Skip to main content

SIGCSE mailing-list workflows

Topics/tags: Overcommitment, SIGCSE, email, long, rambly

One of my professional responsibilities is co-moderator [1] of the SIGCSE [2] mailing lists. As a moderator, I have only a few primary responsibilities: I approve or reject posts, I handle new member requests, and I read and address bounce messages. But it’s not quite as simple as it sounds. Let’s consider the various workflows I follow. I assume my co-moderator follows similar workflows.

Posts. To handle an incoming post, I skim through it and see whether or not it seems to meet our policies. What are those policies? First, it should be relevant to the list. The SIGCSE-members mailing list is for discussion of computer science education. Most members realize that. But we get the occasional strange conference announcement. For those, I look for the keywords teach, learn, and educa… [3]. A conference announcement with none of those keywords is unlikely to be relevant to our membership [4]. There are a few exceptions for conferences that focus on students, but not many.

Next, I check for the secondary policies. Does it rely on an attachment? If so, it is not appropriate for our mailing list [5]. Does it quote an inordinate amount of material? If so, it makes the reading experience much less good for our subscribers who read in digest form [6].

There are also a few other cases, such as posts that seem irrelevant or intended only for the original poster [7].

I am then left with three primary options. I can approve the post, reject the post, or ask the author to repost. The first option is the easiest. I click a link in the email message I received, the system processes the approval, and the posting goes off to the list. The second option is a bit harder. I need to log in to the moderation site, find the post, and click the reject button. The third option is the most time consuming. In that case, I also write a polite message to the poster. For those who seem to have violated the policies, I explain those policies and ask them to repost. For those who may have inadvertently carbon-copied the mailing list, I check to see if that was their intent. In the few other cases, I do whatever seems most appropriate.

It is rare that I simply reject postings; I generally think our posters deserve the courtesy of an email message. But some messages are relatively easy to reject without additional followup. As I noted, I typically reject conference announcements that are irrelevant to SIGCSE. But there are a few that might be relevant; in those cases, I try to send followup. One-line messages that are intended for another poster and not for the list rarely warrant a followup. And, of course, spam just gets deleted.

Unfortunately, not all requests to approve messages on the listserv come to me through email. The listserv only forwards messages for approval if they come from current members who submit using their corresponding email address. However, if someone posts from an address not on the mailing list, the only way to know that is by logging in to the administrative interface [8]. About half the mail that comes in that way is from subscribers. The rest is a mixture of spam, messages from members who are using a different mailing address [9]. The interface is a bit different, but I handle these in the same way: I approve, reject outright, reject with followup message, request clarification, or check with the SIGCSE Chair [10].

Of course, those processes are only for the SIGCSE-MEMBERS mailing list. I also moderate the SIGCSE-ANNOUNCE mailing list. SIGCSE-ANNOUNCE is primarily for messages from the SIGCSE Board or the chairs of SIGCSE-affiliated conferences. Most people understand that. I end up rejecting a bit of spam (with no followup message) and some message from normal members who don’t understand the structure (with a followup message) [11].

Subscription requests. As a list moderator, I am also responsible for subscriptions to both the SIGCSE-MEMBERS and the SIGCSE-ANNOUNCE mailing lists. At present, there are two ways that people request subscriptions. Some email us directly at infodir@sigcse.org. Others click a link somewhere that automatically generates an email address for us. The process is much the same for both cases: I check whether they are current members and either subscribe them or send a message of the form These mailing lists are only for members of SIGCSE. You are not a member. Please consider joining.

Unfortunately, checking membership is not always a trivial operation. I keep a local copy of a relatively recent membership list on my computer, so my first approach is to grep [12] that list. Next, I check the ACM Database. Unfortunately, the ACM’s database is clunky. I need to log in twice. If I know the person’s ACM ID [14], the search is fast. If I have to search by last name, it’s slow. It’s been better these past few months, but I recall searches taking as much as a minute.

Fortunately, adding someone to the list is comparatively painless. If it’s one of those automatically generated subscription messages, I can just click a link. Otherwise, I log in to the Listserv interface [15], navigate to the subscribers page, and enter their subscription information. Fortunately, it’s a sensible UI: I can use one of two formats, both of which are easy to pull off a mail message: user@university.edu First Last or the more traditional First Last <user@university.edu>. In contrast, the Lyris list I manage at work requires that I enter name and email address separately, which is less convenient [16].

Bounce messages. Once in a while, we get bounce messages. It’s then our responsibility to read through those messages and try to determine appropriate responses. In some cases, the mail handler has decided that a message is spam and is informing us of that decision. That happens most frequently with job advertisements. There’s also at least one system that refuses to accept messages with shortened URLs. Those cases are easy; we just ignore the bounce [17].

Some cases are more complicated. If a message says something like This address is no longer active, I end up doing a bit of homework. First, I make sure the person is still a member of SIGCSE [18]. Next, I try to track down where they might be. If I’m lucky, they’ve updated their email address with ACM, even if they haven’t done so with SIGCSE. Those situations are rare. More often, I do a Web search and I look on LinkedIn. On occasion, I’ll write to someone at the department asking for a new address. Then I either update the address or write to the person at the new address asking whether I should update. It’s a bit of work, but I think it’s the friendly thing to do.

Questions and special requests. As you might expect, we get a variety of questions and special requests. Did my post appear? I’ve changed jobs; can you update my email address? I’m overwhelmed by email; please switch me to digest mode. Most of those requests are easy to handle and the listserv UI suffices. Once in a while, we need to do something that doesn’t quite fit within the UI. For example, some posters would prefer both an acknowledgement of their post and a copy. The UI assumes that you want only one (or neither). Hence, for people with those special requests [19], we get to figure out the instructions for command mode. Here’s one to achieve that last goal.

