Multiplication signNeuroinformatikPresentation of a groupComputer animation
21:04
Presentation of a groupInstance (computer science)Meeting/Interview
21:36
Group actionRepository (publishing)Open sourceOpen setProxy serverWeb serviceInternetworkingMeeting/Interview
22:01
InternetworkingProxy serverIP addressSoftwareMeeting/Interview
22:29
Domain nameCoefficient of determinationBookmark (World Wide Web)1 (number)Search engine (computing)Meeting/Interview
23:15
DeterminismMeeting/Interview
23:39
Web browserMedical imagingMeeting/Interview
24:34
Medical imagingWeb serviceQuicksortMeeting/Interview
25:09
User interfaceMatrix (mathematics)Element (mathematics)Multiplication signWeb pageReading (process)ModemLimit (category theory)Revision controlSoftware bugWebsiteTransport Layer SecurityPlanningMeeting/Interview
25:55
outputMeeting/Interview
26:24
YouTubeMeeting/Interview
26:49
Web serviceVideoconferencingGoodness of fitMeeting/Interview
27:19
Meeting/InterviewComputer animation
Transcript: English(auto-generated)
00:07
Hello, my name is Sean Moleseed and it is such an honor to have been invited to speak to you today here at FOSTM22 about my little retro computing pet project, which is frogfind.com
00:21
and 68k.news, two kind of web services that I built targeted towards really old computers and ancient web browsers with the goal of making as much of the internet as possible available to those old machines as if they were still first-class internet citizens.
00:42
So first a little bit about me, I really love to tinker with old computers and I especially have a soft spot for old Macintosh and Apple hardware and I do a lot of this tinkering on my YouTube channel Action Retro where really my favorite thing to do is take old computers
01:01
and try to cram as many upgrades as possible into them to see what kind of stuff they can do. So stuff like putting a CD-ROM drive into an old compact Mac that only ever had a floppy drive or getting a Quadro 950 up and running with Apple's old Unix and trying to host modern services from it. And the project that I'm going to talk about today
01:23
is kind of an extension of that, it's two web services that I built. There is 68k.news which is a front-end to Google News that tries to make all the articles accessible on vintage machines by breaking them down into very basic HTML 1.1 with some nice
01:41
basic HTML formatting and frogfind.com which is a search engine that uses DuckDuckGo under the hood which tries to expand that same concept to pretty much the entire internet. And when I built these services what I really mean is that I took the incredible work of some really smart
02:01
open source developers and some open source projects and kind of hamstrung them together with my own extremely average amateurish coding to create these services which is kind of what I think the point of open source is which again is why I'm so excited to be sharing it with you. When everybody has access to the code people can build the software that they need, they can tweak
02:26
software to do new things and create entirely new pieces of software. Everyone from experienced developers to enthusiastic amateurs such as myself. Without open source silly projects like this just wouldn't happen. So I'm going to split my talk today into three sections. The first
02:46
is what frogfind and 68k.news are for and kind of my thought process behind building them and then we'll do a demonstration on real live ancient Apple hardware and then I'll go into a little bit more detail about exactly how I built them and which open source projects I used
03:04
to make them happen. So what are they and why are they? The main purpose of the sites is to take web content down to a format that ancient web browsers are able to not just display but display quickly and provide a good user experience and to that end it strips out
03:24
all the extraneous junk and bloat of the modern internet that you know nobody really asked for anyway. So all the JavaScript is gone, the CSS is gone, inline images are stripped out and just made available as links at the top of the article because not all browsers can do inline
03:41
images. There's no SSL security of course because if there's one thing old browsers can't do it's SSL security and I really tested this the most with Netscape 1, 2, and 3 on my Macintosh SE 30 which is a 16 megahertz black and white compact Mac. And the project really
04:03
began as just I wanted to build a simple HTTP HTML site that I could test these old computers with as I got them onto my home network and I started out just playing around with looking for an RSS feed that I could import and display as basic HTML. I found that Google news actually
04:24
has a really great super fast RSS feed with all of the headlines and it's broken out into all the different categories and all the different locales so I was able to use that as kind of a starting point and then I thought well is there some way that I could easily
04:42
take the articles that the RSS feed links to and strip those down so that they're also readable on very old machines. So what I wound up building is a front end that takes Google news RSS feeds and then a proxy on the back end that strips out all the extraneous stuff from
05:03
a given web page. So all the links in the Google news RSS feed are replaced with proxified versions of them so when you click on them it goes to the proxy strips out all the junk and returns just basic formatted HTML. And the effect of it is really cool actually so you
05:23
can just open up 68k.news and see all the headlines choose whatever section you want to read and then click on any of the articles and more than likely that article will very quickly appear as just basic HTML and I'd say it works on probably 80% of the articles.
