Profile cover photo
Profile photo
sankarshan mukhopadhyay
Husband, friend, book reviewer and, an all round nice guy
Husband, friend, book reviewer and, an all round nice guy
sankarshan's posts

I logged into the website; selected from the restaurants nearby; picked a few items and attempted to place the order. At this point I had enough "Swiggy Money" in my account to cover for the order value. The payment flow for took me over to Paytm to pay 0 INR; it failed. I retried and this time I opted for Credit Card/Netbanking; it failed again with the "Transaction failed due to incorrectly calculated hash parameter". As I am not excited with the prospect of CoD/Cash-on-Delivery, I attempted to resolve this via a chat with an agent. I was advised to try again. As this was going along the path of the "Robert Bruce" bits of old, I sent an email to the support@ handle for the startup. I am yet to hear from them. However, this transaction failure thing is not new - it has happened in the last couple of orders as well.

Additionally, I noticed that the PayU Money gateway has been taken out but I do not recall any notification to the users of the system that this has happened. I'd have preferred it that way. Over on Twitter, I was advised to use the app with the hope that it will resolve itself. That is equally mysterious isn't it? It used to work via the browser, now it does not and yet it is expected to work on the app.

Post has attachment
Sometimes the photo app springs a surprise. This is one - this "panorama" was not intended, the app stitched it in and made a few corrections. The last vacation in 2015 was at the eastern most range of the Kaziranga National Park. The area behind the berm is the start of the forest. This is early in the morning, looking outside from the cottage at the resort.

Except for re-reads, of which there are a few, is the list of books covered in 2015. In terms of numbers that's a bit less than 2014. In terms of a proliferation of topics and authors, that's probably as varied as any of the other years. There are a few which deserve mention (and perhaps recommendations). I'll edit this post and create a list.

If an event is consistently drawing 1000+ crowds, the safest thing the organizers can do is not fix what is not broken. However, if the event happens to be PyCon India, the best bet going forward is to think through the need for the event and if required, tailor it to really be a celebration of the passion of the contributors and participants in the Python world.

PyCon India 2015 had around 1300+ attendees, a good mix of sponsors and a portfolio of talks that were mostly a let-down. The hardest part now is to figure out how to create a 11 month plan for PyCon India instead of a 4 day event plan. In other words, PyCon India would need to be more PSSI than PyCon India organizers (or, simply put, don't miss the woods for the trees). Over the years the event has always had this tug-of-war along the depth/complexity of the talks on offer and the intended audience. There have been multiple discussions on the list about the need to be mindful of an audience who are beginners. Similarly, there have been conversations around using the event to provide a platform to first time speakers. Both of these are laudable goals and are in tune with the idea that a flagship event should contribute to an increase of knowledge and proficiency. The small downside of the plan is that a build-up to the event is impacted. There is nothing wrong with having 101 talks. The problem is when they come at a cost - possible interesting topics are traded off to include these. I do not have visibility into the selection, curation and arbitration process for talks at PyCon India. But if the final schedule is something to go by, then it is an indictment of sorts - the committee would do well to publicly introspect on the decisions and propose a new path.

The implicit assumption that local chapters would enthusiastically participate in the process is not based on reality. It is incredibly difficult to sustain a momentum of events/activities leading to a big event. There are far too many orphaned groups which provide a clear evidence that organic growth appealing to the "community building" aspect do not last. A "community" is really a solid group of passionate contributors who have agreed upon some common goals and objectives. Everything else around this concept is merely infrastructure and enablers. To be able to gain from the meetings of these passionate users, PSSI does have to consider a variety of options beyond the traditional "monthly meetup somewhere". This may just be an anecdote-of-one, but the traffic around Python topics is really low on the local lists while it isn't that looking at upstream community lists. The availability of "Poster Sessions" at the 2015 event is a good first step to encourage those interested to highlight their work; draw in alliances from other upstream communities (eg. OpenStack etc). Also, a fantastic start in providing child care facilities at the event.

It behoves the PSSI to make the search for the Kenneth Gonsalves Award recipient a year long process instead of the uncommonly hasty last minute scramble. This statement is not to demean or, disrespect the recipients, all of whom have done stellar work. But, this is to emphasize a larger participation from everyone possible. It is also the duty of the PSSI to be planful about the messaging and content leading up to the event. In 2015 there have been a number of instances which can (gently) be termed as faux pas.

I was slightly unhappy about the lack of communication around the AGM of PSSI which is usually organized on the last day of the event. Perhaps it was just not widely publicized.

