Ever since approximately July of 2025 (two months, give or take), we fell into a chain of events that, taken together, meant at first that orders could come in but could not be filled, and then even this stopped working, and we could no longer so much as take an order. In the eyes of the general public, at least, the development of HERMES Eudoramail was at a standstill.
Privately, alternative arrangements had absolutely been made with our clients (or rather, they continued to function as designed, being based on a time-tested, resilient platform), but our communication with the public was unfortunately rather limited, for reasons both of confidentiality and of manpower.
Now, as all the foreseeable and some unforeseeable problems have been remedied, and now we no longer rely on any part of our legacy toolchain, we feel it is incumbent upon us to give you, the public, a reasonably full accounting. We hope that, having a view of our successes as of late, you will be understanding of our failures.
Communication with our clients was originally planned to follow a simple template: one form of asynchronous communication for news, updates, and sales, together with one form of synchronous communication for technical support and customer relations. Given that our flagship product was an eMail client, the asynchronous communication medium of choice would of course be electronic mail—which comes with the fringe benefit of allowing the product to be sent as an attachment on a periodic basis. For our synchronous method of communication, we hemmed and hawed for a little bit between two secure instant messengers, Signal and Telegram; in the end, Signal won out.
This is still a live high-level plan; in fact, it is more live now than it was, for example, in 2024. Why? Quite simply because, on a practical level, sending out as many form messages as we have clients (a number which remains confidential), with the blanks customised with various data extracted from a table or spreadsheet, and with a file attached, requires a function called mail merge. As of yet, Eudoramail, a traditional IMAP/SMTP client, lacks mail merge functionality. Had this not been the case, we would have configured Eudoramail with a thoroughly routine combination of IMAP and SMTP, written a message, attached a file, wired in our client data spreadsheet, and called it a night.
Instead, we were forced to use Gmail, a specific brand of webmail (i.e. a cloud computing service rather than a software product), which we outfitted with a specific mail-merge plugin (Taro Tsukagoshi's Group Merge), and a specific brand of groupware (i.e. Google Workspace, comprising Drive, Pages, Sheets, Docs, and Slides). One limitation, discovered early on, was that Group Merge simply refused to mail-merge a message that included an attachment; it would send the first message properly, and the rest would arrive bereft of all attachments. As partial mitigation against this problem, we used links to Google Drive. An even worse limitation was that Group Merge would sometimes (often) go through a Google Sheets spreadsheet, fail midway through with a cryptic error, with a fraction of the messages already having been sent, and the process would have to be repeated from the beginning (some people would receive multiple messages, others none at all).
With Team HERMES having seemingly run out of options, it was decided that only commercial software would solve the problem, but after the purchase (at that point, of access to a cloud computing service, not of the licence to a software product) was finally made, it was discovered (to our horror) that Gmail disabled the sending of any executable code at all in attachments, whether bare or in a ZIP, RAR, or tarball—for, allegedly, “security reasons” (an excuse which was, at best, unconvincing, not least because Gmail scans all outgoing attachments for viruses). This was too bad: Digital Inspiration's mail merge tool is actually worth the money, so long as one is not duty-bound to contend with arbitrary overreach by monopolists (United States v. Google LLC, 2023).
Ultimately, it dawned on us that a bespoke problem needed a bespoke solution—one that was authored, supported, and used in-house only—and that the plan first articulated all those years ago had to be followed with reckless abandon, no detours. We shan’t go into this overmuch (confidentiality), but for the sake of clarity: the problems mentioned above demonstrably no longer exist, though making them go away took significant devolopment time.
For the majority of our existence, Team HERMES has affirmatively chosen a pricing strategy that is brutally simple and more honest for its simplicity: we total up the client’s consolidated contributions, divide by our headline price per seat, we give the client a licence to use Eudoramail (or, now, UltraTerminal) at that many seats, we verify whether or not there is a remainder, and if so, we give the client an extra seat for the remainder. We prefer not to have defined pricing bands; in addition, all customer relations are conducted via the methods outlined above.
From this it can be seen that, fundamentally, Team HERMES has no need for the dashboards, defined “perqs”, or the off-the-rack workflows of Indiegogo and Kickstarter. There is no need for explicit funding goals, either, except inasmuch as Indiegogo requires a number to be put in, even if it’s more or less ad hoc. We are also entirely capable of doing our own CRM; “Project Updates” and even comments are just a bonus. The three baseline requirements are to take the customer’s personal details and his money, and to produce a spreadsheet every month for CRM purposes. Everything in addition to that is, put simply, going above and beyond the call of duty.
For some time, Indiegogo (unlike Kickstarter) actually fulfilled this requirement, even though the button was hard-wired to read “Donate” with a disclaimer saying, in effect, “Donations are not associated with a perq” (which we had to go out of our way to debunk, given that in our case they very much are—granted, though, our deliverable is a product, not a perq). This all changed in July 2025, when they were acquired by Gamefound and, not coïncidentally, embarked on a wide-reaching Web site redesign.
At about that time, we began receiving messages that the “donate” button had disappeared and, worse, that out of those customers who had “donated” in the past, they were no longer able to do the same again. This was unsettling, to say the least, especially given that the comms and CRM parts of our fourfold plan had also proven to be unworkable in their form at the time. In fact, having three quarters of our workflow become almost useless within approximately a month was so psychologically distressing to some of our team members that going on cognitive strike in order to fix the problem and develop more products was considered the better part of valour!
The eventual decision was to switch to Buy Me a Coffee. We have nothing against Indiegogo; it has served us well since 2022, and those creators (particularly manufacturers) who want forward-looking goals, defined perqs, and staged funding, even if imposed, might find it to their liking. That having been said, for the last nine months, it has failed at the three things that we expect it to facilitate, and it has succeeded at a panoply of other things which are, in our case, superfluous to requirements. This is a matter of control vis-à-vis scaffolding, and we find ourselves regrettably forced to conclude that, this time, Gamefound got it wrong.
We acknowledge and regret that, of our many mistakes and undesirable habits, a predilection to undue silence is certainly one (paired with its twin, an insatiable yen for control). Notwithstanding that fact, we strenuously implore the reader to have some consideration for the thought that losing three-quarters of the payment-to-product pipeline infrastructure within a matter of months may be psychologically exhausting and a drain on mental bandwidth, which, in this situation, is optimally exercised by solving all the problems rather than by writing about them. Quod fecimus.
We concede that, perhaps, we might have been more communicative with the broader public, rather than offering nothing but radio silence to constructive status inquiries for the better part of a year, but we judged that being communicative and ostensibly helpful (rather akin to the synchronised clapping of circus seals), while ‘fixing’ the cascading failure in an objectively worse way due to sheer burnout would not be the best use of our manpower. In the words of the late, great Ralph Klein (peace be upon him), “any self-respecting team lead would have shot, shovelled and shut up”. Omnia in temporem.