05:42
It really only has problems with sites that are unfortunate with their use of too much front end JavaScript. So with 68k.news working way better than I expected it to I thought what if I could take that same idea and kind of extrapolate it to the entire internet. So I built frogfind.com
06:03
so it's just a very simple HTML front page that takes your search query submits it to duck duck go and then takes back the search results and strips out all the results and presents them in very basic HTML and then all of those links are replaced with a proxified version that strips them out so you can click on a link and it will load the page through the proxy
06:26
and attempt to make it readable on extremely antique web browsers and then it goes even further it takes all the links on that page you're viewing replaces them with the proxified versions with the net result being that you can click on any link on the page that you're
06:42
viewing and it'll open that page up in the proxy so you can for example go down a wikipedia rabbit hole on your ancient compact mac or your apple 2e and it's a pretty convincing result you know you can just sit and browse the internet on your vintage machine with just
07:01
these couple pieces of open source software powering it under the hood in PHP on inexpensive web hosting so then the question is why put all this effort into this kind of service for extremely old machines well i mean because i love them and uh there's really there's two
07:23
methods that you can take to try to get an a very old machine to be usable on the modern web the first is to upgrade them as much as physically possible which is my favorite thing to do by shoving all sorts of ram upgrades and accelerators and shoehorning an SSD in there
07:41
but that has a very definite cap there's a ceiling both in technology and the cost because those some of those accelerators are extremely rare hard to find and thus expensive the second option is to downgrade web content to a level that the old browsers can interpret and there's
08:02
been actually a bunch of great projects to do so stuff where you you can use like a raspberry pi as a man in the middle which will strip down web content and return simple html to your browser or there's some out there that will you know take a screenshot of the web page
08:20
and return an image map to the browser so it actually looks like the page that you're viewing on the old computer and all that stuff is very cool and my solution here is very much in line with those except i put all of that stuff up onto the server so the experience to the user is the same as it was back in the 90s where your vintage computer is just connected
08:46
directly up to the internet and you can browse these pages and websites as a first class internet citizen and what i found is that it's actually really nice to just be able to sit down and use an old machine that you love again just as as it was when it was new read the news browse
09:06
wikipedia with no distractions no javascript nothing tracking you just simple content that you can consume on your favorite vintage machine so let me show you on a couple of very old apple
09:22
and macintosh machines this is a macintosh lc 575 from 1994 one of the first macs that i ever used as a kid with a 33 megahertz 68 lc 040 processor lacking an fpu this was a mainstay in many elementary school computer labs back in the 90s it originally came with five or eight
09:47
megs of ram though i've upgraded this one to the unofficial maximum 132 megs which is pretty ludicrous for 1994 and i think this machine is really great for today's demonstration because of its beautiful sony trinitron display and because i put an ethernet card in here
10:04
so let's fire up netscape navigator 3.01 and check out how 68k news and frog find work on a real machine now if you try to go to a real modern website you'll get an error like this
10:22
it can't communicate because they have no common encryption algorithms but there is no such encryption on 68k news and i won't edit this for speed either let's look at this in real time to see just how fast it is so everything on this page below this section here is powered by
10:48
just google news rss feed and i've split it up into the different sections for google news and i have links to the different editions that you can choose for different locations so
11:03
i generally like to stick to technology news and science news because everything else is depressing but you can click through to any of these articles and just read the article right here as if it was made for an old machine and all the images are stripped from inline on the
11:26
page and just linked up here so if you want to you can view them this one didn't work here
11:41
we go here's a nice little image of tux because there's a new security issue that's in the news yeah so you know this works surprisingly well and again it's just it's a really nice experience to just read the news on this old machine and be able to just sit down and use it again there's
12:07
no distractions there's no tracking it's i think quite reasonably fast and it's all happening on the server side all of that grunt is taken up by just php okay so that's 68k news now let's