SET SIGCSE-MEMBERS ACK REPRO FOR rebelsky@grinnell.edu

Everything else. A variety of other tasks come up from time to time. For example, when I started, I needed to clean out the mailing list to eliminate people whose subscriptions had lapsed. I still do that once in a while. I’ve made some policy suggestions to the Board and spent some time discussing them via email. I’ve also participated in some policy discussions that stem from Board suggestions. For example, the Board suggested moving SIGCSE-ANNOUNCE from opt-in to opt-out. I thought we’d accomplished that last year, but it turns out that a miscommunication with ACM Headquarters left the change in limbo. It’s now gone through, but that’s raised new complications, such as members who want to get SIGCSE-ANNOUNCE with a different email address than their ACM membership [20]. In most of these situations, I end up composing and sending a variety of messages, not just as part of the discussion, but also messages to the various listservs.

And there you have it, my life as SIGCSE Information Director, or at least as SIGCSE listserv moderator. It’s not particularly thrilling, but it seems to support our community well.


Postscript: Here are some of the templates I use.

Request a repost

Dear,

Thank you for your submission to the SIGCSE-members mailing list. Unfortunately, I cannot accept it in this present form. Our digest subscribers continue to ask the information directors to try to limit the amount of quoted material from previous messages in postings to SIGCSE members. It does not appear that your posting (appended) needs as much quoted material as you used.

Please resubmit with only the essential information quoted.

Thanks!

-- SamR

Samuel A. Rebelsky
SIGCSE Information Co-Director
Professor of Computer Science, Grinnell College
rebelsky@grinnell.edu
infodir@sigcse.org

I usually add something about the topic of the submission. I may also add a suggestion on cutting.

After subscribing someone

I have subscribed you to the SIGCSE-Members mailing list. Welcome to the list. You should receive a confirmation notice from the LISTSERV.

Be sure to check out the policy statement [1] and the list archives [2,3].

Regards,

-- SamR

[1] http://sigcse.org/sigcse/membership/mailing-lists

[2] https://listserv.acm.org/SCRIPTS/WA-ACMLPX.CGI?A0=SIGCSE-MEMBERS

[3] https://listserv.acm.org/SCRIPTS/WA-ACMLPX.CGI?A0=SIGCSE-ANNOUNCE

Samuel A. Rebelsky
SIGCSE Information Co-Director
Professor of Computer Science, Grinnell College
rebelsky@grinnell.edu
infodir@sigcse.org

A response to a subscription request from a non-member

Thank you for your interest in the SIGCSE-members mailing list. I am sorry, but I cannot find a record of your SIGCSE membership. As our list policy [1] states, subscriptions to the SIGCSE-members and SIGCSE-announce lists are restricted to SIGCSE members. Although ACM manages SIGCSE memberships, SIGCSE memberships are separate from ACM memberships; you can be an ACM member but not a SIGCSE member (or vice versa). Could you check your ACM membership page to verify that you are a SIGCSE member?

If you are not a member but would like to become one, please see the SIGCSE membership page [2]. I would encourage you to consider SIGCSE membership. It’s relatively inexpensive ($25/year for electronic-only access and $40/year with printed materials; $12 for student members), and the SIGCSE mailing lists can be quite useful.

Regards,

-- SamR

[1] http://www.sigcse.org/sigcse/membership/mailing-lists

[2] http://www.sigcse.org/sigcse/membership

Samuel A. Rebelsky
SIGCSE Information Co-Director
Professor of Computer Science, Grinnell College
rebelsky@grinnell.edu
infodir@sigcse.org


[1] Officially, my title is Information Director or Co-Information Director or perhaps Information Co-Director. The Information Directors primary responsibility is to manage the SIGCSE mailing lists. We also do a few other things, such as suggest and implement policy.

[2] SIGCSE is the Association for Computing Machinery (ACM) special interest group for Computer Science Education.

[3] Searching for educa handles educate and education.

[4] I don’t think there’s been an ACM Federated Conference while I’ve been a moderator, but I’d approve that. I’m not sure that my predecessors would have.

[5] The SIGCSE board has approved some specific exemptions to that rule, such as the SIGCSE history

[6] Many mail readers hide the quoted material, so not all of our readers know when they are quoting a lot of material. But our typical digest readers don’t have the same experience; they have to wade through too much repeated material. We’d like to make their experience better.

[7] I’m surprised at the number that seem to be a reply-all. Some clearly address a broader audience than the original sender, but many do not.

[8] No one told me that when I started; it took me about six months to figure it out.

[9] E.g., name@university.edu rather than name@cs.university.edu or vice versa.

[10] Didn’t I just say that there are three primary actions? Now I seem to have listed five. But that’s okay, I’ve split reject into two parts, and I’ve added the rare, but still important, check with SIGCSE chair.

[11] At one time, I was one of those members who did not understand the structure. I think that’s how I ended up in this position.

[12] search

[14] Only in the cases in which they’ve provided it via their email message.

[15] In most cases, I’m already logged in. I spend too much time on the Listserv.

[16] It doesn’t really matter when you are adding one or two folks. But the additional time and effort builds up as you add more and more.

[17] Well, I ignore the bounce. I believe some past holders of this position didn’t like us being identified as spam and sent followup messages.

[18] When I started, we had not cleaned out the lists for some time. We’re better now, but don’t always get to it.

[19] Yes, I’m one of those people who wants both a copy of my messages and an acknowledgement.

[20] Nope, not possible, at least at present.


Version 1.0 of 2018-11-17.