Lastly, while sponsors have every right to decide what to do with the booth space once they have paid up the monies, the organization does need to be cognizant of the fact that the events of 2015 (a particular Gold sponsor doing stall tear down on first day) create an interesting precedence. How it spins out in future iterations of the event would be left to those who attend them. #pyconindia  

There are times when reminds me of another startup - Both of them made plays to get market share and hopefully customer loyalty. During the early days this was amply demonstrated by exemplary service, a focus on listening to and addressing customer issues, trying out new things etc But over a period of time, the issue of "scaling up" catches up with all. My peeves with are fairly well documented everywhere and now I don't place any order from them at all. is displaying similar patterns as well - infrequent delays in delivery of orders, plastic cutlery that is brittle and has deteriorated in quality (they should really include a choking hazard warning), quality of meals seeing a progressive decline. The most telling piece is something different. In the early days, providing feedback to Flipkart would ensure that you receive a discount voucher or, similar. The other day I took a bit of time to write in reasonable detail to the team about things which can improve. The end result - a discount voucher. This entire throw-away tokenism is ugly. A sort of "oh! you took time to write, here's a bone for you!".

Post has attachment
In the Global South especially, people who get scholarships actively or unconsciously suppress the development of their local Wikimedia community so that they retain a leadership role and remain the most eligible people to receive scholarships, grants, attention from Wikimedia community leaders, and other privileges.

Post has attachment
This was on a browser tab for a couple of days now. Stole some time to watch the talk/presentation. Very well put together one.

Post has attachment

Post has attachment
If you've been with an organization for a significant duration, the perspective and understanding is often sharply different than what outsiders have. I've been at #RedHat  for the greater part of 11 years and as is usually expected, faced the "How is it to work at Red Hat?" question a number of times.

The question is simple and perhaps the audience demands a pithy response. The truth is that such an answer is pretty much impossible to craft. How do you begin to explain a company that has as its mission statement the line "To be the catalyst in communities of customers, contributors, and partners creating better technology the open source way." (cf. ?

When I joined, it was truly a "small" company (cf. A year or so before that date almost the entire company assembled for a form of "all hands". And that's pretty much the only/last time it happened. I joined a Red Hat that allowed a somewhat rank newcomer to absorb and assimilate information at a rate that was (at least for me) truly drinking from a firehose. The mailing lists had epic conversations around technology (one would either have Subhendu asking about a tricky implementation or, providing a concise response to a somewhat hand-wavy set of queries) and general things (memo list email threads from Rik, Arjan, Mike Harris, Alan Cox and Tiemann were a given). It was the Red Hat where the then CEO went up on stage (was it Linuxworld?) to talk about patents/IP and cite "Britney and her clones" in his spiel. It was also the company where the intranet had a page from James Morris that had the most click-bait-y title one could ever think of.

And locally, the then-JV which I joined was taking large inroads into the business landscape of the country. A decade later, sitting back and thinking about the "good old days", it is incredibly satisfying to see how things have changed. The "enterprise Linux" company which always punched above its weight now has a footprint that puts it right out there with infrastructure plays across nearly all aspects of a modern enterprise IT set up.

Anyway, the point is, that to an extent, the "What is Red Hat and how is it" side of the question was often answered through innovations in technology, spectacular work in upstream communities and good solid hard work. In some form, the absence of a canon, so to speak, enabled a lot of commentators to write about the company and shape the narrative. While reading through Jim Whitehurst's book (cf. I realized that this would perhaps be the first step in developing the narrative from the perspective of an insider. And including the points of view of individuals I've seen, known or, met within the company. And that's why it is a great read.

Any narrative about Red Hat tends to work around the theme of "enterprise open source company" with typical emphasis on the "open source" parts of how things might work. While this is largely true, the very notion that an agile organization structure can be created drawing upon the fundamental principles is something this book provides a number of citations about. It is not easy work and the surprises which Jim tripped into are well narrated. The part I did like is eschewing the "Red Hat is special" construct of an argument. Instead, it draws upon numerous experiments (successful and failures), conversations, incidents to demonstrate that there are a set of basic guidelines which emerge. A template which can be consciously or, deliberately adopted.

The principle that leadership can and will emerge often from the unlikeliest instances and individuals can come together to make a difference is a powerful one. That arc is a strong complement to the values of the company as well as the singular aspect of an open source community - collaboration. The book does well to curate and collate such instances which the (non-RHT) readers can relate well to. But, for the associates (and alumni) there are daily reminders of such events and the fact that this is "just how we get things done". 

The book weaves in the values (Freedom, Courage, Commitment, Accountability) as a practice and provides the scaffolding to the various decisions which have been made by Jim and his leadership team. However, the part I like most is that it provides a basis to create a well thought answer to the "How is it like to work there?" question. And with the material being generally available, the top tier talent seeking to join the company will have a well defined perspective to their choice. That's a fantastic thing!

The happy anecdote from a product release - 

"This update fixes the following bug: A race condition in the malloc API family of functions could cause a deadlock leading to gluster NFS and Fuse mounts becoming unresponsive while running large amounts of I/O. The race condition in malloc has been removed and gluster NFS and Fuse mounts no longer hang in the described situation."

A number of fine folks came together to make this happen :)
Wait while more posts are being loaded