12:26
check out frog flying.com which is the same concept just extrapolated to try to function on every website out there and of course we have our nice frog logo again by the incomparable
12:40
steve from acuity ford so this is just a simple front page i built and if we put in a search query it just sends it off to duck duck go and then parses the results here again i think it does it very quickly and the point here is that you can click through to a search
13:00
result and wikipedia is a great example and you can just read the article but even further every link on here has been replaced with a link that passes the resulting page through that mozilla readability proxy so you can just go down a whole wikipedia rabbit hole
13:22
clicking on any of these links and i'll tell you it really feels like we're just browsing the internet just like the old days okay so this is a reasonably powerful 33 megahertz
13:42
machine why don't we try it on something a little bit older how about an apple 2e booted off of a single floppy disk this is kon tiki a fairly modern operating system really built for embedded devices but they made a version that works on the apple 2 along with versions
14:04
for like the commodore 64 and this has a built-in tcp ip stack and this talks to my other net 2 card which is installed in here and it has a built-in minimalistic text-based web browser
14:24
which even has very nice mouse control so we can go to 68k.news and here we are already loading google news on this uh 1983 apple 2e and we can browse through the page with
14:48
the down button here but all of this news is depressing so let's look at technology news yeah all of these links are clickable by the way
15:08
so click on technology and let's read about that new linux exploit so of course the article
15:32
images don't work but honestly this is a perfectly readable experience and it's pretty cool to just sit down and read the modern news on an apple 2 i don't know if i like green or white better
16:00
what's next for the james webb telescope yeah a totally serviceable experience okay let's check out the frog find search engine if i can spell correctly
16:27
frogfind.com and here we go this is duck duck go search on an apple 2e and here's the search results and again it's all clickable so let's see the wikipedia page
16:50
on apple 2e yeah a totally serviceable experience and now i can go down a wikipedia rabbit hole
17:07
on my apple 2e so how i built these first again let me stress that i am not a professional
17:21
developer everything i know i've just learned on the go just by tinkering but again that's the promise of open source i mean that's really the thing to love about it is that whether you're a professional developer or whether you're just again like me an enthusiastic amateur if you have a little bit of interest you can take the source code of some software and tweak
17:45
it to do whatever you need it to do and then you can take your creations and further contribute them back into the software zeitgeist the building blocks that i use specifically first was simple pi which is a php rss reader and that's what i first used to power google news rss feed headlines
18:06
on 68k news and then for the articles i found a php port of mozilla's readability library which is the software that powers reader mode in firefox and that's what simplifies the pages
18:21
down to very basic format while still keeping it nicely formatted in html and then i wrote some additional code to take it down even further and just get rid of all the css and turn everything into very basic html for duck duck go i wrote my own kind of janky parser and software to send
18:41
the search query out take the result back in and parse it into simple html and then of course i wrote my own janky php code to tie everything together and use the readability library as a proxy and i can't forget steve from mac 84 contributed that adorable frog logo
19:00
and i put all this stuff back out there it's all up on github under gpl3 and it's been really amazing to see people contribute to the code send in commits and even forking into new projects the first versions of this stuff that i put up there had all sorts of little security issues with my php code and people very quickly contributed fixes to it and that
19:25
was just it was amazing absolutely amazing to see and i also heard from people who were using these services in kind of unexpected ways for example someone told me that they had used 68k news when they had a power outage and limited cell phone connectivity because it was very fast
19:43
to download the text content as opposed to trying to reach full google news on their cell phone and of course there's people who want to share articles without any tracking information because of course most of the evils of tracking happen through javascript which is stripped out on these sites and i've even had somebody fork frog find to make newt find
20:06
specific for apple's newton hardware so apple's newton pdas and the quasi e-mate newton computer from back in the 90s so it's just been really incredible to
20:21
