[Author, Committee]s guides to [IEEE-CIS, INNS] conferences

Authors & Publish chair page, blog	Paper formatting page, blog	Initial paper submission chair, page, blog	Final paper submission chair, page, blog	Problematic papers - corrections page, blog
	Author [PDF,CrossCheck] tests page, blog	IEEE PDF eXpress - paper format chair, page, blog	IEEE electronic Copyright (eCf) chair, page, blog	Attendee downloads of papers page, blog
	Conference registration page, blog	Travel visas to Hungary page, blog	Conference presentations page, blog	HELP contacts WCCI2020, system
Non-Author actions	Paper reviews - authors' perspective page, blog	IEEE CrossCheck text similarity chair, page, blog	IEEE Xplore web-publish chair, page, blog
	IEEE Conference Application chair	IEEE Letter of Acquisition chair	IEEE Publication Form chair	Software systems page

Conference Guides for [IEEE-CIS, INNS] conferences : Although this guide was initially set up for IJCNN2019, and was used for IEEE WCCI 2020, hopefully some of the basic information may still help [organising committee members, authors] up to 2023-2025 (assuming a half-life of links and process information of 3-5 years?). The Authors' and Publications menus have been combined, allowing authors to see the Publications Chair perspective as well. I am no longer keeping this up to date. Bill Howell, 11Dec2020
Summary : Attendee downloads of conference papers

Summary of the sftp server [setup, operation, usage, problems]
Key recommendations for future online paper access
Comparison : DVD, USB memory stick, Online access
User access versus security
Common access problems
Users rarely say what the problem was
Attendee survey of sftp usage and results
Server accesses - cumulative data
User access versus security, [Open, Google]-Docs
Security-driven complications, future improvements
Supporting software
A rough idea of the work involved
Setup and hosting of the sftp server

This [very incomplete, random, scattered] summary is targeted at future [General, Publications] Co-Chairs, but it may be of interest to some of the users as well, especially those who would like to understand more about the details. It is a separate web-page from the main "Attendees downloads of conference papers", as it really isn't useful to attendees needs when it comes to using the sftp server.

Many comments below are "hard-nosed". I've never been an IT specialist or tech support person, but [suspicions, aversions] that I already had about those roles have been [hugely, over-whelmingly] confirmed by my experiences with the email exchanges for all of the Publications Chairs' duties, but especially with the sftp server.

Bill Howell, 08Aug2019

Summary of the sftp server [setup, operation, usage, problems]

For this second year of online paper access (after WCCI2018 with Whova), user familiarity was probably even more of an issue with the sftp server. ~55% of 951 attendees (plus [INNS, IEEE-CIS, IEEE-MCE] contacts) who accessed the system. Most accessed the system within a week of the start of the conference. I assume that the vast majority downloaded the papers that they were seeking (most often all papers, but sometimes just a few?), or that was convenient in light of the effort required (which would have excluded some).

As the sftp server was set up in my own ssh server beyond a reasonable last date for getting started, everything was done in an extreme rush, and many normal capabilities were not addressed for lack of time. Still, neveer having run an sftp server before, it was [easy, robust, reliable] to run, and offered maximum flexibility and power, beyond the [Open, Google]-Docs standard setups. This was key to attaining a level of security that IEEE required, in my not-so-sure interpretation.

Overall, administering the sftp system was a BIG (excessive) investment of my time (~300 hours - see below). This is perhaps the greatest barrier to adopting it, but as long as :

attendees become aware that papers will be online
attendees become familiar over the years with whatever system is used to provide online access
information provided to attendees is adequate
[software, planning, instructions] are passed on to subsequent Publications Co-Chairs

then the time requirement will drop to only a fraction!

Preferably, an online access system will be provided by the IEEE-MCE group, that meets their needs, gives them direct [monitoring, control] if necessary, and avoids requiring each new Publications Co-Chair to be a quasi-expert.

Key recommendations for future online paper access

Online access should be the primary means of providing future conference papers in my opinion. Many, and perhaps most, will strongly disagree with this statement. However, the arguments that have been brought forth against onlin access seem mostly to be based on the understandable preference to stick with [familiar, easy] USB technology. Those arguments neglect key reasons for going with online access, and ignore the great advantages that are offered (pre-conference availability, incremental updates of edited pdfs, etc). Perhaps my opinion will change when full costing figures come in, including sftp [setup, operating, bandwidth] costs.

USB memory sticks might be an option, although IEEE instructions state that only one or the other can be used. If attendees request USB sticks, that should be done at the time of online registration, at least 6 weeks prior to the start of the conference. An extra full-cost fee would apply, including personnel time, ordering-shipping costs, volunteer bag stuffing, cost of the media and company to put the papers and program onto the USB stick, etc, etc]. This could be VERY high for small volumes. On-site and late registrations would not have the USB memory stick option, but again, could access papers online starting 2 weeks after the end of the conference. USB pricing should also be potentially considered as an extra source of profit to help fund the general conference expenses.

ONE Publications Co-Chair should have online paper access as their SOLE responsibility (not the USB memory sticks - which is done by the management company) - The Publications area has greatly expanded. It requires several dedicated people, as suggested elsewhere.

