If you are keen, got a job the boss wants done concurrent to the ongoing dev that I think would suit you. (unlike last time we spoke about a module this one doesn't depend on the management to get moving 😒 so you could get started tomorrow if you like)
tldr: make backend to monitor, parse, store, and retrieve emails
Core part of the planned CRM system is centralised communication tracking, and a core part of that is sent/received email integration. Visit a customer's card to see all interaction between $businessUsingStockman and the customer in question, easy to catch up on the relationship with confidence. Sent/Received emails is the part we want to get built now, rest to be done in the future.
How this is achieved is largely dealers choice given the following:
Firm reqs
- Process new received mail within 2 minutes of arrival
- Associate with company (tblCompanyDetails.CompanyIdentifier) automatically (should be a simple match of from address to company contact email address (tblPerson.email)
- Searchable/filterable on some common fields ie recipient, sender, date, subject
- Mail data must be immutable ... (this is the real meat of the task right here)
- Some emails may be sensitive, need a way to prevent for example a standard sales user from seeing mail between CEO and customer
Soft reqs
- ability to reassemble into an email (edited)
12:48
I've already got the gsuite/gmail integration working, extremely basic in the backend though but the main point is that the domain wide delegation stuff is all tested and working
12:49
so the whole thing can be achieved using just the GCP service account, no need for users to give authorisation, no need for any user input or action whatsoever
jcro 12:57
the gmail go lib makes this stuff pretty good to work with
12:59
and it includes subscription/watch methods so the mailbox monitoring should also be a relatively easy thing to achieve
13:00
the big problem to solve is how to store and serve the information
jcro 13:08
my first thought is to store the headers and MIME text part in the database since they'll be small enough to be reasonable and then we can use them to search/filter and importantly to display in a performant manner
13:09
and then store the attachments in a sub-bucket of that company-card's file bucket
13:09
link the two groups of data together via the email message's UUID
13:10
fetch attachments on demand
13:10
has a few minor bonus side effects too (edited)
13:15
anyway I'll stop talking at you, I'm sure you get the idea - we can discuss more if you're interested 👍
s1へ 13:29
reads
13:31
yeah your idea of storing the email content is spot on - that's exactly what popped to mind reading the spec
13:35
- SLA in terms of processing time is fine
- company association might require some additional functionality (e.g. if someone emails from their personal account, maybe that should go into some kind of reconciliation queue?)
- search/filter is easy peasy
- immutable is fine - i propose two levels: 1) is that it's immutable for everyone except database administrators (i.e. us) via not exposing write mechanisms, 2) is that it's immutable via some kind of hard-to-falsify-after-the-fact log (think blockchain - a string of hashes)
- sensitive content is obviously subjective, so might also require some kind of queue for confirming they contain no sensitive information (this would be an interface used by people who are allowed to see such information maybe?)
jcro 13:35
schweet, I haven't had a look at whether or not the gcp file buckets are suitable for this kind of profile but I'm sure gcp would have something that suits
s1へ 13:36
not sure what the "reassemble into an email" thing means - is this like a "send response" feature where we retain the email thread in the reply?
jcro 13:36
naw, just that people will want to "see the email" if asked to produce it
s1へ 13:37
oh right, that's fine
jcro 13:37
i.e. if asked to by lawyers
s1へ 13:37
yep, so that's more for audit
13:37
all good
jcro 13:39
(e.g. if someone emails from their personal account, maybe that should go into some kind of reconciliation queue?)
if no mail address match found I'm thinking it goes into a 'spam' style collection that purges at 30days of age
13:40
users can go look at that list of mails (showing only those addressed to or from their own email address of course) and click a button on a message to send that message through processing again
13:41
so they could see it didn't attach, go create a contact on the customer card, tell the system to process/re-evaluate that message, and this time it'll attach as expected because it now matches a contact
s1へ 13:42
that'd work
jcro 13:42
that stuff is easy to tack on though
13:43
so the fine user interactions aren't too important as long as we have the structure required
s1へ 13:44
sounds reasonable
jcro 13:44
2) is that it's immutable via some kind of hard-to-falsify-after-the-fact log (think blockchain - a string of hashes)
Is this because you want provable immutability?