build these projects and have so many people use them and to feel like i'm contributing something back to just not only retro and vintage computing but also open source software but thank you very much for taking the time to listen to my presentation today i hope you enjoyed it and
20:41
don't forget to pull out your vintage computers once in a while and you know try to plug them into the internet and see what they can do thank you very much and we are live hello
21:11
hello yeah welcome to our question answer to the talk frog find and 68k news with me here is sean who did this great presentation on uh is and we have a lot of
21:26
questions so we have to go through them quickly our first question is it easy slash possible to host your own copy instance of frog find yeah so i put it up on github it's all open source github.com slash action retro so you'll find both repositories out there it's all
21:45
php and uh yeah you can put it anywhere php runs and have your own running copy and actually some people have done that great i will try that out myself nice have you seen any misuse of your open proxy service well unfortunately on the internet people sometimes tend to misuse
22:04
things uh so really what i've seen is people trying to crawl or or use the proxy in some software so if i see you know a million hits from one ip address sometimes i have to block it but you know i'm not really tracking usage so much other than just a mount so
22:24
yeah uh it is an unfortunate reality of the internet though okay but yeah yeah it hasn't brought you any problems so far no no no problems just uh just small hiccups okay oh where does the name frog find come from uh i'm so glad you asked that's my favorite question
22:44
so way back in the 90s we used to have fun names for search engines i remember my favorite was dog pile so frog find you know i was trying to think of a very old 90s kind of name for it so uh you know most domain names are taken up at least all the good ones these days
23:04
so i was just trying to think of something to do with animals something to do with searching or finding and uh frog find was available and uh is nice and alliterative so that's why i picked it great um there's a question i'm not sure what i should do with that with that so i just asked does it run doom yeah so i know it's a joke question but i have a real answer
23:26
for it uh technically if you've read the news most of the news these days is doom and gloom so i'd say that 68k news is is mostly doom great answer
23:42
what is the most exotic machine anyone tested for it find on so i've seen pictures of people who have run it on some really strange things i mean i've seen people running on old mobile phones with you know wap wap browsers and uh you know it seemed to work pretty well and i've seen
24:04
actually several people have sent me pictures of them using it on their nintendo we the original nintendo we on the web browser channel which of course is discontinued from nintendo but they were able to get uh some use out of it with 68k news i've seen trs 80s running it which is very
24:24
cool um yeah so i don't know most exotic i'd say nintendo we thank you another question why not use leap kaka for images uh i think someone else suggested
24:41
that too um i i do want to continue working on these services uh the handling of images i mean i just kind of you know winged it sort of so there are some issues with the images i mean it actually it doesn't really preserve the aspect ratio correctly even uh so fixing how
25:02
the images work is something that i definitely want to do uh thank you another question will matrix i guess the the element web interface to matrix that we are using here for first time will that work unfortunately uh if if it's not uh unmoving text on a page it's not going to work
25:23
so yeah reading text is about the limit okay so most likely no yeah yeah so fox modem is giving a million hand claps so he was very excited oh thank you and i have one question myself i figured out that um frog find currently doesn't seem to support websites that only have
25:45
tls 1.3 and not the older tls versions um have you any plans to to do something about that yeah i'd like to fix any i mean that sounds like it would be a bug then and uh i'm not quite sure why that would happen uh but yeah um and especially you know one of the advantages of
26:06
having it up on github is we can you know discuss these kinds of things i mean i'm not very good at using github uh but i i do try to be someone active on there but yeah so that's something that i'd love to fix and uh if anybody has any ideas on how to fix it i'd definitely appreciate
26:22
any input yeah i i open an issue on github for that nice so i'm out of questions let's see if there's other questions uh coming up here okay now we are going through any more questions
26:41
from the audience people are here in the chat please let me know if you have more questions for sean if you don't please subscribe to his youtube channel oh yes really really good thank you i appreciate that okay i think there's uh i think there's no more questions coming in
27:11
thank you very much it was a a very nice video good talk and a nice service for everyone well thank you so much i'm so again honored to be invited to share it with everyone
27:24
it was really fun to do so thank you see you around