Earlier this week I migrated www.freecalypso.org to a new server, and that migration had a side effect of taking all of my Mercurial code repositories offline. (It happened this way because only the static file archive was migrated initially.) I am now pleased to report that all FreeCalypso and Themyscira Wireless Hg repositories are back online at this new location. Old URLs still work as well, handled via HTTP(S) redirect.

The new hosting setup for FC and ThemWi should be much more responsive than our old server that was overrun to death by AI scraper bots, and unlike the band-aid solution that was applied between May 11 and now, the new server is set up much more thoughtfully and properly. Because of the way Mercurial works, its hgweb facility is required for all public read-only access to hosted repositories: not only for casual browsing akin to cgit, but also for native clone and pull operations over HTTP(S). And because hgweb is implemented in Python like most of Mercurial, all operations that invoke it are expensive in terms of server load. Our new server has the following improvements over the old one in this regard:

1. Instead of using simple-minded CGI between the web server front end and hgweb (a new process spawned for each and every request), we now run a dedicated back end server that consists of long-lived worker processes, Green Unicorn specifically for hgweb.
2. The proxy path from lighttpd front end to Gunicorn/hgweb back end passes through Anubis - hence all incoming HTTP(S) requests get their souls weighed, and malignant bots get blocked. It is the same solution that has been adopted by Osmocom (our primary inspiration), Linux kernel cgit and many others.

All static content is still served by lighttpd like before, although we got an update to a newer version of this light daemon. Because so much of our published content is in the form of "raw" FTP-like trees of directories and files, I felt it was very important to preserve the look of directory listings and the assignment of MIME types to files - this consideration was the main impetus for staying with lighttpd, as opposed to the other considered option of moving to nginx.

As another point worthy of note, we no longer force HTTPS onto users against their will! All HTTP(S) resources are equally accessible at both http:// and https:// URLs, and neither redirects to the other - so use whichever you prefer. My personal philosophy is that HTTPS "security" is utterly useless and constitutes nothing but security theater, especially for a site like ours that has no login mechanism, no ability to collect anything from users or visitors, just a public read-only resource made free to the world - but we have a ton of legacy URLs with https in them, plus some people will complain loudly if we went "insecure" only. Hence our solution: use whichever protocol you like, let each user choose for herself.

And finally, never forget good old FTP! On our new server both /pub and /members static file resources are accessible via both FTP and HTTP(S) - so now you have a choice of 3 protocols for downloading our static content 🙂.

American 2G Cooperative (A2GC) is the company that "owns" the present Lemmy community in the sense of holding the top moderator role. The purpose of this post is to provide the necessary legal info and contact details.

A2GC is a non-profit corporation seeking federal tax-exempt status under Internal Revenue Code section 501(c)(12) as a mutual or cooperative telephone company (please note it is a different type of tax-exempt non-profit entity from the usual 501(c)(3) - see IRS publication 557), and like all tax-exempt non-profit entities, we are required to publicly disclose our management structure, bylaws, financial details and so forth. While our website is still under construction, the key basics are as follows:

We are a Delaware corporation, our incorporation file number with the state of DE is 10180871, and our federal EIN is 33-4896531. Our physical office address is 16870 West Bernardo Drive, Suite 400, San Diego, CA 92127 - however, because we don't employ any full-time staff, only part-time volunteers, in-person meetings at our office location are available only by appointment. Contact details for our current corporate officers are as follows:

Ms. Mychaela Falconia, President & CTO: falconia@usa2g.org
Ms. Jessica Westerman, Secretary: secretary@usa2g.org
Ms. Shannon Hogan, Treasurer: treasurer@usa2g.org

To see our official Certificate of Incorporation, bylaws and other documents which we are required to make public, please look in our documents directory.

Please see other posts in this Lemmy community for more information about our actual work and how we seek to build a new GSM/2G network in select rural and remote locations in USA.

Questions and comments

If you would like to get in touch with us, you can do so either publicly or privately as you prefer. To contact us privately, please email any of our officers at the addresses listed above. Of course you can also send postal mail to our office address, but email will be much faster. Alternatively, if you have some suggestions or questions or comments that would be of benefit to the larger community, and may form a basis for discussion, please create an account in Lemmy fediverse (on any instance of your liking), join this community and post here. We are not on Reddit or Discord, just here on Lemmy.