13:45
Immutability I had in mind was essentially preventing std user actions from changing the data
13:46
most common and obvious desire being: we don't want them to be able to delete emails
13:46
delete it from your mailbox if you like Mr. Andy Mc'Saleserson, it won't save you
s1へ 13:47
yeah, with the goal of changing the risk profile - if there are questions of reliability in terms of provenance, a blockchain-like string of hashes approach would make it very hard to falsify past emails, and it would get harder as time goes by
13:47
right, so it just depends on the risk profile
13:47
if the business wants to protect itself against end users, simply restricting their ability to modify the content is fine
jcro 13:48
yeah the only desire expressed to me was preventing end users from destructive acts (edited)
s1へ 13:48
if the business wants to protect itself against the people running the system (us), we'll need to take some more steps, and give management a way to confirm that specific records haven't been modified
13:49
cool, sounds easy then
13:49
might be worth clarifying though, if it hasn't been explored already internally
jcro 13:49
the reason I reckoned it'd be the meaty part was because that requirement alone means we must take control of the parsing, storing, and querying instead of leaving it up to gmail
13:50
otherwise we could just retrieve message from the users mailbox
s1へ 13:50
oh i'd be storing it locally anyway, just to make sure a gmail outage doesn't take down the company's comms history
jcro 13:50
if the business wants to protect itself against the people running the system (us), we'll need to take some more steps
oh they have absolutely no concern about that direction at all
s1へ 13:51
nice, easy then
jcro 13:51
so yeah dodged that one
jcro 13:56
RE the queue type system to allow some pre-filtering of sensitive info:
I'm thinking this sort of approach
Standard users: Mail is attached and visible by default, with 30 minute window from insertion to rescind the public display flag
Manager/sensitive users: Mail is attached and invisible by default, with unlimited window to grant the public display flag
s1へ 13:57
right, so depending on the recipient of the email, we'll mark it as public or not public at the start
13:58
that'd work
jcro 13:58
messages that are attached but invisible could display the To and From and Subject but omit the body
13:58
but that's nonstructural I guess (edited)
14:00
right, so depending on the recipient of the email, we'll mark it as public or not public at the start
hmm, yeah a flag for public and a field for a list of users granted specific access
14:02
id: 123
message: { ...message object... }
public: false
explicitViewers: "AdamA, BarryB, CeciliaC"
s1へ 14:03
how would the user list be populated? a default for each recipient with per-message override or something? (edited)
jcro 14:03
nil by default
s1へ 14:03
makes sense
jcro 14:04
and user who 'owns' the email (To or From them) or someone who has administrative privs can go to that message and "Add view user"
14:04
kinda like sharing on a google doc (edited)
14:06
2 files
image.png
image.png
s1へ 14:10
nod
jcro 14:12
I can't really see a way it'd be a factor in this structure but I intend to integrate this into gmail later down the track, in the style of "click a button/icon on the message to change it's visibility/which users have view permission" type stuff
14:12
so maybe we want to record the gmail message UID
s1へ 14:12
is that a gmail thing? or would that be a browser extension or something?
14:12
we can certainly record whatever data is necessary
jcro 14:14
I haven't looked properly but I think you can do it 2 ways, full on gsuite addon which integrates into the gmail interface (surprising amount of agency they give you there), or a "lite" version of an addon (appears in the sidebar and is kinda seperate from the rest of gmail UI)
s1へ 14:14
neat
jcro 14:15
image.png
image.png
14:15
the Keep stuff there is the second type
s1へ 14:15
so ballpark (off the record) i want to say two weeks delivery time, $10k budget
jcro 14:16
yupyup
s1へ 14:16
is that roughly aligned with business objectives?
jcro 14:17
can't say, didn't actually give me an expectation there - but my gut feel is that'd be fine
14:17
90% confidence
s1へ 14:17
i'll have to write up a set of completion criteria to make sure everyone knows what they're getting into, which i'll cost into the end price, and which i think will fit into $10k just fine
14:18
but let's keep that a teensy bit flexible for the moment
14:18
feel free to float the number around, but don't hold me to it until i've written that specification and we've made sure it meets all the required objectives
jcro 14:19
Yep, this can be pretty fluid I'd say as well, since it's not a 'sealed and finished product' type thing
s1へ 14:19
i would be very surprised if it went further than $15k without some serious curveballs
14:19
and even that is like, if someone asked for some really complex buisness rules
14:20
(something that hasn't already been mentioned i mean)
jcro 14:21
yeah I don't see any requirements jumping out
14:21
can't imagine what one could even be realistically
14:22
the scope creep with this sorta thing would surely be the user interaction layer
s1へ 14:22
something nebulous like "anyone i've ever talked to should be counted as one of my contacts, except if they mention the n-word"
jcro 14:23
and the aim is to punch out a super basic version of that
14:24
and I'll apply the fiddly coat of paint and dressings like filter/search/permission/sharing/blahblah later when I start designing the CRM properly
s1へ 14:25
yeah - of course i'd try to structure things to make that as easy as possible
14:25
in other news, did david ask you about his "phpbb for business" concept?
