Slickgrid expert? - Contract Role - Fully Remote

[ryanhamilton]

Hi all,

I’ve built a commercial analytics product that relies heavily on SlickGrid:
https://www.timestored.com/pulse/

We’ve implemented pivoting, sparklines, custom formatting, etc., but users are now asking for much more:

With growing customer demand, I’m looking for a UI engineer with deep grid / data-heavy frontend experience to help drive major improvements across both SlickGrid and the wider React app. Users want grouping, aggregation, checkboxes, better interactions—the full feature set done well.

I’m open to a flexible contract arrangement for the right person who can deliver high-quality work. Full role details here:
https://www.timestored.com/job/react-2025

I would really like to update to the latest slickgrid (without jquery) and feed any improvements we make back upstream to the original project.

[6pac]

Interesting. I have been moonlighting recently with a commercial company to add minor features to Slickgrid. However, I am (a) up to my eyeballs in work and (b) not familiar with React and modern tooling (mainly by choice, although I'm slowly upgrading to the tooling).

This does look like it could be an opportunity for Slickgrid to work on some of its building blocks though. I'd be happy to be involved in terms of giving architectural advice, and perhaps helping to build new features.

I also note that the example page contains grouping, aggregation, checkbox management and multicolumn pivoting examples.

For more sophisticated controls, such as column filter drop-down menus and good column auto sizing, a degree of integration integration with the backend or design framework is needed. Hence I have a number of these type of controls that are part of my personal development framework, and I notice that @ghiscoding also has a number of these type of controls in his Slickgrid-Universal monorepo.

[ryanhamilton]

Hi 6pac.

Thanks for replying and for all the work you've done to move slick grid forward.

Mostly the react part is a thin wrapper carefully created to avoid costly redraws. After that most the work has been slick grid and Pulse specific. Tying how user-config + data-frame updates + slick-grid combines to render the content. One pain has been that as I add one feature e.g. filters, I need to go back and re-work other parts e.g. table footers to make them work well together. That and getting pixel perfect alignment of cells and controls. I'm not very good at this myself so it takes me too long.

You are right. One tough choice I'll need to make together with any implementer is how much to integrate the backend (java/duckdb) and frontend (js). Currently the pivot is performed entirely on the backend and the filtering entirely on the frontend. This is likely to create inconsistencies as we add aggregations etc. unless careful choices are made. One problem particular to Pulse is that this data is often very large and fast updating so it's unlikely a purely client solution will scale.

If you’re open to discussing more, feel free to message me at ryan @@ timestored.com. To be transparent, I didn’t reach out directly because I wasn’t sure I could afford your level of expertise!

Thanks again for such a great component,
Ryan

[6pac]

Cheers, I'll email you soon.

Also wanted to acknowedge the huge amount of work @ghiscoding has put in. I was just treading water, really, keeping the project patched and introducing minor features, and he came along and (1) removed jQuery dependencies (2) implemented pinned rows and columns and (3) did the typescript conversion. That was massive.

It's gotten to the point that I am only just able to keep up with the new project. Ghislain also has a number of forks that are possibly more active than this repo. Thanks @ghiscoding !

My personal fork diverged probably 10 years ago and is heavily customised for my personal development framework, so I don't even use this repo in my own projects. One day I'll get to merge the two, but it's going to be a long and painful process.

[ghiscoding]

We don't have Pivot Table, that was already asked multiple times already. I never personally quite understood how Pivot works in Excel (it seems like some kind of grouping, is that it?).

For the React side of things, I have Slickgrid-React which I would suggest to use, even if I'm not a React developer myself. It was created by another user and I have been supporting it since couple years already and some large projects are starting to use it, one of the newest and largest that is starting to use Slickgrid-React is the VSCode extension mssql (since last week actually, so quite recent).

Yes I have been adding quite a few features, plugins and updated the project quite a lot with modern code (ES6, TypeScript, ...). Because I like the project and also because a few of these features were things that were needed for projects at works (like Tree Data). I learned a lot of things with this project and all of mine related to SlickGrid. It's a big world :)

I don't typically do contracts, Ben does, but if you're really interested into something very particular then perhaps some kind of features sponsoring might be doable (that is if it's doable).

[ryanhamilton]

| year | city | population |
| 2000 | NewYork | 200 |
| 2000 | London | 300 |

Pivot means values within a table become columns. Pivot on city:

| year | NewYork | London |
| 2000 | 200 | 300 |

I'll raise an issue with my target state and you can see for yourself.

I understand your position. I've contributed to a less successful open source project the last 3-4 years. I wouldn't accept contract work, it would ruin why I do it. If I started to calculate the monetary return rather than the joy it brings, I might grow depressed. :)

[ghiscoding]

There's already an open issue for Pivot Table #1020, no need to open another one. I have received some monetary contributions in the past via my GitHub profile and Ko-Fi. It's nice to receive when it comes in, but I won't get rich with that for sure, it's just like receiving a coffee once in a while (like once in a year of two). Regardless, I'm not sure that I would even know where to start with that kind of features, there was Ben that added this example-pivot a while ago but it's mostly static I think!? Ben wrote about it in that issue #1020

[6pac]

I've had a lot of experience with Pivot tables, due to a dark past history with MS Access and MS Reporting Services. The one I designed for the example is fully client side, and is perhaps a little cryptic if you don't know much about pivot tables (or Crosstabs, as they are called in Access). I tend to design for the most complex type, which allows several columns under a pivot value (I'd do a double header column but markdown doesn't support it), eg.

year	city	population	rabbits
1980	NewYork	145	63
1980	London	320	6393
2000	NewYork	200	45
2000	London	300	3458

becomes

year	NY - pop	NY - rabbits	Lon - pop	Lon - rabbits	Total - pop	Total - rabbits
1980	145	63	320	6393	465	6454
2000	200	45	300	3458	500	3503

I'd be happy to create a much better documented architecture, and possibly add editing ability to it. Interest in pivots has only been very minor.
The fact that you have a very large dataset means that you may be much better to pivot on the back end, but I'd suspect you can also condense the data down a lot before the pivot. In that case, it may be better to keep the pivot client side.

Regardless, the pivot point could be optionally on the server or client. Most of my libraries I try to create near-identical versions for both client and server so that tasks can be handed off from one to the other at any point with a minimum of fuss.

[6pac]

@ryanhamilton just a note I did email you a couple of weeks ago. No reply, so thought I'd let you know in case it went to junk.