Some people on That Other Site are asking about the status of my work with regard to the various GSM/2G-related projects I lead, most important of which is A2GC. I am still unable to post on that evil site, so I am going to post this summary here and then rely on other people to post a link in r/vintagemobilephones.

Here are my project-related activities at the present moment:

1. A lot of my time and energy went into the still-ongoing migration of old FreeCalypso archives and repositories to my new server. While I cannot share the full details without violating other people's privacy, our old server setup that goes back to 2015 had some fundamental flaws in its design that went too deep to be fixed organically via small incremental changes - hence wholesale migration to an entirely new server was the only solution I could come up with. The current status of this migration is that all static files have been moved over, but there is no public access to Hg repositories at the moment. Restoring public access to these Hg repos is a remaining high-priority task on this front.
2. The latest word from T-Mobile is that they are going to shut down what is left of their unloved GSM/2G network on August 3. As of the present moment I still rely on that network for within-family communication during times away from home. Given this imminent threat of shutdown, I have to put a higher priority on a side project that has been on my mind for some years now: set up a portable GSM BTS in my car, a sort of personal "cell on wheels", with a Raspberry Pi serving as the BSC, using an LTE mobile Internet link to connect to the fixed server that acts as the MSC. I have a separate post for discussion of that idea.
3. Once I get past the two urgent tasks listed above (both of which count as "putting out fires"), my main goal toward A2GC proper is to implement the new MSC that has been an outstanding need for a while now. We need to have this MSC implemented before we can operate a proper, above-ground GSM network, and I have set a personal goal to implement it before the end of 2026.
4. Macrocell BTS for above-ground GSM deployment: just like I said in my last days on Reddit, we (A2GC) have settled on Nokia BTS hardware, specifically Flexi Multiradio family. We have radio modules for both GSM850 and PCS1900 bands, as well as system and transport modules to go with them. Multiradio concept means that a single radio module can put out any combination of 2G, 3G and/or 4G RATs (radio access technologies), either by connecting multiple system modules (one for each RAT) to the same radio module(s), or by using the newer FSMF (Multiradio 10) system module that does all 3 RATs in one. Since our primary focus is on GSM/2G RAT only, we have stocked up on 2G-specific ESMB system modules, but we can also switch to FSMF if need be. The setup of ESMB driving a Multiradio RFM, with Osmocom BSC driving the ESMB via icE1usb, has already been proven working in our lab. Some further improvements to OsmoBSC software are needed: in particular, we need to teach OsmoBSC how to drive all 3 sectors on the same BCF, which is the config required for a standard 3-sector site. I plan to get back to these OsmoBSC improvements after Hera MSC, which is the previous point.

One detail which I need to re-emphasize with each of these periodic status updates is that we are not going to begin the process of applying for a license with FCC, or even make a final decision as to exactly which geographic location we are going to pursue first, until after we complete a whole bunch of preparatory work. The prep work which we need to complete before applying for our first license consists of Hera MSC (see item 3 above), OsmoBSC improvements for 3-sector macrocell BTS (see item 4 above), and also some FPGA work which I will write about later, creating a low-latency, production-quality interface from OsmoBSC and tw-e1abis-mgw to E1 Abis.

The reason for doing things in this order has to do with FCC rules section 22.946: once we get our first GSM850 license, a clock will start ticking, and we will need to bring our first GSM system fully live within 1 year of license grant. Needless to say, we should do as much of our homework as possible before we start this clock! All work on BSC and MSC components is location-independent, in that all technical requirements remain exactly the same no matter which location we deploy in, and also the same between GSM850 and PCS1900 bands - hence it makes the most sense to complete all of this work before beginning negotiations with property owners at our envisioned candidate locations.

In conclusion: Rome was not built in one day, and we still have a long road ahead of us before we go to the steps of finalizing on a specific location and applying for a license there. But we are not giving up, and the necessary background work is still ongoing.

For several years now I operate a sysmoBTS 1002 at my home, along with a server that runs the requisite suite of osmo-* processes. The result is my own self-made GSM signal at my home; I call this network Themyscira Wireless, named after the fictional home of Wonder Woman. This home network of mine is interconnected with real PSTN via SIP, such that my ThemWi phone has a real phone number, can call any outside numbers and can receive calls from the outside world. For the past 3 years or so I've been using this home GSM system as my primary personal phone - while I do have a traditional POTS land line as well and will never give it up willingly either, ThemWi GSM is more convenient most of the time.