jcro 14:25
yeah robust structure is the main focus at this point for sure
14:26
Don't think so? phpbb for business sounds like a terrible idea though
s1へ 14:26
well, i'm paraphrasing
jcro 14:27
is this david banham?
s1へ 14:27
he was basically saying that the slack-only monoculture in some startups is actually stifling to effective communication
14:27
yep
jcro 14:27
ah, haven't opened VS in about a year
s1へ 14:28
i have it as a tab in my slack client, but i only open it when i get a notification heh
14:28
he had kind of a cool thought process around focussing more on discussion threads
14:28
basically slack tried to make that happen, but they fumbled bad on it
jcro 14:28
what's the main premise of the slack-only is bad theory?
s1へ 14:29
basically that the concept of topic permanence is almost void in slack
14:29
if you chat about something important and reach a conclusion, it can be pushed out by shitposting
jcro 14:29
discussion threads in business would actualise as entity-based wouldn't it?
14:30
so like a much evolved version of the comment system in stockman
14:30
topic = business entity
s1へ 14:30
so without judicious use of threading (annoying in slack) or lots of channels (annoying maybe in general?), a real-time chat system has flaws that make it unsuitable for business use past the point of maybe 5 active users
14:30
yeah, something like that
jcro 14:31
yeah I agree with that
14:31
you see enough evidence in devanz, it's just that it doesn't matter much
s1へ 14:31
he has some interesting thoughts on the topic, and i had a good discussion with him about it the other day
14:31
oh yeah but devanz is just 24/7 shitposts
14:31
so it works fine for that
jcro 14:31
I'm pretty interested in this actually
s1へ 14:32
you should ask him about it - you've got the same level of context i do with him
14:32
it sounds like he actually wants to take a stab at it
jcro 14:32
communication and singularity of direction/intent/goals is a problem in this business
s1へ 14:33
i kind of want to think of it as "4chan for business" - i think a lot of the properties of a *chan-style board would translate well to business
14:33
forced thread topic, forced attachment for initial post, bump (thread size) limit
14:33
but i expect he doesn't want to couch it in those terms
jcro 14:34
I would never have thought getting everyone on the same page would be such a difficult and important task, but turns out a bunch of 50 year olds with no tech inclination will passively resist it
s1へ 14:35
they sure as fuck know how to use facebook though
14:35
and that has a similar model
jcro 14:37
hmm, forced attachment places a barrier to discussion
s1へ 14:38
it does - but an attachment in this context could be a client email, or a specific sale, etc
jcro 14:38
which might actually be desirable in some cases, but dampen comms through increased effort cost
s1へ 14:38
a discussion should have a context
jcro 14:39
"Next FY budget discussion"
14:39
entity that doesn't exist yet
14:40
but I'm probably being too narrow sighted on the "attachment" meaning there
s1へ 14:41
the attachment could also be a meme of donald trump on a motorbike
14:41
so you know
14:41
it's fluid
jcro 14:41
I can see the benefits for sure, but since it's 3rd party the lack of integration might be a killer
14:42
fluid
triggered
s1へ 14:44
heh
14:44
i told him his biggest hurdle would be SSO
14:44
but i like the concept
jcro 17:04
just got off the phone, all seems good, no new requirements and they're happy to move ahead
17:05
they did have one "can you make it so"
s1へ 17:05
cool, i'll write up a more formal document this weekend, then if we can get it approved early next week we can get cracking
jcro 17:05
"A button or tickbox when you're writing the email so make it public or private"
s1へ 17:06
hrm
jcro 17:06
so if you could just inject some custom code and interface into every email client on every platform
17:06
that'd be great
s1へ 17:06
i guess we could put replies into their own queue
17:06
give them a longer timeout
jcro 17:06
obviously I told them 'no not really'
s1へ 17:06
make sure it's effectively communicated to staff
jcro 17:07
I offered that we could make any email that is BCC'd to $systemAccount be flagged private etc
s1へ 17:07
"if you want your email replies to customers to be private, please log into stockman within $x time of sending them"
17:07
that would work too
jcro 17:07
or include a code in subj or body like [!priv]
s1へ 17:07
oh sure, or even the other way
jcro 17:08
make it a macro or whatever for ease
17:08
they didn't like the idea of having to go into the customer card and toggle the email they just sent
s1へ 17:08
make it private-by-default, and require a tag (or follow-up activity in stockman) for making it "public"
jcro 17:09
ah but then we fall foul of the 'path of least resistance'
17:09
and have no public data
17:09
these slaves secondaries respond to the lash
s1へ 17:10
or an email subject tag - user receives this steel rod gave me internal bleeding, replies with sales: re: this steel rod gave me internal bleeting
17:10
we can definitely shoehorn some magic into the email content
jcro 17:10
hmm
17:11
that'd require outgoing interception
17:11
so relaying through a proxy we build
s1へ 17:11
if we can fetch the inbox we can fetch the outbox, right?