Better automated scripts are required to control "excessive downloads" on the basis of the volume of download for each user (Gbytes).

No technical support - There should be ALMOST NO volunteer technical support for online paper downloads by attendees, with the possible exception of providing [username, password]s to NEWLY registered attendees during the conference. Attendees can pre-register, so if they fail to do so, they should NOT expect volunteers to waste their time during the conference! Instead, access can be provided within 2 weeks of the end of the conference. Overwhelmingly, attendees did not require support, and the root causes for support requests clearly indicate that almost all are due to :

users not finding it important enough to modest time into [reading instructions, setup, downloading]. If it's not worth their time, volunteers should not be putting any time into doing their work for them.
poor basic skills, in the most surprising ways

User support should therefore ONLY come from :

the enormous amount of internat blog information available for sftp and other servers
their own colleagues
their organisations' IT support groups
As a last recourse, via volunteer grad students, a 50 $US starting support fee (50% to students, 50% to cover general conference costs), plus 50 $/hr per additional hour of work, with no guarantee of a fix. Fixes cannot be guaranteed as support specialist outside of IT groups cannot generally access details of user systems, nor can they bypass [software, security] controls.

Lastly, testing of the system via Organising Committee members is a very important step! Fix things before they get out to the users!

Attendees' rights to papers - A number of attendees keep refering to their "magic rights", so these should be made clear :

The ONLY right that users have is to download papers, without support, during the pre-stated period of accessibility of the server. Extensions of the download period are a grace, not a right.
If they miss the online download period, they have no further right to paper access.
Technical difficulties do not give additional rights to the users (eg mail a USB with the files) - they are responsible for handling their own problems with the system.
User rights are subordinate to their responsibilities to protect the copyrights. Failure to do so also jeopardizes the inclusion of their paper in the IEEE Xplore online system.

Comparison : DVD, USB memory stick, Online access

... to be added later ...
WCCI2016 Vancouver - I never did bother to extract USB files.

Access versus security

... to be added later ...

Common access problems

...paste list here... plus link to blog!
http://www.billhowell.ca/Neural%20nets/Conference%20guides/Author%20guide%20website/Attendee%20downloads%20of%20conference%20papers%20blog.html

Note that negative comments domninate. I have only a few positive comments about the online access. But my guess is that happy people don't complain?

Users rarely say what the problem was

In spite of the number of people whose problems were resolved, I have almost no information on that the [cause, solution]s were! Worse, in spite of repeatedly requesting it, I received basically NO real information about what the problems were, and how attendees solved them.

I suspect that a main problem is embarrassement over having taken so long, even with support, to get it to work.

In the end, user feedback is typically lacking in substantive detail. Substantive information might only be available if exact [keyboard, mouse] information is available, with a system capable of distilling down the details (or someone with an enormous amount of time to spend on boring things).

Attendee survey of sftp usage and results?

Robert Kozma raised the issue of doing a survey of users of the sftp server, which I think could be useful for future conference planners. It should be done by a person who was not part of the Organising Committee (and especially not by me!).

The bane of [static, multiple-choice ] surveys - As a cautionary statement, I did professional surveys in the past, and I have little use or patience for :

questions & answer selections (especially multiple choice), without free-format text responses
static [questions, approaches]

UNLESS they only a part of a far more [powerful, general] approach. Statistical significance is a key argument for using such surveys, as is the comfort level of doing such surveys in an area that has already been carpet-bombed with far superior surveys in the past, and which already benefits from insights of a few rare individuals who have insider [knowledge, experience].

Unless your are using surveys as a form of very common "conclusions driven research", in which the results are a given, such surveys are almost a guarantee of being blind to significant results and insights that don't conform to belief systems.

Server accesses - cumulative data

Attendee and hackers accesses (08Aug2019)

IJCNN2019 sftp server cumulative accesses

# users total	951
# users accessed	520
% users accessed	54.7%
user logins	2,796
hacker attempts	118,709
# of hacker IPs	3,533

I had initially expected ~75% of ~951 attendees, INNS, IEEE-CIS, other] to access the server, but that was not the case. There have definitely been complaints about the system, but at a low %. Many attendees might not have bothered to access papers because of the sftp arrangement.

Distribution of the number of accesses per user

There is no way at present of knowing which users completed their downloads, as I haven't tried to build scripts for that. But the following gives some idea of how many accesses people require. As you can see, most are below 7 accesses, above which users are considered to have "excess accesses"even though many uses legitimately require many accesses due to connection problems, etc.

# of accesses per user	% of users
0	45.3%
1	11.0%
2	7.0%
3	18.3%
4	5.3%
5	4.5%
6	1.9%
7	1.7%
8	0.8%
9	0.4%
10	0.3%
11	0.6%
12	0.1%
13	0.1%
14	0.4%
15	0.0%
>15	2.1%
---	---
n/a	100.0%

Total number of authorized attendees : 951

Top ten users by number of accesses

Whenever there are more than one access for an attendee, there is a chance that several persons are using one attendee's [username, password] to download the conference papers, in violation of the conditions of access for that attendee. But of course, that concerns grows as the number accesses rises to surprising levels. However, due to connection problems, especially when the pdfs were packaged in zip files of 200 Mbytes, there are legitimate access attempts in the list below.