But of course it works only when I am at home, effectively an exotic type of cordless home phone. For the past 2 years or so I have got into the routine where every time I leave my home for an errand, I turn off my ThemWi phone (to save battery), boot up my T-Mobile 2G phone, then drive away from home - then do the same in the reverse when I come back home. A royal pita, plus having to manage two different phone numbers for home stay vs being out and about.

But now there is word going around in Discord and Reddit vintage mobile phone communities that the evildoers at TMO are going to fully shut down their unloved GSM service on August 3. But if they think for even one nanosecond that I am going to roll over and accept a "modern" (VoLTE-based) phone into my life, they are very mistaken... I have never used a "modern" smartphone and never will!

So what do I do now? For some years now I have had the idea of operating a BTS in my car, as a kind of cell on wheels that travels with me wherever I go. Have another little sysmoBTS 1002 (see product link above) reside in the back seat of my car, with a little Raspberry Pi next to it, interconnected with an Ethernet cable. The Raspberry Pi is to be equipped with an LTE data modem - initially in USB stick form factor, later I will explore a more permanent RPi "hat" solution. The software running on the RPi will maintain a WireGuard VPN connection (running over LTE mobile Internet) to one of my servers acting as the MSC, while osmo-bsc will run on the RPi itself inside the car. The resulting arrangement can be seen as a mobile (on wheels with a gasoline engine) LTE to GSM converter, using the local LTE network wherever I go and giving me GSM.

A Digi-Key order for the necessary RPi components has been placed a few hours ago; when the RPi 3B arrives, I will start playing with Devuan on it, and turning this idea into reality. Stay tuned!

P.S. If you are going to comment about legalities, don't bother. If someone without a clue were to do what I describe, then yes, they will very likely cause interference with other networks, FCC will throw the book at them, and they will get what they deserve. However, I know what I am doing - I know where unused blank spots are in GSM-usable spectrum, and I know how to operate my BTS (both home-based and vehicle-based within my known travel range) in such way that it never interferes with anyone and therefore there will never be any basis for anyone to complain to FCC - there is literally nothing to complain about when I don't interfere with anyone or anything.

Since this community is open to an international audience and since many people online (whether they live in USA or not) are woefully misinformed about the state of GSM/2G services in this country, this post is being made to set the record straight.

For a long time, since at least 2001 (and even earlier if we include these companies' predecessors) and through the end of 2016, there were two nationwide GSM/2G networks available in most of the country: Cingular Wireless (rebranded to AT&T in 2007) and T-Mobile USA, former VoiceStream. Then on 2017 January 1 AT&T shut down their GSM network (the one that used to be Cingular), leaving T-Mobile as the only available GSM network in most locations.

Current state of T-Mobile 2G

TMO have been waging a war against 2G since at least 2018 - in that year I (Mother Mychaela, author of this post) encountered for the first time a specific location in my area (a neighborhood in San Diego) in which T-Mobile turned off their 2G signal (leaving only 3G and 4G) and refused to turn it back on when I called to complain. Then in 2019 I discovered for the first time that T-Mobile started issuing maliciously modified SIM cards: their newly issued SIM cards have GSM 11.11 SIM application removed (DF_GSM directory missing in the SIM file system), causing their SIMs to not work in phones that don't speak the newer USIM-ME interface protocol that was introduced for 3G.

In the present time (as I make this post on 2026 June 4) T-Mobile GSM/2G network is still up in certain locations. I live in Ramona, CA, a small town in San Diego County, and both in our little town and in San Diego proper I experience the same situation: some spots have perfectly strong signal while other spots are dead zones. However, the quality of service on this network is very severely degraded:

• Clear voice quality is attainable only if you intentionally seek out a spot with the very strongest signal level, which may require some driving. Everywhere else outside of these select strong spots, your phone will typically show some bars of signal and will have good enough service to send and receive SMS - but when you try to talk with voice, it will be very choppy, making conversation very difficult.
• Talking while driving or riding in a moving vehicle is impossible - the call will drop. Instead you have to pull over in order to talk on the phone, and be completely stationary.
• Even in idle mode (outside of a call), as soon as you start moving, your phone will very likely drop to no service and will only reacquire service after some delay when you become stationary again, assuming that the destination you drove to has service.