jcro 17:11
moving parts
s1へ 17:11
via gmail?
jcro 17:12
yeah the gmail api subscribe method feeds you every event that occurs if it changes the mailbox state (edited)
17:12
so incoming, outgoing, draft, delete
17:12
etc etc
s1へ 17:12
yeah i figured we'd be keeping any ingoing/outgoing mails
jcro 17:12
but if they write an email in outlook and hit send
17:12
we can't alter the email the customer gets right?
17:12
I assume
s1へ 17:12
no we can't, i don't think
17:13
but as long as it triggers a change in gmail, we can log it
17:13
(and do things based on it)
jcro 17:13
yeah
17:13
oh I've gone down wrong headpath here
s1へ 17:13
oh?
jcro 17:13
How does this identify an email as desired private?
s1へ 17:14
if they put some magic in the email body, subject, or headers, we can react to it
jcro 17:14
ah yeah, that was my original thought
s1へ 17:14
or if they quickly (within $x minutes) log into stockman and change it
17:14
if they forget to put it in the email
jcro 17:14
either a BCC address or a code in subj/body
s1へ 17:14
yeah
jcro 17:15
it's just not really possible to make that user friendly
s1へ 17:15
bcc to secret-cock-story@fagstore.com
jcro 17:15
but that's their problem, if they want it private, they'll do it
s1へ 17:15
yep
jcro 17:15
you won't believe the last picture!
s1へ 17:16
we can hold the email as private for some specified period
jcro 17:16
hmm yeah actually that's a very good idea
s1へ 17:16
release it as public when that period ends if they don't otherwise tell us to retain it
jcro 17:16
just for the outgoings
s1へ 17:16
yeah, like 30m or something
jcro 17:17
push event to stockman, alert modal asks them to give a YES/NO
s1へ 17:17
if you send an email at 170km/h and go "oh fuck" that gives you enough time to pull over and get your laptop out
jcro 17:17
could even go further and if no push event subscriber is present, send an email and extend the grace period to 12 hours
17:17
or 6 working hours, w/e
s1へ 17:18
yeah that'd work
jcro 17:19
trying to imagine if that would be incredibly annoying or not
17:20
I think the avg sales person sends ~50 emails a day
s1へ 17:20
not sure if we keep state of subscribed viewers, but it doesn't sound hard to do
jcro 17:20
well they poll right
17:20
and it's an authenticated request
s1へ 17:20
if we don't have a record of the user having pinged any application server in the last 5 minutes, send them a push notification
17:20
seems easy enough
17:20
yeah
jcro 17:20
so we can just use lastActivity column in tblLogon
s1へ 17:21
something like that
jcro 17:21
no work needed that way
17:21
active in last 10 minutes = he's there
17:21
not as accurate but this ain't a sniper rifle
s1へ 17:21
seems like a reasonable assumption
jcro 17:22
if we don't have a record of the user having pinged any application server in the last 5 minutes, send them a push notification
it'd be if we do wouldn't it?
17:22
if no activity then they don't have stockman open, so send an email and give more grace period
s1へ 17:23
oh, right
jcro 17:23
ok so that kinda logic - much work?
s1へ 17:24
nope, sounds fine to me
jcro 17:24
righto cos it's not a hard req that one
s1へ 17:24
mmm
17:24
if it's desired though, it should go into the requirements
jcro 17:24
I'll go find out if a notification/modal/whatever for each sent email is annoying or not
s1へ 17:25
i don't want to disappoint is all
jcro 17:25
yeah of course
s1へ 17:25
if that functionality is required for the business, i'd like to make sure it's not left out
jcro 17:25
I told them that we could do it via a BCC or string code
17:26
and they were happy enough
s1へ 17:26
so if it's important, i'll consider it in writing up the specification
jcro 17:26
I'll give them a ring now
s1へ 17:26
nod
jcro 17:35
oh boy. True to form we have a slight change. Bucking the trend however - it means less work!