1	115
2	100
3	92
4	85
5	61
6	58
7	48
8	36
9	32
10	30

List of hacker IPs by number of attempts

Click the title link to see a text file of "hacker" IPs ranked by number of attempts. Of course, there's no way for me to know if they are really [hacker, cracker]s or not.

User access versus security, [Open, Google]-Docs

Online document access is NOT set up to be easiest to use, it is a compromise between [ease of use, security, managability (i.e how much volunteeer time and when?), etc]. Clearly, most attendees just want to click on a link in their [email, browser] and enter a password. They don't want to [learn, use] sftp programs! For example, my experience is that few people are [familiar, good] with [public, private] keys, in spite of routinely handling proprietary data. Sure, their IT systems handle that, but I see a major problem with security skills that are far below minimum.

IEEE security needs for online paper access of a conference are very fuzzy - one could easily put together a very poor system that meets most of their "specifications", but that does not protect the copyright in a meaningful way. Then again, IEEE has the right to [approve, reject] approaches. I doubt they have done so?

In my view, [Google Docs, Dropbox, etc] are INADEQUATE systems as commonly used from a security point of view, insofar as I interpret IEEE [needs, specifications]. They are better than posting the papers publicly, but not by much. I only went a small step beyond that by capping accesses in a reactionary mode, but the scrutiny was direct, and affected many.

I understand that conferences are using [Google Docs, Dropbox, etc], so perhaps that precedence will become a specific IEEE policy? It would have been much easier for me to take that route, but I know perhaps too little about the details to be comfortable with them. There is a lot of information available, but I didn't frind concrete examples for going beyond the [simple, basic] level (probably scripting etc is available). In contrast, sftp has a huge amount of [information, power, flexibility, simplicity, programmability].

There are many very powerful proprietary systems for doing the job, but they are targeted to large scale commercial systems, and I suspect that costs would mostly be unacceptable to a conference.

Security-driven complications, future improvements

01Aug2019 Howell : This is a very [incomplete, random, shallow] scratchpad of points to consider for the future. The assumption here is that an extremely simple ssh-sftp setup is used.

Cron-automated bash scripts to analyse user accesses and deactivate excessive users.
User deactivation - should be based on download volume cutoff (eg 1.1 Gigabytes for IJCNN2019), not the # of accesses! This is because it's hard to find the time to follow-up, especially during the conference where volunteers want to attend sessions.
Conference management company email addition of new registrants - automated system.
Rather than emailing permanent passwords, it would be more secure for users to email the system for a 24 hour password
Tons of other ideas would help....

Of course, the IEEE (Tomasz Cholewo) paper review system would be ideal for hosting the online paper system, as it ties in the [papers, copyright, security, IEEE Xplore] in a manner that would greatly facilitate work between the various groups. The conference management company (Association Resources for IJCNN2019) handles attendee registrations, so it is key for updates to be used for attendee registrations. The main headache are the small number of on-site registrations, which takes valuable time away from the volunteer doing this work, at exactly the wrong time!

Supporting software

Software status as of 08Aug2019. There's no documentation other than very simple comments in the [scripts, programs], and the "substantive names" of the variables which helps.

Click to see a directory of sftp-related bash scripts that were used for the analysis of the sftp site.
Nifty bash script to help with re-formatting emails to html blog format.
QNial program to update "raw" conference guides to "full" format with [menues, sponsors, explanations] that is ready to upload to Conference Guides website.

A rough idea of the work involved

My work on the sftp server started in a panic 21Jun2019 when othere options fell through, and ended 09Aug2019 (7 weeks duration). I was working most of my time on the server over that period, other than 4 days of the conference (the other 2 days I worked on the server!), 3 days travel, 1 day volunteer firefighting training, and perhaps 4 days of visits/other, for an estimated 30 to 37 days work (* 8 ~= 240 to 300 hours) on the sftp server requirement.

Another idea of the time and topic distribution is from the number of emails involved :

Sum of email folders	2587
individualised emails to attendees	1840
Overall total email count for Bill Howell	4427

The full table can be viewed at email counts for Bill Howell. This also gives a pretty good idea of author inquiries, and a range of non-sftp responsibilities for IJCNN2019.

Setup and hosting of the sftp server

The sftp server for online access by conference attendees to IJCNN2019 papers was set up by Michael Rae and William Astle of www.lexicom.ca, following my panic request to them. I have been using them for years for my IJCNN email server and my personal [public, ssh] servers. I am mentioning this out of appreciation for their servic, without any form of compensation from them.

REMINDERS : This Authors' Guide is specific to the IEEE conference paper system, and for the most part links to, or repeats, information that is COPYRIGHTED by the IEEE. Keep in mind that the IEEE processes something like 1,800 conferences per year, and delays exceeding a week are to be expected, with longer delays during high-activity parts of the year. So don't wait too long before starting each process!
Author preparations for the conference include (roughly in squence) :