In technical terms, there are two factors contributing to this abysmal quality of service:

• TMO operate their unloved GSM service in the guard bands around their 15 MHz wide LTE signal in PCS1900 band. In itself this arrangement would be perfectly fine - one can operate a high-quality GSM service in guard bands around LTE. However, the problem with how TMO do it is that they emit their guard band GSM signals at much lower power level than the LTE "curtain" they surround. The result is that GSM phones receive signal with a very poor C/I (carrier to interference) ratio, which leads to choppy voice connection.
• Every GSM cell continuously emits so-called system information messages on BCCH (broadcast control channel); among other important info, these SI messages provide a list of neighboring cells. Phones use these broadcast neighbor cell lists in order to reselect from cell to cell in idle mode: when your phone is in idle mode outside of a call, it is camped on a particular cell (called the serving cell), but it has to constantly monitor signal strength of neighbor cells in order to detect movement and thus know when to reselect to a different cell, and to which. The problem with today's state of T-Mobile 2G is that they broadcast completely bogus neighbor cell ARFCNs in their SI messages, causing phones to look in the wrong place when movement occurs. Furthermore, those bogus neighbor cell ARFCNs they broadcast aren't empty, instead they correspond to spots in PCS1900 band that are now occupied by wideband LTE signals - confusing GSM phones that expect valid GSM signals on those ARFCNs where the serving cell told them to look for neighbor cells.

Artificial blocking of new users and phone switchers

One of the most sinister hostile actions done by today's TMO against 2G users is artificial blocking. The following anti-human behaviors have been observed so far:

• While SIM cards issued by TMO themselves are maliciously modified to remove GSM 11.11 SIM application, some TMO-based MVNOs still provide SIM cards that have GSM 11.11 support. However, at that point a network-level block kicks in: TMO's core network sees a non-VoLTE phone connecting, they check the account and see that it is new (not "grandfathered"), and they boot that phone off the network. The block appears to be based on IMEI: VoLTE-capable phones configured to operate in 2G mode don't get disconnected, certain old phone IMEIs that didn't make the blocklist get by, and those users who are able to change their IMEI have also reported success.
• Some GSM phone enthusiasts are not content with using just one particular old phone, instead they have a large collection and like to move their legacy ("grandfathered") T-Mobile SIM between phones. Recent reports indicate that TMO started blocking such users as well!

Foreign SIM roaming workaround

Quite interestingly, when someone roams on T-Mobile's unloved 2G network with a foreign-issued SIM, no artificial blocking of any kind gets applied! The most popular source of foreign SIMs for this trick is Mexican Telcel: there are some private sellers on Amazon who sell pre-activated Telcel SIM cards with already-assigned phone numbers, such that the prospective American user does not have to physically travel to Mexico, and there are English-language websites that allow top-up payments on these Mexican Telcel prepaid lines, eliminating the Spanish requirement.

There are some downsides with this method:

• You will have a Mexican number (country code 52), not USA. Convincing other people to call you at a Mexican number may be difficult, as they may be subject to international call charges.
• You will probably receive a steady flood of marketing SMS in Spanish, and it may be difficult to tell what is junk which you can simply delete vs potentially important communication from Telcel regarding your service.
• Telcel SIMs are made in the same triple-cut form factor as all other "modern" SIMs, thus mechanically weak in 2FF-only usage and thinner than the original 1FF/2FF/3FF standard, potentially resulting in less reliable contact.

Despite these limitations, however, this method is the most practical option available today for those who don't have a legacy T-Mobile account going back decades, or those who desire a SIM card which they can freely and safely move between different phones.

Small regional carriers

AT&T (former Cingular) and T-Mobile were the only two nationwide GSM operators in USA; with AT&T out of the picture since 2017, T-Mobile is the last one. However, there may be some small regional carriers in some corners of the country who still operate GSM/2G independently from TMO - finding these small carriers, if they exist, remains an active area of research.

American 2G Cooperative

A2GC is the non-profit company that created this Lemmy community. Our cooperative was founded as a response to the problems described earlier in this post, the very bleak state of GSM/2G services in USA today. Our cooperative seeks to do the following:

• Operate our own physical GSM/2G services in some very small select locations in USA. By small we mean just a handful of cell sites, an extremely tiny footprint - and our service locations will have to be very far away from civilization, in very remote rural areas with minimal population, as those areas are the only ones where we can get access to GSM spectrum bands.
• Operate a roaming-capable core network, and then use FCC rules section 20.12(d) to compel T-Mobile into letting us roam on them. The resulting effect we seek to achieve would be equivalent to roaming on TMO with a Mexican Telcel SIM (which works today), but with A2GC instead of Telcel as the home operator. The advantages will be USA phone numbers, SIM cards specifically made for retro mobile phones (2FF cut, traditional thickness) and 2G-centric customer service.