17:36
Sandy prefers typing [priv] or something into the email somewhere, so no need for the notification/push event stuff
17:37
furthermore, she doesn't want anyone to have the ability to change a message to private except a couple of select people
17:37
so for the standard sales staff, all public, all of the time
s1へ 17:37
alright, sounds good to me
jcro 17:38
which frankly I think is not going to be popular
s1へ 17:38
2bad
17:38
you want a different environment, get a different job
jcro 17:38
no more talking shit with your customers/suppliers bois
s1へ 17:38
or go back to 1965, fucking boomer
jcro 17:39
and I've seen some emaill chains "check this video out" that are going to come up in meetings...
s1へ 17:39
trump-on-harley.gif
jcro 17:39
this is a famously boys club industry
17:39
without the as much sexism (edited)
s1へ 17:40
where's the sexism if everyone is a boy
17:40
equal footing
jcro 17:40
"the pair on the new receptionist mate phwoar"
s1へ 17:41
"ha ha thanks for this correspondence, unfortunately i have no way to mark it as private"
:joy:
1
jcro 17:41
anyways, not our problem, we just fashion the bayonets
s1へ 17:42
mayonez was not designed to be spread with the bayonet, but the bayonet is perfect size for spreading mayonez
17:42
thank stalin 🙏
jcro 17:42
this means we can axe the whole grace period thing for public flag users
s1へ 17:43
yep
17:43
management override only
17:43
does that mean we need a way to alert management? or does that happen out of band?
jcro 17:44
and management want it to be public by default
17:44
and they'll do the code or toggle in stockman for priv
17:44
so essentially no need for a distinction between the two users anymore
17:44
private toggle ability can just be another permission in the existing permission system
s1へ 17:44
"fuck, fuck, sandy, help, i just told the guy at tasco that their receptionist was fuckin 9/10, yeah, yeah, no, yeah no don't worry what 9/10 means, can you just mark it as private"
jcro 17:45
"sure we can fix this Steven ... so what am I out of 10?"
s1へ 17:46
"sandy if you can bump me up $5k next year you're an 11 easy"
jcro 17:46
and then he just jumped out the window, broke 7 bones, crazy
s1へ 17:46
that's just the way the cookie crumbles
17:47
super weird, totally unpredictable
jcro 17:47
https://devanz.slack.com/archives/D53A04Z6Y/p1602744237124600 out of band
s1へ
does that mean we need a way to alert management? or does that happen out of band?
Direct message | 15 Oct | View conversation
s1へ 17:47
poifect
17:47
so users with that permission will have a way to mark things as private
jcro 17:48
can happen either via code in email, or by toggle in stockman interface for emails in customer card (edited)
s1へ 17:48
and it'll be up to the people without those permissions to get on the horn if they forget the code
17:48
does this apply to a whole thread? or just individual replies?
jcro 17:48
yep
17:48
oooo
17:48
hmm
17:48
I think the whole thread
s1へ 17:48
and is it retroactive?
