Any user may read, or otherwise em-ploy, any published document on the sys-tem, or any private document to which hehas legitimate access. He can make anykind of links to it from his own cocu-ments, private or not.
Looks like yet another foundation of the open source community. Today, we have open source license as a way to encourage free distribution of software and protect the original author by bounding users of the open source software under defined terms and conditions.
What this conventionreally does is stress the singularityof each document, its external and in-ternal borders. Thus, we focus on theintegrity of the "document" as we'veknown it.
It is highly resemblant of how we store and access files today. But removing boundaries between documents and merge everything into one homepage seems to be an increasing trend in productivity tools, pushed by companies like Miro and Notion. Although that is only a front end style change, I wonder if such alternative could have a place in back-end system as well, understanding that the boundary of documents is indispensable when it comes to containment and distribution of certain damage or changes.
If you don't like itthere, hit some sort of a Retuxn But-ton and it pops your previous addressfrom a stack
It's once again amazing to see how Nelson envisioned the "return button", he had everything right, from user scenario, methodology and functionality.
unlessitcanshowyou,wordforword, whatpartsoftwoversions arethesame.
This reminds of many academic honesty tools such as Turnitin, where they are able to scour an enormous dataset, look for similar ones, and compare word for word their differences.
visualizethesealternativeversions asatreeintheongoingbraid,afork ingar-rangementwherebyonedocument becomest wo,eachof these daughterdocumentsmayinturnbecomeothers,etc.
I guess this is the basis of modern code version control and colab, or what now commonly known as github. Although, the programming fields seem to adopt it more than "law and public relations" as Nelson imagined.
Prismatic
I wonder how this "scroll horizontally in space" and "scroll vertically in time" went. This looks like a mind blowing direction that allows users to access any data with a timeseries view.
get ~ part you ~ ~;
This is the foundational idea behind so many protocols we use today. It makes "reading data" a much simpler and modular task than if we were to load an entire page or file all at once.
so that when you ask for agiven part of a given version at agiven time, it comes to your screen.
I wonder what would Nelson's reaction be if he gets to experience github, what we use for version controls, sql, what we use for data storage and query, and google docs, a user facing tool that allows version tracing amongst multiple users.
As a user makes changes, thechanges go directly into the storagesystem; filed, as it were, chronologi-cally.* Now with the proper sort ofindexing scheme, the storage facilitywe've mentioned ought also to be ableto deal with the problem of historicalbacktrack.
Again, this is exactly how data recovery is done today, through a system log that keeps a chronological log of all transactions made to the system which allows for historical backtrack.
A more fundamental use is to keeptrack of former states of the work, incase mistakes or wrong decisions needto be undone. This need, backtrack,is serious and important.
It is amazing to see how Nelson envisioned data recovery half a century ago. And his opinion is spot on, we do use system log to keep track of where and when unwanted transaction occurs.
The past is con-tinually changing - - or at least seemsto be, as we view it.
This marks an interesting viewpoint on access of past data. Since the moment it is created, it could be updated at any time, hence each time it is read, there could be drastic differences. How to keep track the timestamp of these "creates", "updates" and "reads" lies in the center of relational database design.
Sotheproblemistocreateageneral representationandstoragesystemthat willpermitauto-matic storage ofallstructuresausermightwanttoworkon,andthefaithfulaccountingoftheiJ: develop-ment
I guess this is the essence of modern CRUD protocol. A system that allows users to create, read, update and duplicate information from the back-end.
Thehuman mind does not work that way. It operates by asso-ciation
This reminds me of the book I read in the summer called "Algorithms to live by". Indeed our mind stores data not by categorization but by association with a LIFO queue. Recognizing that human mind is an exquisite work of arts, perhaps we can learn more from how the human minds deal with information and feed it back to the research of how machine best handles data.
the author of the future ceasewriting by hand or typewriter and talk directly to therecord
Reminds me of a takeaway from CSCI2370 Scientific Visualization--the optimal channel for machine2mind is graphics and the optimal channel for mind2machine is speech.
Graphics input to the mind is stuck in the endless loop of UIUX optimization and needs groundbreaking findings from cognitive science to help us understand how to better feed information to the eyes.
Speech input to the machine is stuck in constantly growing speech models that rely on semantic understanding with lack of logical understanding. What could incorporate logic into semantics? This could lead to the paradigm shift in chatbot and NLP.
A record, if it is to be useful to science, must becontinuously extended, it must be stored, andabove all it must be consulted
Bush's viewpoint on the extensibility of data is spot on. Data is only useful as long as it scales up.
Even today, the study of how data can be collected, integrated and reduced is at the front of data science.
invented a calculatingmachine which embodied most of the essential features ofrecent keyboard devices, but it could not then come intouse.
This feels analogous to the thrive of deep learning due to new GPUs though essential findings are made decades ago.
Often times, the development of new tools opens up new doors for scientific research. If there can exist a tool where researchers are notified instantly when a new tool that might fit their use gets developed, the productivity of research community can multiply ten folds.
because his publication did not reach the few who werecapable of grasping and extending it; and this sort of cata-strophe is undoubtedly being repeated all about us
Accessibility and identification are two of the most important issues facing researchers doing interdisciplinary work in my opinion. Making sure that whoever wants to find it can access it is one thing, and how to identify papers that are relevant to me is like finding a needle from a haystack.
With recommendation software that accurately pushes us the kittens we find the cutest, why is it that we can't be recommended papers that might inspire or enable our research?
The investigator is staggered by thefindings and conclusions of thousands of other workers—conclusions which he cannot find time to grasp, much lessto remember, as they appear.
It's worth noting that commercial softwares that assist business operations are ubiquitous and constantly evolving, whereas specialized tools for researchers are often outdated and user unfriendly.
We indeed need more platforms and visualization tools to help researchers make sense of related fields' work as they often might be the key to a new breakthrough.
it has provided a record of ideas andhas enabled man to manipulate and to make extracts fromthat record so that knowledge evolves and endures through-out the life of a race rather than that of an individual.
As an electrical engineer by training, his synopsis of what was later categorized as data mining, information analytics and scientific visualization seems very on point. I wonder how he envisioned the era of big data and how he would critique the reality of big data today.
The scientists, burying their oldprofessional competition in the demand of a commoncause
Perhaps in the brief decades post WWII, scientists were more united and working towards a common goal of making the world a better place.
However, I don't think this phenomenon lasted much long. We see researchers withholding discoveries and scandalizing competitors so that they themselves publish first, especially in less economically developed nations.
If researchers truly want to bury their old-time habits, perhaps when all of them have contributed to publishing failures and miss-findings in their research, I would be convinced.
But in the end, it is a game theory trap. Doing so decreases their own utility while increases utility for all.
Is there anyway we can overcome this?
inventions have extended man’s physicalpowers rather than the powers of his mind
One can look at this remark in two perspectives. 1: before the age of AI, inventions enhance physical powers, but since the age of AI, many more inventions amplify the power of brains. 2. intelligence amplifying inventions can always be misused in a way that harms innocent civilians, i.e. facial recognition that discriminates against certain ethnic groups.
which one do you agree?