The above description is just a quick summary; a separate post will be made later to describe the mission and work of A2GC in greater detail.

For the purpose of this community, a retro cellular phone is one that requires a 1G, 2G or 3G network, as opposed to 4G or newer. Why would someone wish to use such a phone? Different people may have different reasons:

• Some people don't want any kind of smartphone: we want to use a plain mobile phone that makes and receives phone calls, and nothing more. We don't want a device that has a smartphone architecture inside, we want a traditional dumbphone from 1990s or early 2000s. Modern "dumbphones" are cheap imitations, while we want the real deal.
• Some people may be very loyal to specific brands or models of old phones. A person may have their favorite Ericsson or Nokia or Motorola phones, or any other brand of olde, and (we say) has a natural right to keep using them forever.
• Another type of person may be less partial to specific brands or models of old phones, but much more partial to the underlying cellular technology - for example, GSM. A retro-telecom engineer may have given years or even decades of his or her life to GSM (or D-AMPS TDMA or CDMA or iDEN) technology, and has a natural right (in our view) to live out the rest of his or her life enjoying said technology.

Different motivations, common interest

While different members of this community may have totally different personal reasons and motivations, we have one goal and interest in common: we want to keep our beloved phones working. Keeping retro mobile phones working means not only maintenance and repair on the phones themselves (including batteries), but also maintaining cellular network services for these phones to be functional, and providing our own replacements when traditional commercial carriers fail in their duty to do so.

Scope of this community

In the olden eras of interest to us, prior to today's LTE monoculture, different countries had different 1G, 2G and 3G cellular standards. The founder of this community (and author of this post) is based in USA (grew up here since age 15.5), hence my own knowledge and expertise is mostly Amero-centric. However, my own interest centers mostly around GSM technology, a form of 2G, and this particular technology originates from Europe, with North American version (with different frequencies, but otherwise identical) created as a follow-up. Therefore, I feel qualified to act as an expert on GSM globally, and not just in USA.

However, irrespective of the individual area of expertise of each given leader or moderator or subject matter expert, this community welcomes all users, devotees and appreciators of all 1G, 2G and 3G cellular technologies and of phones based on said technologies - no matter where you are on the globe, and no matter which specific 1G or 2G or 3G technology is of most interest to you.

Who runs this community

This community was created and is owned (top moderator role) by American 2G Cooperative. A2GC is a non-profit corporation in USA (legal entity details will be provided in a separate post) with a mission to build and operate a new GSM/2G cellular network in this country.

You are not required to live in USA or to be a member-patron of A2GC in order to participate in this Lemmy community - and vice-versa, when A2GC launches its core network and starts accepting dues-paying members who get GSM services in exchange for their membership dues, you will not be required to join this Lemmy community in order to sign up for A2GC membership and services. However, the present community and our company's services are intended for exactly the same target audience (except for the community being world-wide while A2GC is limited in scope to USA), hence it is our vision that the two will remain closely linked.

Other retro cellular technologies

While A2GC is focused on GSM/2G technology in USA, a truly diverse retro mobile phones community needs to include these other technologies too:

• 1G is an umbrella term for all analog technologies. In USA we had AMPS; internationally some other countries also used AMPS, while others had their own standards such as C-Netz and NMT.
• The successor to AMPS was D-AMPS, also called TDMA in USA. It is a digital technology in common with 2G, but because it was designed to coexist with AMPS, it is sometimes called 1.5G.
• In terms of pure 2G technologies, our beloved GSM existed in a state of fierce competition against CDMA and iDEN. Those competing 2G technologies have their own devotees, who are just as welcome in this community as us GSM queens.
• 3G successor to GSM is UMTS, while 3G successor to original 2G CDMA was/is CDMA2000.

While we as in A2GC are subject matter experts only in GSM, we warmly welcome devotees of all other 1G, 2G and 3G cellular technologies (and phones based on those technologies) into our community. Anyone who joins here as a subject matter expert for any of the non-GSM technologies will be officially recognized as such!