jcro 17:49
safer
s1へ 17:49
i.e. mark thread private -> whole thread disappears
jcro 17:49
shite, good point retro is tricky
s1へ 17:49
in case a customer says something fucked
jcro 17:49
I reckon theres a good case for the first half of thread being 👍 public and the rest being 😮 not so
17:50
so how much of a bitch is it to private a thread from this email onward
s1へ 17:50
really it'd be more like a specific message forward
17:51
or maybe "the message subsequent to my last reply" forward
17:51
imagine you've got EmployeeA and CustomerA having a chat
17:51
CustomerA says "fukken alice is hot shit hey"
jcro 17:52
if thread == hasSomePriv { for each email in sorted arrray { if email date > emailContainingPrivDate then hideThisCriminalShit()
s1へ 17:52
EmployeeA wants to put the kibosh on that shit
17:53
then wants to reply with "fuck man i know she's asian but she's not instantly a slut"
jcro 17:53
lol
s1へ 17:53
how do we allow the illustrious fagersta employee to retain both his dignity and that of the customer
17:54
i'm probably thinking too hard on this
17:55
maybe if the customer makes bigoted remarks they deserve to be cancelled
jcro 17:55
public announcement that he's "chosen to explore other opportunities" and has "enjoyed his time at fagersta"
s1へ 17:56
"'s incredible journey"
jcro 17:56
the likely case would be a continuation of the discussion at a higher executive level
s1へ 17:56
automatically delete from salesforce
17:56
https://ourincrediblejourney.tumblr.com/
ourincrediblejourney.tumblr.comourincrediblejourney.tumblr.com
Our Incredible Journey
Putting the ack! in acquihire
jcro 17:57
so for some people the thread ends at n-4 and for some it's at n messages (edited)
s1へ 17:57
right, so again, allow management to choose which messages are private
17:58
nothing to hide, nothing to fear
jcro 17:58
yep, but the feature here is partial threading
s1へ 17:58
that policy has worked for the US government for 70 years
jcro 17:58
as in, first 3 emails are public, subsequent are not (edited)
s1へ 17:58
hrmm right
jcro 17:58
it's for their own good
s1へ 17:58
so customer contact, follow-up, response - public by default
17:59
subsequent messages - private by default, optionally public at management discretion, or at employee discretion within 30m (of each message) (edited)
jcro 18:00
we don't want the kind of people who use obscene language like "mate you guys got fucking slaughtered on the weekend, wooden spoon again!", "I swear our on ballers were a pack of useless cunts"
18:01
meanwhile, I'll have my email address in the whitelist to skip the fuck out of this system
18:01
that's an actual requirement that I didn't think to mention btw
18:02
we don't want every single account in the domain monitored
18:02
otherwise stockman@fagersta.com.au and fagerstasystem@fagersta.com.au will flood everything with spam
s1へ 18:03
wait so does that mean that for each thread, message n>3 with non-management recipient or sender should be public by default? or private by default? or does it not matter?
18:03
assuming that thread with n=1 is customer-related
jcro 18:05
https://devanz.slack.com/archives/D53A04Z6Y/p1602745181141200 yes except for this part "or at employee discretion within 30m (of each message)" - that's what I was saying before we no longer need that 30min part because management have decided that no one except special permission holders can make a message private
s1へ
subsequent messages - private by default, optionally public at management discretion, or at employee discretion within 30m (of each message)
Direct message | 15 Oct | View conversation
s1へ 18:05
okey dokey
jcro 18:08
if the $privateKeywordString is in email 4 of email chain 1, 2, 3, 4, ..., n then x >= 4 = private (edited)
s1へ 18:09
can we make that string the n-word
18:10
if a fagersta employee wants to make a conversation private, they have to say the same word method man does
jcro 18:10
however let me say - if this is more than 30 minutes effort then I'd say just make the entire email chain private once a $privateKeywordString is included
18:10
including retro
18:10
hahah
s1へ 18:11
"what, you're better than method man? wow you really are a racist"
jcro 18:11
"i solemnly vow I'm up to no good"
s1へ 18:11
https://www.youtube.com/watch?v=WJOl3vWPfbc
YouTubeYouTube | UPROXX Video
Wu-Tang Clan - Method Man (Uncut)
18:12
but sure, we can define a content/subject/to: trigger
18:12
if you trip that switch, the whole thread becomes private, pending management intervention
jcro 18:13
shweet
18:15
one more minor thing to look out for - I have not confirmed that an email with 2 recipients say bill@fagersta.com.au and ted@fagersta.com.au possesses a common identifier in each recipient's mailbox
18:16
I presume it does, but I haven't yet checked
18:16
if it didn't that'd make filtering duplicates an absolute cock
18:16
but y'know, surely not
s1へ 17:48
hey look, notifications
jcro 17:48
DING
s1へ 17:48
right so i'm pretty comfortable with the description as it is
17:49
so i guess the ball is in my court to write it up and send over a more detailed spec
17:49
i'll slide it by you for review first, so we can work out the details
jcro 17:50
that all sounds fine to me
17:52
my idea of it isn't too detail heavy as far as 'spec' goes
17:53
largely 'need this feature, works or doesn't' (edited)
17:54
so shout out if you start getting bogged down in details, it might be the case that it's not needed
17:55
(just don't want you to spend time writing it up and later think, "damn, rather not have done that") (edited)
s1へ 17:56
nod
jcro 17:59
oh and you might need some info on how to unfuck the environment into workingness so yell out before you intend to jump in and I'll probably save you some temple-rubbing
17:59
did a 4am "looks like I need to move to webpack4 and friends" on friday
s1へ 18:01
i did one of them recently